Principles and Practice of Stereotactic Radiosurgery

Principles and Practice of Stereotactic Radiosurgery Lawrence S. Chin, MD • William F. Regine, MD Editors Principles...

Author: Lawrence Chin

109 downloads 1379 Views 14MB 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

Principles and Practice of Stereotactic Radiosurgery

Lawrence S. Chin, MD • William F. Regine, MD Editors

Principles and Practice of Stereotactic Radiosurgery

Editors Lawrence S. Chin, MD Professor and Chairman Department of Neurosurgery Boston University School of Medicine Boston, MA, USA

William F. Regine, MD Professor and Chairman Department of Radiation Oncology University of Maryland Medical School Baltimore, MD, USA

ISBN: 978-0-387-71069-3 e-ISBN: 978-0-387-71070-9 DOI: 10.1007/978-0-387-71070-9 Library of Congress Control Number: 2007931622 © 2008 Springer Science+Business Media, LLC. All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identiﬁed as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper. 9 8 7 6 5 4 3 2 1 springer.com

Foreword hen ﬁrst asked to write a foreword to Principles and Practice of Stereotactic Radiosurgery, I hesitated. There have been so many books and peer-reviewed papers written on this subject that I questioned whether another book would add much. However, after Larry and Bill shared the contents of this book with me, I had to change my mind. From my point of view, this book signals the completion of decades of hard work. Pioneering Gamma Knife surgery during the 1970s and 1980s was often a lonely endeavor, with long ﬂights to innumerable meetings on all continents in order to speak about what we were doing in Stockholm. These talks were usually met with polite skepticism, sometimes even outright hostility, initially from neurosurgeons and later by other specialties as well. During the 1970s and 1980s, a large number of foreign colleagues came through Stockholm. The neurosurgical department at the Karolinska Hospital had a good reputation in stereotactic and functional neurosurgery, and many of the visitors later became prominent proponents of radiosurgery. In the mid-1980s, the adapted linear accelerator was pioneered by Federico Colombo in Italy and Osvaldo Betti in Argentina. Later, others joined the ranks. Nevertheless, it would take until the time of the ﬁrst U.S. Gamma Knife installation in 1987 for the concept of noninvasive brain surgery to gain credibility. Slowly, the veracity of our claims from the 1970s began to take hold. By then, we already knew what the next steps would be for us; namely, the further reﬁnement of the Gamma Knife in parallel with the incorporation of stereotactic principles, concepts of precision and accuracy, and imaging into the practice of radiotherapy in the rest of the body. In 1989, we called this stereotactic radiation therapy, or SRT. We believed that there was a gray zone between radiosurgery and conventional radiotherapy that was worthy of attention. The idea was to use increased precision as a way to allow higher doses and maybe fewer fractions in radiotherapy. This could, we thought, improve the treatment of lesions too large for radiosurgery and too small for radiotherapy. I tried to establish a collaboration with one of the major suppliers of linear accelerators in order to explore this gray zone between radiosurgery and conventional radiotherapy, but there was no interest at all at the time. With the rate of development seen over the past 10 years, one wonders what lies ahead for radiation medicine. My guess is that we will see a somewhat slower rate of development in the radiation delivery systems themselves but an increasing emphasis on the integration of radiation delivery systems with software systems such as planning, imaging, and cancer registry systems. On the clinical side, we will see the continued reemergence of radiosurgery in the treatment of functional brain disorders, including epilepsy, movement disorders, obsessivecompulsive states, and possibly severe endogenous depression. In ophthalmology, there is already exploratory work being done in, for example, glaucoma, macular degeneration, endocrine orbitopathy, and uveal melanomas. We will also see the application of stereotactically guided radiation therapy for disorders that currently are not part of standard practice. These will include the precise targeting of intra- and extraaxial spine lesions, as well as disease in the paranasal sinuses and the larynx. Radiation therapy for, for example, lung and prostate cancer will beneﬁt from the increased precision, allowing higher doses to be delivered despite the close proximity of heart muscle and colon. This book is a very good illustration of the term helicopter perspective. It is particularly impressive in that it really approaches the whole spectrum of disease in a very thorough manner. The title of the book is actually quite humble, belying as it does the fact that all available treatment modalities are represented, compared, and put in perspective. It epitomizes the word comprehensive!

W

v

vi

foreword

This textbook contains a wealth of information and truly encompasses the whole ﬁeld of radiosurgery, regardless of technology and regardless of which disease the reader wants to learn more about, be it in the brain, in the spine, in the eye, or elsewhere in the body. For residents and newcomers to the ﬁeld and for the experienced clinician, this volume will represent an invaluable source of information as you strive to design the best therapeutic approach to your individual patients. This is a book that deserves a prominent—and easy to reach—place on our bookshelves. Dan Leksell, MD

Preface he practice of stereotactic radiosurgery has developed from an unprecedented degree of collaboration among practitioners of neurosurgery, radiation oncology, and medical physics. Because the patients and the diseases that we treat are at the intersection of surgery, radiation, and medical therapy, we felt that a full description of this ﬁeld required a comprehensive and global approach to the subject. Not only would we discuss the main diseases that comprise the typical intracranial radiosurgery practice, such as brain metastases and AVMs, we would also cover the fast-growing ﬁeld of extracranial radiosurgery, as well as more unusual indications such as epilepsy and psychiatric disease. We also wanted to avoid a bias toward Gamma Knife radiosurgery, which tends to dominate most publications. Therefore, we made sure that all major stereotactic radiosurgery techniques were represented. In selecting contributing authors, we felt it critical to enlist the help of our international colleagues who have been at the vanguard of expanding radiosurgery indications. After all, stereotactic radiosurgery was invented by a Swedish neurosurgeon. We organized this book into ﬁve main sections, with the ﬁrst few providing important background for the rest of the book. Part One covers the history of radiosurgery, basics of neuroimaging, and a general overview of key concepts in radiosurgery. Part Two concentrates on the principles of radiation physics and radiobiology that explain the noninvasive, yet powerful, nature of stereotactic radiation treatments. Other topics covered include treatment planning and the designing of a radiosurgery unit. We think this portion of the book will be of particular interest to medical physicists, as it is intended to be a practical guide for the running and maintenance of a radiosurgery center. Part Three contains reviews of the major techniques of stereotactic radiosurgery by physicians who are considered by most to be the leading ﬁgure in their disciplines. We hope you ﬁnd their insights as valuable to your practice as we did. Part Four includes eighteen chapters that describe the major disease types treated by practitioners of stereotactic radiosurgery. In each chapter, we asked the authors to provide case reports of actual patients that illustrated the approach, treatment plan, and outcome of their treatment, thus providing a blueprint to follow for those new to the specialty. One of the more unusual aspects of this book is the inclusion of “perspective” chapters that follow a main topic chapter. We felt that having minichapters written by experts in the ﬁeld who might have a differing viewpoint would provide the most balanced approach to diseases that often have more than one effective treatment. The last part of this book presents topics related to patient care and the often ignored but critical socioeconomic side of stereotactic radiosurgery. The diverse subjects tackled include complication management, cost-effectiveness and quality of life, building a radiosurgery practice, and nursing issues. We also included a few topics that have controversial aspects: regulatory and reimbursement issues, medicolegal pitfalls, and radiosurgery semantics. In these chapters, the reader will ﬁnd that some author opinion is unavoidable but does not necessarily reﬂect the views of the editors and the publisher. Our mantra for this book was to be comprehensive and balanced, but we recognize that there will always be disagreements on many of the topics discussed in this book. We hope that this book will be informative but also stimulate a healthy and constructive dialog among its readers. We must continually examine our results in this critical manner to provide the best care for our patients. This book has been the culmination of several years of planning and execution by a large number of very talented individuals. First, we are indebted to the authors of the individual chapters, who provided their time and expertise in the creation of this project. We would like to thank the editors and staff at Springer who brought dedication and excellence to this project: Beth Campbell, Paula Callaghan, Barbara Lopez-Lucio, and Brad Walsh. We thank Barbara Chernow who rounded this book into its ﬁnal form. We thank our assistants Debbie

T

vii

viii

preface

Redmon, Yvette Green, and Michele Murphy who kept our practices humming while dealing with manuscripts, contributors, editors, and Fed-Ex. Our professional lives owe a debt to the mentors who brought us into neurosurgery, radiation oncology, and the world of radiosurgery, Buz Hoff, Martin Weiss, Michael Apuzzo, Steven Giannotta, Howard Eisenberg, Simon Kramer, Larry Kun, and Jay Loefﬂer. Most importantly, we thank our wives Rita and Julie, along with our children, and the rest of our family and friends for their constant love and support. Lastly, we thank our patients, colleagues, trainees, and students who provided the inspiration for this book. Lawrence S. Chin, MD William F. Regine, MD

Contents Foreword by Dan Leksell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PART I

The Fundamentals

1

The History of Stereotactic Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . Michael Schulder and Vaibhav Patil

2

Neuroimaging in Radiosurgery Treatment Planning and Follow-up Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clark C. Chen, Paul H. Chapman, Hanne Kooy, and Jay S. Loefﬂer

3

Techniques of Stereotactic Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . Chris Heller, Cheng Yu, and Michael L.J. Apuzzo

PART II

v vii xv

3

9

25

Radiation Biology and Physics

4

The Physics of Stereotactic Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . Siyong Kim and Jatinder Palta

5

Radiobiological Principles Underlying Stereotactic Radiation Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David J. Brenner

33

51

6

Experimental Radiosurgery Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ajay Niranjan and Douglas Kondziolka

61

7

Treatment Planning for Stereotactic Radiosurgery . . . . . . . . . . . . . . . . . David M. Shepard, Cedric Yu, Martin Murphy, Marc R. Bussière, and Frank J. Bova

69

8

Designing, Building and Installing a Stereotactic Radiosurgery Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lijun Ma and Martin Murphy

PART III

91

Stereotactic Radiosurgery Techniques

9

Gamma Knife Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ajay Niranjan, Sait Sirin, John C. Flickinger, Ann Maitz, Douglas Kondziolka and L. Dade Lunsford

107

10

Linear Accelerator Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . William A. Friedman

129

ix

x 11

contents

Proton Beam Radiosurgery: Physical Bases and Clinical Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Georges Noel, Markus Fitzek, Loïc Feuvret, and Jean Louis Habrand

141

12

Robotics and Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cesare Giorgi and Antonio Cossu

163

13

CyberKnife Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . John R. Adler Jr., Alexander Muacevic, and Pantaleo Romanelli

171

PART IV

Treatment of Disease Types

14

Brain Metastases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . John H. Suh, Gene H. Barnett, and William F. Regine

181

15

Metastatic Brain Tumors: Surgery Perspective . . . . . . . . . . . . . . . . . . . . Raymond Sawaya and David M. Wildrick

193

16

Brain Metastases: Whole-Brain Radiation Therapy Perspective . . . . . . Roy A. Patchell and William F. Regine

201

17

High-Grade Gliomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David Roberge and Luis Souhami

207

18

Malignant Glioma: Chemotherapy Perspective . . . . . . . . . . . . . . . . . . . . Roger Stupp and J. Gregory Cairncross

223

19

Meningioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carlos A. Mattozo and Antonio A.F. de Salles

233

20

Meningioma: Surgery Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lawrence S. Chin, Pulak Ray, and John Caridi

249

21

Intracranial Meningioma: Fractionated Radiation Therapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Leland Rogers, Dennis Shrieve, and Arie Perry

257

22

Meningioma: Systemic Therapy Perspective . . . . . . . . . . . . . . . . . . . . . . . Steven Grunberg

271

23

Acoustic Schwannoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . William M. Mendenhall, Robert J. Amdur, Robert S. Malyapa, and William A. Friedman

275

24

Acoustic Neuroma: Surgical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . Indro Chakrabarti and Steven L. Giannotta

283

25

Acoustic Neuromas and Other Benign Tumors: Fractionated Stereotactic Radiotherapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . David W. Andrews, Greg Bednarz, Beverly Downes, and Maria Werner-Wasik

289

contents

xi

26

Pituitary Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kintomo Takakura, Motohiro Hayashi, and Masahiro Izawa

299

27

Pituitary Adenomas: Surgery Perspective . . . . . . . . . . . . . . . . . . . . . . . . . William T. Couldwell and Martin H. Weiss

309

28

Pituitary and Pituitary Region Tumors: Fractionated Radiation Therapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jonathan P.S. Knisely and Paul W. Sperduto

317

Pituitary and Pituitary Region Tumors: Medical Therapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mansur E. Shomali

327

29

30

Pediatric Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andrew Reisner, Nicholas J. Szerlip, and Lawrence S. Chin

31

Pediatric Brain Tumors: Conformal Radiation Therapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thomas E. Merchant

331

341

32

Pediatric Brain Tumors: Chemotherapy Perspective . . . . . . . . . . . . . . . . Amar Gajjar

351

33

Pineal Region Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gregory P. Lekovic and Andrew G. Shetter

355

34

Pineal Region Tumors: Surgery Perspective . . . . . . . . . . . . . . . . . . . . . . . Alfred T. Ogden and Jeffrey N. Bruce

365

35

Pineal Tumors: Fractionated Radiation Therapy Perspective . . . . . . . . Steven E. Schild

371

36

Pineal Region Tumors: Chemotherapy Perspective . . . . . . . . . . . . . . . . . Barry Meisenberg and Lavanya Yarlagadda

377

37

Skull Base Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stefanie Milker-Zabel, Young Kwok, and Jürgen Debus

383

38

Skull Base Tumors: Surgery Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . James K. Liu, Oren N. Gottfried, and William T. Couldwell

393

39

Skull Base Tumors: Fractionated Stereotactic Radiotherapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . René-Olivier Mirimanoff and Alessia Pica

401

40

Head and Neck Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Daniel T.T. Chua, Jonathan Sham, Kwan-Ngai Hung, and Lucullus Leung

411

41

Head and Neck Tumors: Surgery Perspective . . . . . . . . . . . . . . . . . . . . . Gregory Y. Chin and Uttam K. Sinha

421

xii 42

contents

Head and Neck Malignancies: Chemotherapy and Radiation Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mohan Suntharalingam, Kathleen Settle, and Kevin J. Cullen

425

43

Spinal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Robert L. Dodd, Iris Gibbs, John R. Adler Jr., and Steven D. Chang

431

44

Spine Tumors: Surgery Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gabriel Zada and Michael Y. Wang

443

45

Spinal Metastases: Fractionated Radiation Therapy Perspective . . . . . Eric L. Chang and Almon S. Shiu

455

46

Arteriovenous Malformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bruce E. Pollock

459

47

Arteriovenous Malformations: Surgery Perspective . . . . . . . . . . . . . . . . Ricardo J. Komotar, Elena Vera, J. Mocco, and E. Sander Connolly Jr.

473

48

Cerebral Arteriovenous Malformations: Endovascular Therapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Felipe C. Albuquerque, David Fiorella, and Cameron G. McDougall

479

49

Cavernous Malformations and Other Vascular Diseases . . . . . . . . . . . . Ajay Niranjan, David Mathieu, Douglas Kondziolka, John C. Flickinger, and L. Dade Lunsford

491

50

Cerebral Cavernous Malformations: Surgical Perspective . . . . . . . . . . . Robert L. Dodd and Gary K. Steinberg

503

51

Cavernous Malformations and Other Vascular Abnormalities: Observation-Alone Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sepideh Amin-Hanjani and Frederick G. Barker II

513

52

Trigeminal Neuralgia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lawrence S. Chin, Shilpen Patel, Thomas Mattingly, and Young Kwok

519

53

Trigeminal Neuralgia: Surgical Perspective . . . . . . . . . . . . . . . . . . . . . . . . David B. Cohen, Michael Y. Oh, and Peter J. Jannetta

527

54

Trigeminal Neuralgia: Medical Management Perspective . . . . . . . . . . . Neil C. Porter

535

55

Movement Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sangjin Oh, Ajay Niranjan, and William J. Weiner

541

56

Movement Disorders: Deep-Brain Stimulation Perspective . . . . . . . . . . John Y.K. Lee, Joshua M. Rosenow, and Ali R. Rezai

549

57

Movement Disorder: Medical Perspective . . . . . . . . . . . . . . . . . . . . . . . . Sangjin Oh and William J. Weiner

559

58

Psychiatric and Pain Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jason Sheehan, Nader Pouratian, and Charles Sansur

563

contents

xiii

59

Intractable Epilepsies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jean Régis, Fabrice Bartolomei, and Patrick Chauvel

573

60

Epilepsy: Surgery Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Keith G. Davies and Edward Ahn

583

61

Ocular and Orbital Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gabriela Šimonová, Roman Liscˇák, and Josef Novotný Jr.

593

62

Stereotactic Body Radiation Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laura A. Dawson

611

63

Stereotactic Body Radiation Therapy: Fractionated Radiation Therapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gordon W. Wong, Rafael R. Mañon, Wolfgang Tomé, and Minesh Mehta

64

Stereotactic Body Radiation Therapy: Brachytherapy Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Caroline L. Holloway, Desmond O’Farrell, and Phillip M. Devlin

PART V

635

643

Patient Care and Socioeconomic Issues

65

Complications and Management in Radiosurgery . . . . . . . . . . . . . . . . . . Isaac Yang, Penny K. Sneed, David A. Larson, and Michael W. McDermott

649

66

Cost-Effectiveness and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . Minesh Mehta and May N. Tsao

663

67

Regulatory and Reimbursement Aspects of Radiosurgery . . . . . . . . . . Rebecca Emerick

673

68

Medicolegal Issues in Stereotactic Radiosurgery . . . . . . . . . . . . . . . . . . . April Strang-Kutay

681

69

The Semantics of Stereotactic Radiation Therapy . . . . . . . . . . . . . . . . . . Louis Potters

687

70

Building a Radiosurgery Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . N. Scott Litofsky and Andrea D’Agostino-Demers

691

71

Patient Care in Stereotactic Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . Terri F. Biggins

699

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

709

Contributors John R. Adler Jr., MD Professor of Neurosurgery, Stanford University Medical Center, and Attending Neurosurgeon, Stanford University Medical Center, Stanford, CA, USA Edward Ahn, MD Fellow in Neurosurgery, Department of Neurosurgery, Children’s Hospital of Boston, Boston, MA, USA Felipe C. Albuquerque, MD Assistant Director of Endovascular Neurosurgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, AZ, USA Robert J. Amdur, MD Professor of Radiation Oncology, University of Florida College of Medicine, Gainesville, FL, USA Sepideh Amin-Hanjani, MD Assistant Professor, Neurosurgery, University of Illinois at Chicago, Chicago, IL, USA David W. Andrews, MD Professor and Vice Chairman, Chief, Division of Neuro-Oncologic Neurosurgery & Stereotactic Radiosurgery, Thomas Jefferson University, Philadelphia, PA, USA Michael L.J. Apuzzo, MD Edwin M. Todd and Trent H. Wells Professor of Neurosurgery, Radiation, Oncology, Biology and Physics, Department of Neurological Surgery, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA Frederick G. Barker II, MD Associate Professor, Department of Neurosurgery, Harvard Medical School; Associate Visiting Neurosurgeon, Brain Tumor Center, Massachusetts General Hospital, Boston, MA, USA

Gene H. Barnett, MD, FACS Professor of Surgery, Cleveland Clinic Lerner College of Medicine; Director, Brain Tumor Institute, Cleveland Clinic, Cleveland, OH, USA Fabrice Bartolomei, MD, PhD Service de Neurophysiologie Clinique, Université de la Méditerranée, Marseille, France Greg Bednarz, PhD Medical Physicist, Department of Radiation Oncology, Thomas Jefferson University, Jefferson Hospital for Neuroscience, Philadelphia, PA, USA Terri F. Biggins, RN, BSN Patient Care Coordinator, University of Maryland, Gamma Knife Center, Baltimore, MD, USA Frank J. Bova, PhD Professor of Neurosurgery, University of Florida, Gainesville, FL, USA David J. Brenner, PhD, DSc Professor of Radiation Oncology and Public Health, Center for Radiological Research, Department of Radiation Oncology, Columbia University Medical Center, New York, NY, USA Jeffrey N. Bruce, MD Professor of Neurological Surgery, Department of Neurosurgery, Columbia University—College of Physicians and Surgeons, New York, NY, USA Marc R. Bussière, MSc, DABR Medical Radiation Physicist, Department of Radiation Oncology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA J. Gregory Cairncross, MD, FRCPC Department of Clinical Neurosciences, University of Calgary, Foothills Hospital, Alberta, Canada

xv

xvi

contributors

John Caridi, MD Resident, Department of Neurosurgery, University of Maryland, Baltimore, MD, USA

E. Sander Connolly Jr., MD Associate Professor, Department of Neurological Surgery, Columbia University, New York, NY, USA

Indro Chakrabarti, MD, MPH Neurosurgery Chief Resident, University of Southern California, Los Angeles, CA, USA

Antonio Cossu, MTE 3DLine Medical Systems, Milano, Italy

Eric L. Chang, MD Associate Professor, Department of Radiation Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA Steven D. Chang, MD Assistant Professor of Neurosurgery, Department of Neurosurgery, Stanford University School of Medicine, Stanford, CA, USA Paul H. Chapman, MD Professor, Department of Neurosurgery, Massachusetts General Hospital, Boston, MA, USA Patrick Chauvel, MD Service de Neurophysiologie Clinique, Université de la Méditerranée, Marseille, France Clark C. Chen, MD, PhD Fellow, Radiosurgery, Department of Neurosurgery, Massachusetts General Hospital, Boston, MA, USA Gregory Y. Chin, MD Attending Physician, Department of Head and Neck Surgery, Kaiser Permanente Walnut Creek Medical Center, Walnut Creek, CA, USA Lawrence S. Chin, MD Professor and Chairman, Department of Neurosurgery, Boston University School of Medicine, Boston, MA, USA Daniel T.T. Chua, FRCR Associate Professor, Department of Clinical Oncology, Queen Mary Hospital, The University of Hong Kong, Pokfulam, Hong Kong David B. Cohen, MD Functional Neurosurgery Fellow, Department of Neurosurgery, Allegheny General Hospital, Pittsburgh, PA, USA

William T. Couldwell, MD, PhD Professor and Joseph J. Yager Chair, Department of Neurosurgery, University of Utah, Salt Lake City, UT, USA Kevin J. Cullen, MD Director, University of Maryland Greenebaum Cancer Center, Professor of Medicine, University of Maryland Medical Center, Baltimore, MD, USA Andrea D’Agostino-Demers, MSN, EdD, CS, APRN, BC, NP Clinical Coordinator, Stereotactic Radiosurgery, Image-Guidance, and Functional Neurosurgery Programs, Division of Neurosurgery, UMASS Memorial Healthcare, Worcester, MA, USA Keith G. Davies, MD, FRCS Associate Professor, Department of Neurosurgery, Boston University School of Medicine, Boston, MA, USA Laura A. Dawson, MD Associate Professor, Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada Jürgen Debus, MD, PhD Department of Radiation Oncology and Radiation Therapy, University of Heidelberg, Heidelberg, Germany Antonio A.F. de Salles, MD, PhD Professor, Department of Surgery, Division of Neurosurgery, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA Phillip M. Devlin, MD Assistant Professor, Department of Radiation Oncology, Harvard Medical School; and Chief, Division of Brachytherapy, Department of Radiation Oncology, Dana Farber/Brigham & Women’s Cancer Center, Boston, MA, USA

xvii

contributors

Robert L. Dodd, MD, PhD Endovascular Fellow, Department of Neurosurgery, Stanford University, Stanford, CA, USA

Cesare Giorgi, MD Neurosurgeon, Department of Computer-assisted Neuro and Radiosurgery, Ospedale S. Maria, Terni, Italy

Beverly Downes, MS Chief Medical Physicist, Stereotactic Radiosurgery Units, Jefferson Hospital for Neuroscience, Philadelphia, PA, USA

Oren N. Gottfried, MD Resident, Department of Neurosurgery, University of Utah, Salt Lake City, UT, USA

Rebecca Emerick, MS, MBA, CPA Executive Director, International RadioSurgery Association (IRSA), Harrisburg, PA, USA

Steven Grunberg, MD Professor of Medicine, Department of Medical Oncology, University of Vermont, Burlington, VT, USA

Loïc Feuvret Centre de protonthérapie d’Orsay-Institut Curie, Campus universitaire, Orsay, France

Jean Louis Habrand CPO-Institut Curie, Orsay, France

David Fiorella, MD, PhD Staff Neuroradiology, Department of Neuroradiology and Neurosurgery, Cleveland Clinic Foundation, Cleveland, OH, USA Markus Fitzek, MD Radiation Oncology Center, Tufts—New England Center, Tufts University School of Medicine, Boston, MA, USA John C. Flickinger, MD Professor, Department of Neurological Surgery, University of Pittsburgh, Pittsburgh, PA, USA William A. Friedman, MD Professor and Chair, Department of Neurosurgery, University of Florida College of Medicine, Gainesville, FL, USA Amar Gajjar, MD Professor of Pediatrics, University of Tennessee, Director, Division of Neuro Oncology; Co-leader Neurobiology and Brain Tumor Program, Member and Co-Chair Department of Oncology, St. Jude Children’s Research Hospital, Memphis, TN, USA

Motohiro Hayashi, MD, PhD Lecturer of the Department of Neurosurgery, Chief of Gamma Knife Center, Tokyo Women’s Medical University, Shinjuku, Tokyo, Japan Chris Heller, MD Department of Neurological Surgery, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA Caroline L. Holloway, MD, FRCPC Radiation Oncologist, Department of Radiation Oncology, BCCA—Centre for the Southern Interior, Kelowna, BC, Canada Kwan-Ngai Hung, FRCS Consultant Neurosurgeon, Department of Surgery, The University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong Masahiro Izawa, MD, PhD Assistant Professor, Department of Neurosurgery, Tokyo Women’s Medical University, Shinjuku, Tokyo, Japan

Steven L. Giannotta, MD Chairman, Department of Neurosurgery, University of Southern California, Los Angeles, CA, USA

Peter J. Jannetta, MD Professor, Department of Neurosurgery, Drexel University School of Medicine; Vice-Chairman, Department of Neurosurgery, Jannetta Center for Cranial Nerve Disorders, Allegheny General Hospital, Pittsburgh, PA, USA

Iris Gibbs, MD Assistant Professor of Radiation Oncology, Stanford University, Stanford, CA, USA

Siyong Kim, PhD Department of Radiation Oncology, Mayo Clinic, Jacksonville, FL, USA

xviii Jonathan P.S. Knisely, MD, FRCPC Associate Professor, Department of Therapeutic Radiology, Yale University School of Medicine; and Yale Cancer Center, Yale–New Haven Hospital, New Haven, CT, USA Ricardo J. Komotar, MD Resident, Neurosurgery, Department of Neurological Surgery, Columbia University, New York, NY, USA Douglas Kondziolka, MD, FRCS, FACS Professor of Neurological Surgery, University of Pittsburgh, Pittsburgh, PA, USA Hanne Kooy, PhD Research Associate, Department of Radiation Oncology, Massachusetts General Hospital, Boston, MA, USA Young Kwok, MD Department of Radiation Oncology, University of Maryland Medical Center, Baltimore, MD, USA David A. Larson, PhD, MD, FACR Professor of Radiation Oncology and Neurological Surgery, Director, CyberKnife Radiosurgery Program, Co-Director, Gamma Knife Radiosurgery Program, Department of Neurological Surgery and Radiation Oncology, University of California San Francisco, San Francisco, CA, USA John Y.K. Lee, MD Assistant Professor, Department of Neurosurgery, University of Pennsylvania; Medical Director, Penn Gamma Knife at Pennsylvania Hospital, University of Pennsylvania, Philadelphia, PA, USA

contributors

Roman Liscˇák, MD 3rd Faculty of Medicine, Clinical Department of Neurosurgery, Charles University; Department of Stereotactic and Radiation Neurosurgery, Na Homolce Hospital, Prague, Czech Republic N. Scott Litofsky, MD, FACS Associate Professor, Director of Neuro-Oncology, Director of Radiosurgery, Division of Neurological Surgery, University of Missouri-Columbia School of Medicine, Columbia, MO, USA James K. Liu, MD Resident, Department of Neurosurgery, University of Utah, Salt Lake City, UT, USA Jay S. Loefﬂer, MD Chief Radiation Oncology, Department of Radiation Oncology, Massachusetts General Hospital, Boston, MA, USA L. Dade Lunsford, MD, FACS Professor and Chairman, Department of Neurological Surgery, University of Pittsburgh, Pittsburgh, PA, USA Lijun Ma, PhD Associate Professor, Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA Ann Maitz, MSc Assistant Professor, Department of Neurological Surgery, University of Pittsburgh, Pittsburgh, PA, USA Robert S. Malyapa, MD, PhD Assistant Professor, Department of Radiation Oncology, University of Florida College of Medicine, Jacksonville, FL, USA

Gregory P. Lekovic, MD, PhD, JD Resident Neurological Surgery, Division of Neurological Surgery, Barrow Neurological Institute, Phoenix, AZ, USA

Rafael R. Mañon, MD Department of Human Oncology, University of Wisconsin Hospital, Madison, WI, USA

Lucullus Leung, PhD Physicist, Department of Clinical Oncology, Queen Mary Hospital, Pokfulam, Hong Kong

David Mathieu, MD, FRCS(C) Visiting Assistant Professor, Department of Neurological Surgery, University of Pittsburgh, Pittsburgh, PA, USA

xix

contributors

Thomas Mattingly, MD Resident, Department of Neurosurgery, University of Maryland, Baltimore, MD, USA Carlos A. Mattozo, MD Professor, Department of Surgery, Division of Neurosurgery, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA Michael W. McDermott, MD, FRCSC Professor in Residence of Neurological Surgery, Halperin Endowed Chair, Vice Chairman, Department of Neurological Surgery, University of California San Francisco, San Francisco, CA, USA Cameron G. McDougall, MD Chief of Endovascular Neurosurgery, Barrow Neurological Institute— Neurosurgery, St. Joseph’s Hospital and Medical Center, Phoenix, AZ, USA Minesh Mehta, MD Professor and Chairman, Department of Human Oncology, University of Wisconsin Hospital, Madison, WI, USA Barry Meisenberg, MD Professor of Medicine, Chief Division of Hematology and Oncology, Department of Medicine, University of Maryland School of Medicine, Baltimore, MD, USA William M. Mendenhall, MD Professor, Department of Radiation Oncology, University of Florida College of Medicine, Gainesville, FL, USA

Alexander Muacevic, MD CyberKnife Center Munich, Munich, Germany Martin Murphy, PhD Associate Professor, Department of Radiation Oncology, Virginia Commonwealth University, Richmond, VA, USA Ajay Niranjan, MBBS, MS, MCh Assistant Professor of Neurological Surgery, University of Pittsburgh, Pittsburgh, PA, USA Georges Noel, MD Centre de lutte contre le Paul Strauss, Department of Radiotherapy, Strasbourg, France Josef Novotný Jr., MSc, PhD Assistant Professor, Department of Radiation Oncology, University of Texas Southwestern Medical Center, Dallas, TX, USA Desmond O’Farrell, CMD Senior Dosimetrist, Division of Brachytherapy, Department of Radiation Oncology, Dana Farber/Brigham & Women’s Hospital, Boston, MA, USA Alfred T. Ogden, MD Resident, Department of Neurological Surgery, Columbia University, New York, NY, USA

Thomas E. Merchant, DO, PhD Member and Chief, Division of Radiation Oncology, St. Jude Children’s Research Hospital, Memphis, TN, USA

Michael Y. Oh, MD Assistant Professor, Department of Neurosurgery, Drexel University School of Medicine; Co-Director, Stereotactic and Functional Neurosurgery, Department of Neurosurgery, Allegheny General Hospital, Pittsburgh, PA, USA

Stefanie Milker-Zabel, MD Departments of Radiation Oncology and Radiation Therapy, Hospital of Heidelberg, Heidelberg, Germany

Sangjin Oh, MD Fellow, Department of Neurology, University of Maryland School of Medicine, Baltimore, MD, USA

René-Olivier Mirimanoff, MD Professor, Department of Radiation Oncology, University Hospital of Lausanne (CHUV), Lausanne, Switzerland

Jatinder Palta, PhD Professor and Chief of Physics, Department of Radiation Oncology, University of Florida, Gainesville, FL, USA

J. Mocco, MD Resident, Neurosurgery, Department of Neurological Surgery, Columbia University, New York, NY, USA

Roy A. Patchell, MD Chief of Neuro-oncology, Professor of Neurology and Neurosurgery, University of Kentucky Medical Center, Lexington, KY, USA

xx

contributors

Shilpen Patel, MD Assistant Professor, Department of Radiation Oncology, University of Washington Medical Center, Seattle, WA, USA

Ali R. Rezai, MD Director, Brain Neuromodulation Center, Jane and Lee Seidman Chair in Functional Neurosurgery, The Cleveland Clinic Foundation, Cleveland, OH, USA

Vaibhav Patil, BA Department of Neurosurgery, New Jersey Medical School, Newark, NJ, USA

David Roberge, MD Assistant Professor, Department of Oncology, Division of Radiation Oncology, McGill University, Montreal, Quebec, Canada

Arie Perry, MD Washington University, Division of Neuropathology, St. Louis, MO, USA Alessia Pica, MD Doctor, Department of Radiation Oncology, University Hospital of Lausanne (CHUV), Lausanne, Switzerland Bruce E. Pollock, MD Professor, Department of Neurological Surgery and Radiation Oncology, Mayo Clinic College of Medicine, Rochester, MN, USA Neil C. Porter, MD Assistant Professor, Department of Neurology, University of Maryland School of Medicine, Baltimore, MD, USA

Leland Rogers, MD Radiation Oncologist, GammaWest Radiation Therapy, Salt Lake City, UT, USA Pantaleo Romanelli, MD Clinical Assistant Professor, Department of Neurology, State University of New York, Stony Brook, NY, USA; Consulting Assistant Professor, Department of Neurosurgery, Stanford University, Stanford, CA, USA; Director, Functional Neurosurgery, Department of Neurosurgery, IRCCS Neuromed, Pozzilli, Italy

Nader Pouratian, MD, PhD Resident Physician, Department of Neurological Surgery, University of Virginia, Charlottesville, VA, USA

Joshua M. Rosenow, MD Assistant Professor of Neurosurgery, Director of Stereotactic and Functional Neurosurgery, Department of Neurosurgery, Feinberg School of Medicine, Northwestern University; Assistant Professor of Neurosurgery, Director of Stereotactic and Functional Neurosurgery, Northwestern Memorial Hospital, Chicago, IL, USA

Pulak Ray, MD Resident, Department of Neurosurgery, Temple University, Philadelphia, PA, USA

Charles Sansur, MD, MHSc Resident, Department of Neurosurgery, Hospital of the University of Virginia, Charlottesville, VA, USA

William F. Regine, MD Professor and Chairman, Department of Radiation Oncology, University of Maryland Medical School, Baltimore, MD, USA

Raymond Sawaya, MD Professor and Chairman, Department of Neurosurgery, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA

Jean Régis Professor, Departement de Neurochirurgie Centre Hospitalier, Er Universitaire La Timone, Marseille, France

Steven E. Schild, MD Professor, Department of Radiation Oncology, Mayo Clinic, Scottsdale, AZ, USA

Andrew Reisner, MD, FACS, FAAP Neurosurgeon, Department of Pediatric Neurosurgery, Children’s Healthcare of Atlanta, Atlanta, GA, USA

Michael Schulder, MD Professor and Vice-Chairman, Department of Neurosurgery, New Jersey Medical School, Newark, NJ, USA

Louis Potters, MD, FACR South Nassau Communities Hospital, Oceanside, NY, USA

xxi

contributors

Kathleen Settle, MD Chief Resident, Department of Radiation Oncology, University of Maryland Medical Systems, Baltimore, MD, USA Jonathan Sham, MD Professor, Department of Clinical Oncology, Queen Mary Hospital, The University of Hong Kong, Pokfulam, Hong Kong Jason Sheehan, MD, PhD Assistant Professor of Neurological Surgery and Neuroscience, Department of Neurological Surgery and Neuroscience, University of Virginia, Charlottesville, VA, USA David M. Shepard, PhD Director of Medical Physics, Swedish Cancer Institute, Seattle, WA, USA Andrew G. Shetter, MD, FACS Chairman of Functional Stereotactic Neurosurgery, Division of Neurological Surgery, Director of Pain Research Laboratory, Barrow Neurological Institute, Phoenix, AZ, USA Almon S. Shiu, PhD Professor, Department of Radiation Physics, University of Texas M.D. Anderson Cancer Center; Director Stereotactic Services, Department of Radiation Physics, M.D. Anderson Cancer Center, Houston, TX, USA Mansur E. Shomali, MD, CM Clinical Assistant Professor of Medicine, University of Maryland School of Medicine, Division of Endocrinology, Union Memorial Hospital, Baltimore, MD, USA Dennis Shrieve, MD, PhD Department of Radiation Oncology, University of Utah Medical Center, Salt Lake City, UT, USA Gabriela Šimonová, MD, PhD Department of Stereotactic Radioneurosurgery, Hospital Na Homolce, Prague, Czech Republic

Uttam K. Sinha, MD Associate Professor, Chief and Program Director, Department of Otolaryngology—Head and Neck Surgery, University of Southern California, Keck School of Medicine, Los Angeles, CA, USA Sait Sirin, MD Visiting Assistant Professor, Department of Neurological Surgery, University of Pittsburgh, Pittsburgh, PA, USA Penny K. Sneed, MD, FACR Professor in Residence, Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA Luis Souhami, MD Professor and Associate Director, Department of Oncology, Division of Radiation Oncology, McGill University, Montreal, Quebec, Canada Paul W. Sperduto, MD, MAPP Co-Director, Gamma Knife Center, University of Minnesota Medical Center, Minneapolis, MN, USA Gary K. Steinberg, MD, PhD Bernard and Ronni Lacroute–William Randolph Hearst Professor of Neurosurgery and the Neurosciences; Chairman, Department of Neurosurgery, Stanford University School of Medicine, Stanford, CA, USA April Strang-Kutay, JD Attorney, Goldberg Katzman, P.C., East Petersburg, PA, USA Roger Stupp, MD Multidisciplinary Oncology Center, University of Lausanne Hospitals (CHUV), Lausanne, Switzerland John H. Suh, MD Chairman, Department of Radiation Oncology, Cleveland Clinic, Cleveland, OH, USA Mohan Suntharalingam, MD Professor and Vice Chairman, Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, MD, USA

xxii Nicholas J. Szerlip, MD Resident, Department of Neurosurgery University of Maryland School of Medicine, Baltimore, MD, USA Kintomo Takakura, MD, PhD President, Tokyo Women’s Medical University, Shinjuku, Tokyo, Japan Wolfgang Tomé, PhD Department of Human Oncology, University of Wisconsin Hospital, Madison, WI, USA May N. Tsao, MD, FRCP(C) Assistant Professor, Department of Radiation Oncology, University of Toronto, Toronto-Sunnybrook Regional Cancer Centre, Toronto, Ontario, Canada Elena Vera, BS Department of Neurological Surgery, Columbia University, New York, NY, USA Michael Y. Wang, MD Assistant Professor, Department of Neurological Surgery, University of Southern California, Los Angeles, CA, USA William J. Weiner, MD Professor and Chairman, Department of Neurology, University of Maryland School of Medicine; Professor and Chairman, Department of Neurology, University of Maryland Medical Center, Baltimore, MD, USA Martin H. Weiss, MD Professor of Neurological Surgery, Department of Neurological Surgery, USC; Attending Physician, Department of Neurosurgery, USC University Hospital, Los Angeles, CA, USA

contributors

Maria Werner-Wasik, MD Associate Professor, Department of Radiation Oncology, Jefferson Hospital for Neuroscience, Philadelphia, PA, USA David M. Wildrick, PhD Surgery Publications Coordinator, Department of Neurosurgery, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA Gordon W. Wong, MD Department of Human Oncology, University of Wisconsin Hospital, Madison, WI, USA Isaac Yang, MD Resident, Department of Neurological Surgery, University of California San Francisco, San Francisco, CA, USA Lavanya Yarlagadda, MD Department of Medicine, University of Maryland, Baltimore, MD, USA Cedric Yu, PhD Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, MD, USA Cheng Yu, PhD Professor and Director of Radiation Oncology Physics, Department of Radiation Oncology, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA Gabriel Zada, MD Resident Physician, Department of Neurosurgery, University of Southern California, Los Angeles, CA, USA

PA R T I

The Fundamentals

1

1

The History of Stereotactic Radiosurgery Michael Schulder and Vaibhav Patil

The Early Years The history of radiosurgery can be said to begin with the discovery of X-rays by Wilhelm Konrad Roentgen on November 26, 1895. His report, “Uber eine neue art von strahlen” (“On a new kind of ray”), appeared 6 weeks later [1]. By January 1896, X-rays were being used to treat skin cancers. The discovery of radioactivity by Becquerel in 1896, and of radium by the Curies soon after, provided another means for the use of therapeutic ionizing radiation. Neurosurgical applications were not long in following. X-rays were used to treat patients with pituitary tumors as early as 1906, and radium brachytherapy was applied to treat similar conditions at about the same time [2]. Harvey Cushing, the father of American neurosurgery, had extensive experience with both X-ray and brachytherapy treatments, although he remained skeptical of the utility of either [3]. Other neurosurgeons continued to explore the uses of ionizing radiation throughout the ﬁrst half of the 20th century [4]. In 1951, Lars Leksell coined the term stereotactic radiosurgery (SRS) [5]. A ceaseless innovator, his goal was to develop a method for “the non-invasive destruction of intracranial . . . lesions that may be inaccessible or unsuitable for open surgery.” The ﬁrst procedures were done using an orthovoltage X-ray tube, mounted on an early model of what is now known as the Leksell stereotactic frame, for the treatment of several patients with trigeminal neuralgia. After experimenting with particle beams and linear accelerators, Leksell and his colleagues ultimately designed the Gamma Knife (GK), containing 179 cobalt sources in a hemispheric array (Fig. 1-1). The ﬁrst unit was operational in 1968. The potential of the GK to treat lesions was recognized by Leksell and colleagues early on. In the era before computed tomography (CT), these treatments were limited to patients with arteriovenous malformations (AVMs) [6] and acoustic neuromas, which could be imaged either on angiography or by polytomography, respectively [7]. At the same time, work was continuing elsewhere with focused heavy particle irradiation. Ernest Lawrence, one of the great ﬁgures of 20th century physics and a professor at the University of California Berkeley, invented the cyclotron in

1929, winning the Nobel Prize in 1939 (Fig. 1-2). In the 1950s, his brother John began a decades-long investigation of the use of heavy particles (proton beams, then helium ion beams) for the treatment of patients with pituitary and other intracranial disorders (Fig. 1-3) [8, 9]. Raymond Kjellberg, a neurosurgeon at the Harvard/Massachusetts General Hospital facility, spearheaded the use of proton beam treatments (Fig. 1-4) [10]. A large series of patients with arteriovenous malformations and pituitary tumors was amassed. Similar efforts were carried out in California with helium ions [11]. Particle beams have the advantage of depositing their energy at a distinct point known as the Bragg peak, with minimal exit dose. In practice, the beams must be carefully shaped and spread in order to treat patients with intracranial lesions. The expense of building and maintaining a cyclotron has limited the use of heavy-particle SRS to a few centers.

Acceptance The advent of CT in the mid-1970s, and magnetic resonance imaging some 10 years later, opened up the possibility of direct targeting of tumors and other “soft tissue” targets inside the skull. The 1980s saw the evolution of SRS from an esoteric technique, available at the original GK in Stockholm (and as fractionated treatments at a few heavy-particle accelerators around the world), to an emerging technology of increasing utility. As the potential horizons of SRS broadened, other investigators were able to adapt linear accelerators (“linacs”) for SRS. These devices were more available (and less expensive) than GKs or heavy-particle accelerators [12]. Working independently, in Buenos Aires, Argentina, and in Vicenza, Italy, respectively, Betti and Colombo reported the successful adaptation of linacs for SRS [13, 14]. Their systems allowed for the rotation of the linac gantry in a single plane. After several years of hacking through mounds of red tape, Lunsford and colleagues completed the installation of the ﬁrst American GK at the University of Pittsburgh [15]. This group was instrumental, via an ongoing series of peer-reviewed

3

4

m. schulder and v. patil

FIGURE 1-3. Particle beam accelerator, 1947. (Photo courtesy of the Lawrence Berkeley National Laboratory.)

FIGURE 1-1. Lars Leksell and his physicist colleague, Borje Larsson, preparing a patient for SRS with a particle beam accelerator in 1958. (Photo courtesy of L. Dade Lundsford, MD.)

publications, in placing the technique and clinical indications for SRS on a sound scientiﬁc basis. At about the same time, Winston and Lutz described the use of a commercially available stereotactic frame for linac radiosurgery [16]. Following in their footsteps, Loefﬂer and Alexander demonstrated how a linac dedicated to SRS could be a practical alternative to a GK [17]. In the late 1980s, Friedman and Bova elected not to install the second American GK unit, preferring to develop a new linac SRS system [18]. Other advantages of these linac systems, besides ubiquity and

FIGURE 1-2. Ernest Lawrence at the controls of a cyclotron. (Photo courtesy of the Lawrence Berkeley National Laboratory.)

lower cost, included the availability of collimators in a much greater variety of diameters than provided with the GK. This allowed for the use of single isocenters when treating patients whose targets were more than 18 mm in diameter, the width of the largest GK collimator. However, at around the same time, several GKs were installed in several sites around the world. As clinical experience increased, publications appeared, indications broadened, and vendors became increasingly interested, a debate emerged regarding the merits of the GK versus linac-based SRS. By now, clinical and physics studies seem to have settled the issue in that SRS can be delivered effectively

FIGURE 1-4. Raymond Kjellberg with a frame for proton beam therapy of a patient with an AVM. (Photo courtesy of Richard Wilson, Mallinckrodt Research Professor of Physics, Harvard University.)

1.

the history of stereotactic radiosurgery

5

and accurately with either method [12, 19]. Numerous reports demonstrating the efﬁcacy of SRS with few if any short-term complications and lower costs led to the proliferation of GK and linac units around the world.

Fractionation Linac-based systems also opened up the possibility of SRS without an invasive frame. In 1992, the relocatable Gill-ThomasCosman (GTC) frame was introduced. This device relied on an attached bite block, custom molded for each patient, and was shown to have a stereotactic accuracy of just over 2 mm [20]. Although not sufﬁciently accurate and precise for single-session SRS, the GTC frame opened up the era of fractionated stereotactic irradiation [21, 22]. This in turn began a debate that has not been settled: what to call this new method? fractionated SRS or rather stereotactic radiation therapy (SRT)? This semantic question reﬂects two different underlying views of SRS. The neurosurgeon views it as a type of minimally invasive surgery, whereas the radiation oncologist sees SRS as a technique of small-volume irradiation. Advocating for “FSRS” were neurosurgeons who attached importance to the stereotactic concept, which they viewed as being “neurosurgical.” On the other hand, radiation oncologists claimed that patients were being treated with the standard fractionation schemes that practitioners knew and had been employing for decades. The GTC or similar devices were merely another means of achieving three-dimensional conformality. Confounding this controversy was the introduction of new fractionation schemes. For instance, patients with vestibular schwannomas were treated with 2500 cGy in ﬁve fractions. Other regimens have been used, including frame-based GK to treat hospitalized patients over a 5-day period [23]. Whereas SRT generally was accepted as referring to a stereotactically focused treatment using a conventional fractionation scheme, some neurosurgeons and radiation oncologists insisted that there was nothing sacrosanct about the single-fraction treatment. Who was to say that 3 or 5 doses (i.e., far fewer than usual for radiation therapy, and potentially risky to the patient if not planned and delivered with great precision) were not SRS? Different new technologies made all these options possible, but the argument was honed most precisely by the introduction of a new, robotic device. John Adler, a neurosurgeon who trained at the Brigham and Women’s Hospital in Boston, spent a fellowship year with Lars Leksell in 1985 (Adler JR, personal communication). Excited by his exposure to the GK, Adler saw the potential of SRS being extended to other areas of the body. This required a method of delivering focused radiation without a stereotactic frame. Partnering with engineers at Stanford University and with private ﬁnancial backing, the CyberKnife ultimately came into being in 1994 (Fig. 1-5). The CyberKnife delivers SRS via an X-band linac with an output of 6 MV. It is nonetheless small enough to be mounted on an industrial robot, allowing for a theoretically inﬁnite number of beams to be aimed at the target. Treatments are fashioned using an inverse planning method; to allow for practical computation times, the number of beam origins (“nodes”) and robot angles are limited. Peer-reviewed publications have

FIGURE 1-5. The ﬁrst CyberKnife treatment, 1994. (Photo courtesy of John R. Adler, MD.)

demonstrated the acceptance of the CyberKnife [24–26]. These and other articles have fostered a useful debate regarding the concept of hypofractionation in SRS and indeed if such treatments are still “radiosurgical” [27, 28].

Extracranial Radiosurgery SRS was invented as a means of minimally invasive brain surgery and was expanded with the aid of digital imaging to include extracerebral, intracranial targets. Still, the concept of a highly focused, single- or several-session radiation treatment had obvious appeal for extracranial targets. The ﬁrst radiosurgical moves out of the intracranial compartment were in the logical direction of the skull base and past that into the paranasal sinuses, using either GK [29, 30] or linac units [31]. Creative modiﬁcations of standard stereotactic frames were described to allow for treatment of “lower” targets [32]. The adaptation of available equipment for SRS could go only so far. Hamilton and colleagues described the ﬁrst truly extracranial radiosurgical unit. This prototypical system did not rely on rigid frame ﬁxation to the skull and was designed to provide spinal SRS [33]. The need to surgically place a clamp on a spinous process, and to treat the patient in a prone position, limited the appeal of this groundbreaking concept. With the advent of newer technologies, spinal SRS has become a reality. Reports to date have employed the CyberKnife [34] or other linac-based systems [35]. More recently still, the inevitable and logical extension of SRS to non-CNS targets has begun. Work on CyberKnife treatment of tumors of the lung [36] and prostate [37] has been published. Despite the neurosurgical origins of SRS, all advocates of this concept, in its various forms, can only welcome its spread to other specialties in which neurosurgeons will have little role to play. Table 1-1 summarizes the historical landmarks in the development of SRS.

6

m. schulder and v. patil

TABLE 1-1. Historical landmarks in the development of SRS. Year

Author

Device/Event

1951

Leksell

1954

Lawrence

1962

Kjellberg

1967 1970 1980 1982 1984 1986

Leksell Steiner Fabrikant Betti/Columbo Bunge Winston/Lutz

1991 1992 1994 1997

Friedman Loefﬂer/Alexander Adler Krispel

Invention of SRS with rotating orthovoltage unit Heavy-particle treatment of pituitary for cancer pain Proton beam therapy of intracranial lesions Invention of GK GK SRS of AVMs Helium ion treatment of AVMs Linacs adapted for SRS Installation of commercial GK Linac SRS based on common stereotactic frame Linac system for highly conformal SRS Dedicated linac for SRS developed First CyberKnife treatment Rotating cobalt unit

Other Linac Systems and the Role of Industry The convergence of image-guidance technology and radiation delivery devices has encouraged the entry of multiple vendors into the SRS marketplace. This has reﬂected the undeniable logic of stereotactic localization and the resulting ability to focus radiation treatments on the smallest possible volume. Initially, in addition to the GK, there were a variety of framebased systems designed to provide single-fraction SRS. Vendors included Radionics (X-Knife), Zmed (the University of Florida system), BrainLAB, and Fischer-Leibinger. The acceptance of stereotactic fractionation by radiation oncologists, their reluctance to apply stereotactic frames, and to some extent patients’ preference for avoiding frame use have shifted the focus toward frameless systems. Long-established purveyors of linacs have begun to market stereotactic devices aimed primarily at radiation oncologists but usually with a nod toward neurosurgeons who often will prefer to treat patients with a single fraction, or at most several. Thus, Varian and Phillips (now a division of Elekta) have developed systems with integrated stereotactic localization (Trilogy and Synergy). At the same time, Radionics and BrainLAB have adapted their linac-based SRS devices for frameless use and have marketed directly to radiation oncologists. And to square the circle, American Radiosurgical, Inc., has as its sole product a modiﬁcation of the GK, using a limited number of cobalt-60 sources in a rotating helmet. This industrial involvement in the advancement of SRS and related techniques results from the expense of the equipment and the need for support personnel to ensure their proper functioning. From the days of the ﬁrst GK and on up to the emerging era of frameless, fractionated SRS, companies have played an invaluable role. Without them, SRS would never have come to deﬁne a new standard in patient care, as it so clearly has.

Organized Radiosurgery Neurosurgeons’ interest in SRS was slow to develop but has increased exponentially over time. In 1987, the year that the ﬁrst American GK was installed at the University of Pittsburgh and early work on linac SRS had been published, there were no SRS-related presentations at the meeting of the American Association of Neurological Surgeons (AANS). By 1998, there were 31 such abstracts in addition to practical courses and seminars devoted to the topic. SRS has remained a key item of interest at the major annual meetings of the AANS and of the Congress of Neurological Surgeons. In addition, the meetings of the American and World Societies for Stereotactic and Functional Neurosurgery feature SRS as one of the main topics. The International Stereotactic Radiosurgery Society (ISRS) was founded in 1993 and held its ﬁrst biannual meeting that year in Stockholm. At ﬁrst, the papers presented dealt entirely with the treatment of intracranial conditions. As SRS has moved below the skull base, studies regarding patients with such conditions as tumors of the spine, lung, pancreas, and prostate have been included in the ISRS program. Thus, the expertise of clinicians in ﬁelds completely unrelated to neurosurgery is being applied to the study of SRS. Neurosurgeons comprise the single biggest specialty group in the organization, followed by radiation oncologists and medical physicists. As interest in extracranial and indeed nonneurosurgical SRS inevitably increases, the membership of ISRS no doubt will evolve to reﬂect this broadening of interest. The ISRS publishes a peer-reviewed collection of selected manuscripts from each meeting, entitled Radiosurgery.

Conclusion Acceptance by neurosurgeons, surgical specialists, and radiation oncologists means that as SRS evolves, it will not be a technique for “radiosurgeons” but one of the methods available to treat patients with a wide variety of disorders. At the same time, the historical role of neurosurgeons in the development of SRS, their leadership in its reﬁnement and expansion over the last half century, their knowledge of neuroanatomy, and their understanding of central nervous system pathology and its treatment will ensure the continued active role of neurosurgeons in the ongoing growth of stereotactic radiosurgery.

References 1. Mould R. A Century of X Rays and Radioactivity in Medicine. Philadelphia: Institute of Physics Publishing, 1993. 2. Hirsch O. Uber methoden der operativen behandlung von hypophysistumoren auf endonasalem Wege. Arch Laryngol Rhinol 1910; 24. 3. Schulder M, Loefﬂer J, Howes A, et al. The radium bomb: Harvey Cushing and the interstitial irradiation of gliomas. J Neurosurg 1996; 84:530–532. 4. Schulder M, Rosen J. Therapeutic radiation and the neurosurgeon. Neurosurg Clin N Am 2001; 12(1):91–100, viii. 5. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316–319. 6. Steiner L, Leksell L, Greitz T. Stereotaxic radiosurgery for cerebral arteriovenous malformations. Acta Chir Scand 1972; 138: 459–464.

1.

the history of stereotactic radiosurgery

7. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry 1983; 46:797–803. 8. Kirn TF. Proton radiotherapy: some perspectives. JAMA 1988; 259:787–788. 9. Skarsgard LD. Radiobiology with heavy charged particles: a historical review. Phys Med 1998; 14(Suppl 1):1–19. 10. Kjellberg RN, Abe M. Stereotactic Bragg Peak proton beam therapy. In: Lunsford LD, ed. Modern Stereotactic Neurosurgery. Boston: Martinus Nijhoff, 1988:463–470. 11. Fabrikant J, Lyman J, Frankel K. Heavy charged particle Bragg peak radiosurgery for intracranial vascular disorders. Radiat Res Suppl 1985; 8:S244–258. 12. Podgorsak E, Pike G, Olivier A, et al. Radiosurgery with high energy photon beams: a comparison among techniques. Int J Radiat Oncol Biol Phys 1989; 16:857–865. 13. Betti O, Derechinsky V. Hyperselective encephalic irradiation with a linear accelerator. Acta Neurochir 1984; Suppl 33:385– 390. 14. Columbo F, Benedetti A, Pozza F, et al. External stereotactic irradiation by linear accelerator. Neurosurgery 1985; 16:154–160. 15. Lunsford LD, Flickinger JC, Linder G, et al. Stereotactic radiosurgery of the brain using the ﬁrst United States 210 cobalt-60 source gamma knife. Neurosurgery 1989; 24:151–159. 16. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for stereotactic radiosurgery. Neurosurgery 1988; 22:454–464. 17. Loefﬂer J, Shrieve D, Wen P, et al. Radiosurgery for intracranial malignancies. Semin Radiat Oncol 1995; 5:225–234. 18. Friedman W, Bova F. The University of Florida radiosurgery system. Surg Neurol 1989; 32:334–342. 19. Luxton G, Petrovich Z, Joszef G, et al. Stereotactic radiosurgery: principles and comparison of treatment methods. Neurosurgery 1993; 32:241–259. 20. Gill SS, Thomas DG, Warrington AP, et al. Relocatable frame for stereotactic external beam radiotherapy. Int J Radiat Oncol Biol Phys 1991; 20:599–603. 21. Andrews DW, Silverman CL, Glass J, et al. Preservation of cranial nerve function after treatment of acoustic neurinomas with fractionated stereotactic radiotherapy. Preliminary observations in 26 patients. Stereotact Funct Neurosurg 1995; 64:165–182. 22. Combs SE, Volk S, Schulz-Ertner D, et al. Management of acoustic neuromas with fractionated stereotactic radiotherapy (FSRT): long-term results in 106 patients treated in a single institution. Int J Radiat Oncol Biol Phys 2005; 63:75–81.

7

23. Noren G. Gamma knife radiosurgery of acoustic neurinomas. A historic perspective. Neurochirurgie 2004; 50:253–256. 24. Chang SD, Murphy M, Geis P, et al. Clinical experience with image-guided robotic radiosurgery (the CyberKnife) in the treatment of brain and spinal cord tumors. Neurol Med Chir (Tokyo) 1998; 38:780–783. 25. Ishihara H, Saito K, Nishizaki T, et al. CyberKnife radiosurgery for vestibular schwannoma. Minim Invasive Neurosurg 2004; 47: 290–293. 26. Mehta VK, Lee QT, Chang SD, et al. Image guided stereotactic radiosurgery for lesions in proximity to the anterior visual pathways: a preliminary report. Technol Cancer Res Treat 2002; 1:173– 180. 27. Adler JR Jr, Colombo F, Heilbrun MP, et al. Toward an expanded view of radiosurgery. Neurosurgery 2004; 55:1374–1376. 28. Pollock BE, Lunsford LD. A call to deﬁne stereotactic radiosurgery. Neurosurgery 2004; 55:1371–1373. 29. Firlik KS, Kondziolka D, Lunsford LD, et al. Radiosurgery for recurrent cranial base cancer arising from the head and neck. Head Neck 1996; 18:160–165; discussion 166. 30. Kondziolka D, Lunsford LD. Stereotactic radiosurgery for squamous cell carcinoma of the nasopharynx. Laryngoscope 1991; 101:519–522. 31. Kaplan ID, Adler JR, Hicks WL Jr, et al. Radiosurgery for palliation of base of skull recurrences from head and neck cancers. Cancer 1992; 70:1980–1984. 32. Samblas JM, Bustos JC, Gutierrez-Diaz JA, et al. Stereotactic radiosurgery of the foramen magnum region and upper neck lesions: technique modiﬁcation. Neurol Res 1994; 16:81–82. 33. Hamilton A, Lulu B, Fosmire H, et al. Preliminary clinical experience with linear accelerator-based spinal stereotactic radiosurgery. Neurosurgery 1995; 36:311–319. 34. Gerszten PC, Welch WC. CyberKnife radiosurgery for metastatic spine tumors. Neurosurg Clin N Am 2004; 15:491–501. 35. De Salles AA, Pedroso AG, Medin P, et al. Spinal lesions treated with Novalis shaped beam intensity-modulated radiosurgery and stereotactic radiotherapy. J Neurosurg 2004; 101(Suppl 3):435–440. 36. Whyte RI, Crownover R, Murphy MJ, et al. Stereotactic radiosurgery for lung tumors: preliminary report of a phase I trial. Ann Thorac Surg 2003; 75:1097–1101. 37. King CR, Lehmann J, Adler JR, et al. CyberKnife radiotherapy for localized prostate cancer: rationale and technical feasibility. Technol Cancer Res Treat 2003; 2:25–30.

2

Neuroimaging in Radiosurgery Treatment Planning and Follow-up Evaluation Clark C. Chen, Paul H. Chapman, Hanne Kooy, and Jay S. Loefﬂer

Introduction Radiosurgery refers to the precise delivery of a large, single dose of radiation to a focal target. The focal dose distribution allows for a positive therapeutic gain. Because of the opacity of the cranial vault, target volume deﬁnition relies entirely on the anatomic accuracy of the available imaging modalities. The need for accurate anatomic visualization is magniﬁed by the use of higher radiation dose delivered in radiosurgery as compared with radiotherapy. To achieve maximal spatial accuracy in radiosurgical planning, an understanding of the basic principles underlying neuroimaging as well as the limitations associated with each imaging modality is mandatory. Additionally, the optimal management of patients after radiosurgery requires knowledge of the expected neuroimaging changes as they relate to clinical outcome. These issues will be reviewed in this chapter.

Imaging Modalities Since its inception with the discovery of X-rays in 1895, radiology has played a pivotal role in the diagnosis and treatment of various neurosurgical lesions. The advent of computed tomography (CT) imaging in the 1970s marked a major step forward in the application of imaging in radiotherapeutic planning by allowing improved anatomic resolution as well as calculation of electron density maps. Improved soft tissue resolution was achieved with the introduction of magnetic resonance imaging (MRI), a technique based on differential nuclear interaction rather than differential density. Advances made in computational technology in the past decade have enabled the superposition of CT and magnetic resonance (MR) images in order to maximize anatomic delineation. More recently, signiﬁcant strides in functional imaging have further reﬁned target deﬁni-

tion in radiosurgical planning (Fig. 2-1). The following section will review the basic principles underlying the various neuroimaging modalities as well as limitations associated with each modality.

Computed Tomography Imaging Computed tomography provides cross-sectional images of the body using mathematical reconstructions based on X-ray images taken circumferentially around the subject. In practice, X-ray transmissions through the subject from a rotating emitter are detected and digitally converted into a grayscale image. Because CT images are ultimately a compilation of X-ray transmissions, the physical principles underlying the two modalities are identical; that is, structural discrimination is made based on the relative atomic composition, and therefore the electron density, of the tissue imaged. CT images, however, offer improved anatomic resolution because each image represents the synthesis of information from multiple X-ray images (Fig. 2-1a). Besides improved anatomic delineation, CT imaging aids radiosurgical planning in another way. Because the pixel intensity on a CT image reﬂects the electron density of the tissues imaged, the pixel intensity can be mathematically converted into electron density maps (electrons per cm3). This information can be used to deﬁne isodose lines in radiosurgical planning. Without this information, actual radiation dose delivered can deviate from the desired dose by as much as 20% as a result of tissue inhomogeneity [1]. Despite yielding improved anatomic resolution as well as electron density information, delineation of soft tissue structures by CT imaging is suboptimal, even with the aid of intravenous contrast agents. For the most part, delineation of soft tissue structures is achieved by the use of MRI, especially for targets in the cranial base.

9

10

c.c. chen et al.

FIGURE 2-1. CT, MR, and MRS images from a patient with a left cerebellar tumor. (a) CT imaging without intravenous contrast shows a poorly deﬁned left cerebellar mass with effacement of the fourth ventricle and displacement of the brain stem. (b) Intravenous contrast administration improves the anatomic resolution of the left cerebellar mass, revealing a densely enhancing mass with surrounding edema. (c) The same lesion is visualized using T1-weighted MRI. (d) MRI after gadolinium administration reveals a heterogeneously enhancing mass. The homogeneously enhancing tissue on CT is further resolved into tissues of varying intensity on MRI, demonstrating the superiority of MRI over CT in soft tissue resolution. The numbered grid corresponds with the MR spectral arrays shown in (e). The grid is placed over

normal-appearing tissue. (e) The various chemical peaks are as indicated in box 9. The thick arrow indicates the choline peak. The arrowhead represents the creatine peak. The thin arrow designates the N-acetylaspartate (NAA) peak. The MRS in box 9 is typical of normal tissue, with comparable choline and creatine peaks and a notable NAA peak. (f) The numbered MRS grid is placed over the diseased tissue. The MRS is shown in (g). (g) The various chemical peaks are labeled in box 1. The diseased tissue shows an elevated choline peak (thick arrow) relative to a diminished creatine peak (arrowhead). The NAA peak is also decreased (thin arrow) relative to normal tissue. The accumulation of lactate (double arrow) is another signature of diseased tissue.

Magnetic Resonance Imaging

source of error involves the imperfection of the input magnetic ﬁeld. The input magnetic ﬁeld in MRI is produced by electric currents passing through sets of mutually orthogonal coils. Ideally, the magnetic ﬁeld generated should be uniform such that a linear relationship between space and resonance frequency can be established [3]. However, such uniform ﬁelds cannot be easily achieved in practice. This phenomenon is referred to as gradient ﬁeld nonlinearity and tends to escalate with increasing distance from the central axis of the main magnet. For the most part, gradient ﬁeld nonlinearity can be corrected computationally. Prior to correction, gradient ﬁeld nonlinearity can induce spatial distortions as large as 4 mm. After computational correction, the distortion is minimized to <1 mm [4]. A more complex MR distortion that is more difﬁcult to correct computationally involves electromagnetic interactions between the imaged tissue and the input magnetic ﬁeld. This distortion is often referred to as resonance offset. Resonance offset occurs because hydrogen atoms carry with them an inherent magnetic ﬁeld. Thus, placement of hydrogen-bearing tissues in a magnetic ﬁeld necessarily induces a perturbation in the input magnetic ﬁeld. This perturbation disrupts the linear relationship between space and resonance frequency as to produce

The human body consists primarily of fat and water, both having a high content of hydrogen atoms. MRI exploits the nuclear spin property of these hydrogen atoms as a means to attain soft tissue resolution. In MRI, a radiofrequency pulse is applied to the imaged subject. As a result, the nuclear spin states of these atoms shift from that of equilibrium to that of excitation. To return to their equilibrium state, the law of energy conservation dictates that an energy equal to that absorbed must be emitted. The energy release between nuclear spin state transitions can be measured and analyzed. Because the process of energy absorption and emission is affected by the local chemical environment, hydrogen atoms in soft tissues of varying chemical composition will absorb and emit differential energy. Mathematical transformation of this information yields ﬁne-resolution maps of soft tissue structures (Fig. 2-1c, d). Because tumor and normal tissues often differ in chemical composition [2], the same principle allows delineation of these tissue types. Because of the complexity of the nuclear interactions involved in MRI, the modality is subject to many sources of error, resulting in distortion of the image obtained. One such

2.

neuroimaging in radiosurgery planning and evaluation

geometric distortions. The physics of this perturbation is complex because it depends on the inherent magnetic properties as well as the volume and shape of the imaged object. Resonance offset distortions tend to be largest at the interface of materials that differ in magnetic properties, such as at the airwater interface. In anatomic imaging, this translates into large distortions at the air-bone or air-tissue interfaces. Studies reveal that distortions at these interfaces can be as large as 2 mm [3, 5, 6]. Although the development of higher-strength magnets has allowed for improved resolution of soft tissue structures as well as minimization of geometric distortions related to gradient ﬁeld nonlinearity, higher-strength magnets do not address the issue of resonance offset. Because resonance offset is a product of the input magnetic ﬁeld and the local ﬁeld imposed by the imaged tissue, increasing the strength of the input magnetic ﬁeld will magnify the effects of resonance offset [7]. The accuracy of MR as a stand-alone imaging modality has been determined by a number of investigators [8–12]. Most investigators report a localization uncertainty of 2 to 3 mm [8– 11], but maximal absolute errors of 7 to 8 mm have also been reported [12]. These studies reveal that error in ﬁducial localization is ampliﬁed by subsequent mathematical transformation. Though the degree of localization uncertainty varies between studies, the reported uncertainty consistently remains greater than 1 mm, failing to achieve the current radiosurgical standard set forth by the American Society of Therapeutic Radiology and Oncology (ASTRO) [13–15]. Another downside of MRI as it pertains to radiosurgical planning is the absence of electron density information (see earlier “Computed Tomography Imaging” section). Contrary to CT imaging where the image is derived based on differential electron density, pixel intensities in MR images bear no correlation with electron density. For radiosurgical planning using MR as the only imaging modality, image processing and assignment of hypothetical electron density values are required. Such strategies have led to suboptimal radiosurgical plans [16]. Motion artifact is another consideration affecting spatial accuracy in MRI. The prolonged duration required for image acquisition increases the potential for patient movement. Even with a cooperative patient, motion artifact occurs with breathing and internal physiologic motions. The resultant motion compromises the accuracy of spatial resolution. Though MRI is inadequate as a stand-alone modality in radiosurgical planning, combining MR and CT images has led to radiosurgical plans that are superior to plans derived from each modality alone [17–20]. For example, Shuman et al. reported that the incorporation of MR information into CTbased radiotherapy plans resulted in better deﬁnition of tumor volume in 53% of the cases [18]. These observations have led to the development of algorithms for superimposing MR and CT images.

CT-MR Image Integration The differences between CT and MRI illustrate the conceptual distinction between geometric and diagnostic accuracy. Although CT imaging is geometrically accurate due to absence of spatial distortion effects, disease tissues are often missed by this modality. As such, CT imaging is diagnostically inaccurate.

11

On the other hand, due to enhanced soft tissue resolution, MRI affords enhanced diagnostic accuracy; however, the spatial accuracy is limited due to MR distortion effects. Algorithms have been developed to maximally utilize the different types of information afforded by CT and MRI (Fig. 2-2). Simple approaches to image integration involve manual superposition of equivalent views of MR and CT images, using bony landmarks as correlation points. Such approaches, however, are labor intensive and error-prone with uncertainties of up to 8 mm [21]. Advances in computational technology have allowed for the development of automated algorithms for superposition of CT and MR images in three-dimensional space. One way of integrating CT and MR images requires that the patient be placed in an immobilization device, such as the stereotactic frame. The immobilization device minimizes motion artifacts and ensures that the images are acquired in a predetermined manner. Fiduciary markers are used to establish the spatial relationship between the target and the head frame. Additionally, they serve as coregistration points between the MR and CT images. Because image acquisition and correlative points are ﬁxed in space in a predetermined way, this mode of image fusion is sometimes referred to as prospective image coregistration [14]. Alternatively, image coregistration can be done with images that are not acquired in a predetermined manner. This mode of image fusion is also known as retrospective coregistration. Retrospective image coregistration relies on matching corresponding anatomic landmarks instead of ﬁduciary markers. The CT and MR images are integrated on the basis of aligning these anatomic landmarks [22]. Various computational techniques, including point matching [23], line matching, and iterative matching [24], have been developed for retrospective image superposition. Whether one method is superior to another remains an area of research. In general, with proper training and quality control, most current algorithms will coregister MR and CT images to an uncertainty of 1 to 2 mm using prospective registration and of 2 to 3 mm using retrospective registration [14].

Contrast Administration Contrast administration takes advantage of the observation that disease processes, such as tumor growth, often result in vascular encroachment or faulty angiogenesis [2]. These processes allow contrast material to escape the vasculature and preferentially accumulate in the diseased tissue. The accumulation of contrast material can be easily visualized on CT or MRI (Fig. 2-1b, d). In malignant gliomas for instance, contrast enhancement correlates with diseased tissue. Kelly et al. evaluated 195 brain tumor biopsies acquired from various locations relative to the contrast-enhancing regions of CT or MRI scans and showed that the regions of contrast enhancement best correlated with regions of tumor burden [25]. Because contrast-enhancing volumes are used for radiosurgery target deﬁnition, diseased tissues without contrast enhancement often escape therapy. Investigators have used various functional imaging modalities to address this issue. Although these modalities hold tremendous promise, they are limited by poor anatomic resolution. As such, functional imaging

12

c.c. chen et al.

FIGURE 2-2. Fusion of MR and CT images in radiosurgical planning. (a) CT image of a patient with left frontal metastatic lesion. The image is selected to illustrate the continuity of the ventricular contour and the cranial vault as landmarks to gauge the spatial discrepancy when comparing MR and CT images. The lesion is not shown in (a). (b) Equivalent T1-weighted MR image of the CT image shown in (a). Again, note the continuity of the ventricular contour and the cranial vault. (c) Superposition of (a) and (b) without correction of MR distortion shows spatial discrepancy as evidenced by the discontinuity of ventricular contour and the cranial vault at the transition point. The CT-derived image is shown on the top-half panel. The MR image is shown on the

bottom-half panel. (d) After computational correction of MR distortion, continuity of the transition point is restored and various anatomic landmarks are coregistered. (e) Three-dimensional view of the metastatic lesion in relation to the stereotactic frame and a surface rendering of the patients head. The lines interconnecting the small red, green, and yellow spheres indicate the planes through which radiation is delivered. (f) Axial, (g) sagittal, and (h) coronal views of the lesion on MRI after correcting for MR distortion effects. The colored lines represent the various isodose contours. The magnitude of the radiation delivered is shown in the right lower corner of each panel.

is most useful in conjunction with traditional anatomic imaging modalities. In many instances, the clinical applications of functional imaging remain investigational.

intense upregulation of glucose metabolism in tumor cells. Once inside the cell, 18F-FDG undergoes phosphorylation to yield an intermediate that cannot undergo further metabolic processing or cellular export. The phosphorylated intermediate is, therefore, preferentially transported into tumor cells and trapped there [26]. Studies investigating the use of 18F-FDG in guiding radiosurgery for treatment of gliomas yielded mixed results. Tralins et al. reported a series of 27 patients who underwent conventional MR or CT scanning as well as 18F-FDG PET. In this study, a multivariate analysis revealed 18F-FDG PET ﬁndings as the only variable that retained statistical signiﬁcance in predicting time to tumor progression and overall survival. Moreover, the 18F-FDG PET deﬁned target volumes differing from those deﬁned by MR or CT imaging by at least 25% in all patients [27]. Gross et al., on the other hand, reported that regions of 18F-FDG abnormal uptake closely correlated with regions of contrast enhancement in their 18 patients. In a minority of patients, 18F-FDG PET did affect target volume deﬁnition. These changes, however, were not associated with improved survival when compared with historical controls [28]. Likewise, Prado et al. reported that the inclusion of PET scan data minimally altered radiation planning in most patients [29]. These conﬂicting data can, in part, be attributed to the variability and subjectivity involved in PET image interpretation.

Positron Emission Tomography and Single-Photon-Emission Computed Tomography One type of functional imaging relies on visualizing tracer molecules that preferentially accumulate in diseased tissues. This type of imaging includes positron emission tomography (PET) and single-photon-emission computed tomography (SPECT). PET is designed to detect the preferential accumulation of positron-emitting radioactive tracer compounds in the diseased tissue. The emitted positron collides with an electron to yield opposing gamma rays. These emissions are detected by a gamma-ray camera, thereby generating images of regional radioactivity (Fig. 2-3). Similarly, SPECT is designed to detect the preferential accumulation of tracer compounds bearing photon-emitting isotopes. Photon emission is detected by a rotating gamma camera detection system and reconstructed into three-dimensional tomographic images. In tumor neuroimaging, the enhanced metabolic state of the tumor cells is often exploited to achieve preferential tracer accumulation in these tissues. For instance, 18-ﬂuorodeoxyglucose (18F-FDG), a commonly used PET tracer, is preferentially transported into tumor cells relative to normal cells due to an

2.

neuroimaging in radiosurgery planning and evaluation

13

FIGURE 2-3. Use of 18F-FDG PET in neuroimaging. (a) A left parietal-occipital mass (arrow) that shows gadolinium enhancement on T1-weighted MR imaging. (b) The same lesion shows increased 18FFDG accumulation on PET imaging (arrow). The signal intensity of the lesion on 18F-FDG PET imaging is comparable with those of the gray matter. (c) Bilateral gadolinium enhancement in a patient undergoing treatment for malignant glioma on a T1-weighted MR sequence. (d)

PET imaging showed preferential 18F-FDG accumulation in the right hemispheric lesion (arrow). Biopsy specimen of this lesion reveals recurrent glioma. (From Davis WK, Boyko OB, Hoffman JM, et al. [18F] 2-ﬂuoro-2deoxyglucose-positron emission tomography correlation of gadolinium-enhanced MR imaging of central nervous system neoplasia. AJNR 1993; 14:515–523. Copyright by American Society of Neuroradiology.)

Because normal gray matter exhibits high physiologic uptake of 18F-FDG, the distinction between normal gray matter and tumor is sometimes subjective, especially in cases of signiﬁcant anatomic distortion secondary to mass effect. Additionally, regions known to be at high risk for tumor inﬁltration, such as regions of edema, often display increased 18F-FDG uptake. In many instances, different thresholds are used for deﬁning abnormal 18F-FDG uptake [27–30]. Despite their differences, the various studies suggest that, in selected patients, the inclusion of PET or SPECT can aid in the deﬁnition of target volumes in radiosurgery for gliomas. The extent of clinical beneﬁt and the criteria for patient selection await future investigations. The likelihood of routine PET or SPECT for radiosurgical planning should increase with the development of tracer compounds that exhibit high speciﬁcity to diseased tissues.

of 46 patients with malignant gliomas, patients with MRS abnormality outside of the MR-deﬁned target volume showed decreased median survival relative to those with MRS abnormality inside the MR-deﬁned tumor volume (10.7 months vs. 17.4 months, p = 0.002) [32]. Other studies have conﬁrmed the correlation between untreated MRS abnormality and worse prognosis [33–37]. These studies suggest that MRS data should be taken into consideration in target volume determination for the treatment of gliomas.

Magnetic Resonance Spectroscopy Another type of functional imaging capitalizes on the ability of MRI to measure the levels of biochemical metabolites. Three metabolites commonly used to distinguish tumor and healthy tissue include choline, creatine, and N-acetylaspartate (NAA) (Fig. 2-1d–g; Fig. 2-4). Choline is an essential component of the cell membrane. The level of choline reﬂects the rapidity of membrane turnover and is increased in rapidly proliferating tumors. Creatine is a metabolic intermediate for the synthesis of phosphocreatine, an energy source for cellular metabolism. The level of creatine corresponds with the level of cellular energy reserve, which is decreased in tumor tissues. NAA is a marker for neuronal differentiation and is decreased in tumors [2]. Using elevated choline and decreased NAA as criteria, Pirzkall et al. compared the magnetic resonance spectroscopy (MRS)-deﬁned tumor volume to that deﬁned by contrastenhanced MRI for malignant gliomas. The authors report that the MRS-deﬁned volume extended outside of the MRI-deﬁned volume by <2 cm in 88% of the patients [31]. In another study

MR Perfusion Imaging Like levels of biochemical metabolites, perfusion parameters such as cerebral blood volume (CBV) can be measured using MRI techniques (Fig. 2-5). CBV is measured by monitoring the transit of a rapid bolus of contrast with respect to time. This parameter is an indirect measure of tissue vascularity, a property often associated with tumor burden. It is, therefore, not surprising that MR-derived measurements of cerebral blood volume correlate with tumor grading and clinical outcome. In a series of 28 patients with gliomas, pretreatment high CBV intensity was associated with shorter median survival [2, 38, 39]. The use of CBV in radiosurgical planning is limited by several factors. CBV values in tumor volumes are often greater than CBVs of normal white matter but comparable with CBVs of normal gray matter. Thus, distinguishing tumor and cortex is problematic, especially in the context of anatomic distortion caused by large tumors. Additionally, regions of increased CBV correlate well with regions of contrast enhancement. As such, incorporation of CBV information will only alter radiosurgical plans in a subpopulation of patients.

MR Diffusion Weighted Imaging The white matter in a normal cerebrum is organized into tracts that allow communication between cortical neurons. As a result of this high degree of organization, water molecules in the

14

c.c. chen et al.

FIGURE 2-4. MRS images from a patient with high-grade glioma. (a) T1-weighted MRI reveals an area of hypointensity in the basal frontal lobe involving the anterior limb of the internal capsule. (b) The lesion does not enhance with gadolinium administration but shows increased signal intensity on a (c) FLAIR sequence. (d) The numbered MRS grid is placed over normal-appearing tissue. The MRS is shown in (e). (e) The various chemical peaks are as indicated in box 1. The

MRS shown is typical of normal tissue, with comparable choline (thick arrow) and creatine peaks (arrowhead) and a notable NAA peak (thin arrow). (f) The numbered MRS grid is placed over the diseased tissue. The MRS is shown in (g). (g) The various chemical peaks are labeled in box 10. The diseased tissue shows an elevated choline peak (thick arrow) relative to a diminished creatine peak (arrowhead). The NAA peak is also decreased (thin arrow) relative to normal tissue.

cerebral cortex diffuse in a highly directional manner. The extent of this directional diffusion can be estimated using specialized MR techniques and is referred to as apparent diffusion coefﬁcient (ADC) (Fig. 2-6). Because gliomas often distort cerebral architecture, regions with altered ADC are expected to correlate with tumor burden. This expectation was demonstrated in several studies [40–42]. These studies revealed that

patients with lower ADC values in the tumor volume showed shorter median survival than patients with normal or nearnormal ADC values (12 months vs. 21.7 months). These studies suggest that ADC maps may be helpful in guiding radiosurgical planning, especially in cases where conventional MRI and ADC maps yield discordant information with regard to tumor volume.

FIGURE 2-5. Application of MR perfusion imaging in tumor diagnosis. (a) T1-weighted MRI shows a heterogeneous right temporal lesion. (b) Gadolinium administration reveals peripheral enhancement and septation of the lesion. A nodular enhancing region is seen in the

right lower corner of the lesion. (c) MR perfusion shows increased cerebral blood volume (CBV) correlating with the contrast-enhancing rim, septation, and nodule. Biopsy of the lesion reveals a grade IV glioma.

2.

neuroimaging in radiosurgery planning and evaluation

15

FIGURE 2-6. MR diffusion imaging in tumor assessment. (a) T1weighted MRI shows a heterogeneous right temporal lesion. (b) Gadolinium administration reveals a multicystic lesion with a central region

of enhancement. (c) MR diffusion imaging shows decreased apparent diffusion coefﬁcient (ADC) signal in the region of central enhancement. Biopsy of the lesion reveals a grade IV glioma.

Combined Imaging Modality

demonstrated by other studies, AVM obliteration occurred 2 to 3 years after radiosurgical treatment. The risk of hemorrhage during this latency interval was the same as that seen in untreated AVMs [45–53]. Radiosurgical treatment of AVMs requires a precise deﬁnition of the nidus in three-dimensional space. This precise deﬁnition can be achieved only by the combination of MRI, CT imaging, and cerebral angiography. Whereas MR and CT images afford anatomic resolution, they cannot discriminate between the AVM nidus and the feeding arteries and draining veins [54]. This distinction is crucial because the goal of AVM treatment lies in the obliteration of the former while sparing the latter. On the other hand, cerebral angiography offers a limited deﬁnition of the nidus margin without the corresponding MR and CT imaging, especially in cases of irregularly shaped AVMs (Fig. 2-7). Conventional angiographic studies yield two-dimensional projections of the AVM nidus, a three-dimensional lesion. The spatial information lost as a result of dimensional reduction represents a source of inaccuracy in nidus deﬁnition [48, 55–57]. For instance, angiographic projections may outline different AVM nidus margins depending on the angle of projection and the nidus geometry [54, 56]. Not surprisingly, studies examining the value of incorporating CT and MR information into conventional angiography–based radiosurgical plans yielded data supporting the superiority of the combined approach. In one study, inclusion of CT imaging in angiography-based plans resulted in a mean isocenter shift of 3.6 mm in 44 of 81 (54%) patients and changes in the diameter of collimator beam in an equal number of patients [55]. Other studies have reported similar ﬁndings [57]. CT- and MR-based angiograms are proposed alternatives to conventional angiogram in radiosurgical planning (Fig. 2-8); however, the spatial accuracy and the resolution of vessel architecture afforded by CT and MR angiograms are insufﬁcient as stand-alone modalities [58, 59]. Tanaka et al. compared AVM resolution by MR angiography (MRA), CT angiography (CTA), and conventional angiogram in terms of feeding vessel and draining vein visualization [59]. In this study, only 20% to 30% of feeding vessels and draining veins detected by conventional angiogram were identiﬁed by MRA or CTA. Combined use of CTA and MRA did not further improve AVM resolution. The work by Aoyama et al. further illustrated the inadequacy of CT and MR angiograms as stand-alone modalities in radiosurgery

Given the complexity of physiology and pathology underlying tumor biology, it is unlikely that any single imaging modality will allow perfect deﬁnition of the diseased volume [43]. The failure to precisely deﬁne tumor volume will result in inadequate or excessive radiation treatment and suboptimal clinical outcomes. Precise tumor volume deﬁnition likely requires a synthesis of information obtained from contrast-enhanced CT or MRI, PET, SPECT, MRS, CBV, and ADC in a meaningful way. The optimal algorithm for the synthesis of this information remains an area of active research.

Cerebral Angiography The primary application of cerebral angiography in radiosurgical planning lies in the treatment of cerebral arteriovenous malformations (AVMs). AVMs represent abnormal communications between vessels of disproportionately unbalanced hydrodynamic stress. The region of abnormal communication is referred to as the nidus. Due to increased hydrodynamic stress, the nidus in an AVM is at high risk for rupture, causing intracranial hemorrhage. The goal of AVM treatment is to eliminate this risk by obliterating the AVM nidus. Estimates of the annual risk for hemorrhage secondary to AVM rupture lie in the range 2.2% to 4.0%, with an associated fatality of roughly 10%. Whereas surgical resection is the treatment of choice for AVMs, radiosurgical treatment is often performed in cases of surgical inaccessibility, patient preference, or severe preoperative morbidity. The risks associated with surgical resection of AVMs are graded by the Spetzler-Martin scale, which reﬂects the importance of AVM size, location, and venous drainage. Higher grades are associated with increased postoperative morbidity and mortality. In most series, complete surgical resection of grade I or II AVMs is associated with minimal surgical complications (0 to 10%). Resection of grade IV or V AVMs is associated with complication rates exceeding 40% [44]. The efﬁcacy of radiosurgery in AVM treatment has been demonstrated in a number of studies [45–53]. For instance, Pollock et al. reported the results of stereotactic Gamma Knife radiosurgery for 65 patients with Spetzler-Martin grade I or II AVMs who opted not to undergo surgery. The series reported a cure rate of 84% and a complication rate of 7.7% [52]. As was

16

c.c. chen et al. FIGURE 2-7. MRI as an adjunct to conventional angiography in deﬁning AVM nidus. (a) Anterior-posterior and (b) lateral views of an AVM during mid–arterial phase (right internal carotid artery injection) revealed a large AVM in the right parietal temporal region supplied by distal right middle cerebral artery, posterior cerebral artery, and external carotid artery branches. (c) Coronal T2-weighted and (d) axial T1-weighted images further reﬁne the anatomic geometry of this complex AVM.

FIGURE 2-8. Comparison of CTA, MRA, and conventional angiography in AVM resolution. (a) Axial CTA image of a patient who presented with a left parietal-occipital intracranial hemorrhage. An AVM nidus is visualized on the superoposterior aspect of the hematoma. (b) Axial image reconstruction affords improved anatomic resolution of the AVM nidus, demonstrating feeder vessels from the middle cerebral artery and the posterior cerebral artery. (c) Sagittal reconstruction of the CTA images shows venous drainage of the AVM into the superior sagittal sinus. (d) Three-dimensional reconstruction of the CTA images

allows visualization of the spatial geometry of the AVM nidus. (e) Cranial-caudal view of an MRA demonstrating the left parietaloccipital hematoma as well as an enlarged, left middle cerebral artery branch. The AVM nidus is poorly visualized. (f) Anterior-posterior and (g) lateral views of the AVM during the mid–arterial phase of a conventional angiogram (left internal carotid artery injection) show detailed view of the arterial feeders and venous drainage of the AVM. The resolution afforded by conventional angiogram in this regard is superior to that of CTA or MRA.

2.

neuroimaging in radiosurgery planning and evaluation

planning for AVM treatment [58]. These authors investigated the spatial discrepancy between AVM targets as deﬁned by MRA, CTA, and conventional stereotactic angiogram. The authors reported a mean discrepancy of 2 to 3 mm in the center of the target volume when MRA and CTA targets were compared with conventional angiography–deﬁned targets. Discrepancies were noted in the left-right, anterior-posterior, and cranio-caudal directions. Independent of the size of the AVM, the discrepancy was greater than 5 mm in one-third of the cases. Despite their limitations, CT and MR angiograms allow for a more precise deﬁnition of the AVM nidus when used in conjunction with conventional angiography [48, 55–57]. In a series of 28 patients, Kondziolka et al. reported that MR angiography provided information on irregularly shaped AVMs that was not visualized by conventional angiography alone. The utility of MR angiography was especially evident in situations where conventional angiography showed different superior and inferior nidus margins on different projections. The improved nidus deﬁnition resulted in modiﬁcation of treatment plans in 16 of 28 (55%) cases [56]. Others have reported similar results [48, 55, 57]. Conventional angiography for the purpose of radiosurgical planning is generally done by placing the patient in an immobilizing stereotactic head frame in order to ensure maximal spatial accuracy. Recent advances in image acquisition and retrospective image coregistration techniques have raised questions about the necessity of such practice. Three-dimensional rotatory angiography is an imaging technique that combines the principles of CT with those of conventional angiography. As the contrast material is injected through the cerebral vasculature, X-ray transmissions from a rotating emitter are detected and digitally converted into a high-resolution view of the cerebral architecture in three dimensions. Radiosurgical planes derived using nonstereotactic three-dimensional angiography have been compared with those derived from conventional stereotactic angiography. In one study, this comparison revealed target coordinate discordance in 5 of the 20 patients. Coordinate discordance ranged from 0.3 to 1 mm with a mean of 0.7 mm [54]. Today, the gold standard for radiosurgical treatment of AVMs remains a combination of stereotactic cerebral angiography, CT-based or MR-based imaging, (including CTA and MRA), and three-dimensional angiography. The resultant information enables a detailed geometric reconstruction of nidus anatomy that is required in complex treatment strategies [60]. With improvement in algorithms for image acquisition and coregistration, nonstereotactic three-dimensional angiography may supplant the need for stereotactic angiography in radiosurgical planning for selected patients.

Radiologic Considerations in Radiosurgical Planning The following section will review pertinent radiologic features that affect radiosurgical planning, including size of lesion and proximity to critical neuroanatomic structures. Nonradiologic factors that affect radiosurgical planning are reviewed elsewhere [61, 62].

17

Size of the Lesion The size of the lesion treated remains an important criterion for determining the appropriateness of radiosurgery versus radiotherapy. As the size of the irradiated target volume increases, undesired radiation of the surrounding nontarget tissue increases in an exponential manner. The clinical impact of this geometric inevitability is magniﬁed by the higher doses of radiation delivered in radiosurgery. Generally, single-dose irradiation of normal cerebral parenchyma should be restricted to <12 to 15 Gy, because doses exceeding this range are associated with increased risks of neurologic deﬁcits [63]. As a result of this dose restriction, lesions with sizes >4 cm are usually not treated radiosurgically.

Critical Neuroanatomic Structures Radiosurgery may be contraindicated for the treatment of lesions in the immediacy of highly radiosensitive neuroanatomic structures. For instance, because of the radiosensitivity of cranial nerve II, radiosurgery is contraindicated in the treatment of lesions in the proximity of or intrinsic to the optic nerve, chiasm, or tracts. The distance between the tumor margin and the optic apparatus should be at least 4 mm before radiosurgery is considered [64, 65]. The dose delivered to the optic apparatus should be restricted to less than 10 Gy to minimize the risk of optic neuropathy. Like cranial nerve II, cranial nerves VII and VIII are more radiosensitive than other neuroanatomic structures [66]. Detailed delineation of these structures on neuroimaging is required for radiosurgical planning. Lesion localization relative to regional cerebral anatomy is another consideration in radiosurgical planning because this spatial relationship is a major predictor for posttreatment complications. Flickinger et al. reviewed 332 patients with AVMs treated with radiosurgery and correlated the risk of posttreatment neurologic injury to the location of the lesion. The risk for neurologic deﬁcit is maximal when the lesions are located in the deep gray matters (thalamus, basal ganglia) and brain stem (pons/midbrain). Minimal risk for deﬁcit was seen in the lesions located in the frontal and temporal lobe [63]. Thus, depending on lesion location, radiation doses should be adjusted to minimize the risk of posttreatment neurologic deﬁcit.

Evaluation of Treatment Efﬁcacy Radiation can induce imaging changes that are unrelated to the underlying disease process. Misinterpretation of these imaging results can lead to inappropriate treatment. For instance, radiation induces cytotoxicity and also disrupts cerebral vascular architecture. These changes, often referred to as radiation necrosis, can lead to imaging ﬁndings that are indistinguishable from tumor recurrence on contrast-enhanced MRI. As another example, radiation can induce inﬂammatory changes, causing a temporary increase in the volume of contrast enhancement that is unrelated to tumor regrowth [67, 68]. Thus, optimal patient management requires an understanding of the imaging ﬁndings as they relate to clinical outcome. This issue will be addressed

18

c.c. chen et al.

in the following section. Means of distinguishing radiation necrosis from tumor recurrence will be discussed in the context of malignant gliomas and intracranial metastasis. Temporary increases in the volume of contrast enhancement will be reviewed in the context of vestibular neuroma. The utility of serial MR in the management of trigeminal neuralgia will also be discussed. Finally, the complexity of posttreatment management after AVM radiosurgery will be reviewed.

Malignant Gliomas and Metastatic Disease For malignant gliomas, radiosurgery represents a second-line treatment option, usually offered in the context of recurrence after primary therapy or as adjuvant therapy. In contrast, radiosurgery is a ﬁrst-line treatment option for intracranial metastasis without signiﬁcant mass effect. In both diseases, typical follow-up regimens include neurologic examination and contrast-enhanced MRI 6 to 8 weeks after radiosurgery and then at roughly 3-month intervals. More frequent imaging is performed with neurologic deterioration or with interval enlargements of the treated lesion [67, 68]. For both malignant gliomas and metastatic diseases, response to radiosurgery is radiologically deﬁned by a decrease in the size of the contrast-enhancing volume on conventional MRI [67, 68]. Complete and partial responses are generally deﬁned as the complete disappearance or a partial decrease in the size of the enhancing lesion, respectively. No change in the contrast-enhancing volume is usually referred to as stable disease. Using these deﬁnitions, Ross et al. described the imaging ﬁndings after radiosurgery for malignant gliomas and intracranial metastasis. Based on imaging obtained roughly 3 months after radiosurgery, 17% of the patients with malignant glioma showed partial or complete response; 10% showed stable disease, and 73% showed an increase in the size of contrast enhancement. In patients with intracranial metastasis, images obtained in the same time frame revealed 51% complete or partial response, 27% stable disease, and 22% contrastenhancement size increase [67]. Comparable results are reported by other studies [69–72]. The use of corticosteroids is common in the treatment of cerebral edema in both malignant gliomas and intracranial metastasis. Questions are often raised as to whether corticosteroid administration affects MRI ﬁndings after radiosurgery. Case series addressing this issue revealed that whereas corticosteroids reduced the extent of peritumoral edema on T2weighted MRI, their administration did not alter the size of the contrast-enhancing volume [67]. The physiology underlying increased contrast uptake is complex and cannot be equated with treatment failure in all cases. Two independent series that detailed the imaging changes after radiosurgical treatment of intracranial metastatic disease revealed that 6% to 12% of all radiosurgically treated lesions displayed a transient increase in contrast uptake [67, 68]. This transient increase occurred at 3 months after treatment (range, 2 to 10 months) and resolved after an additional 6 months (range, 2 to 6 months). Microsurgical resection of these contrast-enhancing volumes revealed hyalinized thrombosis, tumor necrosis, and granulomatous changes [73]. Genuine treatment failures, documented by surgical biopsy and enlargement of

contrast-enhancing volume, on the other hand, were more likely to occur at 6 months after treatment (range, 3 to 24 months). Thus, although contrast-enhancing volume represents an important predictor for therapeutic efﬁcacy [67], it must be interpreted with caution. Another caveat in equating increased contrast uptake with radiosurgical failure involves the phenomenon of radiation necrosis. Radiation necrosis is a term used to describe radiologic changes (primarily visualized in the form of increasing size of contrast-enhancing volume) resulting from radiation-induced cytotoxicity that is unrelated to the underlying disease process. Some reports suggest that radiation necrosis is more likely to produce a fuzzy, indiscrete pattern of enhancement in contrast with the discrete pattern of enhancement seen in tumor recurrence [68]; however, 10% to 20% of post-radiosurgery patients develop radiologic ﬁndings indistinguishable from tumor recurrence [74]. The functional imaging modalities described earlier have been employed as a means to better distinguish radiation necrosis from tumor recurrence. Early results are promising in this regard. In general, whereas functional imaging modalities exhibit high degrees of sensitivity for detecting tumor recurrence, the speciﬁcity remains somewhat poor. For instance, Tusyuguchi et al. compared the methionine PET ﬁndings in eight cases of biopsy-proven glioma recurrence with six cases of biopsy-proven radiation necrosis. They calculated the ratio of methionine accumulation in the regions of contrast enhancement relative to regions of normal gray matter. They found this ratio elevated in cases of tumor recurrence when compared with cases of radiation necrosis. In this study, the sensitivity and speciﬁcity of methionine PET for tumor recurrence detection were 100% and 60%, respectively [75]. Comparable results are reported for thallium 201 SPECT [76]. Measurement of cerebral blood volume (CBV) on MRI represents another proxy for tumor recurrence [38, 39, 77]. Essig et al. described their experience with 18 patients imaged at 6 weeks and 3 months after radiosurgery for solitary metastasis. In this study, a decrease in the CBV of the radiosurgically treated volume at 6 weeks predicted treatment outcome with a sensitivity of 97% and a speciﬁcity of 71% [77]. Perhaps the most promising imaging modality for determining radiation necrosis versus tumor recurrence involves the use of MRS. As previously described, choline is a cell membrane component that reﬂects the extent of membrane turnover. Elevated levels of choline are associated with rapidly proliferating tumors and poor clinical outcomes [2]. The level of creatine is a proxy for cellular metabolism and is decreased in tumor cells. Using measurements of choline and creatine as guides, Rabinov et al. were able to distinguish recurrent tumor from radiation necrosis in 13 of 14 cases [74]. Others have reported similar ﬁndings [35, 78]. Future imaging evaluation after radiosurgery likely will involve the synthesis of information obtained from different imaging modalities. The development of algorithms for data integration requires clinical-imaging correlation. To date, most studies carried out for such purpose are retrospective in design and involve a small number of patients. Ultimately, prospective and randomized studies will be required for the purpose of correlating imaging ﬁndings with clinical outcomes.

2.

neuroimaging in radiosurgery planning and evaluation

Benign Lesions: Acoustic Neuroma Imaging ﬁndings after radiosurgical treatment for acoustic neuromas are discussed as a review of the basic principles underlying the management of benign intracranial lesions after radiosurgery. Acoustic neuromas, also known as vestibular schwannomas, are rare, benign tumors that arise from the Schwann cells associated with the eighth cranial nerve. Their incidence is approximately 1 in 100,000 in the general population. Historically, surgical resection has been the treatment of choice. Two important goals of acoustic neuroma surgery are facial nerve and hearing preservation. The likelihood of achieving these goals is largely a function of the tumor size. In recent years, radiosurgery has emerged as an alternative to microsurgical treatment for small acoustic neuromas (<2 cm). In the largest series reported to date (827 patients over a span of 10 years at the University of Pittsburgh), radiosurgical administration of 12 to 20 Gy to the schwannoma achieved local control in 97% of the cases after 10 years [79]. Other series have reported similar ﬁndings [80–84]. After radiosurgery, the rate of House-Brackman grade I/II facial nerve function preservation ranged from 70% to 95%; Robertson-Garner serviceable hearing preservation ranged from 13% to 40% [80–84]. In multiple retrospective comparisons of radiosurgery and microsurgical resections, no statistically signiﬁcant difference was observed in the rate of tumor recurrence, facial nerve preservation, or hearing preservation [82, 85]. Currently, MRI with contrast enhancement is the gold standard for the evaluation of treatment response [86, 87]. Typically, follow-up neuroimaging and clinical examinations are performed at 6 and 12 months during the ﬁrst year and every 12 months thereafter. In the largest clinical series to date, Nakamura et al. reported their experiences with serial MRI after Gamma Knife radiosurgery for vestibular schwannoma. They classiﬁed the changes in the posttreatment contrast-enhancing volume into four categories. The ﬁrst category consisted of schwannomas that showed initial enlargement followed by sustained regression (25/78, or 32%). This temporary enlargement peaked in roughly 1 year and regressed within an additional 2 years. In many cases, the schwannoma doubled in size before regression. The tumor may not have regressed to the size of the initial lesion. Instead, many tumors regressed to an intermediate size and remained stable at that size. The second category consisted of tumors that showed repeated enlargement followed by regression (8/78, or 9%). Many of these tumors were cystic schwannomas (5/8), with size ﬂuctuations resulting from enlargement or collapse of the cystic component. Size ﬂuctuations in noncystic schwannomas also occurred (3/8). Again, the size enlargement could reach a doubling of the initial lesion size before regression. The third category consisted of schwannomas that remained stable in size or regressed in size (21/78, or 27%). The fourth category consisted of continual tumor enlargement (7/78, or 9%) [87]. Microsurgical resection of tumors that showed radiologic “progression” often revealed hyalinized thrombosis, thickened vascular wall, and granulomatous changes [88, 89]. Thus, some of the cases of radiologic “progression” may have represented inﬂammatory changes rather than genuine treatment failure. Because up to 41% (combined category 1 and 2) of schwanno-

19

mas treated with radiosurgery displayed a temporary increase in the volume of contrast enhancement (sometimes doubling in size), cautious observation may be warranted for at least 2 years after treatment, provided the imaging changes are not associated with clinical deterioration or signiﬁcant compression of the brain stem. Similar temporary enlargement of contrast-enhancing volumes is seen after radiosurgery for pituitary adenomas and other benign diseases [90]. As such, in the management of patients with benign intracranial lesions after radiosurgery, surgical intervention or a second round of treatment should be reserved for cases with continual disease progression after or for cases with neurologic deterioration. Another radiologic ﬁnding pertinent to the management of vestibular schwannoma after radiosurgery is that of hydrocephalus. It is estimated that roughly 10% of patients with vestibular schwannoma develop communicating hydrocephalus after radiosurgical or radiation treatment [91]. In most cases, the hydrocephalus is not associated with tumor enlargement or cerebrospinal ﬂuid (CSF) ﬂow obstruction. The etiology of the hydrocephalus is unclear though many investigators attribute the phenomenon to CSF malabsorption secondary to tumor necrosis. It is also unclear whether radiation contributes to the process, as a comparable percentage of vestibular schwannoma patients without radiation treatment develops communicating hydrocephalus [92]. Regardless of the etiology, prompt identiﬁcation of ventricular enlargement on imaging and clinical ﬁndings associated with hydrocephalus are needed in order to ensure timely neurosurgical intervention.

Trigeminal Neuralgia Imaging ﬁndings after radiosurgical treatment for trigeminal neuralgia are reviewed to illustrate a clinical scenario where routine, serial MRI is not warranted. Trigeminal neuralgia is a facial pain syndrome consisting of paroxysmal, lancinating pain occurring in the distribution of cranial nerve V. Most patients with trigeminal neuralgia are successfully treated with anticonvulsives, antidepressants, neuroleptics, or opioids. Options available for the treatment of medically resistant trigeminal neuralgia include microvascular decompression, thermal, chemical, or radiofrequency ablative procedures, and radiosurgery [93]. In most studies, excellent responses to radiosurgery are reported in 70% to 90% of patients after treatment [94, 95]. Many institutions obtain a MR contrast-enhanced scan approximately 6 months after radiosurgical treatment. The scan is performed primarily for the purpose of target site veriﬁcation. In one report, all patients treated with 45 Gy at the 50% isodose line developed contrast enhancement of the target zone within 6 months of treatment [94]. The detection of contrast enhancement on cranial nerve V after radiosurgery, therefore, served as a conﬁrmation of accurate targeting. Studies correlating clinical responses to contrast enhancement of cranial nerve V have yielded mixed results. Some studies suggest that the exact region of enhancement relative to the pontine edge and along the retrogasserian portion of cranial nerve V correlate well with treatment outcome [96–98]. Others report poor correlation between contrast enhancement and clinical response [99]. In many instances, beneﬁcial clinical

20

c.c. chen et al.

responses or treatment failures are apparent before the onset of contrast enhancement [99]. Despite the lack of consistent correlation between radiologic ﬁndings and clinical response, a case can be made for obtaining a MR contrast-enhanced scan at the 6-month followup evaluation, because it allows target conﬁrmation as well as an assessment of potential brain-stem injuries. Serial MRI in patients without neurologic or radiologic changes at the initial posttreatment scan, however, does not appear warranted. In patients who exhibit good clinical response to radiosurgery without incurring neurologic deﬁcits, it is also reasonable to forgo the MRI at the 6-month follow-up evaluation, because the imaging ﬁndings are unlikely to alter patient management.

Arteriovenous Malformations Therapeutic efﬁcacy of radiosurgery for AVMs depends on radiation-induced endothelial cell and mesenchymal cell proliferation, causing progressive vasoocclusion [100]. Although it is generally agreed that the peak effect of this process occurs between 2 and 3 years after radiation treatment, the actual time course of AVM obliteration varies widely [45–53]. Complete angiographic obliteration, as determined by conventional angiography, remains the gold standard for deﬁning treatment success and can occur as early as 4 months or as late as 5 years after treatment [101, 102]. Because of the variable time course in therapeutic response, the frequency and timing of neuroimaging follow-up evaluation vary widely between institutions [103–105]. Because of the risks associated with conventional cerebral angiography and because most AVMs obliterate between 2 and 3 years after radiosurgery, there is a tendency to delay cerebral angiography until the presumed time of obliteration [103, 105]. Some investigators, however, advocate early cerebral angiography on grounds that ﬁndings on these studies are highly predictive of the ﬁnal treatment outcome. Oppenheim et al. reported their experience with 138 patients radiosurgically treated for AVMs. These patients underwent early angiography 6 to 18 months after treatment. Eighty-four percent of the radiosurgically treated AVMs with evidence of early regression (deﬁned as >75% reduction in size at the time of early angiography) eventually developed complete obliteration on subsequent angiograms. On the other hand, only 10% of the AVMs without evidence of early regression (<50% reduction in size on the early angiogram) developed complete obliteration. The authors concluded that the identiﬁcation of AVMs unresponsive to radiosurgery at a early stage will facilitate planning for subsequent treatment strategies [104]. To avoid the risks associated with conventional cerebral angiograms, investigators have identiﬁed CT and MR imaging ﬁndings that correlate well with treatment response. One such ﬁnding involves contrast enhancement in the region of the AVM nidus after radiosurgery. Contrast enhancement of the AVM nidus on CT imaging was ﬁrst described in two patients with eventual AVM obliteration after radiosurgery [106]. Subsequent studies reported good correlation between nidus enhancement on CT or MR imaging after radiosurgery and therapeutic efﬁcacy [102, 107]. These studies revealed that the volume of contrast enhancement corresponds with the radiosurgical target volume. The degree of enhancement increases

with time and is correlated with a reduction in the nidus size on conventional angiography. The onset for contrast enhancement is typically 6 to 24 months after radiosurgery. Contrast enhancement on CT tends to resolve 1 to 2 years after angiographic demonstration of complete AVM obliteration, whereas enhancement on MRI tends to persist even after disappearance of contrast enhancement on CT [102]. MR and CT angiograms are also used to evaluate cerebral AVMs after radiosurgery. Though the spatial resolution of CT and MR imaging remains poor, AVM changes visualized using these modalities correlate well with those detected using conventional angiography [102, 108–110]. As such, serial MR or CT angiograms are often performed for routine monitoring while the deﬁnitive conventional angiography is postponed until 2 to 3 years after radiosurgery. Radiosurgery of AVMs inevitably results in irradiation of the surrounding nontarget tissues. It is, therefore, not surprising that an estimated 28% to 50% of the patients undergoing treatment develop T2 signal abnormalities in these regions [100, 111]. The onset of these ﬁndings occurs between 1 day and 44 months after treatment. Regression of the signal abnormality is seen in 80% to 90% of the cases and tends to occur 5 to 8 months after the initial onset [112]. As previously discussed, the risk of neurologic deﬁcit expected due to these abnormalities depends on their location. Maximal risk for deﬁcit is expected in cases of signal abnormalities in the deep gray matters (thalamus, basal ganglia) and brain stem (pons/midbrain). Minimal risk of deﬁcit is expected with frontal and temporal lobe abnormalities [63]. Whereas conventional angiography represents the gold standard for deﬁning therapeutic efﬁcacy for AVM treatment, disappearance of the AVM on angiography after radiosurgery does not always indicate disease eradication. Shin et al. followed 236 radiosurgery-treated AVMs between 1 and 133 months after angiographic conﬁrmation of obliteration. The authors identiﬁed four patients who developed intracranial hemorrhage between 16 and 51 months after angiographic conﬁrmation. No evidence of residual AVM was found on retrospective review of the conﬁrmation angiograms. Two of the patients underwent surgical resection. Histologic analysis of the resection specimen revealed evidence of vasoocclusion as well as small residual AVM vessels. The only radiologic ﬁndings associated with these hemorrhages were the persistence of contrast enhancement on CT and MRI after angiographic evidence of AVM obliteration [113]. Given these ﬁndings, yearly followup evaluation after angiographic evidence of AVM obliteration may be warranted. Treatment of patients with AVMs remains one of the most complex and challenging in the ﬁeld of radiosurgery. Optimal patient management requires an understanding of the pathophysiology, neuroanatomy, as well as the radiologic manifestations after treatment. As such, treatment efforts should involve collaborative inputs from experienced radiation oncologists, neurosurgeons, neuroradiologists, and the patient.

Neuroimaging for Radiation-Associated Secondary Tumors The probability of secondary tumors arising from radiosurgery is quite low [114]. Thus, once a patient has achieved a positive

2.

neuroimaging in radiosurgery planning and evaluation

result from treatment (e.g., complete obliteration of an AVM documented by angiography), we do not recommend further imaging to look for secondary tumor formation.

Conclusion There is no doubt that advances in neuroimaging will help to reﬁne all aspects of radiosurgery and improve treatment efﬁcacy. For many modalities, clinical applications remain poorly deﬁned and await further investigation. For radio surgeons and therapists, the challenge lies in understanding the basis and the limitations associated with the various imaging modalities. Ultimately, prospective and randomized studies correlating imaging ﬁndings and clinical outcomes are required for developing guidelines for optimal patient care.

References 1. Battista J, Rider W, Van DJ. Computed tomography for radiotherapy planning. Int J Radiat Oncol Biol Phys 1980; 6:99–107. 2. Nelson SJ, Cha S. Imaging glioblastoma multiforme. Cancer J 2003; 9:134–145. 3. Sumanaweera TS et al. Characterization of spatial distortion in magnetic resonance imaging and its implications for stereotactic surgery. Neurosurgery 1994; 35:696–703; discussion 703–704. 4. Chang H, Fitzpatrick JM. A technique for accurate magnetic resonance imaging in the presence of ﬁeld inhomogeneities. IEEE Trans Med Imaging 1992; 11:319–329. 5. Sumanaweera T et al. MR susceptibility misregistration correction. IEEE Trans Med Imaging 1993; 12:251–259. 6. Sumanaweera T et al. Quantifying MRI geometric distortion in tissue. Magn Reson Med 1994; 31:40–47. 7. Mack A et al. Analyzing 3-tesla magnetic resonance imaging units for implementation in radiosurgery. J Neurosurg 2005; 102(Suppl): 158–164. 8. Bednarz G et al. Evaluation of the spatial accuracy of magnetic resonance imaging-based stereotactic target localization for gamma knife radiosurgery of functional disorders. Neurosurgery 1999; 45:1156–1161; discussion 1161–113. 9. Kondziolka D et al. A comparison between magnetic resonance imaging and computed tomography for stereotactic coordinate determination. Neurosurgery 1992; 30:402–406; discussion 406– 407. 10. Orth RC et al. Development of a unique phantom to assess the geometric accuracy of magnetic resonance imaging for stereotactic localization. Neurosurgery 1999; 45:1423–1429; discussion 1429– 1431. 11. 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; 38:170– 176; discussion 176–178. 12. Walton L et al. Stereotactic localization with magnetic resonance imaging: a phantom study to compare the accuracy obtained using two-dimensional and three-dimensional data acquisitions. Neurosurgery 1997; 41:131–137; discussion 137–139. 13. Task Force on Stereotactic Radiosurgery. Consensus statement on stereotactic radiosurgery quality improvement. The American Society for Therapeutic Radiology and Oncology, Task Force on Stereotactic Radiosurgery and the American Association of Neurological Surgeons. Int J Radiat Oncol Biol Phys 1994; 28:527–530. 14. Khoo VS et al. Magnetic resonance imaging (MRI): considerations and applications in radiotherapy treatment planning. Radiother Oncol 1997; 42:1–15.

21

15. Shaw E et al. Radiation Therapy Oncology Group: radiosurgery quality assurance guidelines. Int J Radiat Oncol Biol Phys 1993; 27:1231–1239. 16. Schad LR et al. Radiotherapy treatment planning of basal meningiomas: improved tumor localization by correlation of CT and MR imaging data. Radiother Oncol 1992; 25:56–62. 17. Just M et al. MRI-assisted radiation therapy planning of brain tumors—clinical experiences in 17 patients. Magn Reson Imaging 1991; 9:173–177. 18. Shuman WP et al. MR imaging in radiation therapy planning. Work in progress. Radiology 1985; 156:143–147. 19. Shuman WP et al. The utility of MR in planning the radiation therapy of oligodendroglioma. AJR Am J Roentgenol 1987; 148: 595–600. 20. Ten Haken RK et al. A quantitative assessment of the addition of MRI to CT-based, 3-D treatment planning of brain tumors. Radiother Oncol 1992; 25:121–133. 21. De Salles AA et al. Transposition of target information from the magnetic resonance and computed tomography scan images to conventional X-ray stereotactic space. Appl Neurophysiol 1987; 50:23–32. 22. Pelizzari CA et al. Accurate three-dimensional registration of CT, PET, and/or MR images of the brain. J Comput Assist Tomogr 1989; 13:20–26. 23. Judnick JW et al. Radiotherapy technique integrates MRI into CT. Radiol Technol 1992; 64:82–89. 24. Evans PM et al. Image comparison techniques for use with megavoltage imaging systems. Br J Radiol 1992; 65:701–709. 25. Kelly PJ et al. Stereotactic histologic correlations of computed tomography- and magnetic resonance imaging-deﬁned abnormalities in patients with glial neoplasms. Mayo Clin Proc 1987; 62: 450–459. 26. Apisarnthanarax S, Chao KS. Current imaging paradigms in radiation oncology. Radiat Res 2005; 163:1–25. 27. Tralins KS et al. Volumetric analysis of 18F-FDG PET in glioblastoma multiforme: prognostic information and possible role in deﬁnition of target volumes in radiation dose escalation. J Nucl Med 2002; 43:1667–1673. 28. Gross MW et al. The value of F-18-ﬂuorodeoxyglucose PET for the 3-D radiation treatment planning of malignant gliomas. Int J Radiat Oncol Biol Phys 1998; 41:989–995. 29. Pardo FS et al. Functional cerebral imaging in the evaluation and radiotherapeutic treatment planning of patients with malignant glioma. Int J Radiat Oncol Biol Phys 1994; 30:663–669. 30. Nuutinen J et al. Radiotherapy treatment planning and long-term follow-up with [(11)C]methionine PET in patients with low-grade astrocytoma. Int J Radiat Oncol Biol Phys 2000; 48:43–52. 31. Pirzkall A et al. MR-spectroscopy guided target delineation for high-grade gliomas. Int J Radiat Oncol Biol Phys 2001; 50:915–928. 32. Chan A, Nelson S. Use of 1H magnetic resonance spectroscopic imaging in managing recurrent glioma patients undergoing radiosurgery. ISMRM Workshop Proceedings on MR of Cancer 2002; 227. 33. Chan AA et al. Proton magnetic resonance spectroscopy imaging in the evaluation of patients undergoing gamma knife surgery for Grade IV glioma. J Neurosurg 2004; 101:467–475. 34. Graves EE et al. A preliminary study of the prognostic value of proton magnetic resonance spectroscopic imaging in gamma knife radiosurgery of recurrent malignant gliomas. Neurosurgery 2000; 46:319–326; discussion 26–28. 35. Graves EE et al. Serial proton MR spectroscopic imaging of recurrent malignant gliomas after gamma knife radiosurgery. AJNR Am J Neuroradiol 2001; 22:613–624. 36. McKnight TR et al. Histopathological validation of a three-dimensional magnetic resonance spectroscopy index as a predictor of tumor presence. J Neurosurg 2002; 97:794–802.

22

c.c. chen et al.

37. Nelson SJ et al. In vivo molecular imaging for planning radiation therapy of gliomas: an application of 1H MRSI. J Magn Reson Imaging 2002; 16:464–476. 38. Cha S. Perfusion MR imaging: basic principles and clinical applications. Magn Reson Imaging Clin N Am 2003; 11:403–413. 39. Cha S. Perfusion MR imaging of brain tumors. Top Magn Reson Imaging 2004; 15:279–289. 40. Castillo M et al. Apparent diffusion coefﬁcients in the evaluation of high-grade cerebral gliomas. AJNR Am J Neuroradiol 2001; 22:60–64. 41. Moffat BA et al. Functional diffusion map: a noninvasive MRI biomarker for early stratiﬁcation of clinical brain tumor response. Proc Natl Acad Sci USA 2005; 102:5524–5529. 42. Nelson S. Imaging of brain tumors after therapy. Neuroimaging Clin N Am 1999; 9:801–819. 43. Benjamin R, Capparella J, Brown A. Classiﬁcation of glioblastoma multiforme in adults by molecular genetics. Cancer J 2003; 9:982–990. 44. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg 1986; 65:476–483. 45. Flickinger JC, Kondziolka D, Lunsford LD. Clinical applications of stereotactic radiosurgery. Cancer Treat Res 1998; 93:283–297. 46. Friedman WA. LINAC radiosurgery. Neurosurg Clin N Am 1990; 1:991–1008. 47. Friedman WA, Bova FJ. Linear accelerator radiosurgery for arteriovenous malformations. J Neurosurg 1992; 77:832–841. 48. Friedman WA, Bova FJ, Mendenhall WM. Linear accelerator radiosurgery for arteriovenous malformations: the relationship of size to outcome. J Neurosurg 1995; 82:180–189. 49. Friedman WA, Bova FJ, Spiegelmann R. Linear accelerator radiosurgery at the University of Florida. Neurosurg Clin N Am 1992; 3:141–166. 50. Kondziolka D, Lunsford LD. The case for and against AVM radiosurgery. Clin Neurosurg 2001; 48:96–110. 51. Lunsford LD et al. Stereotactic radiosurgery of brain vascular malformations. Neurosurg Clin N Am 1992; 3:79–98. 52. Pollock BE et al. Patient outcomes after stereotactic radiosurgery for “operable” arteriovenous malformations. Neurosurgery 1994; 35:1–7; discussion 7–8. 53. Steiner L et al. Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 1992; 77:1–8. 54. Colombo F et al. Three-dimensional angiography for radiosurgical treatment planning for arteriovenous malformations. J Neurosurg 2003; 98:536–543. 55. Blatt DR, Friedman WA, Bova FJ. Modiﬁcations based on computed tomographic imaging in planning the radiosurgical treatment of arteriovenous malformations. Neurosurgery 1993; 33:588–595; discussion 595–596. 56. Kondziolka D et al. Stereotactic magnetic resonance angiography for targeting in arteriovenous malformation radiosurgery. Neurosurgery 1994; 35:585–590; discussion 590–591. 57. Levy RP et al. Computed tomography slice-by-slice target-volume delineation for stereotactic proton irradiation of large intracranial arteriovenous malformations: an iterative approach using angiography, computed tomography, and magnetic resonance imaging. Int J Radiat Oncol Biol Phys 1996; 35:555–564. 58. Aoyama H et al. Comparison of imaging modalities for the accurate delineation of arteriovenous malformation, with reference to stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2005; 62:1232–1238. 59. Tanaka H et al. Initial experience with helical CT and 3D reconstruction in therapeutic planning of cerebral AVMs: comparison with 3D time-of-ﬂight MRA and digital subtraction angiography. J Comput Assist Tomogr 1997; 21:811–817. 60. Kakizawa Y et al. Compartments in arteriovenous malformation nidi demonstrated with rotational three-dimensional digital sub-

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71. 72.

73.

74.

75.

76.

77.

78.

79.

80.

traction angiography by using selective microcatheterization. Report of three cases. J Neurosurg 2002; 96:770–774. Debus J et al. Brainstem tolerance to conformal radiotherapy of skull base tumors. Int J Radiat Oncol Biol Phys 1997; 39: 967–975. Debus J et al. Dose-volume tolerance of the brainstem after high-dose radiotherapy. Front Radiat Ther Oncol 1999; 33:305– 314. Flickinger JC et al. 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; 46:1143–1148. 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 et al. Tolerance of cranial nerves of the cavernous sinus to radiosurgery. Int J Radiat Oncol Biol Phys 1993; 27:215– 221. Foote KD et al. Analysis of risk factors associated with radiosurgery for vestibular schwannoma. J Neurosurg 2001; 95:440– 449. Huber PE et al. Transient enlargement of contrast uptake on MRI after linear accelerator (linac) stereotactic radiosurgery for brain metastases. Int J Radiat Oncol Biol Phys 2001; 49:1339– 1349. Ross DA et al. Imaging changes after stereotactic radiosurgery of primary and secondary malignant brain tumors. J Neurooncol 2002; 56:175–181. Auchter RM 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. Engenhart R et al. Long-term follow-up for brain metastases treated by percutaneous stereotactic single high-dose irradiation. Cancer 1993; 71:1353–1361. Flickinger JC et al. Radiosurgery: its role in brain metastasis management. Neurosurg Clin N Am 1996; 7:497–504. Mehta MP et al. Stereotactic radiosurgery for glioblastoma multiforme: report of a prospective study evaluating prognostic factors and analyzing long-term survival advantage. Int J Radiat Oncol Biol Phys 1994; 30:541–549. Shrieve DC et al. Comparison of stereotactic radiosurgery and brachytherapy in the treatment of recurrent glioblastoma multiforme. Neurosurgery 1995; 36:275–282; discussion 82–84. Rabinov JD et al. In vivo 3-T MR spectroscopy in the distinction of recurrent glioma versus radiation effects: initial experience. Radiology 2002; 225:871–879. 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; 18:291–296. Kimura T et al. Evaluation of the response of metastatic brain tumors to stereotactic radiosurgery by proton magnetic resonance spectroscopy, 201TlCl single-photon emission computerized tomography, and gadolinium-enhanced magnetic resonance imaging. J Neurosurg 2004; 100:835–841. Essig M et al. Assessment of brain metastases with dynamic susceptibility-weighted contrast-enhanced MR imaging: initial results. Radiology 2003; 228:193–199. Wald LL et al. Serial proton magnetic resonance spectroscopy imaging of glioblastoma multiforme after brachytherapy. J Neurosurg 1997; 87:525–534. Lunsford LD et al. Radiosurgery of vestibular schwannomas: summary of experience in 829 cases. J Neurosurg 2005; 102(Suppl): 195–199. Harsh G et al. Proton beam stereotactic radiosurgery of vestibular schwannomas. Int J Radiat Oncol Biol Phys 2002; 54:35–44.

2.

neuroimaging in radiosurgery planning and evaluation

81. Karpinos M et al. Treatment of acoustic neuroma: stereotactic radiosurgery vs. microsurgery. Int J Radiat Oncol Biol Phys 2002; 54:1410–1421. 82. Pollock B et al. Outcome analysis of acoustic neuroma management: a comparison of microsurgery and stereotactic radiosurgery. Neurosurgery 1995; 36:215–224. 83. Prasad D, Steiner M, Steiner L. Gamma surgery for vestibular schwannoma. Neurosugery 2000; 92:745–759. 84. Petit J et al. Reduced-dose radiosurgery for vestibular schwannomas. Neurosurgery 2001; 49:1299–1306. 85. Karpinos M et al. Treatment of acoustic neuroma: stereotactic radiosurgery vs. microsurgery. Int J Radiat Oncol Biol Phys 2002; 54:1410–1421. 86. Linskey ME, Lunsford LD, Flickinger JC. Neuroimaging of acoustic nerve sheath tumors after stereotaxic radiosurgery. AJNR Am J Neuroradiol 1991; 12:1165–1175. 87. Nakamura H et al. Serial follow-up MR imaging after gamma knife radiosurgery for vestibular schwannoma. AJNR Am J Neuroradiol 2000; 21:1540–1546. 88. Hirato M et al. Gamma knife radiosurgery for acoustic schwannoma: early effects and preservation of hearing. Neurol Med Chir (Tokyo) 1995; 35:737–741. 89. Kobayashi T, Tanaka T, Kida Y. The early effects of gamma knife on 40 cases of acoustic neurinoma. Acta Neurochir Suppl 1994; 62:93–97. 90. Tung GA et al. MR imaging of pituitary adenomas after gamma knife stereotactic radiosurgery. AJR Am J Roentgenol 2001; 177:919–924. 91. Sawamura Y et al. Management of vestibular schwannoma by fractionated stereotactic radiotherapy and associated cerebrospinal ﬂuid malabsorption. J Neurosurg 2003; 99:685–692. 92. Pirouzmand F, Tator CH, Rutka J. Management of hydrocephalus associated with vestibular schwannoma and other cerebellopontine angle tumors. Neurosurgery 2001; 48:1246–1253; discussion 53–54. 93. Cheshire WP. Trigeminal neuralgia: diagnosis and treatment. Curr Neurol Neurosci Rep 2005; 5:79–85. 94. Alberico RA, Fenstermaker RA, Lobel J. Focal enhancement of cranial nerve V after radiosurgery with the Leksell gamma knife: experience in 15 patients with medically refractory trigeminal neuralgia. AJNR Am J Neuroradiol 2001; 22:1944–1948. 95. Cheuk AV et al. Gamma knife surgery for trigeminal neuralgia: outcome, imaging, and brainstem correlates. Int J Radiat Oncol Biol Phys 2004; 60:537–541. 96. Kondziolka D, Lunsford LD, Flickinger JC. Gamma knife radiosurgery as the ﬁrst surgery for trigeminal neuralgia. Stereotact Funct Neurosurg 1998; 70(Suppl 1):187–191. 97. Urgosik D et al. Gamma knife treatment of trigeminal neuralgia: clinical and electrophysiological study. Stereotact Funct Neurosurg 1998; 70(Suppl 1):200–209. 98. Young RF et al. Gamma Knife radiosurgery for treatment of trigeminal neuralgia: idiopathic and tumor related. Neurology 1997; 48:608–614.

23

99. Friedman DP, Morales RE, Goldman HW. Role of enhanced MRI in the follow-up of patients with medically refractory trigeminal neuralgia undergoing stereotactic radiosurgery using the gamma knife: initial experience. J Comput Assist Tomogr 2001; 25:727–732. 100. Morikawa M et al. Radiosurgery for cerebral arteriovenous malformations: assessment of early phase magnetic resonance imaging and signiﬁcance of gadolinium-DTPA enhancement. Int J Radiat Oncol Biol Phys 1996; 34:663–675. 101. Lunsford LD et al. Stereotactic radiosurgery for arteriovenous malformations of the brain. J Neurosurg 1991; 75:512–524. 102. Yamamoto M et al. Long-term results of radiosurgery for arteriovenous malformation: neurodiagnostic imaging and histological studies of angiographically conﬁrmed nidus obliteration. Surg Neurol 1992; 37:219–230. 103. Kauczor HU et al. 3D TOF MR angiography of cerebral arteriovenous malformations after radiosurgery. J Comput Assist Tomogr 1993; 17:184–190. 104. Oppenheim C et al. Radiosurgery of cerebral arteriovenous malformations: is an early angiogram needed? AJNR Am J Neuroradiol 1999; 20:475–481. 105. Young C et al. Radiosurgery for arteriovenous malformations: the University of Toronto experience. Can J Neurol Sci 1997; 24:99–105. 106. Loefﬂer JS et al. Stereotactic radiosurgery of the brain using a standard linear accelerator: a study of early and late effects. Radiother Oncol 1990; 17:311–321. 107. Marks MP et al. Intracranial vascular malformations: imaging of charged-particle radiosurgery. Part II. Complications. Radiology 1988; 168:457–462. 108. Abe T et al. Magnetic resonance angiography of cerebral arteriovenous malformations. Neurol Med Chir (Tokyo) 1995; 35:580– 583. 109. Noorbehesht B, Fabrikant JI, Enzmann DR. Size determination of supratentorial arteriovenous malformations by MR, CT and angio. Neuroradiology 1987; 29:512–518. 110. Quisling RG et al. Persistent nidus blood ﬂow in cerebral arteriovenous malformation after stereotactic radiosurgery: MR imaging assessment. Radiology 1991; 180:785–791. 111. Flickinger JC et al. Radiosurgery and brain tolerance: an analysis of neurodiagnostic imaging changes after gamma knife radiosurgery for arteriovenous malformations. Int J Radiat Oncol Biol Phys 1992; 23:19–26. 112. Guo WY et al. Gamma knife surgery of cerebral arteriovenous malformations: serial MR imaging studies after radiosurgery. Int J Radiat Oncol Biol Phys 1993; 25:315–323. 113. Shin M et al. Risk of hemorrhage from an arteriovenous malformation conﬁrmed to have been obliterated on angiography after stereotactic radiosurgery. J Neurosurg 2005; 102:842– 846. 114. Loefﬂer JS, Niemierko A, Chapman PH. Second tumors after radiosurgery: tip of the iceberg or a bump in the road? Neurosurgery 2003; 52:1436–1440.

3

Techniques of Stereotactic Radiosurgery Chris Heller, Cheng Yu, and Michael L.J. Apuzzo

Introduction Stereotactic navigation and radiosurgery share a rich history highlighted by innovative minds, technological advances, and clinical success. Although the two concepts are in many ways intimately joined, radiosurgery could not exist without a thorough understanding of the fundamental principles of stereotaxis. Just as the original concept of radiosurgery was a natural, albeit inspired, application of the ideas of stereotactic navigation, advances in radiosurgery over the years have largely been preceded by innovations in stereotaxy.

History The ﬁrst device intended for experimental and surgical intervention in deep-seated structures of the brain was developed for animal studies by Sir Victor A.H. Horsley and Robert H. Clarke in 1908. Using an early atlas consisting of serial transverse sections of the animal brain, each structure of interest could be assigned three-dimensional coordinates and thus accurately located using the stereotactic device [1]. Almost 40 years passed, however, until Ernest A. Spiegel and Henry T. Wycis introduced a similar device intended for targeting locations within the human cerebrum [2]. Their device was similar to that of Horsley and Clarke in that it was a rectangular design using Cartesian coordinates in an orthogonal frame of reference, but it represented a considerable advancement in target localization. Spiegel and Wycis pioneered the use of intracranial structures rather than skull landmarks for navigational reference points [3, 4]. Plain X-ray and air and contrast ventriculography were used to visualize structures such as the foramen of Monroe, the pineal gland, and the habenular calciﬁcation. In 1952, they published their own brain atlas entitled Stereoencephalotomy, which provided the relative positions of these radiographic landmarks to structures of interest within the brain [3–5]. The ﬁrst clinical use of this device was for making functional lesions within the medial globus pallidus for the treatment of Huntington’s chorea. Early success would pave the way for future treatments of other movement disorders as well as epilepsy, psychiatric disorders, and pain [6] (see Fig. 3-1).

From Stereotaxis to Radiosurgery In 1949, Swedish neurosurgeon Lars Leksell (Fig. 3-2) introduced a stereotactic apparatus for application to intracerebral surgery [4]. Though the instrument developed by Spiegel and Wycis preceded it, Leksell’s design proved more versatile for a number of reasons. His innovative design placed the target at the center of a semicircular arc rather than a rectangular box, thus eliminating the need for trigonometric calculations for angled trajectories. It also allowed for alteration of the trajectory along the arc without the need to recalculate the coordinates of the target. Additionally, he introduced the use of skull pins for rigid ﬁxation of the base frame, which increased the overall precision of the system [7]. Two years after the introduction of his stereotactic frame, Dr. Leksell developed the concept of radiosurgery. With his ability to accurately localize a target in three-dimensional space, he postulated that narrow beams of radiation intersecting at a common point could be used to deliver high doses of energy to a chosen volume in space. Using this approach, lesions deep within the brain or at the base of the skull could be treated with minimal disruption of surrounding normal tissue [7]. Leksell’s original concept used multiple 300-kV stationary collimated photon beams focused on a common point [8]. Further reﬁnement of the idea led to the use of multiple ﬁxed gamma-emitting cobalt-60 (60Co) sources. This arrangement was called the Gamma Knife and was ﬁrst installed for clinical use at Sophiahemmet Hospital in Sweden in 1968 [9]. Designed as a means to ablate epileptogenic foci and to create functional lesions within deep ﬁber tracts and nuclei, the Gamma Knife was eventually applied to small tumors and arteriovenous malformations. The second Gamma Knife was installed at the Karolinska Hospital in Stockholm in 1974. It was capable of creating more spherical isodose distributions compared with the disk-shaped lesions seen with the ﬁrst unit, but with lesion localization still accomplished via diagnostic X-ray, angiography, and air ventriculography, the capabilities of radiosurgery were limited greatly by the imaging technology of the day.

Notable Contributors Although Leksell is viewed as the pioneer of stereotactic radiosurgery, other notable ﬁgures have contributed greatly to the

25

26

c. heller et al.

FIGURE 3-1. Examples of early stereotactic frames: (a) Horsley-Clarke device, (b) Leksell stereotactic device, (c) Todd-Wells stereotactic frame.

principles of target localization and thus the feasibility of radiosurgery. In 1957, Jean Talairach of the Centre Hospitalier Ste. Anne in Paris introduced his own stereotactic instrument, but more important was his publication of the Atlas d’Anatomie Stereotaxique [10]. Using this highly detailed atlas, he proposed the use of the line connecting the anterior and posterior commissurae as a common reference for localizing neural structures [6]. Among the most notable of these contributors were neurosurgeon Edwin M. Todd and engineer Trent H. Wells, who in 1965 introduced the Todd-Wells Stereotactic unit. Using an arc-based design similar to Leksell’s, it became the ﬁrst widely produced, distributed, and utilized stereotactic device of its era and led to the establishment of stereotaxy in mainstream neurosurgery [4].

In the late 1970s, Wells collaborated with Russell Brown and Theodore Roberts of the University of Utah to develop the Brown-Roberts-Wells (BRW) stereotactic device. The system’s innovative “N-bar” ﬁducial system allowed for precise localization of targets using the newly developed computed tomography (CT) imaging technology [4]. Using a handheld computer, the user was able to convert two-dimensional reference points from CT images into three-dimensional target coordinates. The Cosman-Roberts-Wells (CRW) frame was the followup to the BRW with design improvements including a movable arc, which allowed for an inﬁnite number of non-preselected entry points, and a phantom simulator for preoperative conﬁrmation of target coordinates [11]. Early experience with these devices at the University of Utah and the Los Angeles County/ University of Southern California Hospitals conﬁrmed a great leap forward in target localization utilizing modern imaging techniques and ushered in a new era in stereotactic radiosurgery [12]. The ability to precisely localize a target in three-dimensional space using modern imaging techniques expanded the therapeutic abilities of the Gamma Knife and made it possible for conventional radiotherapy devices such as the linear accelerator to be used for radiosurgery. Subsequent evolution of these techniques and the integration of magnetic resonance imaging (MRI), positron emission tomography (PET), and three-dimensional CT have seen the establishment of four widely utilized modalities of stereotactic radiosurgery: Gamma Knife, conventional and robot-assisted linear accelerator–based systems, and charged particle beams.

Gamma Knife

FIGURE 3-2. The original concept of radiosurgery is credited to Swedish Neurosurgeon Lars Leksell.

There are currently two radiosurgical devices that utilize 60Co sources, but the Leksell Gamma Knife is by far the most widely used (Fig. 3-3). The Leksell Gamma Knife differs from all other forms of radiosurgery in that the radiation energy is delivered to a ﬁxed point and the position of the target is manipulated to create the desired volume of treatment.

3.

techniques of stereotactic radiosurgery

FIGURE 3-3. The Gamma Knife treatment area at the USC University Hospital, Los Angeles, California.

Target Localization Target localization begins with the application of the Leksell head frame, which is ﬁxed to the skull using four percutaneous pins. Conscious sedation and local anesthetic minimize the discomfort of this procedure, which is generally well tolerated. Tumor location must be taken into account when placing the head frame. Interactions between the patient’s head, the frame, and the treatment helmet make it imperative that the target is as close to the center of the frame as possible. A variation of the “N-bar” ﬁducial device is then attached to the frame, and magnetic resonance (MR) and/or CT images are obtained and transferred to the Leksell Gamma Plan workstation where the treatment plan is created. With the ﬁrmly afﬁxed frame as a reference point, each pixel of imaging data is assigned an x, y, and z coordinate value. This allows the dose planning software to know the precise relative locations of the target, critical structures, and the focal point of the treatment beams.

patient is set either manually or by an automatic positioning system (APS) so that the x, y, and z coordinates of the target match up with the focal point of the radiation beams. The patient is repositioned throughout the treatment as needed to accommodate all of the shots from the treatment plan and create the ideal treatment volume and dose. Treatment time depends on the size and number of targeted lesions, the prescribed dose of radiation, and the strength of the cobalt source. With a half-life of 5.26 years, treatment times become prolonged as the cobalt source ages. Once the treatment is concluded, the head frame is removed, and the patient is discharged home to resume normal activities. The rigid ﬁxation of the Gamma Knife provides accuracy in the delivery of radiation and allows for high-energy treatment to a discrete volume with minimal damage to adjacent normal tissue. The need for a head frame makes hypofractionation impractical and, along with the limited amount of space inside the treatment helmet, limits the ability of the Gamma Knife to treat certain lesions at the extreme periphery of the intracranial space and precludes treatment of lesions elsewhere in the body.

Linear Accelerator–Based Systems The linear accelerator (Fig. 3-4) uses microwave technology to accelerate electrons in a part of the accelerator called the “wave guide” and then allows these electrons to collide with a heavy metal target. As a result of the collisions, high-energy photons are scattered from the target. A portion of these X-rays passes through a collimator to form a beam that is directed to the target. The linear accelerator (linac) is the most common instrument used for external beam radiotherapy, but with the integration of stereotactic localization, it can be used for radiosurgery as well [8]. There have been a number of recent advancements to linear accelerator radiosurgery such as frameless localization, which will be discussed later, and multileaf collimation. As the highenergy photon source travels along its treatment arc, the shape

Dose Planning When a target is placed at the focal point of the Gamma Knife, the volume of energy delivered is approximately spherical and termed a “shot.” A treatment plan consists of one or more “shots” positioned in such a way to conform to the often irregular volume of the target. Shots can differ from one another by their x, y, z coordinates, volume, and radiation dose. The volume of a shot depends on the size of the collimator used. The Gamma Knife includes four interchangeable treatment helmets with collimators measuring 4, 8, 14, and 18 mm in diameter. The radiation dose delivered to each volume depends on the length of time the target is left at the focal point of the beams. The plan is designed so that the entire lesion receives a high percentage (e.g., 50%) of the maximum treatment dose, and surrounding structures are relatively spared.

Treatment Once the dose planning is complete, the head frame is ﬁxed to the cast-steel helmet of the Gamma Knife. The position of the

27

FIGURE 3-4. A linac treatment facility.

28

c. heller et al.

FIGURE 3-5. The linac treatment gantry and couch rotate on axes that are perpendicular with respect to each other and intersect at the target point.

of the lesion from the beam’s point of view changes. This primarily affects beam conformity and dose distribution within the target. Multileaf collimators modify the cross-sectional shape of the treatment beam to match that of the lesion. Using a single isocenter, a multileaf collimated photon beam can achieve good beam conformity and homogeneous dose distribution. Accuracy of multileaf collimation, however, is a function of the size of each individual leaf and may diminish when applied to smaller lesions. Photon beams passing through circular collimators can likewise achieve good beam conformity through the use of multiple isocenters. This does, however, result in a more heterogeneous dose distribution throughout the target.

Target Localization Linac-based systems, when used for radiosurgery, have traditionally obtained target coordinates using CT and/or MR imaging with a frame afﬁxed to the head. Recent advances in hardware and software technology, however, have introduced frameless target localization and patient position tracking to linac-based radiosurgery. There are several methods for tracking patient and target position without the use of a head frame. TomoTherapy, which delivers high-energy photons from a device similar in design to a diagnostic CT scanner, has the ability to use “step down energy” for position-tracking purposes. During treatment, lower-energy photons can be used to periodically obtain images similar to conventional diagnostic CT imaging to conﬁrm proper position. Some systems utilize reformatted CT data in comparison with real-time diagnostic X-ray images to allow for patient positioning and tracking. Skeletal landmarks and/or radiofrequency or infrared ﬁducials can be used as reference points precluding the need for a rigid skull frame. Other linac-based systems include a lightweight and robotic linear accelerator (CyberKnife; Accuracy, Inc., Sunnyvale, CA) and modiﬁed linear accelerators to allow image guidance, such as Novalis (BrainLAB, Inc., Westchester, IL), Synergy (Eletka Oncology, Stockholm, Sweden), Trilogy (Varian Medical Systems, Palo Alto, CA), and Artiste (Siemens, Concord, CA).

Dose Planning As opposed to the ﬁxed-beam design of the Gamma Knife, linac-based systems produce a single beam that can be positioned in a number of different ways to create an ideal treatment plan. As discussed below, the treatment gantry rotates in a single plane about a horizontal axis. The use of treatment arcs rather than a stationary beam prevents surrounding tissues from receiving unwanted doses of radiation. As the couch rotates about a perpendicular axis with respect to the gantry, multiple non-coplanar treatment arcs are possible. The main goals with any treatment plan are to maximize tumor coverage with a sharp dose fall-off to limit damage to normal tissues. These goals are inﬂuenced primarily by collimator size, the number of isocenters used, and the number of treatment arcs per isocenter. Whereas a larger collimator with fewer isocenters may shorten treatment time, the opposite approach may be necessary to achieve ideal conformity and dose fall-off.

Treatment Once the treatment plan is ﬁnalized, the patient is placed on the treatment couch and the target location is conﬁrmed. In the case of traditional linac-based radiosurgery, the head frame is secured to either a ﬂoor stand or a couch-mounted holder, though frameless localization is becoming more prominent as previously discussed. The x, y, and z coordinates of the target are centered at a point intersected by the vertical axis around which the couch rotates. The linear accelerator rotates in a single plane around a horizontal axis that intersects the vertical axis of couch rotation at the target site [8] (Fig. 3-5). Treatment is typically done as an outpatient. When frameless localization is used, linac-based systems can be used for hypofractionated radiosurgery as well.

Charged Particle Beam Charged particle beam devices (Fig. 3-6) such as proton beam systems at Harvard Medical School and at Loma Linda University Medical Center in California function on the principle of

3.

techniques of stereotactic radiosurgery

29

Treatment The patient lies on the treatment couch and the head frame or body immobilization device is locked into place in preparation for treatment. There are many strategies as discussed previously that can be used to track and conﬁrm appropriate positioning throughout the treatment. In general, the treatment is an outpatient procedure.

Robot-Assisted Linac Radiosurgery

FIGURE 3-6. The Proton Beam at Loma Linda University Medical Center, Loma Linda, California.

The principles of radiosurgery and frameless stereotaxy were brought together in 1994 when the ﬁrst prototype of the CyberKnife stereotactic radiosurgery system (Fig. 3-7) was installed for clinical use at Stanford University. Developed in the early 1990s by neurosurgeon John Adler Jr., the CyberKnife consists of a compact 6-MV linac guided by a robotic arm with 6 degrees of freedom eliminating physical limitations on beam position [13]. It was designed to overcome the limitations that framed stereotactic localization places on radiosurgery.

Target Localization the Bragg peak [8]. When charged particles such as protons or helium ions are delivered to a lesion, the energy they release is related to their velocity. When they slow down and ﬁnally stop at a depth determined by the initial energy of the beam and type of tissue traversed, they release the majority of their energy in what is called the Bragg peak. The release of energy falls off precipitously thereafter. This allows for theoretically precise delivery of large amounts of energy to a well-deﬁned volume but makes it very important to precisely deﬁne the target location and monitor patient movement to avoid unwanted radiation dosage to normal tissues.

Target localization for CyberKnife treatment is both innovative and traditional combining the latest in computer technology with diagnostic X-ray imaging. Treatment planning consists of a ﬁne-cut, contrast-enhanced CT scan of the target area, which is reformatted by the computer software into multiple digital reconstructed radiographs (DRRs) to resemble diagnostic Xray images from a large number of viewing angles. Once the patient is lying on the treatment couch, a customﬁt immobilization device is placed over the treatment area to limit movement. The CyberKnife includes two ceiling-mounted X-ray imaging devices that, by projecting onto amorphous silicon detector plates rather than conventional X-ray ﬁlm,

Target Localization When employed as a tool for radiosurgery, target localization for charged particle beam treatment can be accomplished using a variety of patient immobilization and tracking strategies. Fiducial markers placed either on or beneath the skin surface can be used as reference points for imaging data as well as for patient tracking purposes during treatment. When intracranial lesions are targeted, a skull frame is used and is secured by either percutaneous pins or a custom-molded mouthpiece.

Dose Planning Charged particle beam systems need not utilize treatment arcs because of their unique properties as discussed above. Although a treatment plan may include more than one beam entry site to achieve desired tumor coverage and dose distribution, the beam is stationary while ﬁring. Beam conformity is achieved through the use of custom-designed apertures that are shaped to match the proﬁle of the lesion from the point of view of each beam used. An additional shielding device known as a compensator is custom-designed to match the depth proﬁle of each lesion. This ensures that the charged particles will release their energy at the desired depth throughout the lesion.

FIGURE 3-7. The CyberKnife treatment facility at the USC/Norris Cancer Hospital.

30

c. heller et al.

provide real-time digital images of the target area. Prior to commencing treatment, the baseline X-ray image is compared with the reformatted DRR images and the best match is determined based on skeletal landmarks and/or ﬁducial markers. Any translational or rotational errors are compensated for by the movable treatment couch during the initial set-up. Real-time X-ray images are obtained during treatment, and any movement by the patient is calculated in terms of deviation from the baseline. This information is directed to the robot arm that is guiding the linac, and appropriate corrections are made. The robot is able to correct for patient movements of up to 1 cm in three translational dimensions and 5 degrees in three rotational dimensions.

Dose Planning CT images are primarily used to outline the intended target when planning for treatment with the CyberKnife. When a lesion is not well visualized on CT, however, the CyberKnife software has the ability to merge MR and PET images with the existing CT images to enhance targeting. The target lesion is carefully outlined on the computer workstation along with all adjacent critical structures. The physicist is able to input a number of variables into the computer such as the desired radiation dose to the target lesion and maximum allowable dosage to critical structures. Using this information, the planning software creates a treatment plan consisting of multiple stationary beam positions to satisfy the desired dosage parameters. If a solution cannot be found to satisfy these parameters, the physicist must modify one or more variables such as tumor coverage, dose to critical structures, or collimator size. The CyberKnife’s robotic arm could theoretically position the beam at an inﬁnite number of points. For ease of calculation, however, there are 101 predetermined stopping points termed nodes. At each node, the beam can assume one of 12 positions resulting in a total of 1212 different possible beam locations at the planning software’s disposal [13].

Treatment The patient is placed supine on the treatment couch and ﬁtted with the custom-designed immobilization device to limit excessive movement. Once the baseline position is achieved, the treatment begins. The unique feature of the CyberKnife is its ability to reposition the treatment beam in real time to compensate for patient movement. Because of its frameless targeting capabilities, the CyberKnife is ideal for hypofractionated treatments.

Conclusion The development of radiosurgery from the early 20th century to the present time has been one of the most signiﬁcant instances of interdisciplinary collaboration in the history of medicine. The free exchange of ideas, technology, and innovation among physicians, physicists, and engineers has revolutionized the treatment of some of the most challenging medical disorders. Target localization has always been and continues to be central to the advancement of radiosurgery, which has become the standard of care in the primary and adjuvant treatment of a wide range of neoplastic diseases. Additionally, as the sophistication and clarity of functional brain mapping begins to approach that of anatomic imaging, radiosurgery is poised to realize its originally intended role as a noninvasive means of treatment for epilepsy, movement disorders, and psychiatric illnesses.

References 1. Horsley V, Clarke RH. The structure and function of the cerebellum examined by a new method. Brain 1908; 31:45–124. 2. Spiegel EA, Wycis HT, Marks M, Lee A. Stereotaxic apparatus for operations on the human brain. Science 1947; 106:349– 350. 3. Apuzzo MLJ, Chen JC. Stereotaxy, navigation and the temporal concatenation. Stereotact Funct Neurosurg 1999; 72:82–88. 4. Chen JC, Apuzzo MLJ. Localizing the point: evolving principles of surgical navigation. Clin Neurosurg 2000; 46:44–69. 5. Spiegel EA, Wycis HT. Stereoencephalotomy, Part 1. New York: Grune & Stratton, 1952. 6. Nashold B. The history of stereotactic neurosurgery. Stereotact Funct Neurosurg 1994; 62:29–40. 7. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316–319. 8. Luxton G, Petrovich Z, Jozsef G, et al. Stereotactic radiosurgery: principles and comparison of treatment methods. Neurosurgery 1993; 32(2):241–258. 9. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry 1983; 46:797–803. 10. Talairach J, David M, Tournoux P, et al. Atlas D’Anatomie stereotaxique. Paris: Masson, 1957. 11. Couldwell WT, Apuzzo MLJ. Initial experience related to the use of the Cosman-Roberts-Wells stereotactic instrument. J Neurosurg 1990; 72:145–148. 12. Heilbrun MP, Roberts T, Apuzzo MLJ, et al. Preliminary experience with Brown-Roberts-Wells (BRW) Computerized Tomography Stereotaxic Guidance System. J Neurosurg 1983; 59:217–222. 13. Kuo J, Yu C, Petrovich Z, Apuzzo MLJ. The CyberKnife Stereotactic Radiosurgery System: description, installation, and an initial evaluation of use and functionality. Neurosurgery 2003; 53(5): 1235–1239.

PA R T II

Radiation Biology and Physics

31

4

The Physics of Stereotactic Radiosurgery Siyong Kim and Jatinder Palta

Introduction In radiosurgery, instead of using a surgical knife when treating a patient, high-energy ionizing radiation is the tool of choice. To understand the tumoricidal effects of ionizing radiation, it is important to know how radiation interacts with matter. This chapter describes general concepts and principles in radiation physics, including basic physics that are applicable to stereotactic radiosurgery. Commonly used delivery systems are also brieﬂy reviewed. This chapter is written for non–physics professionals, especially neurosurgeons and radiation oncologists.

The Radiation Source for Radiosurgery Ionizing radiation is any electromagnetic or particulate radiation capable of producing ions either directly or indirectly in its passage through matter [1, 2]. Among the many radiations we have, the photon, deﬁned as a discrete packet of electromagnetic energy, is the most commonly used for patient treatment [3, 4]. A photon is called by several different names depending on its energy level, such as radio waves, infrared, visible light, ultraviolet, and X-rays in the form of energy as illustrated in Figure 4-1. Only the X-ray range of photons is a form of ionizing radiation, and, though the energy range of X-rays is fairly wide, only a high-energy range is used for treatment. Photon treatment beams can be obtained either from radioisotope sources or from X-ray–generating machines. Currently, cobalt-60 (60Co) is the most widely used radioisotope [5–9] for radiosurgery, and the medical linear accelerator is the most popular X-ray– generating machine [10–14]. Protons, deﬁned as positively charged particles found in the nucleus of an atom, are also used for treatment [15–19]. Proton beams can be obtained from particle accelerators such as cyclotrons and synchrotrons [20], which cost a tremendous amount of money; thus, very few proton facilities exist worldwide, and there are currently only ﬁve in the United States. Because the majority of radiosurgery treatments are performed with photon beams, emphasis is placed on the photon beam.

Cobalt-60 Cobalt-60 is the radioactive isotope used as the radiation source in the Leksell Gamma Knife treatment machine (Elekta, Nor-

cross, GA), which is described later. A radioactive isotope is an atom with an unstable nucleus that tries to stabilize itself through a process called radioactive decay by emitting ionizing radiation such as alpha, beta, and/or gamma particles. When 60 Co undergoes radioactive decay, it emits beta particles and two strong gamma radiations, one with 1.17 MeV of energy and the other with 1.33 MeV of energy (Fig. 4-2). MeV, a commonly used unit of energy for particles in the treatment energy range, means 1 million electron-volts, where 1 electron-volt is the energy gained by an electron that accelerates through a potential difference of 1 volt. A real treatment beam of Gamma Knife can include photons of energy different from 1.17 and 1.33 MeV because some photons experience interactions with 60Co itself and/or its capsulation material and lose a part of that energy. Therefore, the effective energy of Gamma Knife is slightly lower than 1.25 MeV (the average of 1.17 and 1.33 MeV). Cobalt-60 ultimately decays to nonradioactive nickel. The half-life of 60Co, that is, the length of time needed for 60Co to lose half of its radioactivity from decay, is 5.26 years. At the end of 1 half-life, only 50% of the original radioactive material remains. This means that treatment time after 5 years is double that at the time of initial installation, and that the source of Gamma Knife needs to be replaced after a certain period of time to avoid long treatment times.

X-Ray Production in a Linear Accelerator One of the interesting characteristics of a fast-moving electron is that when it interacts with a material, it can produce an X-ray. A part or whole of the kinetic energy of the electron is transformed into electromagnetic energy. This principle is used to obtain X-rays in most X-ray production machines in which electrons are accelerated and induced to hit a target and then generate an X-ray. The amount of X-ray particles produced increases as the kinetic energy of the incident electrons increases. Electrons are accelerated by potential differences in diagnostic X-ray tubes up to several hundred kilo-electron-volts (keV). The acceleration mechanism in a linear accelerator, however, is somewhat different because it uses microwave technology similar to that used for radar to accelerate electrons in a linear tube, often called an accelerator tube [21]. In linear accelerators, electrons are usually accelerated to the energy range of 4 to 25 MeV. A 6-MeV electron is most often used to create an X-ray for radiosurgery. When an electron creates a

33

34

s. kim and j. palta

FIGURE 4-1. The electromagnetic spectrum is illustrated. Therapeutic X-ray or ionizing radiation is in the energy range of 105 eV and higher. The energy of photon beams commonly used in radiosurgery are higher than 106 eV.

photon, theoretically, the photon can take any energy from zero to the same as the energy of the incident electron. In other words, the photon beam produced by a 6-MeV electron beam can have an energy spectrum of 0 to 6 MeV. The energy of an X-ray produced in a linear accelerator is denoted MV instead of MeV. The major components and auxiliary systems of typical medical linear accelerators are shown in a block diagram (Fig. 4-3). The role of each part is brieﬂy explained below.

Power Supply The power supply provides direct current (DC) power to the modulator.

Electron Gun The electron gun produces electrons by thermionic emission. In a conducting material, the electrons all exist at or below the baseline electron energy at low temperatures. As the temperature is increased, some of these electrons have sufﬁcient energy to pass over the surface-potential barrier between the material and the vacuum. This process of increasing the temperature of a bulk material to increase the number of electrons that can leave the material is called thermionic emission. These electrons are pulse-injected into the accelerator tube according to the signal from the modulator.

Radiofrequency Source Modulator The modulator generates high-voltage pulses lasting a few microseconds that are simultaneously introduced to the radiofrequency source and the electron gun.

Radiofrequency (RF) is a range of electromagnetic frequencies above sound and below visible light, generally in the 30-kHz to 300-GHz range. RF is used for all broadcast transmissions including AM and FM radio, television, shortwave, microwave,

Bending Magnet 60 27 Co

5.26 y Electron Gun

Accelerator Tube/ Wave-guide System

Modulator

RF Generator (Magnetron)/ Amplifier (Klystron)

b 1– (99.8%) 0.313 MeV

g 1 (99.8%) 1.173 MeV

Power Supply

g 2 (100%) 1.332 MeV

60 28 Ni 60

FIGURE 4-2. For Co, most of the disintegrations (99.8%) are the emission of a b 1− with a maximum energy of 0.313 MeV. This leads to an excitation of 60Ni, which releases its energy quickly by emitting two gamma rays (1.173 MeV and 1.332 MeV) in cascade. In some disintegration (0.12%), a b 2− particle with a maximum energy of 1.486 MeV is released and leads to the lowest excited state of 60Ni.

RF Generator for Klystron

Target Treatment Head

b 2– (0.12%) 1.489 MeV

FIGURE 4-3. The power supply provides DC to the modulator, which provides a high-voltage pulse to the RF generator and electron gun simultaneously. The electron gun introduces the electron in pulse to the accelerator tube. Electron production occurs through the thermionic process. The RF generator introduces a microwave to the accelerator tube. The electron and the microwave are timed exactly to be met in the accelerator tube by the modulator. The accelerator tube accelerates the electrons via the resonance cavity. As the accelerated electron comes out of the accelerator tube, the path of the electron is bent through a bending magnet to hit the target and generates an X-ray.

4.

the physics of stereotactic radiosurgery

35

and satellite transmissions. In a linear accelerator used for radiosurgery, a magnetron is often the RF source. The magnetron is a diode-type electron tube that functions as a high-power oscillator to generate microwave pulses. Pulse duration is several microseconds long, and the repetition rate is several hundred pulses per second. Following the pulses from the modulator, microwave pulses are introduced into the accelerator tube through the wave-guide system. The typical frequency of microwaves used in a linear accelerator is about 3 GHz. In some designs, a klystron is used instead of a magnetron. The klystron is a microwave ampliﬁer rather than a source, therefore, it requires a low-power microwave oscillator called an RF driver as an RF source. In general, a klystron produces microwaves of higher power than the magnetron and is often used with high-energy linear accelerators (>10 MV).

Wave-Guide System The wave-guide system is made up of a pressurized rectangular tube through which microwave pulses generated by either a magnetron or klystron are transmitted into the accelerator tube.

Accelerator Tube An accelerator tube consists of many RF resonant cavities. An RF cavity allows power to be coupled into the particle beam. In an electron linear accelerator, RF accelerating cavities are basically microwave resonators. Resonance is a very common physical phenomenon that can be observed in many types of systems. Imagine a child on a swing, for example. If the child pushes the swing at just the right time, the push can make the swing go higher and higher with very little effort. The swing has a natural frequency of oscillation. Pushing “at just the right time” means that energy is put into the system at a frequency known as the fundamental frequency. Now the system is said to be in resonance condition. The amplitude of the motion can increase rapidly with resonance. An analogous phenomenon can occur in a linear accelerator with RF resonance, which is generally derived from an RF cavity that is typically in a right circular cylindrical shape with connecting holes to allow a charged particle’s beam to pass through for acceleration. Figure 4-4 shows a typical cylindrically symmetric RF cavity. In a fundamental RF mode, the electric ﬁeld is roughly parallel to the beam axis and radically decays to

FIGURE 4-4. A typical cylindrically symmetric cavity.

FIGURE 4-5. As the electron travels into a uniform magnetic ﬁeld, it experiences a downward force. This force causes the electron to travel in a circular path. The crosses in the ﬁgure indicate the uniform magnetic ﬁeld pointing into the page.

zero upon approach to the cavity walls. The pulsed electron particle beam traverses the cavity through the centered hole, creating an accelerating force along the axis of the cavity due to the electric ﬁeld.

Bending Magnet When an electron moves in a magnetic ﬁeld, it experiences force. The direction of the force is at a right angle to both the direction of the magnetic ﬁeld and the direction of the electron motion (Fig. 4-5). Using this principle, the path of accelerated electrons is controlled so that they can hit the target normally (note the electron path, bending magnet, and target in Fig. 4-3). Magnets used for this purpose are called bending magnets. In certain cases, straight-ahead beam designs can be applied without the bending magnet if the acceleration structure is short enough to be placed in a vertical direction. A linear accelerator in straight-ahead beam design is used in the CyberKnife unit, which will be better described later.

Treatment Head The major parts of the treatment head are the target (often called the X-ray target), primary collimator, ﬂattening ﬁlter, ion chamber, and secondary collimators (Fig. 4-6). Accelerated electrons collide with the target to generate an X-ray beam. For a given energy of electrons, more X-rays can be produced if a target material with a higher atomic number (Z number) is used. Tungsten and gold are commonly used as target materials. The production of an X-ray is based on a process called bremsstrahlung (meaning “braking radiation”), which is the result of radiative collisions between a fast-moving electron and a nucleus. When the electron passes near a nucleus, it can be deﬂected from its path by Coulomb forces of attraction and lose its energy as bremsstrahlung radiation (i.e., an X-ray). The Xrays are produced in all directions, but high-energy X-rays are produced in the forward direction (Fig. 4-7). Therefore, in a linear accelerator, higher photon intensity is observed in the area close to the central line (i.e., the extended line following the direction of the incident electron) below the X-ray target.

36

s. kim and j. palta

The amount of radiation produced by the linear accelerator is monitored by an ionization chamber, often called a monitor chamber. The monitor chamber not only controls the amount of radiation delivered to the patient, but it also checks and controls the uniformity of the delivered radiation. Whereas the primary collimator deﬁnes the maximum aperture of the radiation ﬁeld, a real treatment ﬁeld is determined by the secondary collimator, which deﬁnes a rectangular ﬁeld that ﬁts to each patient treatment in routine radiation therapy. In radiosurgery, smaller ﬁelds are needed. Thus, an additional collimator system, either a circular collimator or micro-multileaf collimator [22], is used.

A Photon’s Interaction with Matter FIGURE 4-6. A schematic of the treatment head of a medical linear accelerator is illustrated. Initially, as the accelerated electrons hit an X-ray target, an X-ray beam is generated via the bremsstrahlung process. This X-ray beam is collimated by the primary collimator. As it exits the primary collimator, the X-ray beam proﬁle is ﬂattened out by the ﬂattening ﬁlter. Then, it goes through the ion chamber, and, ﬁnally, it gets collimated by the secondary collimator. In radiosurgery, generally one more collimator is added to make a very small ﬁeld (e.g., circular collimator or micro-multileaf collimator).

The X-ray beam is ﬁrst collimated by the primary collimator, which has a circular aperture through which a photon beam can pass to make a radiation ﬁeld. The primary collimator is typically made of tungsten. For patient treatment, a radiation beam with uniform intensity across the ﬁeld is desirable, and to create a uniform photon beam, a conical ﬁlter called a ﬂattening ﬁlter is used. The central area of the ﬂattening ﬁlter is thicker to attenuate more photons in the central area. In general, a high-Z-number material is used for the ﬂattening ﬁlter to effectively attenuate photons.

FIGURE 4-7. Schematic illustration of the angular distribution of emission of bremsstrahlung X-rays around a target. At the kinetic energy of an electron beam less than 100 keV, X-rays are emitted relatively equally in all directions. As the kinetic energy of the electron beam increases, the direction of the bremsstrahlung X-ray emission becomes signiﬁcantly forward. In megavoltage X-ray machines, the transmission X-rays are therefore used to treat a patient.

The effectiveness of radiation treatment is based on how much radiation energy is deposited in the tumor compared with the energy deposited in the surrounding normal tissues. Energy deposition occurs through interactions between a radiation beam and the human body at either the atomic or nuclear level. When photons traverse a material, they reduce the intensity of the incident photons. The three most common ways for a photon to interact with matter for energies above a few keV are photoelectric absorption, Compton scattering, and pair production. These three mechanisms account for more than 99% of the interactions between photons and matter, and the probability of each depends on the energy of the photon and the material with which it interacts. Because these interactions are quantum in nature, a speciﬁc interaction is never guaranteed. It is simply more or less probable than the others (depending on the energy and the material).

Photoelectric Absorption In photoelectric absorption, an X-ray photon is absorbed by the interacting material and an electron from a shell of the atom is ejected, leaving the atom ionized, as illustrated in Figure 4-8. The ionized atom returns to the neutral state with the emission

FIGURE 4-8. In a photoelectric effect, an incident photon striking a bounded electron is absorbed, and the electron is ejected from the atom with a kinetic energy equal to the difference between the incident photon energy and the binding energy of the electron. The probability of the photoelectric effect increases as the atomic number of the target material increases; however, it decreases as the energy of the incident photon increases. (E is the energy of the incident photon, T is the kinetic energy of the ejected electron, and BE is the binding energy. Photon energy is expressed as hν, where h is Plank’s constant and ν is frequency.)

4.

the physics of stereotactic radiosurgery

of an X-ray, called the characteristic X-ray of the atom. The kinetic energy transferred to the ejected electron during the interaction is the same as the incident photon energy minus the ejected electron’s binding energy to the atomic nucleus. Photoelectric absorption occurs predominately at low energies of photons and for high-atomic-number materials. In water (one of the closest materials to tissue), the probability of this interaction becomes negligible once photon energy reaches above 100 keV. Because the effective energies of photon beams from the Gamma Knife (i.e., 60Co source) and linear accelerator are of the order MeV, the effect of photoelectric absorption is not signiﬁcant in radiosurgery.

Compton Scattering In Compton scattering, also known as an incoherent scattering, the incident X-ray photon ejects an electron from an atom, loses some of its energy to the ejected electron, and continues to move in a direction different from the initial direction. The resulting incident photon is called a scattered photon, whereas it is called a primary photon before the interaction. The energy and momentum are conserved in this process. Kinetic energy transferred to the electron is the energy difference between the primary and scattered photons. The energy of the scattered photon depends on the direction (i.e., the angle with respect to the direction of the primary photon). The scattered photon has minimal energy when it is backscattered (180° from its direction of travel). If it is scattered in the same direction as the primary, the scattered photon energy is the same as the primary photon energy. Compton scattering is important for low-atomic-number materials and photon energies of 100 keV to 10 MeV. Therefore, Compton scattering is the most signiﬁcant interaction for photons used in radiosurgery. Figure 4-9 illustrates the Compton scattering interaction.

37

FIGURE 4-10. When a photon with energy equal to or higher than 1.02 MeV comes very close to the nucleus of an atom, it interacts with the nuclear Coulombic force ﬁeld and can create a pair of an electron and a positron. The total kinetic energy of the electron and position is the energy of the incident photon minus 1.02 MeV. The probability of pair production is proportional to the atomic number of the target material and the incident photon energy.

Pair Production Photons have a quantum of energy but zero mass. In pair production, the incident photon energy is completely absorbed in the medium and an electron and its antiparticle, a positive electron called a positron, are created. This interaction is illustrated in Figure 4-10. Pair production can only occur when the energy of the incident photon is greater than 1.02 MeV, which is the minimum energy needed to create an electron and a positron. Positrons are very short lived and disappear with the creation of two 0.51-MeV photons each, a process called positron annihilation. The total kinetic energy transferred to the electron and positron pair is equal to the incident photon energy minus 1.02 MeV. Pair production is of particular importance when high-energy photons pass through materials of a high atomic number but plays a small role in photon beams used for radiosurgery.

Dose and Dosimetry

FIGURE 4-9. In Compton scattering, the incident photon strikes an electron and ejects it out of the atom to which it was bound. During this process, the incident photon energy is transferred to the ejected electron. The least energy is transferred to the electron when the incident photon has no scattering (i.e., θ = 0°), and the most energy is transferred to the electron when the incident photon is scattered backward (i.e., θ = 180°). Compton scattering becomes most signiﬁcant at an energy range of 100 keV to 10 MeV, and it is almost independent of the atomic number (i.e., Z value) of the interacting material. Thus, for the radiosurgery energy range, Compton scattering is the dominating interaction. (E is the energy of the incident photon, T is the kinetic energy of the ejected electron, and E′ is the energy of the scattered photon.)

The physical quantity that is used in radiation therapy to kill neoplastic cells is the kinetic energy of the incident radiation beam. Once photons enter a medium, they start to interact with the medium with the probability of interactions described earlier. Any photon particle loses a part or all of its energy when it initially interacts with a medium and most of the lost energy is absorbed by the medium around the interaction point. Energy absorbed in a unit mass of the interaction medium during interactions with a radiation beam is called the dose or, more speciﬁcally, the radiation dose. The commonly used unit of dose is the gray (Gy), which is equivalent to J/kg (joules per kilogram), where joule is a unit of energy (1 J is the approximate energy required to lift 1 kg by 0.1 m assuming gravity accelerates at 10 m/s2). Depending on the amount of dose, cGy (centigray, or one-hundredth of a Gy) is also often used. Dosimetry is either the measuring or calculating of dose. As you may have noted, the energy absorption mechanism during photon interaction is a two-step process: (1) energy transfers from the photon to the ejected electron, and (2) energy is deposited from the ejected electron to the medium (Fig. 411). Thus, energy deposition occurs within a volume around the

38

s. kim and j. palta

There are two commonly used dose-calculation methods: measurement-based and model-based. Measurement-based approaches directly use dose distributions measured in phantoms. On the other hand, model-based approaches try to calculate the dose from the ﬁrst principles with limited use of measured data.

Measurement-Based Dose Calculation Dose calculation is mainly based on three parameters: percent depth dose, proﬁle, and output.

Percent Depth Dose FIGURE 4-11. A photon enters a volume of a medium and transfers a portion of its energy, thus giving the electron a kinetic energy. The electron will travel a short distance and deposit its energy in a medium. The original photon will continue to travel within the medium as it transfers more energy to another electron.

spot of interaction rather than at a point of interaction. If the photon still has kinetic energy after the ﬁrst interaction (e.g., Compton scattering), it can travel and interact at another point, then cause another energy deposition. This second interaction is independent of the ﬁrst interaction, thus, third, fourth, and multiple scatterings are possible. The energy absorbed from scattered photons is called the scattered dose, contrary to the primary dose, which is the energy absorbed in the ﬁrst interaction. The same thing happens when a patient is irradiated with a photon beam. Photons interact with human tissue and deposit energy into the tissue. Radiation therapy’s aim is to kill the target tissue (or cell) through enough impact from the energy deposition while minimizing the impact to the normal tissue. In other words, local control of treatment depends on how much dose is delivered to the target (and how accurately, too, of course), and normal tissue complication is dependent on how much unwanted dose is given to the normal structure.

When a photon beam is incident on a medium, it creates a dose distribution within the medium. Relative dose at any depth is called the depth dose (DD) and when it is expressed as a percentile it is called the percent depth dose (PDD). Let us assume a certain number of photons in a hypothetical narrow beam are incident on a thin medium (Fig. 4-12a). The number of photons decreases exponentially because some photons interact with the medium and disappear from the path of the narrow beam. Now think of energy deposition (i.e., dose) at depth. The absorbed dose is proportional to the amount of radiation interactions. Energy, however, is deposited by ejected electrons rather than by photons themselves. Because most of the ejected electrons lose energy continuously while traveling through the medium, there is a distance lag between the photon interactions and the dose. This mechanism also creates a so-called build-up region in the shallow depth of the medium. Although the number of interactions is the highest at the surface of the medium, dose is almost zero (because most ejected electrons move deeper into the medium). Thus, dose is close to zero at the surface, starts to build up with depth, and reaches the maximum. Then, the dose begins to decrease exponentially. Figure 4-13 shows a typical shape of DD (or PDD) for a narrow beam. In reality, the target volume is ﬁnite in size (Fig. 4-12b). In this scenario,

FIGURE 4-12. (a) In the thin line of beam and thin medium, the scattered photons are out of the medium and never return back to the medium. There is thus no contribution by the scattered photons to the dose. Percent depth dose (PDD) is mainly determined by the energy deposition of the primary photons. (b) On the other hand, in the thick medium with the thin line of photon beam, the scattered photons can further interact with the medium and return to the center line of PDD measurement. In general, however, this effect is insigniﬁcant. (c) In the

broad beam and thick medium, the single or multiple scattering of photons can occur at any line of beam. The broad beam is considered as a bundle of thin lines of beam. The scattered photons coming from each beam line to the center line of interest contribute to the total dose. PDD values beyond the dmax (depth of dose maximum) increase with ﬁeld size because there are more line beams for larger ﬁelds. Because most radiosurgery ﬁelds are small, radiosurgery beams are close to case (b).

4.

the physics of stereotactic radiosurgery

39

DD (log scale)

Max

FIGURE 4-15. An ideal dose proﬁle in a plane is illustrated.

Incident Photon Depth (cm) Build-up Region FIGURE 4-13. A typical percent depth dose with a narrow beam is illustrated. The dose is close to zero at the surface and starts to build up with depth until reaching the maximum. This region is called the build-up region. After the depth of dose maximum, it then begins to decrease exponentially. Note that depth dose (DD) is displayed in the “log” scale whereas depth is in the “linear” scale (i.e., log-linear plot). In a log-linear plot, an exponential decrease is expressed as a line.

some of the scattered photons can scatter back to the center of the ﬁeld and contribute to the dose at depth. This effect is minimal, however. The scatter effect is signiﬁcantly enhanced with a broad beam (Fig. 4-12c) because more scattered photons can contribute to the dose. The chance of multiple scattered photons coming to the central line increases with ﬁeld size because more scattered photons exist in a larger ﬁeld, so PDD increases with ﬁeld size (Fig. 4-14). In radiosurgery, however, most ﬁelds are relatively small, so the scatter effect is insigniﬁcant. Thus, most radiosurgery beams are similar to the case shown in Figure 412b, and PDD can be analytically expressed as a simple exponential function.

Proﬁle or Off-Axis Ratio Figure 4-15 shows an ideal dose proﬁle, often called the off-axis ratio (OAR), in a plane where a certain amount of dose is given

100

PDD(%)

5×5 cm2 10×10 cm2 20×20 cm2

uniformly inside the ﬁeld and no dose is deposited outside the ﬁeld. In reality, it is not possible to obtain such a proﬁle. A typical dose proﬁle in a plane at a certain depth for a clinical radiation beam is shown in Figure 4-16a. The dose is uniform in the central area but starts to gradually decrease as it nears the ﬁeld edge. It then drops rapidly at the geometric edge of the ﬁeld and slowly approaches zero dose (<1%) away from the ﬁeld edge. The region of rapid dose fall-off is called the penumbra. When a ﬁeld is large, the ﬂat plateau in the central area is clearly observed. As the ﬁeld size becomes smaller, the plateau disappears as illustrated in Figure 4-16b, which is a proﬁle for a typical radiosurgery beam. In radiosurgery, most radiation ﬁelds (<50 mm in diameter) show either no plateau region or a very narrow plateau region. Because of the large dose variation within a small region, a high-resolution measurement detector is required for dose proﬁle measurement [23–29]. The shape of the individual radiation ﬁeld is deﬁned by the collimator system that is speciﬁcally based on the radiosurgery delivery system.

Output The magnitude of radiation produced by the delivery system is called the output. In general, this is a relative value that is normalized to the output of a reference ﬁeld size. Output varies with ﬁeld size (or collimator size). Larger ﬁeld sizes have higher output. With the PDD, OAR, and output for a given geometry, relative dose at any point is simply a multiplication of these three parameters. Absolute dose is obtained when this value is multiplied with the machine calibration factor and monitor units (MU) in the linear accelerator (or calibration dose rate and treatment time in Gamma Knife).

Model-Based Dose Calculation Most model-based dose-calculation methods deﬁne a kernel, which is a function that describes how the dose is distributed within a medium (mostly water). Two commonly used kernels are point kernel and pencil-beam kernel.

50

RELATIVE DEPTH FIGURE 4-14. Comparison of PDD curves for various ﬁeld sizes. As the depth gets deeper, PDD increases with ﬁeld size because of the increased amount of scatter dose contribution. Note that it is intentionally shown in “linear-linear” scale to illustrate that PDD is often drawn in two ways, “linear-linear” and “log-linear” (see Fig. 4-13).

FIGURE 4-16. (a) A typical dose proﬁle is illustrated for a conventional beam. The region of rapid dose change is called the penumbra. (b) A typical dose proﬁle for a radiosurgery beam.

40

s. kim and j. palta

FIGURE 4-17. A point kernel is illustrated, which is a distribution of energy deposited at any point to total energy released at the primary interaction point.

Point Kernel When a photon particle interacts at a point within a medium, a certain amount of energy is released per mass, which is called the total energy released per unit mass (TERMA). A part of this energy is absorbed at the point of interaction (often called the primary interaction point), but the rest of it is dispersed around the point. The scattered photon can also travel to any other points, interact, and deposit energy around the area of interaction. Therefore, a photon interaction can contribute dose to the whole volume. The point kernel describes how dose is distributed with respect to the interaction point within the medium, as illustrated in Figure 4-17. In general, the distributed dose is expressed as a ratio between the energy deposited at any point and the total energy released during the interaction at the primary interaction point. The dose at any point is a summation of kernels multiplied with the TERMA from all of the primary photon interactions. TERMA is easily calculated using analytical expression of interaction mechanisms.

Pencil-Beam Kernel The pencil-beam kernel includes dose contributions from all of the primary interactions along a ray instead of at one interac-

FIGURE 4-18. A pencil-beam kernel is illustrated, which is a distribution of deposited energy from all primary interactions along a pencilbeam line.

tion point. Figure 4-18 illustrates a pencil-beam kernel. If photon energy ﬂuence (the same as the number of photons times the corresponding energy per area) is known, dose at any point is simply a summation of kernels multiplied with energy ﬂuence. In general, kernel values are preobtained by calculations, and sometimes the pencil-beam kernel can be extracted from measurement data.

Isodose Line and Dose-Volume Histogram Once the dose is calculated in the treatment planning system, each dose distribution is usually displayed in an isodose line, which is made up of all the points of the same dose connected in a line. In radiosurgery, isodose lines are expressed as a percentile of the maximum dose. Dose distribution within the target is much more nonuniform in radiosurgery than in conventional radiotherapy. The maximum dose is usually observed in the center of the target, and it is usual to have very stiff dose fall-offs near the target boundary. Prescription is assigned to isodose lines of 50% to 80% depending on the situation. A prescription isodose line is higher in the case of a single isocenter and lower with multiple isocenters. Figure 4-19 shows

FIGURE 4-19. An isodose line connects all points of the same dose. In this ﬁgure, absolute isodose lines are displayed in axial, sagittal, and coronal planes of the CT image.

4.

41

the physics of stereotactic radiosurgery

Dose Volume Histogram 1.0 0.9 0.8

Norm.Volume

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

500

1000

1500

2000

2500

3000

Dose (cGy) FIGURE 4-20. In a dose-volume histogram (DVH), accumulated volume (i.e., histogram) is described according to dose for any particular volume of interest (e.g., the target). In this case, 95% of the volume of the target will get a dose equal to or more than 2000 cGy.

isodose lines displayed in absolute value on three (axial, sagittal, and coronal) planes of tomographic image. Dose-volume histogram (DVH) describes accumulated volume (i.e., histogram) according to dose for any particular volume of interest (e.g., the target) as illustrated in Figure 4-20. In Figure 4-20, 95% of the volume of the target will get a dose equal to or more than 2000 cGy. Although the DVH does not show spatial information, it is a useful tool for plan evaluation [30, 31]. In radiosurgery, the DVH for the target usually shows a gentle slope on the high-dose side because of a signiﬁcantly nonuniform dose proﬁle of small ﬁelds. With the lower isodose line selected for prescription, a gentler DVH slope on the highdose side is expected. In the case of multiple isocenters, ﬁeld overlap causes a signiﬁcant dose increase in the central volume of the target, resulting in a high dose tail in DVH.

a set of four collimator helmets, and a control unit. Figure 4-21 and Figure 4-22 show a photograph of a Gamma Knife unit and a simple schematic diagram, respectively. The 60Co radioactive source is in pellet form, and 20 pellets are encapsulated in a steel capsule. Each steel capsule is 1 mm in diameter and 20 mm in height. Sources are aligned with three collimator systems, the precollimator, the stationary collimator, and the ﬁnal collimator of a helmet. Both the precollimator and the ﬁnal collimator of the helmet are made of tungsten alloy whereas the stationary collimator is made of lead. The total

Photon Beam Delivery Systems Gamma Knife In the late 1960s, Leksell introduced a prototype of Gamma Knife, which is a dedicated radiosurgery unit [5]. Leksell’s prototype incorporated 179 60Co radioactive sources placed over a 60° × 160° sphere. The modern Gamma Knife models incorporate 201 60Co radioactive sources that converge and focus on the treatment target at a source-to-target distance of about 40 cm [6, 7]. Each source contributes a clinically insigniﬁcant dose following the beam line. However, when many beam lines are converged on the focus where the target is, the therapeutic dose is delivered to the target while the surrounding normal tissue dose remains under the limit. Gamma Knife mainly consists of the radiation unit, the operating table and sliding couch,

FIGURE 4-21. A photograph of a Gamma Knife unit.

42

s. kim and j. palta

Rear shield

Middle shield Front shield

Central body

Upper shielding door

Plug

Treatment table Helmet

Couch

Position indicator Trolley .

.

Lower shielding door

FIGURE 4-23. A photograph of a collimator helmet. A properly chosen helmet, depending on the target size, is connected to the sliding couch.

FIGURE 4-22. A simple schematic diagram of a Gamma Knife unit.

The current state-of-the-art Gamma Knife units are equipped with several auxiliary capabilities. They provide software that tracks the patient’s position 10 times per second. A robotic automatic positioning system repositions each isocenter with submillimeter precision without operator intervention. Radiosurgery using Gamma Knife has been performed for more than 250,000 patients worldwide during the past 30 years. Many clinical studies with longer than 15 years of follow-up have been reported.

activity of the Gamma Knife unit with new 201 sources loaded is of the order 6000 Ci (2.22 × 1014 Bq, where Bq is the same as disintegration per second, and Ci is 3.7 × 1010 Bq). The activity of each source is about 30 Ci and its variation is expected to be within ±5%. With 6000 Ci, a typical dose rate is about 300 cGy/ min at the center of the 16-cm-diameter, spherical water phantom. As described earlier, this value drops down to 150 cGy/min (50% of initial dose rate) 1 half-life (5.26 years) later. Four helmets of different collimator sizes, 4, 8, 14, and 18 mm in diameter, are provided. Figure 4-23 is a photograph of a collimator helmet. A proper helmet is chosen depending on the target size and is connected to the sliding couch. The Leksell stereotactic frame (Fig. 4-24) is ﬁxed to the patient’s skull and connected to the couch. Once the target position is aligned to the focus, the sliding couch is pushed into the mechanically set treatment position. Its mechanical spatial accuracy is claimed to be in submillimeters (about 0.3 mm). In treatment mode, the shielding doors are open, resulting in an increase of scatter dose inside the treatment room. When the sliding couch is withdrawn and the shielding doors are still open, the scatter dose becomes about 1.3 times higher. Because of the unit’s design, however, no primary beam directly escapes the unit even with the doors open. When the doors are closed, exposure rates by leakage radiation are signiﬁcantly reduced.

FIGURE 4-24. Picture of a Leksell stereotactic frame. The frame is ﬁxed to the patient’s skull and is connected to the couch.

.

..

..

.

4.

the physics of stereotactic radiosurgery

43

Isocentric Linear Accelerator Most clinical linear accelerators are isocentric, which means that the gantry, the collimator, and the table can be rotated with respect to one point in the space called the isocenter. In reality, the isocenter may not be a point because three rotation axes (gantry, collimator, and table) often do not intersect at a common point. Instead, they only near each other within a sphere. The center of this sphere is considered the nominal isocenter. The size of the sphere is the key factor in the linear accelerator’s spatial accuracy, and the smaller the better. Common consensus says that the sphere needs to be less than 1 mm in diameter for radiosurgery based on the fact that the resolution of modern imaging modalities is around 1 mm. Another important factor for spatial accuracy is how precisely positioned the patient is. In a linear accelerator, the patient’s positioning is usually determined using the treatment room laser system. Three lasers, one from the ceiling and two from side-walls, are aligned to point to the nominal isocenter. By matching the laser crosshairs and the desired marks on the frame attached to the patient’s skull, the target is aligned to the nominal isocenter. Therefore, laser system accuracy is signiﬁcantly important when used for patient positioning. As described earlier, linear accelerators are equipped with a so-called secondary collimator system that provides a rectangular ﬁeld for conventional radiotherapy, but rectangular ﬁelds are not considered appropriate for radiosurgery. Therefore, additional circular collimators are commonly used as tertiary collimators [10–14]. Circular beams are superior to rectangular beams for reasons like the sharper beam, easier dosimetry calculation, more precise beam delivery, and better ﬁeld deﬁnition for small ﬁelds. Because tertiary collimation happens closer to the patient, the beam is more accurately aligned and the penumbra is reduced. More rapid dose fall-off outside the target can therefore be obtained. Collimator sizes typically range from 5 to 40 mm to treat a variety of lesion sizes and shapes. Most circular collimators are made of tungsten alloy or lead and are 5 to 10 cm thick. Figure 4-25 shows several circular collimators with different diameters used at the University of Florida.

FIGURE 4-25. Several circular collimators in different sizes of diameter (5, 10, 12, 14,20, 24, and 30 mm) used at the University of Florida are shown.

FIGURE 4-26. A setup of the ﬂoor stand used at the University of Florida is shown. The ﬂoor stand is mechanically ﬁxed to the holes in the ﬂoor.

At the University of Florida, developed was a special device called a ﬂoor stand [32, 33] that improves the accuracy of the beam’s focus to the isocenter by adding a set of bearings to the stereotactic collimator system that accounts for imperfections in the gantry rotation. A set of bearings is also attached to the patient and the target area to bypass the inaccuracies of the table rotation. It is reported that the ﬂoor stand can provide mechanical spatial accuracy within 0.2 ± 0.1 mm to deﬁne the isocenter [32, 33]. Figure 4-26 shows how the ﬂoor stand is set up. With the recent introduction of a micro-multileaf collimator (mMLC), which has many collimator leaves that can be individually moved to conform to the shape of the target, it is possible to do two-dimensional conformal radiosurgery for irregularly shaped targets with a single isocenter (Fig. 4-27). In addition, dynamic conformal radiosurgery can be performed with the aid of the automatic ﬁeld shaping capability while the gantry rotates. It is also possible to do intensity modulated stereotactic radiosurgery (IMSR) if the treatment planning system supports the appropriate dose optimization [34]. The collimator leaf is usually made of tungsten. The width of each leaf is an important parameter related to the conformality and it ranges from 1.5 to 6 mm at the isocenter depending on the manufacturer. State-of-the-art linear accelerators incorporate a volumetric imaging system referred to as a cone beam computed tomography (CT) system. A diagnostic X-ray tube is added on a separate gantry at a 90° angle from the treatment gantry, and a ﬂat panel detector is installed on the opposite side as shown in

44

s. kim and j. palta

FIGURE 4-27. A photograph of a commercial mMLC. As can be seen, there are many collimator leaves, and each of them can be individually moved to conform to the shape of the target.

FIGURE 4-29. A CyberKnife unit uses a miniature linear accelerator mounted on a robotic arm.

Figure 4-28. Volumetric imaging can be obtained in a couple of minutes by recording multiple cone beam geometry planar images while the gantry rotates. Currently, it is not clear whether this capability needs to be used for radiosurgery. This advancement opened the door to exploring the possibilities with imageguided radiosurgery using isocentric linear accelerators.

beam. It is much smaller and lighter than a conventional S-band (2856 MHz) linear accelerator. CyberKnife incorporates an image-guided system consisting of two diagnostic Xray tubes mounted on the ceiling and two detectors placed next to the lateral sides of the table. With this X-ray imageguided system, CyberKnife continuously and automatically tracks a tumor’s precise location throughout the procedure, detects any tumor or patient movement, and makes corrections as needed. Because of its image-guidance capability, the radiosurgery procedure is performed without an invasive head frame. Even without the frame, the use of intelligent robotics technology enables treatments with submillimeter accuracy. The robot can position the linear accelerator at any point within a spherical shell of 60 to 100 cm from the target with a precision of ±0.5 mm. In principle, the frameless approach gives more choices in beam-angle selection compared with radiosurgery with a frame. The robotic arm must, however, avoid a collision with the patient, table, and imaging system as well as direct beam incidence into the imaging system. Another concern is the long treatment time compared with other delivery systems like the Gamma Knife and the isocentric linear accelerator. A total of 12 collimators from 5 to 60 mm in diameter are provided. Installing the unit would require a 10-ft ceiling over a 12 × 16 ft2 area.

CyberKnife CyberKnife, developed by Accuray Inc. (Santa Clara, CA) in collaboration with Stanford University (Stanford, CA), uses a miniature linear accelerator mounted on a robotic arm as shown in Figure 4-29 [35, 36]. The miniature linear accelerator operates in the X-band RF (9300 MHz) and provides a 6-MV

Proton Therapy

FIGURE 4-28. In an advanced linear accelerator, a diagnostic X-ray tube is added on a separate gantry at a 90° angle to the treatment gantry, and a ﬂat panel detector is installed on the opposite side. Using an X-ray tube and cone beam geometry reconstruction software, volumetric images can be obtained in the treatment room.

Clinical proton beams are produced by injecting hydrogen atoms with their electron stripped in an accelerating structure; electric ﬁelds accelerate the free protons to the desired clinical energy. The proton accelerators fall into two broad categories [20]. Cyclotrons are the classic proton accelerators in which protons are injected at the center of two halves (called “dees”) of an electrically powered circular dipole magnet with a constant magnetic ﬁeld (Fig. 4-30). The proton acceleration occurs within the narrow gap between the two half-magnets employing a correctly phased alternating electric ﬁeld. The protons gain energy, twice for each revolution, while moving in a constant circular path. Accelerated protons with a ﬁxed energy and a

4.

the physics of stereotactic radiosurgery

FIGURE 4-30. In cyclotrons, protons are injected at the center of two halves (“dees”) of an electrically powered circular dipole magnet with a constant magnetic ﬁeld. The proton acceleration occurs within the narrow gap between the two half-magnets employing a correctly phased alternating electric ﬁeld. The protons gain energy, twice for each revolution, while moving in a constant circular path. Accelerated protons with a ﬁxed energy and a continuous beam current can be extracted.

45

a constant radius (Fig. 4-31). As the proton energy increases, the magnetic ﬁeld strength also increases to keep them in a stable orbit. The acceleration cycle in synchrotrons is typically 2 seconds. Therefore, it is only possible to obtain proton beams in pulses with a repetition frequency of 30 cycles per minute and the energy of the protons can be varied from pulse-to-pulse. Operating the synchrotrons is complex because it requires rapid magnetic ﬁeld variations to operate. However, synchrotrons do not require energy degraders, thus minimizing secondary radiation production in the beam line. An elaborate beam transport system (BTS) carries the protons exiting the accelerator to the patient site. The beam line consists of focusing, bending, and switching magnets that steer a pencil beam of protons to different treatment rooms. The length of the beam line can easily exceed 30 m. A stable and precise BTS is required for reproducible dosimetry and patient treatments. The stability of the centroid of the beam position in the beam line must be better than 1.00 mm. Therefore, the bending and focusing magnets have to be very stable, mechanically and electrically, during the operation of the proton machine.

A Proton’s Interaction with Matter continuous beam current of up to 1 milliampere (mA) can be extracted and transported into an evacuated tube (beam line). Proton energy is changed by inserting an energy degrader in the path of the beam extracted from the cyclotron. The other type of proton accelerator is the synchrotron, which allows preaccelerated protons (up to a few MeV) to be injected into an accelerating chamber equipped with a number of electromagnets. The protons are accelerated in an orbit with

Protons are positively charged elementary particles that continuously lose energy and go through small-angle deﬂections as they travel through matter before they get absorbed. While the radiation dose deposited by photons drops off exponentially with penetration depth, the dose deposited by protons increases very slowly for about three-quarters of its range of travel in the medium before increasing sharply, reaching a maximum value before rapidly dropping off to zero. The depth at which the maximum energy is deposited by protons is called the Bragg peak. Figure 4-32 illustrates a depth dose of a proton beam showing its Bragg peak. The typical peak to entrance dose ratio of narrow proton beams is 4 or higher. The peak’s position is proportional to the energy of the proton beam. The range of

FIGURE 4-31. In synchrotrons, the protons are accelerated in an orbit with a constant radius. As the proton energy increases, the magnetic ﬁeld strength also increases to keep them in a stable orbit.

FIGURE 4-32. The depth dose of a proton beam is almost constant to a certain depth and starts to increase, then it jumps signiﬁcantly at the farthest end of the path. This peak is called the Bragg peak.

46

s. kim and j. palta

230-MeV protons is approximately 33.00 cm, which is sufﬁcient to penetrate any part of the body. The typical range of proton energies used for radiosurgery is 70 to 150 MeV (4 to 15 cm depth). The clinical target volumes are usually thicker than the width of the Bragg peaks, therefore, the range of the proton beam must be modulated to spread out the Bragg peak. By changing the energy of the incident proton beam with a variable range shifter, the Bragg peaks will “stack” at different depths. The variable range shifter can be a binary ﬁlter, a double-wedge variable absorber, or a circular wedge. Scattering foils or active scanning of a narrow beam with scanning magnets will laterally spread the proton beams to achieve uniform intensity across the target volume.

Dose Characteristics in Different Modalities In 1951, Leksell introduced the concept of stereotactic radiosurgery, and the ﬁrst patient was treated with the Gamma Knife prototype in 1967. Since then, Gamma Knife has only seen improvements with the establishment of many linear accelerator–based radiosurgery programs during the past 20 years. The use of proton beams for radiosurgery is also increasing. Some research groups have investigated and compared the dosimetric characteristics of each modality. The linear accelerator is thought to be able to provide dosimetric characteristics similar to Gamma Knife when multiple, non-coplanar arcs are used. In the case of more than ﬁve non-coplanar arcs, the full width at the half maximum (FWHM) of a 5-mm circular collimator ﬁeld is very close to that of Gamma Knife with a 4-mm helmet. The dose fall-off of the linear accelerator is also almost identical when more than ﬁve non-coplanar arcs are used. A recent study for trigeminal neuralgia treatment reports that compared with Gamma Knife, linear accelerator–based deliveries exhibit a ﬂatter top at the high isodose line level (>50%) and faster dose-off at the low isodose line levels (<50%) [37]. The penumbral widths (80% to 20%) were 1 mm for Gamma Knife with a 4-mm helmet and 2.1 mm for the 6-MV linear accelerator with a 5-mm collimator in another study [38]. Radiosurgery with a proton beam is different from both a linear accelerator and the Gamma Knife. Proton radiosurgery is performed with shaped and fully compensated ﬁelds, widely separated in the entrance angle, and with a single isocenter based on the characteristics of the proton beam, that is, the Bragg peak [39]. The proton beam, Gamma Knife, and linear accelerator have been compared in ﬁve clinical cases [3]. The optimal modality for stereotactic radiosurgery depends on target size, shape, and location; however, all modalities are equally good if the target is small and regular. In the case of trigeminal neuralgia, linear accelerator–based stereotactic radiosurgery is effective and can be comparable with Gamma Knife [38]. Gamma Knife is superior to the linear accelerator with regard to the conformality. In contrast, the linear accelerator is superior with regard to the dose fall-off outer region of the target volume [40]. The linear accelerator with mMLC can be better than Gamma Knife when hearing preservation is important during treatment [41].

Quality Assurance In most conventional radiation therapy treatments, prescribed dose is delivered in many fractions, and dose per each fraction is relatively small; however, a signiﬁcantly large dose is delivered in a single treatment during radiosurgery. Therefore, the impact of the inaccurate localization of the target or dose delivery in radiosurgery can be disastrous. To minimize such disasters, stereotactic radiosurgery quality assurance (QA) should be stringent. Both an international QA task group and Task Group 42 of the American Association of Physicists in Medicine (AAPM) have published general recommendations [42, 43]. The Radiation Therapy Oncology Group (RTOG) has also published guidelines [44]. In general, radiosurgery QA consists of common QA for the overall performance of all radiosurgery equipment, valid for a relatively long term, and speciﬁc QA for the calibration and preparation of equipment on each treatment day.

Common QA Common QA procedures need to be set to periodically verify equipment status and performance. It should include target localization QA, basic dosimetry QA, treatment planning QA, and output calibration and delivery QA. A detailed QA schedule must be set based on frequency and consequence: any item that has either high failure rate or severe consequence must frequently be checked. Overall target localization can be tested using a phantom containing a target located in the known geometry. Accuracy of within 1 mm is recommended. The treatment planning system’s dosimetric and nondosimetric conditions should both be checked. Speciﬁc items are well described in the recommendations of the AAPM Task Group 53 [45]. Dosimetry-related QA procedures are speciﬁc to the delivery system. In the Gamma Knife unit, the following are periodically veriﬁed: dose rate at the center of a 16-cm-diameter tissue-equivalent sphere at the focus; shutter error; connections of frame, collimator helmets, and sliding couch; and leak test on collimator helmets. For linear accelerators, output is checked daily. If the laser system is used for patient alignment, its accuracy needs to be veriﬁed rigorously. The most important QA item is the mechanical accuracy of the nominal isocenter. Less than a 1-mm diameter around the isocenter sphere is recommended [14, 46]. If an independent ﬂoor stand is used, like at the University of Florida, the accuracy is directly dependent on the ﬂoor stand. Thus, no additional QA tests on the laser and linear accelerator isocenter, other than those for routine conventional radiotherapy, are required.

Speciﬁc QA The goal of speciﬁc quality assurance is to make sure that the machine is running adequately, that all auxiliary devices are prepared and working properly, and that treatment parameters are correctly set just before and during patient treatments. Before placing the frame into the patient, the treatment streamline should be checked to minimize unnecessary treatment

4.

the physics of stereotactic radiosurgery

delay or cancellation after the frame is on. Figure 4-33 shows a sample of a patient-speciﬁc QA checklist used for linear accelerator–based radiosurgery at the University of Florida. As you can see, check items are listed chronologically. Once the secondary collimator is set as needed, the collimator drives are physically disconnected to prevent an accidental change of the

47

secondary collimator. To prevent the treatment table from accidentally dropping, a mechanical block is applied to the column of the table. Because the patient’s head is connected to the ﬂoor stand while his or her body is on the table, an abrupt drop of the table could cause serious problems as serious as a patient death. For the ﬁrst isocenter treatment, the accuracy of focusing

FIGURE 4-33. A sample of a patient-speciﬁc QA checklist used for linear accelerator–based radiosurgery at the University of Florida. In this sheet, checklist items are listed chronologically.

48

s. kim and j. palta

FIGURE 4-34. An example of veriﬁcation ﬁlm. It can be seen that the steel ball in the calibration jig and radiation beams are well aligned.

the radiation ﬁelds to the isocenter is checked by a WinstonLutz ﬁlming test. A staff member sets the ﬂoor stand to the treatment position. A steel ball in the calibration jig is set to the isocenter coordinates independently by another staff member and attached to the ﬂoor stand. Then, the veriﬁcation ﬁlm is irradiated in four different beam angles with a 24-mm-diameter circular collimator. If both the ﬂoor stand and calibration jig are set to the isocenter correctly, and beam alignment is appropriate in all beam angles, images of the ball must be centered within all the circular ﬁelds. Figure 4-34 shows an example of the veriﬁcation ﬁlm. Once the ﬁlm is conﬁrmed satisfactory, the patient is moved to the ﬂoor. Before the treatment beam is on, the size of the circular collimator and the treatment parameters, such as the table angle, gantry start angle, gantry end angle, and monitor unit, are veriﬁed. In subsequent isocenter treatments, the ﬂoor stand is set by a physicist and coordinates are independently double-checked by two other staff members. Speciﬁc QA in the Gamma Knife unit is relatively less intense. It is basically ready for treatment in most cases, and only a few safety checks need to be performed.

Conclusion Stereotactic radiosurgery is a special radiotherapy technique that holds promise not only radio-therapeutically but also as a neurosurgical tool. The underlying physics principles for radiosurgery are the same as those for conventional radiation therapy; however, the accuracy requirements in dosimetry and patient positioning for radiosurgery are at the millimeter level. There are several dosimetric parameters that must be measured in tissue-equivalent phantoms such as water including dose calibration, percentage depth doses, relative dose output factors, and cross-beam proﬁles. Radiosurgery treatment ﬁelds are often very small, therefore, the spatial resolution requirements for radiation detectors are much more stringent in radiosurgery.

References 1. Evans RD. The Atomic Nucleus. Malabar, FL: Krieger, 1955. 2. Attix FH. Introduction to Radiological Physics and Radiation Dosimetry. New York: Wiley-Interscience, 1986. 3. Khan FM. The Physics of Radiation Therapy, Second ed. Philadelphia: Lippincott Williams & Wilkins, 1992. 4. Johns HE, Cunningham JR. The Physics of Radiology, Fourth ed. Springﬁeld, IL: Charles C Thomas Publisher, 1983. 5. Leksell L. Cerebral radiosurgery I. Gamma thalamotomy in two cases of intractable pain. Acta Chir Scand 1968; 134:585–595.

6. Wu A, Maitz AH, Kalend AM, et al. Physics of gamma knife approach on convergent beams in stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1990; 18:941–949. 7. Simonova G, Novotny J Jr, Vladyka V, et al. Fractionated stereotactic radiotherapy with the Leksell gamma knife: feasibility study. Radiother Oncol 1995; 37:108–116. 8. Poffenbarger B. Viability of an isocentric cobalt-60 teletherapy unit for stereotactic radiosurgery. MSc thesis, McGill University, Montreal, Canada, 1998. 9. Poffenbarger B. Viability of an isocentric cobalt-60 teletherapy unit for stereotactic radiosurgery. Med Phys 1998; 25:1935–1943. 10. Bellerive M, Kooy HM, Loefﬂer J. Linac radiosurgery at the Joint Center for Radiation Therapy. Med Dosim 1998; 23:187–199. 11. Betti OO, Derechinsky VE. Hyperselective encelphalic irradiation with linear accelerator. Acta Neurochir Suppl 1984; 33:385–390. 12. Colombo F, Benedetti A, Pozza F, et al. External stereotactic irradiation by linear accelerator. Neurosurgery 1985; 16:154–160. 13. Hamilton JA, Lulu BA, Fosmire H, et al. Preliminary clinical experience with linear accelerator based spinal stereotactic radiosurgery. Neurosurgery 1995; 36:311–318. 14. Lutz W, Winston KR, Maleki N. A system for stereotactic radiosurgery with a linear accelerator. Int J Radiat Oncol Biol Phys 1988; 14:373–381. 15. Larsson B. Dosimetry and radiobiology of protons as applied to cancer therapy and neurosurgery. In: Thomas RH, Perez-Mendez V, eds. Advances in Radiation Protection and Dosimetry in Medicine. New York: Plenum, 1980:367–394. 16. Frankel KA, Phillips MH. Charged particle method: protons and heavy charged particles. In: Phillips MH, ed. In Physical Aspects of Stereotactic Radiosurgery. New York: Plenum, 1993. 17. Kjellberg RN, Shintani NA, Frantz AG. Proton beams in acromegaly. N Engl J Med 1968; 278:689–695. 18. Larsson B, Leksell L, Rexed B, et al. The high energy proton beam as a neurosurgical tool. Nature 1958; 182:1222–1223. 19. Breuer H, Smit BJ. Proton Therapy and Radiosurgery. Berlin, Heidelberg, New York: Springer-Verlag, 2000. 20. Wieszczycka W, Scharf WH. Proton Radiotherapy Accelerators. Singapore: World Scientiﬁc, 2001. 21. Karzmark CJ, Nunan CS, Tanabe E. Medical Electron Accelerators. New York: McGraw-Hill, 1993. 22. Shiu AS, Kooy HM, Ewton JR, et al. Comparison of miniature multileaf collimation with circular collimation for stereotactic treatment. Int J Radiat Oncol Biol Phys 1997; 37:679–688. 23. Arcovito G, Piermattei A, D’Abramo G et al. Dose measurements and calculations of small radiation ﬁelds for 9-MV x rays. Med Phys 1985; 12:779–784. 24. Bova FJ. Radiation physics. Neurosurg Clin N Am 1990; 1:909– 931. 25. Houdek PV, VanBuren JM, Fayos JV. Dosimetry of small radiation ﬁelds for 10-MV x rays. Med Phys 1983; 10:333–336. 26. Rice RK, HansenJL, Svensson GH, et al. Measurements of dose distributions in small beams of 6 MV x-rays. Phys Med Biol 1987; 32:1087–1099. 27. Heydarian M, Hoban PW, Beddoe AH. A comparison of dosimetry techniques in stereotactic radiosurgery. Phys Med Biol 1996; 41:93–110. 28. Rustgi SN. Evaluation of dosimetric characteristics of a diamond detector for photon beam measurement. Med Phys 1995; 22:567–570. 29. Vatnitsky S, Jarvinen H. Application of natural diamond detector for the measurement of relative dose distribution in radiotherapy. Phys Med Biol 1993; 38:173–184. 30. Schell MC, Smith V, Larson DA, et al. Evaluation of radiosurgery techniques with cumulative dose volume histograms in linac-based stereotactic irradiation. Int J Radiat Oncol Biol Phys 1991; 20: 1325–1330.

4.

the physics of stereotactic radiosurgery

31. Serago CF, Houdek PV, Bauer-Kirpes B, et al. Stereotactic radiosurgery: dose volume analysis of linear accelerator techniques. Med Phys 1992; 19:181–185. 32. Friedman WA, Bova FJ. The University of Florida radiosurgery system. Surg Neurol 1989; 32:334–342. 33. Meeks SL, Bova FJ, Friedman WA, et al. Linac scalpel radiosurgery at the University of Florida. Med Dosim 1998; 23:177–185. 34. Webb S. Optimization by simulated annealing of three-dimensional conformal treatment planning for radiation ﬁelds deﬁned by a multileaf collimator. Phys Med Biol 1991; 36:1201–1226. 35. Adler JR, Cox RS. Preliminary clinical experience with the CyberKnife: image guided stereotactic radiosurgery. In: Kondziolka D, ed. Radiosurgery. Basel: Karger, 1995:317–326. 36. Maciunas RJ, Fitzpatrick M, Galloway RL, et al. Beyond stereotaxy: extreme levels of application accuracy are provided by implantable ﬁducial markers for interactive image guided neurosurgery. In: Maciunas RJ, ed. Interactive Image Guided Neurosurgery. Washington, DC: AANS, 1994:261–270. 37. Ma L, Kwok Y, Chin LS, et al. Comparative analyses of linac and Gamma Knife radiosurgery for trigeminal neuralgia treatments. Phys Med Biol 2005; 50:5217–5227. 38. Gerbi BJ, Higgins PD, Cho KH, et al. Linac-based stereotactic radiosurgery for treatment of trigeminal neuralgia. J Appl Clin Med Phys 2004; 5:80–90. 39. Verhey LJ, Smith V, Sergo CF. Comparison of radiosurgery treatment modalities based on physical dose distributions. Int J Radiat Oncol Biol Phys 1998; 40:495–505. 40. Plowman PN, Doughty D. Stereotactic radiosurgery, X: clinical isodosimetry of Gamma knife versus linear accelerator X-knife for pituitary and acoustic tumors. Clin Oncol 1999; 11:321–329. 41. Perks JR, St. George EJ, Hamri KE, et al. Stereotactic radiosurgery XVI: isodose comparison of photon stereotactic radiosurgery techniques (Gamma knife vs. micromultileaf collimator linear accelerator) for acoustic neuroma-and potential clinical importance. Int J Radiat Oncol Biol Phys 2003; 57:1450–1459. 42. Hartmann G, Lutz W, Arndt J, et al. Quality Assurance Program on Stereotactic Radiosurgery. Berlin: Springer-Verlag, 1995. 43. Schell M, Bova FJ, Larson DA, et al. Stereotactic radiosurgery. Report of Task Group 42. American Association of Physicists in Medicine (AAPM), Report No. 54. Medical Physics Publishing: Madison, WI, 1995. 44. Shaw E, Kline R, Gillin M, et al. Radiation therapy oncology group: radiosurgery quality assurance guidelines. Int J Radiat Oncol Biol Phys 1993; 27:1231–1239. 45. Fraass B, Doppke K, Hunt M, et al. American Association of Physicists in Medicine (AAPM): Radiation Therapy Committee Task Group 53: Quality assurance for clinical radiotherapy treatment planning. Med Phys 1998; 25:1773–1829. 46. Falco T, Lachaine M, Poffenbarger B, et al. Setup veriﬁcation in linac-based radiosurgery. Med Phys 1999; 26:1972–1978.

Glossary AAPM American Association of Physicists in Medicine accelerator a device that accelerates subatomic charged particles or nuclei to high energies alpha particle positively charged particle consisting of two protons and two neutrons Bq Becquerel, the Système International (SI) unit of radioactivity equal to one disintegration per second beta particle high-speed electron or positron binding energy net energy required to remove an atomic electron from its orbit

49

Bragg peak peak can be seen at the end of the depth dose curve of heavy charged particles such as a proton BTS beam transport system, a proton accelerator system that carries the protons exiting the accelerator to the patient site cavity a device through which power is coupled to the accelerating particles in a resonant mode Ci Curie, unit of radioactivity equal to 3.7 × 1010 disintegrations per second collimator a device capable of collimating radiation CT computed tomography cyclotron a particle accelerator in which charged particles are accelerated in a circular path by an alternating electric ﬁeld in a constant magnetic ﬁeld DD depth dose, relative dose level according to the depth of the material when the radiation beam is incident dipole a pair of electric charges or magnetic poles dose energy absorbed per unit mass by radiation DVH dose-volume histogram, accumulated volume according to dose for a volume of interest ﬁlter a device used to modify the intensity of radiation FWHM full width at half maximum gamma particle electromagnetic radiation particle (photon) emitted by radioactive decay Gy Gray, a unit of dose equivalent to joules per kilogram half-life length of time needed for a radioactive substance to lose half of its radioactivity from decay IMSR intensity-modulated stereotactic radiosurgery, a treatment technique in which radiation intensity is modulated within a ﬁeld to obtain an optimized dose distribution ionization chamber a chamber for determining the intensity of ionizing radiation by measuring ionization of the enclosed gas ionizing radiation electromagnetic or particulate radiation capable of producing ions in its passage through matter isocenter a point in the space that is the center of rotation of the gantry, collimator, and table of an isocentric medical linear accelerator isodose line a line made by connecting all points of the same dose kernel a function that describes how the dose is distributed within a medium keV kilo-electron-volts, a unit of energy of particles klystron a microwave ampliﬁer linear accelerator a particle accelerator in which the path of the particles is straight mA milliampere, a unit of electric current magnetron a diode-type electron tube that generates microwave pulses MeV million electron-volts, a unit of energy of particles mMLC micro-multileaf collimator, a special collimator consisting of many small collimator leaves that can be individually moved to conform to the shape of the target mR milliroentgen, a unit of exposure μSv microsievert, the SI unit for an equivalent radiation dose that accounts for biological effect MV million volts, a unit of beam quality of radiation therapy beams produced in linear accelerators OAR off-axis ratio, relative dose distribution following a line on a plane normal to the direction of the radiation beam, same as proﬁle

50

s. kim and j. palta

output magnitude of radiation produced PDD percent depth dose, relative dose level in percentile according to the depth of the material when the radiation beam is incident photon quantum of electromagnetic energy with no mass and charge considered to be both particle and wave positron positively charged particle of the same mass and magnitude of charge as an electron proﬁle relative dose distribution following a line on a plane normal to the direction of radiation beam, same as the offaxis ratio (OAR) proton stable and positively charged subatomic particle QA quality assurance radioactive decay spontaneous disintegration of a radionuclide followed by the emission of ionizing radiation in the form of alpha, beta, or gamma particles

radioactivity capability of radioactive decay radioisotope naturally or artiﬁcially produced radioactive isotope of an element range distance charged particles can travel resonance condition of being resonant in which an increase in the amplitude of a physical quantity can occur RF radiofrequency, a range of electromagnetic frequencies above sound and below visible light synchrotron a particle accelerator in which charged particles are accelerated around a ﬁxed circular path by an electric ﬁeld and held to the path by an increasing magnetic ﬁeld TERMA total energy released per unit mass, amount of energy released from radiation to the unit mass of the interaction material Z atomic number; that is, the number of protons in an element

5

Radiobiological Principles Underlying Stereotactic Radiation Therapy David J. Brenner

Introduction Since the Gamma Knife was ﬁrst conceived in 1968, primarily for arterial and functional lesions [1], single-fractioned stereotactic radiation therapy* has been increasingly used to treat a variety of cerebral lesions. By 1985, an alternative modality was available for stereotactic radiation therapy, using a linear accelerator (linac) and a stereotactic head frame [4, 5]. Recently, the CyberKnife, a frameless robotic system, has been developed for stereotactic radiation therapy [6], and intensity-modulated stereotactic radiation therapy [7, 8] is now entering clinical practice. In its early use, stereotactic radiation therapy was always applied in single fractions (i.e., radiosurgery), so its beneﬁts were entirely related to its ability to irradiate target volumes with excellent dose distributions. By about 1990, however, several groups [9–15] began to consider the potential biological advantages of fractionated stereotactic radiotherapy, stimulated also by the development of relocatable stereotactic head frames for linac-based treatments [11, 12, 16]. As we will discuss, new technology has made it increasingly practical to fractionate a stereotactic treatment, and the use of fractionated stereotactic radiotherapy has indeed increased steadily over the past 15 years, as illustrated by more than 700 publications documented in PubMed/Medline on this modality. We will review here the radiobiological principles underlying stereotactic radiation therapy and their applications to single or multifractioned radiotherapy of the three main types of lesions that might be treated with this modality: malignant tumors, benign tumors, and vascular disorders. In that little is known about the radiobiological rationale behind radiotherapy for functional disorders, this area will not be covered. A complementary review on the clinical aspects of fractionated stereo-

* Note the term stereotactic radiation therapy will be used here to apply both to single-fraction treatment (often called radiosurgery) and to multiple-fractioned stereotactic radiotherapy. There is still debate about the most appropriate terminology [2, 3].

tactic radiotherapy has been published by Tomé and colleagues [17].

The Three R’s of Radiobiology: Reoxygenation, Repair, and Repopulation All these three radiobiological phenomena relate ultimately to cell-killing processes, the assumption being that this is the primary (though not the only) mechanism by which all radiotherapy both produces tumor control and induces side effects. Of course, there are a variety of mechanisms that lead to cell killing, and a variety of relevant target cells, but underlying all radiotherapeutic response remains cell killing [18].

Reoxygenation The ﬁrst radiobiological principle of importance here is that malignant tumors, even those of limited size, almost always contain a proportion of hypoxic cells that, because of their deﬁciency in oxygen, are highly resistant to killing by X- or γrays [19]. Examples are shown in Figure 5-1, showing hypoxic regions in sections of small tumors derived from human glioma xenograph lines. Figure 5-2 shows a dose-response curve, derived from the classic studies of Powers and Tolmach [20], illustrating the fraction of cells surviving in very small tumors in a mouse irradiated in a single fraction in vivo, and subsequently assayed by transplantation to other animals. The cellular survival curve is characterized by two distinct components; the slopes of the two components differing by a factor of 2 to 3. Up to doses of several Gy, the response is dominated by the killing of aerobic cells, whereas, for higher doses, the killing of hypoxic cells dominates. It is apparent that irradiating partially hypoxic tumors with a single large dose of several tens of Gy is a futile exercise if the goal is sterilization, because the hypoxic cells will not be adequately depopulated with a dose of this size. Figure 5-3 gives rough estimates, based on in vitro data, of the single fraction dose to sterilize a 30-mm-diameter tumor, with and without a hypoxic component [21].

51

52

d.j. brenner 100 90 80 20% hypoxic

70

Dose (Gy)

60 50 40 30 Fully oxygenated

20 10

Tumors, however, exhibit a characteristic known as reoxygenation, whereby, between fractionated doses of X- or γ-rays, tumors tend to reestablish their original pattern and proportion of oxygenated and hypoxic cells [22] (Fig. 5-4). In a fractionated

head & neck

ovarian

renal

brain

prostate

breast

medullo

colo-rectal

NSCLC

melanoma

SCLC

0 sarcoma

FIGURE 5-1. Hypoxic regions are present even in small tumors. Imaged here is a section of a small tumor derived from a human glioma xenograft line, with the black areas indicating regions of hypoxia. (From Rijken PF, Peters JP, Van der Kogel AJ. Quantitative analysis of varying proﬁles of hypoxia in relation to functional vessels in different human glioma xenograft lines. Radiat Res 2002; 157:626–632. Used with permission.)

FIGURE 5-3. Rough estimate, based on in vitro data, of the singlefraction dose required to sterilize various 30-mm-diameter tumors, which are either fully oxygenated (lower) or contain 20% hypoxic cells (upper). (Adapted from Leith JT, Cook S, Chougule P, et al. Intrinsic and extrinsic characteristics of human tumors relevant to radiosurgery: comparative cellular radiosensitivity and hypoxic percentages. Acta Neurochir Suppl 1994; 62:18–27. With kind permission of Springer Science + Business Media.)

100 response dominated by aerated cells

10–1

Surviving Fraction

10–2

response dominated by 1–2% hypoxic cells

10–3 10–4 10–5 10–6 10–7 0

5

10 15 20 25 Dose (Gy) FIGURE 5-2. Fraction of surviving cells, after a single radiation dose, in small mouse lymphosarcomas irradiated in vivo. The shallow initial part of the curve is due to the presence of aerated cells, whereas the steeper slope at higher doses is caused by the presence of 1% to 2% of hypoxic cells. (Adapted from Hall EJ, Brenner DJ. The radiobiology of radiosurgery: rationale for different treatment regimes for AVMs and malignancies. Int J Radiat Oncol Biol Phys 1993; 25:381–385. Copyright 1993, with permission from Elsevier.)

regime, therefore, each X- or γ-ray dose predominately kills aerobic cells, and the interval between treatments allows hypoxic cells to reestablish their oxygenated state.

Repair The second radiobiological principle is repair. It has been well established for many decades that protracting or fractionating an acute exposure reduces the level of cell killing. An example is shown in Figure 5-5 [23]. If all tissues were equally affected by changes in protraction or fractionation, then there would be no radiotherapeutic signiﬁcance to fractionation beyond the need to increase the dose to compensate for the increased cellular repair. There is, however, a wealth of experimental evidence indicating that there is a difference in shape between the doseresponse relationship characteristic of early-responding tissues and tumors and late-responding tissues. The inference from experiments in animals, which is conﬁrmed in clinical practice, is that the dose-response relationship for late-responding tissues is more “curvy” than that for early-responding tissues, as shown in Figure 5-6. In mathematical terms, if the dose-survival relationship is expressed in terms of a linear-quadratic relation, S = exp(−αD − βD2),

5.

radiobiological principles underlying stereotactic radiation therapy

FIGURE 5-4. Schematic illustration of the process of reoxygenation: Almost all tumors initially contain a mixture of aerated and hypoxic cells. A single radiation dose will kill a larger fraction of aerated than hypoxic cells because they are more radiosensitive (see Fig. 5-2). Thus, after irradiation, a larger fraction of cells are hypoxic. But after a short time, the tumor tends to return to its original proportion of aerated and hypoxic cells, allowing many of the previously hypoxic (but now aerated) cells to be killed by a second, and subsequent, dose fraction. (Reprinted from Hall EJ, Brenner DJ. The radiobiology of radiosurgery: rationale for different treatment regimes for AVMs and malignancies. Int J Radiat Oncol Biol Phys 1993; 25:381–385. Copyright 1993, with permission from Elsevier.)

Fraction of Initial Colony-Forming Units (3.5 cells/col) Surviving

100 0.36 rad/min 10–1 0.86 rad/min

10–2

16 rad/min

10–3

30 rad/min

10–4 107 rad/min 10–5

500

1000 1500 2000 2500 Dose (rad) FIGURE 5-5. Illustrating that, as a given dose is protracted, either by lowering the dose rate (as here), or by increasing the number of fractions, the biological effect (in this case, cell killing of Chinese hamster cells) decreases, due to repair. Base on data from Bedford and Mitchell [23]. (Reprinted from Hall EJ, Brenner DJ. The dose-rate effect revisited: radiobiological considerations of importance in radiotherapy. Int J Radiat Oncol Biol Phys 1991; 21:1403–1414. Copyright 1991; with permission from Elsevier.)

53

FIGURE 5-6. The γ-ray dose-response curve for late-responding tissues (e.g., AVM obliteration, cerebral necrosis) is “curvy” (i.e., has a small α/β ratio); for early-responding end points such as tumor control, the dose-response curve is straighter (i.e., the α/β ratio is larger). Consequently, dose fractionation spares late-responding tissues more than early-responding tissues. (Reprinted from Hall EJ, Brenner DJ. The radiobiology of radiosurgery: rationale for different treatment regimes for AVMs and malignancies. Int J Radiat Oncol Biol Phys 1993; 25:381– 385. Copyright 1993, with permission from Elsevier.)

then the ratio α/β (the dose at which the cell killing by the linear and quadratic terms is equal) tends to be small for lateresponding tissues (less than ∼3 Gy) and larger (greater than ∼8 Gy) for early-responding tissues [18, 22, 24–27]. The practical consequence of the difference in shape between the dose-response curves for early and late responding tissues is a marked difference in the response to fractionation of these two types of tissues. Late-responding tissues are more sensitive to changes in fractionation than early-responding tissues. The overall outcome is illustrated in Figure 5-6. A fractionated regime spares late-responding normal tissues more than a single acute dose for a given level of tumor damage. It is pertinent to ask what exactly constitute “early-” and “late” responding tissues? Probably, early-responding tissues contain a large proportion of cycling cells, whereas lateresponding tissues contain large proportions of noncycling cells.

Accelerated Repopulation As a tumor shrinks during radiotherapy, the surviving clonogens are stimulated to proliferate at an accelerated rate [28, 29]. This accelerated repopulation typically does not commence for about 30 days after the beginning of treatment as illustrated in Figure 5-7 [30]; however, radiation treatments of duration longer than about 30 days need to increase the total dose to compensate for this clonogen proliferation. Note that it is unrealistic to categorize almost any tumor type as one that is not signiﬁcantly prone to repopulation—if research in predictive assays has taught us anything, it is that intertumor variation can be large; thus, in this context, some tumors will grow slowly, and some will grow very quickly.

54

d.j. brenner

remaining two R’s of radiotherapy (reoxygenation and repair) affect stereotactic radiotherapy for the three main target lesions: malignancies, arteriovenous malformations, and benign tumors.

Malignancies

FIGURE 5-7. Tumor-control probability for T3 laryngeal cancer, as a function of overall time and overall dose. The data points, at the top of the dashed vertical lines, are those reported by Slevin et al], and the surface is the result of a linear-quadratic (with repopulation) ﬁt] to the data, with repopulation effectively starting about 33 days after the beginning of the treatment. (From Brenner DJ. Accelerated repopulation during radiotherapy: quantitative evidence for delayed onset. Radiat Oncol Invest 1993; 1:167–172. Used with permission.)

How Do These Radiotherapeutic Principles Apply to Stereotactic Radiation Therapy? In general, these three phenomena control the dose delivery for cancer radiotherapy: 1. We must fractionate treatment to overcome hypoxia and for differential response with late effects. 2. We would like to prolong treatment to limit early sequelae. 3. We would like to shorten treatment to prevent accelerated repopulation. Prima facie, requirements (2) and (3) are mutually exclusive and so, in most radiotherapeutic situations, the overall treatment time represents a compromise between short treatment times to minimize tumor repopulation and long treatment times to prevent unacceptable early complications. In those protocols that have attempted to utilize shorter overall treatment times, it has been necessary to reduce the overall dose in order to avoid excessive early complications [31]. By contrast, stereotactic radiotherapy represents a situation where the contradictory aspects of overall treatment time can be resolved without the need for compromise. Because of the excellent dose distributions that can be obtained using stereotactic radiotherapy, the need for long overall treatment times to reduce early normal-tissue complications will rarely apply. Speciﬁcally, early skin or mucosal reactions, which are often a limiting factor in radiotherapy, will not be an issue for intracranial stereotactic radiotherapy. Thus, short overall treatment times become a practical possibility, with their attendant advantages in terms of limiting tumor repopulation. Inherently, then, stereotactic radiation therapy is advantageous in terms of repopulation. We discuss here how the

As we have discussed, it is very unlikely that a single-dose fraction could sterilize even a small tumor that contains hypoxic malignant cells. Thus, there is a clear rationale, based on the issue of hypoxia, for stating unequivocally that single-fraction radiotherapy (radiosurgery) is a suboptimal modality when the target is a malignant tumor [32]. In addition, because of the differential α/β ratio of the tumor and the surrounding late-responding normal tissue sequelae (such as delayed cerebral necrosis, or optic neuropathy), even in the rare situation when there is minimal hypoxia, single-fractioned treatment of a malignancy would be expected to give a suboptimal therapeutic ratio between tumor control and late complications. Thus, an optimal stereotactic radiotherapy protocol for a large intracranial malignancy would involve large numbers of fractions over a short time period—20 fractions in 2 weeks might be an appropriate target. For smaller tumors, where the relative dose to normal tissue surrounding the malignancy is smaller—though still, as always, dose limiting—a regime consisting of 5 or 10 fractions over 1 week would combine the clinical advantages of a very short overall time, the radiobiological advantages of reoxygenation and (partial) repair of lateresponding normal tissue damage, and the practical advantages of a small number of treatments. It has been argued that, even for malignancies, “when the treatment volume is small and contains little functioning brain tissue, the need for fractionation may not apply” [33]; however, the dose-limiting factor of any radiotherapeutic treatment of a malignancy must always be the normal-tissue tolerance, so fractionation should always allow a greater tumoricidal dose. Of course, one of the potential advantages of radiosurgery is that it may be able to overcome the radioresistance of tumors such as metastases from melanomas or renal cell sarcomas; however, by keeping the number of fractions low (perhaps to ﬁve or six), the radioresistance of such tumors will still be adequately addressed while maintaining the other biological advantages of fractionation that have been outlined here. The arguments presented here do not exclude the fact that good results may be obtained using radiosurgery for brain malignancies. Indeed, there are reports of good results with this technique. Rather, for any given situation, better results— in terms of therapeutic ratio between tumor control and complications—would always be expected by fractionating than can be obtained with a single fraction, an important consideration in the treatment of cerebral malignancies.

Arteriovenous Malformations In treating arteriovenous malformations (AVMs), the goal is to produce progressive luminal closure through an inﬂammatory reaction in the vessel walls of the malformation, by irradiating the constituent epithelial cells [34]. This is a classic “late”

5.

radiobiological principles underlying stereotactic radiation therapy

response, occurring weeks to months after the radiation treatment and thus, prima facie, one would expect a relevant α/β ratio similar to that for other late-responding tissue. The issue here is that the dose-limiting side effect of the treatment, generally delayed cerebral radionecrosis, is, of course, also a late response [35–39]. Thus, in the treatment of AVMs, the tissue that it is desired to damage and the surrounding normal tissue that it is desired to spare are both of the same radiobiological type (i.e., they are both late-responding tissues). Consequently, one would expect that nothing is to be gained by a fractionated course relative to a single dose. In other words, a change in fractionation pattern will not preferentially produce more damage in the AVM than in the surrounding normal tissues. The ratio of damage between the AVM and the surrounding normal tissue would be the same whether the dose is delivered in a single or in multiple exposures. Given that the issue of hypoxic cells is not relevant here, this is a strong rationale for radiosurgery (i.e., a single, highly localized radiation treatment). In fact, there have been few estimates of α/β ratios reported in the literature corresponding with the end point of AVM obliteration, as few dose-response data have been established. An early estimate [32] of 0.6 Gy (range, 0.2 to 5 Gy) was reported based on limited dose-response results for complete AVM obliteration after stereotactic irradiation with helium ions. Using the more recent dose-response data from Touboul et al. [40] and Flickinger [41], we estimated values of 5.2 Gy from long-term obliteration data and 2.4 Gy (range, 0.9 to 8 Gy) from short-term obliteration data. A recent analysis by Kocher et al. [42] of most available data resulted in an overall best estimate of 3.5 Gy, and an estimate of 4.6 to 6.4 Gy for small AVMs (Fig. 5-8). All these values are consistent with those of a classic late response. In summary, in the treatment of AVMs, there is probably nothing to be gained by a fractionated course relative to a single

FIGURE 5-8. Linear quadratic ﬁt to combined long-term AVM obliteration data from Paris [39] and Pittsburgh [40]. The estimated α/β ratio from this analysis is 5.2 Gy, consistent with a recent estimate of 3.5 Gy by Kocher et al. [41], based on most currently available data, which is in turn consistent with a typical α/β value for a late-responding tissue.

55

dose. Thus, the optimal treatment for AVMs is expected to be single-fractionated stereotactic radiotherapy or radiosurgery; that is, a single, highly localized radiation treatment at, of course, the appropriate dose.

Benign Tumors Less is known about the sensitivity to fractionation of benign tumors. It is less likely that hypoxia will be a signiﬁcant role here; one estimate of the α/β ratio for benign parasellar meningiomas is 3.3 Gy [43], similar to that for late-responding normal tissues. For benign tumors as with AVMs, the most likely scenario is that both the target and the surrounding normal tissue respond in a similar way to fractionation, like that of a lateresponding normal tissue, suggesting that there is little to be gained from fractionation. For example, if the dose-limiting late sequelae were radiation-induced optic neuropathy, which is caused by damage to the arteries supplying the optic nerve, this almost certainly responds to changes in fractionation like a classic late-responding tissue [44, 45]. In summary, in the treatment of benign tumors, there is probably nothing to be gained by a fractionated course relative to a single dose; however we certainly do not have adequate data for a variety of benign tumors to consider this conclusion deﬁnitive.

How Many Fractions Should Be Used in Stereotactic Radiation Therapy of Cerebral Malignancies? As we have discussed, any multifraction scheme (with appropriately chosen doses) would be expected [32] to be superior to a single fraction for treating malignancies. Even apart from the issue of sparing late-responding normal tissue, if a tumor contains hypoxic cells, it would be most unlikely that it could be sterilized by a single dose of radiation. Most tumors, even small ones, probably contain hypoxic cells [22], so why take that risk? Given, then, that one should fractionate, how many fractions is appropriate? Five? Ten? Thirty? Consider the use of 30 fractions: when critical structures such as the optic nerve are likely to receive large doses, large fraction numbers such as 30 are indicated. On the other hand, most institutions treat solely with single fractions, and the use of, say, 10 fractions is certainly preferable to that. When critical structures are not considered to be at risk, however, then accelerated fractionation may become the treatment of choice [46] to avoid accelerated tumor repopulation, and as small a fraction number as ﬁve is probably appropriate to overcome the problem of hypoxic cells. Given that some tumors will show rapid repopulation, when critical structures are not a problem, why take the risk of a prolonged treatment? On a pragmatic note, whereas a few institutions are in a position to use, say, 30 fractions, most of the rapidly increasing number of centers using a Gamma Knife or a linac simply are not. On the other hand, smaller numbers of fractions could realistically become a generally acceptable option, with attendant beneﬁts compared with a single fraction.

56

d.j. brenner

Is Fractionated Stereotactic Radiotherapy Practical? The very earliest fractionated stereotactic radiotherapy treatments were accomplished using an invasive stereotactic head frame that was kept on for the entire course of treatment [5, 9, 15]. This approach, however, was difﬁcult for the patient and led to signiﬁcant infection issues at the pin sites. The ﬁrst major advance, in the early 1990s, was the development of noninvasive relocatable head frames [11, 12, 16], and versions of the Gill-Thomas-Cosman relocatable frame [16], which uses a bite block system, have been widely used [47–51]. Other systems are based on other anatomic features such as the ear canals/nasal bridge (the Laitinen stereoadapter [12, 52–54]) or a facial mask [55, 56]. Another recent system involves the use of one-time invasive placement of screws into the skull [57]. By contrast, the Gamma Knife is rather less suited to fractionated treatment. An invasive method based on inserting four screws into the skull has been described and tested by Walton et al. [58], but no subsequent clinical results have been reported. More recently, frameless, noninvasive optical systems have been introduced for use with linac systems, which show great potential [59–61]. Localization is accomplished through detection of four passive markers attached to a custom biteplate linked to the maxillary dentition of the patient. Translations and rotations of the patient’s isocenter are tracked in real time using a charge-coupled device (CCD)-based optical system detecting infrared light that is reﬂected off the four passive markers. A logical extension of frameless, noninvasive optical systems is the CyberKnife system [6]. The CyberKnife, developed at Stanford University, mounts a lightweight 6-MV linear accelerator on a computer-controlled robotic arm with 6 degrees of freedom, which irradiates the target while being guided in real time by computer image tracking technology in a frameless environment. Tracking is achieved using two ceiling-mounted diagnostic X-ray cameras with corresponding orthogonal and ﬂoor-mounted amorphous silicon detectors for real-time digital imaging. With such a system, a fractionation exposure presents no more technical difﬁculties than a single fraction, and the use of the CyberKnife for fractionated treatments has indeed been recently reported [62–66]. A ﬁnal approach to stereotactic radiation therapy may be the one with the greatest potential: to extend the current technology of intensity-modulated radiation therapy to apply to stereotactic radiotherapy. Intensity-modulated radiation therapy refers to linear accelerator–based radiation therapy in which individual small beams (typically 10 × 10 mm) within the larger ﬁeld are modulated with a multileaf collimator to produce highly conformal dose distributions. By modulating the intensity of large numbers of small beamlets within a larger open ﬁeld, an intensity variation is created that will generate an “optimized” dose distribution. To adapt this widely used technique to stereotactic radiotherapy, beamlet widths and multileaf collimator leaf widths needed to be reduced from ∼10 mm to ∼3 mm [67, 68], and this technology is now commercially available, leading to the use of both single-fraction [7, 68, 69] and multiple-fraction [70–73] intensity-modulated stereotactic radiation therapy.

How to Calculate the Appropriate Multifractioned Dose That Is Biologically Equivalent to a Given Single-Fractioned Dose The standard technique for making such isoeffect calculations is the linear-quadratic (L-Q) model, which was spelled out in detail more than 50 years ago by Lea and Catcheside [74, 75], based on a mechanistic analysis of chromosomal aberration induction in Tradescantia spores. The application of this formalism to radiation therapy has been reviewed extensively [18, 25, 76, 77]. Central to the approach is that radiotherapeutic response is primarily related to cell survival (or perhaps survival of groups of cells). This concept is itself not necessarily true, but there is now a wealth of evidence that cell killing is the dominant determinant of radiotherapeutic response, both for early- and lateresponding end points [18]. In the most basic linear-quadratic approach, cellular survival S at a dose D is written as S(D) = exp(−αD − βD²).

(5-1)

The mechanistic interpretation of Eq. (5-1) is that cell killing results from the interaction of two elementary damaged species (probably DNA double-strand breaks) to produce a species (perhaps a dicentric chromosomal aberration) that may cause lethality. The two terms in Eq. (5-1) indicate that the two elementary damaged species may be produced by the passage of the same track of radiation (linear term in dose) or by two different tracks (quadratic term). Clearly, if some time elapses between the passage of the ﬁrst and second tracks, there exists the possibility of the ﬁrst damaged site being repaired before interacting with the second. This repair will result in the a reduction of the second, quadratic term in Eq. (5-1) (but not the ﬁrst), by a factor denoted G by Lea and Catcheside [75]: S = exp(−αD − GβD²),

(5-2)

where, for acute exposures, G → 1, and for very long exposures, G → 0. In this context, “acute” and “long” are deﬁned relative to the half time for repair of sublethal damage (T1/2). In general, the G factor in Eq. (5-2) will depend on the details of the temporal distribution of the dose, as well as on T1/2. As discussed before, for many simple cases, G can be calculated analytically. For example, for n short, equal fractions, where the separation between fractions is much longer than T1/2, then G ≈ 1/n: S = exp(−αD − βD²/n),

(5-3)

Formulae for G for many other standard radiotherapeutic regimens have also been derived [75, 76], as has a general formalism for any possible regime [78]. It is important to note the mechanistic basis of Eq. (5-2) so that it is not simply an equation that happens to ﬁt cellular survival curves. It has been suggested, for example, that the L-Q model can be considered simply as the ﬁrst two terms (i.e., dose and dose squared) of a power-series expansion [79]. If L-Q were just another empirical model, there would be no good reason for considering the linear dose term to be independent of protraction/fractionation, and the quadratic term in dose to be fractionation dependent. This distinction between the linear

radiobiological principles underlying stereotactic radiation therapy

and the quadratic terms is at the heart of the L-Q model and its application. Based on Eq. (5-2), we can equate schemes; that is, produce a regimen with the same, say, tumor response, as a “tried and tested” regimen. Thus, assuming tumor repopulation (see above) is negligible (which will be true for a 5- or 10-fraction treatment), to match a new fractionation scheme (labeled 2, with n2 fractions) to a given (“old”) fractionation scheme (labeled 1, with n1 fractions), we can calculate the dose (D2) in scheme 2 such that αD2 + βD22/n2 = αD1 + βD12/n1.

57

60 multi-fractioned dose (Gy)

5.

50 40 30 20 10 0

And if the “tried and tested” regime is a single fraction (i.e., n1 = 1) then, dividing by β, we have (α/β)D2 + D22/n2 = (α/β)D1 + D12, which is a simple quadratic equation, easily solved to yield D2, assuming a given value of α/β. Some equivalences are shown in Table 5-1 and Figure 5-9 [13], and similar results have recently been published by Liu et al. [80]. It is important to be cautious in using isoeffect calculations to extrapolate from single-fraction radiosurgical protocols of 15 to 20 Gy to more radiobiologically appropriate multifractioned regimens. Indeed, the results of all isoeffect calculations in radiotherapy should not be accepted uncritically. There is, however, good evidence that the results are reasonable. Figure 5-10, for example, shows some isoeffect results from Van der Kogel [81] for late-responding damage to the rat spinal cord and from Douglas and Fowler [82] for acute damage in mouse skin. The form of the plots is such that, if the L-Q formalism applies, the data would fall on a straight line [82]. Although there are more sophisticated methods available for assessing agreement with the L-Q model [83], given the inherent uncer-

9

8

7 6 5 #o 4 f fra 3 ctio ns

2 1

25 21 23 19 ) 17 Gy 15 se ( 13 n do o i 11 t ac le fr sing

FIGURE 5-9. Gamma-ray doses for multifractioned stereotactic radiotherapy, which are calculated to be equivalent, in terms of tumor control, to single-fractioned radiosurgical doses. See also Table 5-1. (Reprinted from Hall EJ, Brenner DJ. The radiobiology of radiosurgery: rationale for different treatment regimes for AVMs and malignancies. Int J Radiat Oncol Biol Phys 1993; 25:381–385. Copyright 1993, with permission from Elsevier.)

tainties in the data, it is clear that all these data, including those for single fractions of ∼20 Gy, are consistent with the L-Q model. Although it is important to be appropriately critical of the L-Q model and its application to radiotherapy, it is equally important to recognize that it is the best model we have. It is a

TABLE 5-1. Total gamma-ray doses for multifractioned stereotactic radiotherapy, which are calculated [13] to be equivalent, in terms of tumor control, to single-fractioned radiosurgical doses (see also Fig. 5-9). Single-fraction dose (Gy)

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Fx = 2 (Gy)

Fx = 3 (Gy)

Fx = 5 (Gy)

Fx = 9 (Gy)

12.7 14.0 15.4 16.8 18.1 19.5 20.9 22.3 23.6 25.0 26.4 27.8 29.2 30.6 32.0 33.4

14.3 15.9 17.6 19.2 20.8 22.5 24.1 25.8 27.4 29.1 30.8 32.5 34.1 35.8 37.5 39.2

16.5 18.4 20.4 22.4 24.4 26.5 28.5 30.6 32.6 34.7 36.8 38.9 41.0 43.2 45.3 47.4

18.9 21.3 23.7 26.2 28.7 31.2 33.8 36.4 39.0 41.6 44.3 47.0 49.7 52.4 55.1 57.8

FIGURE 5-10. Isoeffect data for late response from three (䊐, 䊊, 䉭) different regions of the rat spinal cord [80], and for acute skin reactions (䉬) in mice [81]. All the points at ≥20 Gy/fraction correspond with single acute exposures. The data are plotted in a form such that, if they follow a linear-quadratic relationship, the points would fall on a straight line. The highlighted data points, which are for single-fractioned exposures of 20 to 25 Gy, are clearly consistent with the linear-quadratic formalism. (Reprinted from Hall EJ, Brenner DJ. In response to Dr. Marks. Int J Radiat Oncol Biol Phys 1995; 32:275–276. Copyright 1995, with permission from Elsevier.)

58

d.j. brenner

mechanistically based model of cell killing [84], and there is a wealth of evidence [18, 22] that cell killing dominates all radiotherapeutic response, including, for example, late vascular damage. Of course, the simplest form of the L-Q model is not necessarily the most appropriate to apply. When repopulation is important, the L-Q formalism can be appropriately modiﬁed [29, 85]. If redistribution or reoxygenation is important, the L-Q formalism can again be appropriately modiﬁed [86]. Similarly, if there is evidence that the L-Q model is underpredicting survival at high doses (though from results such as in Fig. 5-10, it appears that this is not the case at doses ≤20 Gy), appropriate saturation-related modiﬁcations to the L-Q formalism have been described [87].

Conclusion Malignant Tumors • Radiobiological and clinical arguments, based on the need to overcome hypoxia and the potential for a differential response between the tumor and the surrounding normal tissue, strongly suggest that fractionation will result in improved outcome. • For most malignancies, even when good dose distributions can be obtained, single-fraction radiation therapy has not been in general use for more than 70 years. • For most sites, only a few fractions (5 to 10) would be optimal for treating malignant tumors. • Modern technology has made fractionated stereotactic radiotherapy increasingly practical.

Arteriovenous Malformations • The arguments in favor of fractionation probably do not hold, and single-fraction treatment is generally optimal. • There is no hypoxic cell issue to overcome, and both the AVM and the surrounding normal tissue probably respond the same way to changes in fractionation.

Benign Tumors • There is some evidence that benign tumors respond to fractionation like a late-responding normal tissue, in which case fractionation would not improve outcome, unless hypoxic cells were an issue. • Overall, the radiobiological evidence is probably insufﬁcient to make recommendations on appropriate fractionation schemes for benign tumors.

References 1. Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two cases of intractable pain. Acta Chir Scand 1968; 134:585–595. 2. Lunsford LD, Flickinger JC, Larson D. Regarding: Rosenthal DI, Glatstein E. “We’ve Got a Treatment, but What’s the Disease?” Oncologist 1996; 1. Oncologist 1997; 2:59–61. 3. Adler JR Jr, Colombo F, Heilbrun MB, et al. Toward an expanded view of radiosurgery. Neurosurgery 2004; 55:1374–1376.

4. Colombo F, Benedetti A, Pozza F, et al. External stereotactic irradiation by linear accelerator. Neurosurgery 1985; 16:154–160. 5. Houdek PV, Fayos JV, Van Buren JM, et al. Stereotaxic radiotherapy technique for small intracranial lesions. Med Phys 1985; 12:469–472. 6. Adler JR Jr, Chang SD, Murphy MJ, et al. The CyberKnife: a frameless robotic system for radiosurgery. Stereotact Funct Neurosurg 1997; 69:124–128. 7. Benedict SH, Cardinale RM, Wu Q, et al. Intensity-modulated stereotactic radiosurgery using dynamic micro-multileaf collimation. Int J Radiat Oncol Biol Phys 2001; 50:751–758. 8. Leavitt DD, Watson G, Tobler M, et al. Intensity-modulated radiosurgery/radiotherapy using a micromultileaf collimator. Med Dosim 2001; 26:143–150. 9. Pozza F, Colombo F, Chierego G, et al. Low-grade astrocytomas: treatment with unconventionally fractionated external beam stereotactic radiation therapy. Radiology 1989; 171:565–569. 10. Heifetz MD, Whiting J, Bernstein H, et al. Stereotactic radiosurgery for fractionated radiation: a proposal applicable to linear accelerator and proton beam programs. Stereotact Funct Neurosurg 1989; 53:167–177. 11. Hariz MI, Henriksson R, Lofroth PO, et al. A non-invasive method for fractionated stereotactic irradiation of brain tumors with linear accelerator. Radiother Oncol 1990; 17:57–72. 12. Delannes M, Daly NJ, Bonnet J, et al. Fractionated radiotherapy of small inoperable lesions of the brain using a non-invasive stereotactic frame. Int J Radiat Oncol Biol Phys 1991; 21:749– 755. 13. Brenner DJ, Martel MK, Hall EJ. Fractionated regimens for stereotactic radiotherapy of recurrent tumors in the brain. Int J Radiat Oncol Biol Phys 1991; 21:819–824. 14. Souhami L, Olivier A, Podgorsak EB, et al. Fractionated stereotactic radiation therapy for intracranial tumors. Cancer 1991; 68: 2101–2108. 15. Schwade JG, Houdek PV, Landy HJ, et al. Small-ﬁeld stereotactic external-beam radiation therapy of intracranial lesions: fractionated treatment with a ﬁxed-halo immobilization device. Radiology 1990; 176:563–565. 16. Gill SS, Thomas DG, Warrington AP, et al. Relocatable frame for stereotactic external beam radiotherapy. Int J Radiat Oncol Biol Phys 1991; 20:599–603. 17. Tomé WA, Mehta MP, Meeks SL, et al. Fractionated stereotactic radiotherapy: a short review. Technol Cancer Res Treat 2002; 1: 153–172. 18. Thames HD, Hendry JH. Fractionation in Radiotherapy. London: Taylor & Francis, 1987. 19. Overgaard J, Horsman MR. Modiﬁcation of hypoxia-induced radioresistance in tumors by the use of oxygen and sensitizers. Semin Radiat Oncol 1996; 6:10–21. 20. Powers WE, Tolmach LJ. A multicomponent x-ray survival curve for mouse lymphosarcoma cells irradiated in vivo. Nature 1963; 197:710–711. 21. Leith JT, Cook S, Chougule P, et al. Intrinsic and extrinsic characteristics of human tumors relevant to radiosurgery: comparative cellular radiosensitivity and hypoxic percentages. Acta Neurochir Suppl 1994; 62:18–27. 22. Hall EJ. Radiobiology for the Radiologist. Philadelphia: Lippincott Williams & Wilkins, 2000. 23. Bedford JS, Mitchell JB. Dose-rate effects in synchronous mammalian cells in culture. Radiat Res 1973; 54:316–327. 24. Brenner DJ, Hall EJ. Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy. Int J Radiat Oncol Biol Phys 1991; 20:181–190. 25. Fowler JF. The linear-quadratic formula and progress in fractionated radiotherapy. Br J Radiol 1989; 62:679–694.

5.

radiobiological principles underlying stereotactic radiation therapy

26. Thames HD, Bentzen SM, Turesson I, et al. Fractionation parameters for human tissues and tumors. Int J Radiat Biol 1989; 56: 701–710. 27. Withers HR, Thames HD Jr, Flow BL, et al. The relationship of acute to late skin injury in 2 and 5 fraction/week gamma-ray therapy. Int J Radiat Oncol Biol Phys 1978; 4:595–601. 28. Withers HR, Taylor JM, Maciejewski B. The hazard of accelerated tumor clonogen repopulation during radiotherapy. Acta Oncol 1988; 27:131–146. 29. Brenner DJ. Accelerated repopulation during radiotherapy— evidence for delayed onset. Radiat Oncol Invest 1993; 1:167– 172. 30. Slevin NJ, Hendry JH, Roberts SA, et al. The effect of increasing the treatment time beyond three weeks on the control of T2 and T3 laryngeal cancer using radiotherapy. Radiother Oncol 1992; 24:215–220. 31. Saunders MI, Dische S, Grosch EJ, et al. Experience with CHART. Int J Radiat Oncol Biol Phys 1991; 21:871–878. 32. Hall EJ, Brenner DJ. The radiobiology of radiosurgery: rationale for different treatment regimes for AVMs and malignancies. Int J Radiat Oncol Biol Phys 1993; 25:381–385. 33. Loefﬂer JS, Kooy HM, Wen PY, et al. The treatment of recurrent brain metastases with stereotactic radiosurgery. J Clin Oncol 1990; 8:576–582. 34. Ogilvy CS. Radiation therapy for arteriovenous malformations: a review. Neurosurgery 1990; 26:725–735. 35. Kjellberg RN, Hanamura T, Davis KR, et al. Bragg-peak protonbeam therapy for arteriovenous malformations of the brain. N Engl J Med 1983; 309:269–274. 36. Marks LB, Spencer DP. The inﬂuence of volume on the tolerance of the brain to radiosurgery. J Neurosurg 1991; 75:177–180. 37. Nedzi LA, Kooy H, Alexander III E, et al. Variables associated with the development of complications from radiosurgery of intracranial tumors. Int J Radiat Oncol Biol Phys 1991; 21:591– 599. 38. Sheline GE, Wara WM, Smith V. Therapeutic irradiation and brain injury. Int J Radiat Oncol Biol Phys 1980; 6:1215–1228. 39. Wowra B, Schmitt HP, Sturm V. Incidence of late radiation necrosis with transient mass effect after interstitial low dose rate radiotherapy for cerebral gliomas. Acta Neurochir (Wien) 1989; 99:104–108. 40. Touboul E, Al Halabi A, Buffat L, et al. Single-fraction stereotactic radiotherapy: a dose-response analysis of arteriovenous malformation obliteration. Int J Radiat Oncol Biol Phys 1998; 41:855–861. 41. Flickinger JC, Kondziolka D, Maitz AH, et al. An analysis of the dose-response for arteriovenous malformation radiosurgery and other factors affecting obliteration. Radiother Oncol 2002; 63:347– 354. 42. Kocher M, Wilms M, Makoski HB, et al. Alpha/beta ratio for arteriovenous malformations estimated from obliteration rates after fractionated and single-dose irradiation. Radiother Oncol 2004; 71:109–114. 43. Shrieve DC, Hazard L, Boucher L, et al. Dose fractionation in stereotactic radiotherapy for parasellar meningiomas: radiobiological considerations of efﬁcacy and optic nerve tolerance. J Neurosurg 2004; 101(Suppl 3):390–395. 44. Goldsmith BJ, Rosenthal SA, Wara WM, et al. Optic neuropathy after irradiation of meningioma. Radiology 1992; 185:71–76. 45. Bhandare N, Monroe AT, Morris CG, et al. Does altered fractionation inﬂuence the risk of radiation-induced optic neuropathy? Int J Radiat Oncol Biol Phys 2005; 62:1070–1077. 46. Brenner DJ, Hall EJ. Stereotactic radiotherapy of intracranial tumors—an ideal candidate for accelerated treatment. Int J Radiat Oncol Biol Phys 1994; 28:1039–1041; discussion 1047.

59

47. Kooy HM, Dunbar SF, Tarbell NJ, et al. Adaptation and veriﬁcation of the relocatable Gill-Thomas-Cosman frame in stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 1994; 30:685–691. 48. Kitchen ND, Thomas DG. Minimally invasive stereotaxy: clinical use of the Gill-Thomas-Cosman (GTC) repeat stereotactic localiser. Minim Invasive Neurosurg 1994; 37:61–63. 49. Fairclough-Tompa L, Larsen T, Jaywant SM. Immobilization in stereotactic radiotherapy: the head and neck localizer frame. Med Dosim 2001; 26:267–273. 50. Burton KE, Thomas SJ, Whitney D, et al. Accuracy of a relocatable stereotactic radiotherapy head frame evaluated by use of a depth helmet. Clin Oncol (R Coll Radiol) 2002; 14:31–39. 51. Kumar S, Burke K, Nalder C, et al. Treatment accuracy of fractionated stereotactic radiotherapy. Radiother Oncol 2005; 74:53–59. 52. Miranpuri AS, Tome WA, Paliwal BR, et al. Assessment of patient-independent intrinsic error for a noninvasive frame for fractionated stereotactic radiotherapy. Int J Cancer 2001; 96:320–325. 53. Kalapurakal JA, Ilahi Z, Kepka AG, et al. Repositioning accuracy with the Laitinen frame for fractionated stereotactic radiation therapy in adult and pediatric brain tumors: preliminary report. Radiology 2001; 218:157–161. 54. Ashamalla H, Addeo D, Ikoro NC, et al. Commissioning and clinical results utilizing the Gildenberg-Laitinen adapter device for X-ray in fractionated stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 2003; 56:592–598. 55. Alheit H, Dornfeld S, Dawel M, et al. Patient position reproducibility in fractionated stereotactically guided conformal radiotherapy using the BrainLab mask system. Strahlenther Onkol 2001; 177:264–268. 56. Lopatta E, Liesenfeld SM, Bank P, et al. Improved patient repositioning accuracy by integrating an additional jaw ﬁxation into a high precision face mask system in stereotactic radiotherapy of the head. Strahlenther Onkol 2003; 179:571–575. 57. Salter BJ, Fuss F, Vollmer DG, et al. The TALON removable head frame system for stereotactic radiosurgery/radiotherapy: measurement of the repositioning accuracy. Int J Radiat Oncol Biol Phys 2001; 51:555–562. 58. Walton L, Hampshire A, Roper A, et al. Development of a relocatable frame technique for gamma knife radiosurgery. Technical note. J Neurosurg 2000; 93(Suppl 3):198–202. 59. Kai J, Shiomi H, Sasama T, et al. Optical high-precision threedimensional position measurement system suitable for head motion tracking in frameless stereotactic radiosurgery. Comput Aided Surg 1998; 3:257–263. 60. Kamath R, Ryken TC, Meeks SL, et al. Initial clinical experience with frameless radiosurgery for patients with intracranial metastases. Int J Radiat Oncol Biol Phys 2005; 61:1467–1472. 61. Meeks SL, Bova FJ, Wagner TH, et al. Image localization for frameless stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 2000; 46:1291–1299. 62. Mehta VK, Lee QT, Chang SD, et al. Image guided stereotactic radiosurgery for lesions in proximity to the anterior visual pathways: a preliminary report. Technol Cancer Res Treat 2002; 1:173–180. 63. Ishihara H, Saito K, Nishizaki T, et al. CyberKnife radiosurgery for vestibular schwannoma. Minim Invasive Neurosurg 2004; 47: 290–293. 64. Pham CJ, Chang SD, Gibbs IC, et al. Preliminary visual ﬁeld preservation after staged CyberKnife radiosurgery for perioptic lesions. Neurosurgery 2004; 54:799–810; discussion 810–792. 65. Kajiwara K, Saito K, Yoshikawa K, et al. Image-guided stereotactic radiosurgery with the CyberKnife for pituitary adenomas. Minim Invasive Neurosurg 2005; 48:91–96.

60

d.j. brenner

66. Chang SD, Gibbs IC, Sakamoto GT, et al. Staged stereotactic irradiation for acoustic neuroma. Neurosurgery 2005; 56:1254– 1261. 67. Xia P, Geis P, Xing L, et al. Physical characteristics of a miniature multileaf collimator. Med Phys 1999; 26:65–70. 68. Cosgrove VP, Jahn U, Pfaender M, et al. Commissioning of a micro multi-leaf collimator and planning system for stereotactic radiosurgery. Radiother Oncol 1999; 50:325–336. 69. Georg D, Dieckmann K, Bogner J, et al. Impact of a micromultileaf collimator on stereotactic radiotherapy of uveal melanoma. Int J Radiat Oncol Biol Phys 2003; 55:881–891. 70. Baumert BG, Norton IA, Davis JB. Intensity-modulated stereotactic radiotherapy vs. stereotactic conformal radiotherapy for the treatment of meningioma located predominantly in the skull base. Int J Radiat Oncol Biol Phys 2003; 57:580–592. 71. Tobler M, Leavitt DD, Watson G. Optimization of the primary collimator settings for fractionated IMRT stereotactic radiotherapy. Med Dosim 2004; 29:72–79. 72. Jin JY, Yin FF, Ryu S, et al. Dosimetric study using different leaf-width MLCs for treatment planning of dynamic conformal arcs and intensity-modulated radiosurgery. Med Phys 2005; 32:405– 411. 73. Fuss M, Salter BJ, Sadeghi A, et al. Fractionated stereotactic intensity-modulated radiotherapy (FS-IMRT) for small acoustic neuromas. Med Dosim 2002; 27:147–154. 74. Lea DE. Actions of Radiation on Living Cells. Cambridge: Cambridge University Press, 1946. 75. Lea DE, Catcheside DG. The mechanism of the induction by radiation of chromosome aberrations in tradescantia. J Genet 1942; 44:216–245. 76. Brenner DJ, Herbert DE. The use of the linear-quadratic model in clinical radiation oncology can be defended on the basis of empirical evidence and theoretical argument. Med Phys 1997; 24:1245–1248.

77. Dale RG. The application of the linear-quadratic dose-effect equation to fractionated and protracted radiotherapy. Br J Radiol 1985; 58:515–528. 78. Brenner BJ, Huang Y, Hall EJ. Fractionated high dose-rate versus low dose-rate regimens for intracavitary brachytherapy of the cervix: equivalent regimens for combined brachytherapy and external irradiation. Int J Radiat Oncol Biol Phys 1991; 21:1415–1423. 79. Yaes RJ, Patel P, Maruyama Y. On using the linear-quadratic model in daily clinical practice. Int J Radiat Oncol Biol Phys 1991; 20:1353–1362. 80. Liu L, Bassano DA, Prasad SC, et al. The linear-quadratic model and fractionated stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 2003; 57:827–832. 81. van der Kogel AJ. Chronic effects of neutrons and charged particles on spinal cord, lung, and rectum. Radiat Res Suppl 1985; 8: S208–216. 82. Douglas BG, Fowler JF. The effect of multiple small doses of x rays on skin reactions in the mouse and a basic interpretation. Radiat Res 1976; 66:401–426. 83. Tucker SL. Tests for the ﬁt of the linear-quadratic model to radiation isoeffect data. Int J Radiat Oncol Biol Phys 1984; 10:1933– 1939. 84. Sachs RK, Hahnfeld P, Brenner DJ. The link between low-LET dose-response relations and the underlying kinetics of damage production/repair/misrepair. Int J Radiat Biol 1997; 72:351–374. 85. Travis EL, Tucker SL. Isoeffect models and fractionated radiation therapy. Int J Radiat Oncol Biol Phys 1987; 13:283–287. 86. Brenner DJ, Hlatky LR, Hahnfeldt PJ, et al. A convenient extension of the linear-quadratic model to include redistribution and reoxygenation. Int J Radiat Oncol Biol Phys 1995; 32:379–390. 87. Zaider M, Rossi HH. Saturation effects for sparsely ionizing particles. New York: Columbia University, Radiological Research Laboratory Annual Report (COO-4733-3:126-134), 1980.

6

Experimental Radiosurgery Models Ajay Niranjan and Douglas Kondziolka

Introduction The ﬁeld of stereotactic radiosurgery represents one of the fundamental shifts in neurologic surgery over the past two decades. Compared with conventional invasive surgery techniques, radiosurgery is minimally invasive and relies on biological response of tissues in order to eradicate or inactivate them. Radiosurgery is conceptually different from fractionated radiation therapy. The efﬁcacy of large-ﬁeld fractionated radiotherapy to treat brain tumors is dependent on biological differences between normal and tumor cells. Fractionated radiotherapy exploits these differences to limit the risk of normal tissue injury in patients with malignant brain tumors, thus it can increase the therapeutic ratio, which is equivalent to the rate of tumor control divided by the rate of complications. Radiosurgery, in contrast with conventional radiotherapy, uses a single high dose of radiation. Normal tissue effects are limited by the highly focused nature of the radiosurgical beams. In addition, unlike radiotherapy, radiosurgery manages smallvolume targets using much higher doses. Finally, whereas fractionated radiotherapy is generally most effective in killing rapidly dividing cells, radiosurgery induces biological responses irrespective of the mitotic activity, oxygenation, and inherent radiosensitivity of target cells. Considering the unique biological response of tissues to radiosurgery, it is important to study the biological effects of radiosurgery in both normal and pathologic nervous system tissues in animal models. Information gained from radiosurgical research studies would be useful in devising strategies to avoid, prevent, or ameliorate damage to normal tissue without compromising treatment efﬁcacy. As radiosurgery evolves from a treatment speciﬁcally for brain tumors into a widely available treatment modality for a variety of intracranial lesions, understanding of biological responses using animal investigations becomes crucial.

proton beam from a 230-cm synchrocyclotron) [1]. The early histologic results (3rd to 8th day) showed complete transection of the rabbit spinal cord with 40,000 rads (400 Gy) using 1.5-mm beam diameter and with 200 Gy using 10-mm beam diameter. These investigators also used a goat brain model to document sharply deﬁned lesions in deep parts of brain 4 to 7 weeks after 200 Gy of stereotactic multiple-port proton beam radiation. Rexed et al. performed proton beam radiosurgery on rabbit brain to study the long-term effects (2 to 56 weeks) of irradiation [2]. Using a 1.5-mm collimator, 20,000 rads (200 Gy) was delivered to the anterior part of rabbit brain. Serial histology up to 3 months showed a well-demarcated lesion in the path of the beam. After 3 months, a lesion broader than the beam size was noted. Leksell et al. investigated the features of a radiolesion in the depth of brain produced by cross-ﬁred irradiation with a narrow beam of high-energy [3]. Their results showed that with 200 Gy, well-circumscribed intracerebral lesions of appropriate size and shape could be created. Andersson et al. performed protons radiosurgery on goat brain to study the late histologic effects of the cross-ﬁred beams [4]. These investigators did not detect untoward changes in or around the lesion (e.g., elements resembling neoplasm, hemorrhage, or telangiectasis) 1.5 to 4 years after 200 Gy radiosurgery. Nilsson et al. irradiated (100 to 300 Gy) the basilar artery of cats by stereotactic technique using a 179-source cobalt-60 prototype gamma unit [5]. Histology demonstrated vascular lesions such as vacuolization, degeneration, and desquamation of the endothelium and hyalinization and necrosis of the muscular coat. The reparatory reactions were relatively sparse, and thrombosis was completely absent. These investigations demonstrated that radiosurgery could potentially be used to create sharply deﬁned lesions in deep parts of the brain (Table 6-1).

Experimental Models for Investigations into Effects of Normal Brain Radiosurgery Experimental Models to Investigate Radiosurgery as a Neurosurgical Tool In the initial studies involving focused radiation as a neurosurgical tool, rabbit and goat central nervous system (CNS) models were used. Larsson et al. used a rabbit spinal cord model to investigate the effect of the high-energy proton beam (185-MeV

The effect of fractionated radiation therapy on the nervous system depends on both the dose delivered and the time elapsed [6–8]. Although all brain regions are affected, the radiation response tends to be most severe in white-matter regions [9–11]. A dose-related variable latency period after irradiation can last from months to years [6–8]. A single radiation dose of 20 Gy to

61

62

a. niranjan and d. kondziolka

TABLE 6-1. Radiosurgery as a neurosurgical tool. First author and year

Maximum dose (Gy)

Region(s) irradiated

Irradiation technique

Animal model

Results

Larsson, 1958 [1]

20 krad

Internal capsule

Proton beam

Rabbit and goat

Rexed, 1960 [2]

20 krad

Forebrain

Proton beam

Rabbit

Leksell, 1960 [3]

20–38 krad

Internal capsule

Proton beam

Goat

Anderson, 1970 [4]

15–20 krad

Basal ganglia, internal capsule, optic chiasm, optic tract

Proton beam

Goat

Stereotactic proton beam radiation can produce restricted brain and spinal cord lesions in 4–6 weeks High-energy proton irradiation can cause a welldeﬁned lesion in 2–56 weeks Well-circumscribed radiolesions of appropriate size and shape can be produced with a suitable dose (20 krad) No untoward late effects (neoplasia, telangiectasis, hemorrhage) were noted 1.5 to 4 years after proton beam radiosurgery

rat brain creates lesions that are primarily conﬁned to the vasculature at latencies of more than 12 months [12, 13]. Radiation-induced vascular changes in the CNS include perivascular ﬁbrosis and ﬁbrinoid necrosis of vessel walls, hyaline degeneration, edema, telangiectasia, thrombosis, and hemorrhage [14]. At higher doses of around 25 Gy, white-matter lesions predominate in irradiated rat brain at latencies of less than 12 months. Radiation-induced lesions in the white matter can range from demyelination to malacia [14]. Radiosurgery models used to study the biologic response of radiosurgery on the CNS are listed in Table 6-1. In initial studies, Lunsford et al. [15] and Kondziolka et al. [16] studied the radiobiological effects of stereotactic radiosurgery using a baboon model. These investigators delivered a central dose of 150 Gy using an 8-mm collimator to the caudate, thalamus, or pons regions using the Gamma Knife. No changes were noted by computed tomography (CT) or by T1-weighted, T2-weighted, and gadolinium-enhanced magnetic resonance imaging (MRI) at 4 weeks after irradiation. A circumscribed, contrast-enhanced lesion was seen by 6 to 8 weeks, which was characterized by demyelination, microvascular damage, hemorrhage, and astrocytosis. The edema was ﬁrst evident at 8 weeks. Frank necrosis appeared in the irradiated region by 24 weeks. Kondziolka et al. [17] used a rat brain model to study the histologic changes in rat brain 90 days after radiosurgery (ﬁxed latency, variable dose). The frontal lobe of rats was irradiated with maximal doses of 30 to 200 Gy using a 4-mm collimator. Detectable histologic alterations were noted with doses of more than 70 Gy. Necrosis was seen only in tissues irradiated with more than 100 Gy. Blatt et al. [18] evaluated serial tissue changes after 125 Gy linear accelerator (linac) radiosurgery of internal capsule of cats (variable latency, ﬁxed dose). MRI and histopathologic evaluations were performed serially for 1 year starting at 3.5 weeks after irradiation. Tissue necrosis was evident in the cat brain by 3.5 weeks and was accompanied by vascular proliferation and edema. The lesions showed increased vascularity and microglial inﬁltration that resolved by 12 to 29 weeks. In a study evaluating the effects of radiation dose and time after treatment on the radiosensitivity of brain, Kamiryo et al. irradiated rat brain with maximum doses of 50, 75, or 120 Gy and

analyzed for histologic changes and blood-brain barrier integrity up to 12 months (variable latency, variable dose) [19], Whereas higher doses (120 Gy) induced alterations in astrocytic morphology by just 3 days after treatment, such changes were not observed until 3 months with lower doses (50 Gy). Bloodbrain barrier breakdown as assessed by Evans blue leakage was evident within 3 weeks of 120-Gy irradiation but was not seen across 12 months after 50 Gy. These ﬁndings indicate that the latent period between irradiation and detection of pathologic alterations is dependent on both the dose and the biological end point used. Such ﬁndings are consistent with the results of studies using a more conventional radiation source (60Co) to irradiate the rat spinal cord. In this model of radiation-induced CNS injury, latency to paralysis after irradiation of an 8- or 16mm segment of cervical spinal cord decreased as dose increased [20]. Also, the ED50 for paralysis after 4 mm of spinal cord irradiation was 51 Gy, whereas the ED50 for vascular damage was only 25.6 Gy. The impact of dose and biological end points on latency was also reported by Karger et al. [21], who evaluated the rat brain using T1- and T2-weighted, gadoliniumenhanced MRI at 15, 17, or 20 months after treatment with 26 to 50 Gy of linear accelerator–based radiosurgery. A 3-mm collimator was employed to deliver the radiosurgical dose using a convergent arc technique and resulted in an 80% isodose distribution of 4.7 mm in diameter. No radiation-induced affect was noted on MRI at any time point for doses less than 30 Gy. After 40-Gy radiosurgery, the latency of detectable MRI changes was approximately 19 to 20 weeks, whereas the latency after 50 Gy was 15 to 16 weeks. In addition, T1-weighted changes in the MRI signal had a shorter latency than T2-weighted changes. Considering that the changes in T1-weighted images are due to leakage of gadolinium-DTPA across the blood-brain barrier, the results of this study point to the likely role of vascular damage in radiation-induced injury. The importance of the vasculature in radiation-induced brain injury is well recognized; a prevalent hypothesis regarding the pathogenesis after conventional radiotherapy is that damage to capillary endothelium and/or supporting cells ultimately interrupts blood ﬂow resulting in secondary ischemic necrosis. In a report focusing on vascular changes after a maximum dose of 75 Gy delivered

6.

63

experimental radiosurgery models

TABLE 6-2. Dose and latency effects of radiosurgery on central nervous system. First author and year

Maximum dose (Gy)

Region(s) irradiated

Irradiation technique

Animal model

Results

Lunsford, 1990 [15]

150

Caudate, thalamus, pons R. frontal lobe

Gamma Knife, 8-mm collimator Gamma Knife, 4-mm collimator

Baboon

Internal capsule

Linac, 10-mm collimator

Cat

75

Parietal cortex

Gamma Knife, 4-mm collimator

Rat

26–50

Parietal cortex

Linac, 3-mm collimator

Rat

MRI and histology showed lesion 45–60 days posttreatment Histopathology at 90 days showed tissue changes at doses lower than 60 Gy, necrosis at doses of 100 Gy MRI and serial histopathology showed mass effect and neurologic deﬁcits at 3.5–4.5 weeks, some necrosis 12–29 weeks, and late resorption of necrosis Electron microscopy at 3.5 months showed decreased vascularity and increased capillary diameter in irradiated regions; basement membrane changes precede vascular damage MRI showed contrast enhancement at 15 weeks after 50-Gy and 19 weeks after 40Gy radiosurgery

Kondziolka, 1992 [17]

Blatt, 1994 [18]

Kamiryo, 2001 [22]

Karger, 2002 [21]

30–200

149

to the rat brain using a Gamma Knife, it was noted that vascular changes, speciﬁcally alterations in the basement membrane, preceded changes in necrosis [22]. This ﬁnding suggests that vascular damage is also an important component in biological response after radiosurgery. Although radiosurgery generally involves the use of higher single doses and smaller treatment volumes than conventional irradiation, the histologic effects of these two methodologies appear similar. The biggest differences are that the latency period after radiosurgery is shorter, and the major histologic ﬁnding is vascular damage (Table 6-2).

Experimental Models Exploring Strategies to Enhance Tumoricidal Effect of Radiosurgery Although benign tumor radiosurgery is associated with high tumor control rates, malignant glial tumors often recur. Additional strategies to improve cell kill of malignant brain tumors and to protect normal surrounding brain tissue are needed. A few strategies for radioprotection of normal tissue and radiosensitization of tumor tissue have already been explored.

Rat

breaks and initiate lipid peroxidation of vascular membranes, ultimately leading to membrane lysis and cell death. As a lipid antioxidant and free radical scavenger, 21-AS inhibit oxygen radical–initiated peroxidation of vascular membrane. 21Aminosteroids also block the release of free arachidonic acid from cell membrane, thereby inhibiting activation of the proinﬂammatory cyclooxygenase pathway. These properties of 21AS are thought to protect cerebral vessels from injury and prevent cerebral edema. The effects of the 21-AS compound U-74389G on radiation injury have been evaluated in both rat and cat models. Bernstein et al. [26] reported that U-74389G reduced brachytherapy-induced brain injury in the rat. Buatti et al. [27] found that this same agent also protected the cat brain from injury due to radiosurgery and was signiﬁcantly more effective than corticosteroids. In our own studies, 15 mg/kg but not 5 mg/kg U-74389G was effective at reducing brain injury in the rat when administered 1 hour before radiosurgery. U74389G ameliorated vasculopathy and regional edema and delayed the onset of necrosis while gliosis was unaffected [28]. Preliminary results suggest that this agent may act through reduction of the cytokines induced by brain irradiation. An alternative strategy for radiation protection seeks to repair radiation-induced brain damage.

Radiation Protection and Repair The initial strategies included use of cerebral protective agents while delivering a high dose to tumor cells. Oldﬁeld et al. [23] noted protection from radiation-induced brain injury using pentobarbital. The 21-aminosteroids (21-AS) have been evaluated as potential radioprotective agents. The 21-AS, commonly known as lazaroids, have been advocated as cerebral protective agents in patients with head trauma or subarachnoid hemorrhage [24]. The 21-AS act as antioxidants [25], and much of the damage from radiation is due to the production of oxygen free radicals, which can induce DNA modiﬁcations and strand

Radiation Potentiation We studied the synergistic effect of tumor necrosis factor-α (TNF-α) on enhancing the tumor response to radiosurgery. TNF-α can act as a tumoricidal agent with direct cytotoxicity mediated through binding to its cognate cell-surface receptors and a variety of activities triggering a multifaceted immune attack on tumors [29–34]. In addition, locally produced TNF-α has been reported to enhance the sensitivity of tumors to radiation in nude mice [29]. We employed a replication-defective herpes simplex virus (HSV) as a vector to deliver thymidine

64

a. niranjan and d. kondziolka

TABLE 6-3. Effects of radiosurgery on malignant brain tumors. First author and year

Tumor margin dose (Gy)

Tumor model

Animal model

Irradiation technique

Experimental treatment

Results

Kondziolka, 1992 [38]

15–50

C6 glioma

Rat

Gamma Knife, 4-mm collimator

Radiosurgery

Kondziolka, 1999 [39]

35 Gy

C6 glioma

Rat

Gamma Knife, 4-mm collimator

Radiosurgery + 21-aminosteroid

Niranjan, 2000 [35]

15

U 87 MG

Nude mouse

Gamma Knife, 4-mm collimator

Radiosurgery + HSV TK-GCV+TNF

Nakahara, 2001 [40]

16

MADB 106 cells

Rat

Gamma Knife, 4-mm collimator

Radiosurgery + cytokine transduced tumor cell vaccine

Niranjan, 2003 [41]

15

9L gliosarcoma

Rat

Gamma Knife, 4-mm collimator

Radiosurgery + HSV TK-GCV+TNF+ Connexin

Treated animals survived 39 days (control 29 days). Treated tumors had hypocellular appearance with cellular edema. 21-Aminosteroid exhibited a radioprotectant effect on normal brain tissue, but did not protect the tumor The combination treatment enhanced median survival (75 days) with 89% animal surviving The combination treatment signiﬁcantly prolonged animal survival and protected animals from a subsequent challenge by parental tumor cells placed in the CNS The combination of radiosurgery and multigene therapy enhanced median animal survival (150 days) with 75% of animals surviving

kinase (TK) and/or TNF-α genes to U-87 MG tumors in nude mice. Radiosurgery was performed 48 hours after gene transfer using 15 Gy to the tumor margin (21.4 Gy to the center). Daily ganciclovir (GCV) therapy was started after gene transfer and continued for 10 days. The combination of radiosurgery with TNF-α or with HSV-TK-GCV (suicide gene therapy) and TNFα signiﬁcantly improved median survival of animals [35]. In additional experiments, the connexin-43 gene was added to enhance the formation of gap junctions between tumor cells, which should facilitate the intercellular dissemination of TKactivated GCV from virus-infected cells to noninfected surrounding cells. This creates a bystander effect that can improve tumor cell killing [36]. Addition of connexin-43 gene to this paradigm (TK-GCV + TNF-α + radiosurgery) further improved survival (90% survival in tumor-bearing mice). We also studied this strategy in a 9L rat glioma model and found that addition of radiosurgery to suicide gene therapy (SGT) signiﬁcantly improved animal survival compared with SGT alone. The combination of HSV-based SGT (TK-GCV), TNF-α gene transfer, and radiosurgery was more effective than SGT or radiosurgery alone. The combination of SGT with radiosurgery was also more effective than SGT or radiosurgery alone. Although the exact mechanism of this effect is unclear and remains the subject of future investigations, these experiments indicate that gene therapy could be an effective strategy for enhancing the radiobiological impact of radiosurgery. In other studies, tumor sensitization to radiation was apparently mediated by extracellular TNF-α promoting the destruction of tumor vessels, whereas HSV-vector–mediated TNF-α–enhanced killing of malignant glioma cell cultures is presumably a consequence of an intracellular TNF-α activity (Table 6-3) [34, 37].

Experimental Models for Functional Brain Radiosurgery Radiosurgery is rapidly expanding beyond its use as a treatment of brain tumors and arteriovenous malformations (AVMs). It has been found effective for other neurologic disorders, such as epilepsy, movement disorders, and trigeminal neuralgia. The promise of “functional” radiosurgery has led to a need to investigate its efﬁcacy, limitations, and potential drawbacks.

Hippocampal Radiosurgery for Epilepsy The potential efﬁcacy of radiosurgery for the treatment of epilepsy has been evaluated using rat models. Kainic acid reproducibly induces epilepsy in rats when injected into the hippocampus. Mori et al. [42] treated kainic acid–induced epilepsy in rats with doses of 20 to 100 Gy radiosurgery using Gamma Knife. The efﬁcacy of the treatment on epilepsy was evaluated by direct observation and scalp EEG for 42 days. Even 20 Gy signiﬁcantly reduced the number of seizures, and the efﬁcacy improved with increasing dose. Only doses >60 Gy induced histologic changes. Maesawa et al. [43] treated epileptic rats with a single dose of 30 or 60 Gy. Both doses signiﬁcantly reduced EEG-deﬁned seizures. The latency to this effect was less after the higher dose (5 to 9 weeks for 60 Gy vs. 7 to 9 Gy for 30 Gy). Whereas kainic acid injection alone reduced performance of rats on the water maze task, the performance of rats that were treated by radiosurgery after kainic acid administration was not different from controls. Liscak et al. [44] evaluated the effects of radiosurgery on normal hippocampus in an effort to identify potential normal

6.

65

experimental radiosurgery models

tissue complications and determine dose limits for hippocampal radiosurgery. This study employed four separate 4-mm isocenters to irradiate the entire hippocampus with 25 to 100 Gy. Doses <50 Gy did not cause any perceptible changes based on histology, MRI, and Morris water maze testing. In contrast, the performance on the Morris water maze was signiﬁcantly worse for animals treated with >50 Gy. These investigations support the concept that radiosurgery may be an effective method for treating epilepsy, but they also suggest that doses to the hippocampus should be limited to reduce potential effects on learning and memory.

Thalamic Radiosurgery for Movement Disorders The effect of radiosurgery on potential targets for the treatment of movement disorders has been evaluated. De Salles and colleagues [45] used a linear accelerator and 3-mm collimator to deliver a maximal dose of 150 Gy to the subthalamic nucleus of one vervet monkey and to the substantia nigra of another. Follow-up MRI detected a 3-mm lesion that did not increase in size throughout the course of the study. Kondziolka et al. [46] examined the effects of thalamic radiosurgery in a baboon model and reported that a dose of 100 Gy (central dose using 4-mm collimator) was sufﬁcient to induce contrast enhance-

ment of magnetic resonance images and coagulative necrosis as evaluated by histology.

Trigeminal Nerve Radiosurgery for Trigeminal Neuralgia Radiosurgery has signiﬁcant potential as an effective, noninvasive method for treatment of trigeminal neuralgia, and the effect of Gamma Knife irradiation on the trigeminal nerve has been evaluated in the baboon [47]. We irradiated the normal proximal trigeminal nerve with 80 or 100 Gy using a 4-mm collimator. A 4-mm region of contrast enhancement was visible by MRI at 6 months after treatment. Both large and small ﬁbers were affected with axonal degeneration occurring after 80 Gy and necrosis after 100 Gy. Neither dose was effective at selectively damaging ﬁbers responsible for transmission of pain while maintaining those responsible for other sensations, which would be optimal for effective treatment of trigeminal neuralgia. Nevertheless, this study does demonstrate that it is possible to noninvasively and precisely affect speciﬁc nerves using the Gamma Knife. Whether other dose regimens might cause selective damage to pain ﬁbers will require further investigation (Table 6-4).

TABLE 6-4. Central nervous system response to functional radiosurgery. First author and year

Ishikawa, 1999 [48]

Maximum dose (Gy)

200

Region(s) irradiated

Irradiation technique

Animal model

Results

Medial temporal lobe

Gamma Knife, 4-mm collimator

Rat

Gamma Knife, 4-mm collimator Gamma Knife, 4-mm collimator

Rat

Gamma Knife, 4-mm collimator Gamma Knife, 4-mm collimator

Baboon

Linac, 3-mm collimator

Monkey

Baboon

Caudate-putamen complex

Gamma Knife, 4-mm collimator Gamma Knife, 4-mm collimator Gamma Knife, 4-mm collimator

Hippocampus

Proton beam

Rat

Medial hypothalamus

Gamma Knife, 4-mm collimator

Rat

Sequential MRI and histopathology showed consistent necrosis at 2 weeks after 200-Gy radiosurgery Seizure frequency decreased after ≥20-Gy radiosurgery Seizure frequency decreased after 30–60 Gy radiosurgery. Shorter latency after higher dose. Learning and memory unaffected MRI, histology at 6 months showed axonal degeneration at all doses Subnecrotic (20–40 Gy) radiosurgery substantially reduces seizure frequency and duration MRI and histology showed that necrotic lesion remained at <3-mm size MRI, histology showed necrosis at 6 months More than 50 altered memory performance 6-OHDA–induced hemiparkinsonian behavior was signiﬁcantly reduced. Necrotic lesions were surrounded by regions that were highly positive for GDNF Doses 90 CGE or higher resulted in adverse behavioral effects and necrosis in 3 months 30 or 60 CGE radiosurgery led to marked increase in HSP-72 staining but no necrosis Signiﬁcant and sustained reductions in weight set-point after a latency of 7 weeks was noted

Mori, 2000 [42]

20–100

Hippocampus

Maesawa, 2000 [43]

30–60

Hippocampus

Kondziolka, 2000 [47] Chen, 2001 [49]

80–100

Trigeminal nerve

20–40

Hippocampus

De Salles, 2001 [45]

150

Kondziolka, 2002 [46] Liscak, 2002 [44]

100

Zerris, 2002 [50]

140

Brisman, 2003 [51]

Vincent, 2005 [52]

25–150

5–130 CGE

40 Gy

Subthalamic nucleus, substantia nigra Thalamus Hippocampus

Rat

Rat

Rat Rat

66

a. niranjan and d. kondziolka

Hypothalamic Radiosurgery for Obesity Vincent et al. studied the effect of subnecrotic hypothalamic radiosurgery on body weight set-point [52]. These investigators performed hypothalamic radiosurgery on genetically obese Zucker rats using a total dose of 40 Gy delivered to two nearby targets in the medial hypothalamus. These investigators noted signiﬁcant and sustained reductions in weight set-point for animals that received radiosurgery compared with sham-treated animals after a latency of 7 weeks. No gross behavioral abnormalities were noted. Histopathologic analysis showed no abnormalities except a small area of necrosis in one animal. These investigations indicate the potential role of a single dose of irradiation to the hypothalamus in producing sustained reduction in the weight set-point. At the University of Pittsburgh, the authors are currently investigating the effects of hypothalamic radiosurgery on obesity using a primate model.

Future Models of Experimental Radiosurgery Rapid developments in neuroimaging, stereotactic techniques, and robotic technology in the past decade have contributed to improved results and wider applications of radiosurgery. The role of radiosurgery has expanded well beyond its initial application for functional neurosurgery, pain management, AVMs, and selected skull base tumors. The clinical spectrum now includes a wide variety of rare skull base neoplasms, serves as the primary treatment of metastatic brain cancer, and provides adjuvant management of malignant primary brain tumors. Although radiosurgery provides survival beneﬁts in diffuse malignant brain tumors, cure is still not possible. Microscopic tumor inﬁltration into surrounding normal tissue is the main cause of recurrence. Additional strategies are needed to speciﬁcally target tumor cells. In the future, gene transfer to sensitize malignant tumor cells to radiosurgery may provide enhanced tumor cell kill while radioprotective agents will prevent damage to surrounding normal tissue. Although the nature of brain injury after radiosurgery appears similar to that seen after conventional radiation treatments, there remain a number of questions concerning the effects and the pathogenesis of such effects after both forms of radiation. At present, the cellular target that is primarily responsible for radiation-induced breakdown of normal tissue is unclear. The white matter and the cerebral vasculature appear to be particularly susceptible to radiation, which suggests that oligodendrocytes and endothelial cells may be critical targets of radiation. Recent studies have also implicated a potential role for neural progenitors in radiation-induced brain injury [53, 54]. The role of neural and endothelial precursors in repairing radiation-induced brain damage is under evaluation. Neural stem cells can be isolated from normal adult mammalian brain and can be induced to differentiate into neurons or glia. In the future, if implanted neural stem cells could prevent or repair radiation-induced damage to normal brain, then the tumors could be targeted with higher radiosurgery doses. These higher doses may prove lethal to tumor cells while stem cells will prevent damage to surrounding normal tissue. While radiosurgery usage continues to expand as we sort out the roles of precision radiation, we must strive to understand the mechanism of biological response of CNS tissues to

radiation as well as the potential of long-term adverse effects including the risk of delayed oncogenesis. Radiosurgery can affect the cerebral microenvironment. The role of radiosurgery in altering the local immune response by activating microglia and stimulating cytokines needs to be studied in order to develop strategies to treat brain tumors. Further research to answer these questions is needed to maximize the effectiveness of radiosurgery on target regions and to minimize injury to other areas.

References 1. Larsson B, Leksell L, Rexed B, et al. The high-energy proton beam as neurosurgical tool. Nature 1958; 182(6):1222–1223. 2. Rexed B, Mair W, Sourander P, et al. Effect of high-energy protons on the rabbit. Acta Radiol Oncol Radiat Phys Biol 1960; 53:289–299. 3. Leksell L, Larsson B, Anderson B, et al. Lesions in the depth of the brain produced by a beam of high-energy protons. Acta Radiol Ther Phys Biol 1960; 53:251–264. 4. Andersson B, Larsson B, Leksell L, et al. Histopathology of late local radiolesions in the goat brain. Acta Radiol Ther Phys Biol 1970; 9(5):385–394. 5. Nilsson A, Wennerstrand J, Leksell D, et al. Stereotactic gamma irradiation of basilar artery in cat. Preliminary experiences. Acta Radiol Oncol Radiat Phys Biol 1978; 17(2):150–160. 6. Calvo W, Hopewell JW, Reinhold HS, et al. Time- and doserelated changes in the white matter of the rat brain after single doses of X rays. Br J Radiol 1988; 61:1043–1052. 7. Fike J, Gobbel G. Central nervous system radiation injury in large animal models. In: Gutin P, Leibel S, Sheline G, eds. Radiation Injury to the Nervous System. New York: Raven Press; 1991:113– 135. 8. van der Kogel A. Central nervous sytem radiation injury in small animal models. In: Gutin P, Leibel S, Sheline G, eds. Radiation Injury to the Nervous System. New York: Raven Press; 1991:113– 135. 9. Calvo W, Hopewell J, Reinhold H, et al. Dose-dependent and time-dependent changes in the choroid plexus of the irradiated rat brain. Br J Radiol 1987; 60:1109–1117. 10. Hopewell J, Calvo W, Campling D, et al. The role of the vasculature in normal tissue responses. In: Fielden E, Fowler J, Hendry J, Scott D, eds. Radiation Research, Proceedings of the 8th International Congress of Radiation Research. London: Taylor & Francis; 1987:789–794. 11. Zeman W, Samorajski T. Effects of irradiation on the nervous system. In: Berdjis CC, ed. Pathology of Irradiation. Baltimore: Williams & Wilkins, 1971:213–277. 12. Hopewell JW, Wright EA. The nature of latent cerebral irradiation damage and it modiﬁcation by hypertension. Br J Radiol 1970; 43:161–167. 13. Reinhold H, Hopewell J. Late changes in the architecture of blood vessels of the rat brain after irradiation. Br J Radiol 1980; 53: 693–696. 14. Schultheiss T, Stephens L. Permanent radiation myelopathy. Br J Radiol 1992; 65:737–753. 15. Lunsford LD, Altschuler EM, Flickinger JC, et al. In vivo biological effects of stereotactic radiosurgery: a primate model. Neurosurgery 1990; 27(3):373–382. 16. Kondziolka D, Lunsford LD, Altschuler EM, et al. Biological effects of stereotactic radiosurgery in the normal primate brainstem. In: Lunsford LD, ed. Stereotactic Radiosurgery Update. New York: Elsevier; 1992:291–294. 17. Kondziolka D, Lunsford LD, Claassen D, et al. Radiobiology of radiosurgery: Part I. The normal rat brain model. Neurosurgery 1992; 31(2):271–279.

6.

experimental radiosurgery models

18. Blatt DR, Friedman WA, Bova FJ, et al. Temporal characteristics of radiosurgical lesions in an animal model. J Neurosurg 1994; 80(6):1046–1055. 19. Kamiryo T, Kassell NF, Thai QA, et al. Histological changes in the normal rat brain after gamma irradiation. Acta Neurochir (Wien) 1996; 138(4):451–459. 20. Hopewell JW, Wright EA. The effects of dose and ﬁeld size on late radiation damage to the rat spinal cord. Int J Radiat Biol Relat Stud Phys Chem Med 1975; 28(4):325–333. 21. Karger CP, Munter MW, Heiland S, et al. Dose-response curves and tolerance doses for late functional changes in the normal rat brain after stereotactic radiosurgery evaluated by magnetic resonance imaging: inﬂuence of end points and follow-up time. Radiat Res 2002; 157(6):617–625. 22. Kamiryo T, Lopes MB, Kassell NF, et al. Radiosurgery-induced microvascular alterations precede necrosis of the brain neuropil. Neurosurgery 2001; 49(2):409–414; discussion 14–15. 23. Oldﬁeld EH, Friedman R, Kinsella T, et al. Reduction in radiation-induced brain injury by use of pentobarbital or lidocaine protection. J Neurosurg 1990; 72(5):737–744. 24. Smith SL, Scherch HM, Hall ED. Protective effects of tirilazad mesylate and metabolite U-89678 against blood-brain barrier damage after subarachnoid hemorrhage and lipid peroxidative neuronal injury. J Neurosurg 1996; 84(2):229–233. 25. Braughler JM. Lipid peroxidation-induced inhibition of gammaaminobutyric acid uptake in rat brain synaptosomes: protection by glucocorticoids. J Neurochem 1985; 44(4):1282–1288. 26. Bernstein M, Ginsberg H, Glen J. Protection of iodine-125 brachytherapy brain injury in the rat with the 21-aminosteroid U-74389F. Neurosurgery 1992; 31(5):923–927; discussion 7–8. 27. Buatti JM, Friedman WA, Theele DP, et al. The lazaroid U74389G protects normal brain from stereotactic radiosurgery-induced radiation injury. Int J Radiat Oncol Biol Phys 1996; 34(3):591– 597. 28. Kondziolka D, Somaza S, Martinez AJ, et al. Radioprotective effects of the 21-aminosteroid U-74389G for stereotactic radiosurgery. Neurosurgery 1997; 41(1):203–208. 29. Staba MJ, Mauceri HJ, Kufe DW, et al. Adenoviral TNFalpha gene therapy and radiation damage tumor vasculature in a human malignant glioma xenograft. Gene Therapy 1998; 5(3):293– 300. 30. Cao G, Kuriyama S, Du P, et al. Complete regression of established murine hepatocellular carcinoma by in vivo tumor necrosis factor alpha gene transfer.[comment]. Gastroenterology 1997; 112(2):501–510. 31. Han SK, Brody SL, Crystal RG. Suppression of in vivo tumorigenicity of human lung cancer cells by retrovirus-mediated transfer of the human tumor necrosis factor-alpha cDNA. Am J Respir Cell Mol Biol 1994; 11(3):270–278. 32. Ostensen ME, Thiele DL, Lipsky PE. Enhancement of human natural killer cell function by the combined effects of tumor necrosis factor alpha or interleukin-1 and interferon-alpha or interleukin-2. J Biol Response Modiﬁers 1989; 8(1):53–61. 33. Owen-Schaub LB, Gutterman JU, Grimm EA. Synergy of tumor necrosis factor and interleukin 2 in the activation of human cytotoxic lymphocytes: effect of tumor necrosis factor alpha and interleukin 2 in the generation of human lymphokine-activated killer cell cytotoxicity. Cancer Res 1988; 48(4):788–792. 34. Gridley DS, Archambeau JO, Andres MA, et al. Tumor necrosis factor-alpha enhances antitumor effects of radiation against glioma xenografts. Oncol Res 1997; 9(5):217–227. 35. Niranjan A, Moriuchi S, Lunsford LD, et al. Effective treatment of experimental glioblastoma by HSV vector-mediated TNF alpha and HSV-tk gene transfer in combination with radiosurgery and ganciclovir administration. Mol Ther 2000; 2(2):114– 120.

67

36. Marconi P, Tamura M, Moriuchi S, et al. Connexin 43-enhanced suicide gene therapy using herpesviral vectors. Mol Ther 2000; 1(1):71–81. 37. Moriuchi S, Oligino T, Krisky D, et al. Enhanced tumor cell killing in the presence of ganciclovir by herpes simplex virus type 1 vectordirected coexpression of human tumor necrosis factor-alpha and herpes simplex virus thymidine kinase. Cancer Res 1998; 58(24): 5731–5737. 38. Kondziolka D, Lunsford LD, Claassen D, et al. Radiobiology of radiosurgery: Part II. The rat C6 glioma model. Neurosurgery 1992; 31(2):280–287; discussion 7–8. 39. Kondziolka D, Mori Y, Martinez AJ, et al. Beneﬁcial effects of the radioprotectant 21-aminosteroid U-74389G in a radiosurgery rat malignant glioma model. Int J Radiat Oncol Bio Phys 1999; 44(1):179–184. 40. Nakahara N, Okada H, Witham TF, et al. Combination of stereotactic radiosurgery and cytokine gene-transduced tumor cell vaccination: a new strategy against metastatic brain tumors. J Neurosurg 2001; 95(6):984–989. 41. Niranjan A, Wolfe D, Tamura M, et al. Treatment of rat gliosarcoma brain tumors by HSV-based multigene therapy combined with radiosurgery. Mol Ther 2003; 8(8):530–542. 42. Mori Y, Kondziolka D, Balzer J, et al. Effects of stereotactic radiosurgery on an animal model of hippocampal epilepsy. Neurosurgery 2000; 46(1):157–165; discussion 65–68. 43. Maesawa S, Kondziolka D, Dixon CE, et al. Subnecrotic stereotactic radiosurgery controlling epilepsy produced by kainic acid injection in rats. J Neurosurg 2000; 93(6):1033–1040. 44. Liscak R, Vladyka V, Novotny J Jr, et al. Leksell gamma knife lesioning of the rat hippocampus: the relationship between radiation dose and functional and structural damage. J Neurosurg 2002; 97(5 Suppl):666–673. 45. De Salles AA, Melega WP, Lacan G, et al. Radiosurgery performed with the aid of a 3-mm collimator in the subthalamic nucleus and substantia nigra of the vervet monkey. J Neurosurg 2001; 95(6):990–997. 46. Kondziolka D, Conce M, Niranjan A, et al. Histology of the 100 Gy thalomotomy in the baboon. Radiosurgery 2002; 4(4):279–284. 47. Kondziolka D, Lacomis D, Niranjan A, et al. Histological effects of trigeminal nerve radiosurgery in a primate model: implications for trigeminal neuralgia radiosurgery. Neurosurgery 2000; 46(4):971–976; discussion 6–7. 48. Ishikawa S, Otsuki T, Kaneki M, et al. Dose-related effects of single focal irradiation in the medial temporal lobe structures in rats—magnetic resonance imaging and histological study. Neurol Med Chir (Tokyo) 1999; 39(1):1–7. 49. Chen ZF, Kamiryo T, Henson SL, et al. Anticonvulsant effects of gamma surgery in a model of chronic spontaneous limbic epilepsy in rats. J Neurosurg 2001; 94(2):270–280. 50. Zerris VA, Zheng Z, Noren G, et al. Radiation and regeneration: behavioral improvement and GDNF expression after Gamma Knife radiosurgery in the 6-OHDA rodent model of hemi-parkinsonism. Acta Neurochir Suppl 2002; 84:99–105. 51. Brisman JL, Cole AJ, Cosgrove GR, et al. Radiosurgery of the rat hippocampus: magnetic resonance imaging, neurophysiological, histological, and behavioral studies. Neurosurgery 2003; 53(4):951– 961; discussion 61–62. 52. Vincent DA, Alden TD, Kamiryo T, et al. The baromodulatory effect of gamma knife irradiation of the hypothalamus in the obese Zucker rat. Stereotact Funct Neurosurg 2005; 83(1):6–11. 53. Tada E, Yang C, Gobbel GT, et al. Long-term impairment of subependymal repopulation following damage by ionizing irradiation. Exp Neurol 1999; 160(1):66–77. 54. Tada E, Parent JM, Lowenstein DH, et al. X-irradiation causes a prolonged reduction in cell proliferation in the dentate gyrus of adult rats. Neuroscience 2000; 99(1):33–41.

7

Treatment Planning for Stereotactic Radiosurgery David M. Shepard, Cedric Yu, Martin Murphy, Marc R. Bussière, and Frank J. Bova

Introduction This chapter provides an introduction to treatment-planning procedures for stereotactic radiosurgery (SRS). The chapter begins with a brief history of SRS planning. Next, the basic steps followed in the development of an SRS treatment plan are described including imaging, contouring, selection of plan parameters, and evaluating treatment plan quality. The remainder of the chapter focuses on treatment planning for speciﬁc SRS delivery techniques including the Gamma Knife, linear accelerator–based SRS, CyberKnife, and proton radiosurgery.

Background on SRS Planning During the past three decades, radiosurgery has been transformed from a curiosity available at a single institution to a routinely used therapy available at medical facilities throughout the world. During this increase in utilization, radiosurgical planning and delivery techniques have undergone signiﬁcant technical improvements including (1) availability of true threedimensional image data, (2) improved plan delivery techniques, and (3) increased sophistication of planning algorithms. The most critical improvements have come about through the availability of true three-dimensional image data provided primarily through computed tomography (CT) and magnetic resonance imaging (MRI) scanning techniques. Today’s practicing radiosurgeon has the advantage of viewing target and nontarget tissues with increased contrast resolution and with dramatically improved spatial resolution. Clinicians have also beneﬁted from the development of tools for registering or fusing single and multimodality data sets. The second area that has undergone signiﬁcantly improvement is that of plan delivery. For several years, the only radiosurgical device available was the Gamma Knife. The Gamma Knife has now been joined by several designs that use linear accelerators as the radiation source. Whereas the initial linear accelerator (linac)-based systems were less accurate than the Gamma Knife, devices were developed in the late 1980s that provided signiﬁcant improvement in the delivery accuracy of linac-based radiosurgical units. Recent years have seen additional advancements in basic linac technology

resulting in highly accurate base linear accelerators. These devices provide efﬁcient radiosurgical delivery and have made it possible to extend the beneﬁts of radiosurgery to targets outside the cranium. The third area of technological advancement in radiosurgery has been the increased sophistication of the planning algorithms that has been made possible by the tremendous increase in computer processing power over the past three decades. The ability to process images, register images, and develop plans has gone from a process that required many hours to one that can be accomplished in a matter of seconds. This has allowed radiosurgical teams to not only develop and evaluate their ﬁrst best guess at a plan but also to iterate through many different plans in order to arrive at a more optimal plan. Although these planning and delivery tools help provide improved treatment plans, the basics principles set forth by Lars Leksell in the 1950s remain the foundation for radiosurgery [1–4]. Leksell’s basic idea was to place the target tissues in the center of a large number of beams of radiation. The arrangement of beams was designed so that the beams only intersected over the target tissues and they quickly diverged from the target in all directions. The majority of clinical data published on radiosurgery is based on a planning technique known as sphere packing. This technique was initially developed by Leksell and constitutes the basis of the Gamma Knife planning process. It has also been used in conjunction with many of the linac-based delivery systems. In this technique, a set of beams are aimed at a point in space, know as the isocenter. The beams are selected so that they approach and leave the isocenter through unique paths, providing both geometric concentration and a high-dose gradient. Much of the accuracy discussed in the radiosurgery literature regards the measure of delivery systems’ ability to precisely target radiation beams at this “point” in space. The resultant dose distribution is approximately spherical and therefore provides a reasonable plan for spherical target volumes. For more complex target shapes, a technique known as sphere packing is employed. In this technique, the planner places the initial dose sphere inside the target volume. The initial dose sphere is typically selected to be the largest sphere that the system can produce and that can ﬁt “inside” the target volume. The planner then continues to pack the target volume with spheres of equal

69

70

d.m. shepard et al.

from basic radiosurgical principles and begins to depend more upon radiologic principles. While this may be appropriate in certain clinical situations, a cautious approach is warranted.

Basic Steps in Developing an SRS Treatment Plan The basic steps followed in radiosurgery treatment planning are illustrated in Figure 7-1.

Stereotactic Localization Because SRS involves the delivery of a very high dose of radiation in either a single delivery or a small number of fractions, it is essential to achieve a high degree of dose conformity and accurate patient positioning. This critical dosimetric requirement is accomplished by stereotactically focusing many convergent radiation beams on the target. By cross-ﬁring a large number of beams, a high dose is delivered in the region of intersection with a rapid fall-off in dose outside of the target. One approach to accurate stereotactic localization is achieved with ﬁxation frames such as the frame shown in Figure 7-2. On the day of the procedure, this ﬁxation frame is ﬁxed to the patient’s skull under local anesthetic. The frame serves two important purposes. First, it provides a rigid ﬁxation system that ensures that the patient cannot move during the delivery. Second, it provides a frame of reference whereby the tumor location can be determined relative to the frame and relative to the delivery unit. FIGURE 7-1. The basic steps in SRS treatment planning.

or smaller diameter until adequate target coverage is achieved. An advantage of a linear accelerator–based system over the Gamma Knife is the ability to produce larger spheres and therefore pack a speciﬁc volume with fewer total spheres or isocenters. These spheres of dose are created using beam sets that deliver radiation from a large number of beam angles. The Gamma Knife makes use of 201 individual beams whereas multiple arcs are used in the linear accelerator systems. It has been shown that in order to produce the concentration and gradient, which is responsible for a vast majority of all published radiosurgical clinical outcomes, at least 15 beams must be spread over approximately 2π space [5]. In recent years, SRS delivery techniques have expanded beyond the Gamma Knife and linac-based delivery using circular collimators to include the CyberKnife, tomotherapy, and the adaptation of linacs with micro-multileaf collimators. Although each of these systems offers the radiosurgeon special features, they are all bound by the same principle that a large number of beam paths are required to concentrate the dose in the target and create steep dose gradients. It is important to note that as new SRS delivery techniques emerge, clinicians must examine the impact on dose conformity compared with historical SRS delivery techniques using the Gamma Knife or linac-based delivery with circular collimators. A loss in dose conformity may force the user to adopt a treatment schedule that deviates

FIGURE 7-2. Gamma Knife coordinate frame. (Image Courtesy of Elekta, Inc.)

7.

treatment planning for stereotactic radiosurgery

71

FIGURE 7-3. (a) The ﬁducial indicator box for CT imaging. (b) The patient is imaged with a ﬁducial box attached to the frame. (Image Courtesy of Elekta, Inc.)

One approach for facilitating the registration of the patient images is to attach a ﬁducial box (Fig. 7-3) to the stereotactic head frame before patient imaging. The box includes tubes ﬁlled with a high-contrast solution that provides easy to delineate ﬁducial marks on all of the patient’s images (Fig. 7-4). After the imaging is complete, the patient images are loaded into the planning system. The images are registered with respect to the ﬁducial marks. The markings on the ﬁducial box provide a coordinate system that makes it possible to precisely determine the position of the treatment volume with respect to the stereotactic head frame. In addition to the ﬁxation frames described above, frameless SRS techniques have been developed that maintain a highly precise delivery [6–12]. For example, researchers at the

FIGURE 7-5. Repeat ﬁxations system that allows the biteplate to be reinserted a number of times and judges the reproducibility against a ﬁxed head band. (Image courtesy of Varian Medical Systems. Copyright 2006, Varian Medical Systems. All rights reserved.)

Patient’s Positional Indicator (left/anterior in this axial image) Fiducials

Third Plate

Image

University of Florida have developed and commercialized a biteplate system (Fig. 7-5). Frameless SRS is typically the preferred option in fractionated radiosurgery where the invasive nature of ﬁxation frames makes them a less-viable option. Fractionation radiosurgery, known as stereotactic radiotherapy (SRT), is commonly delivered to patients where the likelihood of late toxicity makes it inadvisable to treat a large target volume in a single high-dose fraction [7, 12–14].

Patient Imaging

Fiducials Right Left Fiducial Fiducial Markers Markers FIGURE 7-4. The ﬁducial marks appear as white dots on this MRI image of a Gamma Knife patient. (Image Courtesy of Elekta, Inc.)

The three primary imaging techniques for SRS are magnetic resonance imaging (MRI), computed tomography (CT), and planar angiography. MRI is an excellent technique for imaging soft tissue contrast. MRI, however, does not provide attenuation coefﬁcients and must be fused with CT images if one wishes to make appropriate heterogeneities in the dose calculation. It should be noted that in most MRI units, the spatial uniformity of the images degrade as the radius of the images increase. Because most ﬁducial systems place the ﬁducial markers at the outermost extent of the MRI image volume, the accuracy obtained in mapping the imaged voxels to stereotactic space can be compromised. This is one of several reasons

72

d.m. shepard et al.

FIGURE 7-6. One approach to image registration.

some radiosurgical practitioners have opted for image fusion technologies as opposed to direct stereotactic MRI scanning (Fig. 7-6). CT imaging provides inferior soft tissue contrast compared with MRI scanning; however, the spatial resolution of CT is extremely reliable. CT also provides attenuation coefﬁcients that make it possible to make accurate heterogeneity corrections in the dose calculation. Planar angiography is an imaging option that can be used for imaging patients with arteriovenous malformations (AVMs). Although plane ﬁlm angiography was the gold standard for AVM nidus identiﬁcation, cut ﬁlms have almost been fully replaced by electronically derived images. For the past few decades, the primary mode of electronic image acquisition has been the image intensiﬁer. Although this device is know for its excellent image contrast, it is equally known for its poor spatial accuracy. Most image intensiﬁers have multiple levels of image distortion. More recently, solid-state image acquisition systems have been developed. These systems have all but eliminated concerns regarding spatially nonuniform images. An additional problem with angiography is that although plane ﬁlm angiography can provide images in multiple planes, the basic image set that the clinician has to work with is twodimensional. This two-dimensional image data set can lead to inaccuracies in the target deﬁnition. It is for this reason that most targeting for vascular radiosurgery targets is obtained either through CT angiography, MRI angiography, or both.

Registering the Images The general planning procedure starts with the acquisition of three-dimensional image information and the import of patient images into the treatment planning system via a computer network or data storage media. On the planning computer, these images are typically displayed in two-dimensional slices along the axial, sagittal, or coronal planes [15]. A screen capture from the GammaPlan system shown in Figure 7-4 illustrates the process of aligning the ﬁducial marks in the SRS planning system. The planning system must be able to accurately identify the ﬁducial marks and have internal mechanisms for validating their consistency. Once the ﬁducial markers are identiﬁed, the patient anatomy shown on the images is placed in the coordinate system deﬁned by the stereotactic localization device. Some systems use ﬁducial markers that fully deﬁne the stereotactic reference system whereas others require speciﬁc information from the scanner. For example, a fully deﬁned system

does not require information on the axial slice position. This is because it can derive this information from the actual scan image. Consequently, this approach eliminates the need for QA measurements to guarantee accurate scanner information such as linearity of table movement or degree of gantry tilt. Frameless systems also need to be able to map the threedimensional data set into a rigid and deﬁnable coordinate system. To achieve this, stereotactic systems have been designed that incorporate surrogate ﬁducial markers into the CT or MRI scans. These markers can also be referenced at the time of therapy through mechanical, electrical, or optical means. A critical feature of frameless systems is the ability to demonstrate that the reference can in fact be reliably and precisely ﬁxed to the patient. Many systems offer testing only at the time of design and development. These tests are usually at the hands of the developer, an expert with extensive experience. It is critical for a repeat ﬁxation system to also provide a methodology of testing its accuracy for each patient. Figure 7-5 shows the RadioCameras Treatment Guidance System (Varian, Palo Alto, CA). This system provides patient-speciﬁc statistics on the device’s ability to be precisely reapplied. Two basic approaches are used in the registration of images. The ﬁrst is to use direct stereotactic MRI and CT scanning. In this procedure, ﬁducial markers are incorporated into both image data sets and each data set is independently mapped into stereotactic space. In the second approach, stereotactic CT scanning is performed using a ﬁducial system when acquiring the images. Nonstereotactic MRI scans are also obtained (see Fig. 7-6). The MRI scans are then mapped into stereotactic space by registration of the MRI data set to the CT data set. In most cases, rigid registration techniques are used. Typically, the clinician visually inspects the results of the registration to verify the quality of the ﬁt. An advantage of this second approach is that it allows the MRI scan to be obtained prior to the placement of any stereotactic ring. Consequently, the clinician can exam the actual scan that will be used for scanning prior to deciding that radiosurgery is the appropriate mode for treatment.

Contouring Structures After the images have been imported into the planning system and the ﬁducial marks have been registered, the target to be treated is then identiﬁed. To aid target delineation, fusion of images from different imaging modalities is possible with many SRS planning systems. Whereas some planning systems may not require the users to outline the target and critical structures, contouring is required if plan evaluation tools such dose-volume histograms (DVHs) are to be used. Contouring is also required by systems that employ inverse planning. The patient’s external contour can either be manually contoured or identiﬁed automatically by the planning system based on grayscale information from the images. For Gamma Knife planning, the skin contour or scalp of the patient’s head is measured with a skull-scaling device (Fig. 7-7). The location of the surface is important for computing the penetration depth for each beam during the dose calculation. The skull-scaling device used for Gamma Knife planning makes it possible to know the penetration depth for each photon beam even if the entire skull has not been imaged.

7.

treatment planning for stereotactic radiosurgery

73

80% and 40% isodose surfaces. Therefore, prescribing to the 80% isodose line minimizes the integral dose to all normal tissues (the tissues outside the target volume). Compared with the Gamma Knife, single-isocenter plans are used more frequently for radiosurgery treatments on linear accelerators. Gamma Knife units are limited to four focusing helmets (shot sizes), which is signiﬁcantly fewer than the typical number of cones that are available on a linear accelerator–based system. Additionally, multileaf collimator–based deliveries generally use only one isocenter. For multiple-isocenter plans, this optimal surface shifts to the 70% isodose surface and at times can be extended to isodose surfaces as low as the 50% dose surface, as is often the case in Gamma Knife plans.

Designing the Treatment Plan

FIGURE 7-7. Skull scaling device used with the Gamma Knife to provide mapping of the skull. (Image Courtesy of Elekta, Inc.)

Deﬁning the Prescription The dosimetric characteristics of the treatment plan vary from one delivery approach to another. With the Gamma Knife and linac-based SRS, the use of multiple isocenters means that multiple spherical isodose distributions have to be packed into the target volume. Overlaps of such high-dose spheres are inevitable. As a result, the dose is highly nonuniform in the target. The isodose surface that encloses the target (which is often taken as the prescription dose) is typically 50% of the maximum dose in the target. For linac-based SRS, the target dose is highly uniform when a single isocenter is used. The isodose line that encompasses the target can be greater than 80% of the maximum dose in the target. In SRS, the prescription dose is normally set to the dose level that conforms to the target, or the minimum target dose. Compared with fractionated radiation treatments, radiosurgery treatment plans have signiﬁcantly less normal tissue included in the prescription isodose volume. For this reason, the typical restrictions on dose uniformity (a foundation of fractionated radiation therapy) do not apply in radiosurgery cases. The radiosurgeon is more concerned with the dose to normal tissue outside the target volume. Consequently, the goal in treatment planning is to achieve maximum normal tissue sparing rather than maximum target dose uniformity. The radiosurgeon can minimize the dose to normal tissue by designing a plan where the isodose that just covers the target surface is along the steepest portion of the dose gradient. It can be shown that for single-isocenter plans, this places the optimal dose prescription at the 80% isodose surface [16]. This is because the dose gradient falls off most quickly between the

For optimal planning, the neurosurgeon, radiation oncologist, and physicist or dosimetrist should perform the treatment planning as a team and bring all aspects of expertise to bear on the problem. After the images have been imported and registered and contouring of structures has been completed, one can proceed with the task of formulating a treatment plan that meets the dosimetric requirements speciﬁed by the physician. The planning steps that are followed differ from one delivery technique to the next. For the Gamma Knife and for linacbased SRS with circular collimators, the task of planning is to ﬁnd a set of isocenter locations, the size of the collimators to use for each location, and the weights of the isocenters. Multiple isocenters are commonly used in a Gamma Knife treatment plan due to the limited number of collimator sizes to choose from and the relative efﬁciency with which each isocenter can be delivered [17–24]. For linac-based SRS with circular collimators, the planner strives to minimize the number of isocenters in a plan to compensate for the relative inefﬁciency of the beam delivery [25, 26]. Linac systems, however, have the advantage of a much wider range of collimators usually extending from 5 mm to 50 mm in diameter. For linac-based SRS with micro-multileaf collimators, only one isocenter is used [25, 27–37], and the planning methods are similar to those used for external beam radiation therapy. The selection of beam directions and ﬁeld shapes is aided by the use of tools such as beam’s-eye-view (BEV) visualization and digitally reconstructed radiographs (DRRs).

Dose Calculation A variety of dose calculation techniques have been employed for SRS treatment planning [38–48]. Simple empirical dose calculation methods were employed with many of the earlier SRS planning systems [17]. For Gamma Knife radiosurgery, a standard set of beam data is included with the system, and the user veriﬁes the dose proﬁles of individual beams for each of the focusing helmets. The use of a standard data set is possible because all Gamma Knife systems use the same design with only slight variations in the single-beam dosimetry. The dose distribution in the patient is calculated by adding the dose distributions for all 201 beams [20]. For linac-based SRS with circular collimators, the tissue-maximum ratios (TMRs) and dose proﬁles can be measured for each collimator size as a function of depth. For a given arc, the mean TMR for all beam directions

74

d.m. shepard et al.

is calculated and the total dose contributed by the arc can be calculated. Dose distributions can be computed by approximating an arc as a series of ﬁxed beams and summing the dose distributions from all of the beams. For intracranial SRS, heterogeneity corrections are generally not applied in the dose calculation. Corrections to account for variations in the attenuating properties of tissue, bone, and air are not needed because of the simple geometry that is employed [44, 49]. The magnitude of the resulting error can be estimated by assuming that the average beam passes through 5 mm of skull with an average density of 1.2 g/cm3. Assuming an attenuation of 4% per centimeter for a 6-MV beam, the maximum error in absolute TMR that would result from ignoring heterogeneity corrections is less than 1%. However, when SRS is extended to extracranial applications, traditional empirical dose calculation methods are no longer adequate for accurate dose calculations due to the presence of bone and air. Therefore, many treatment-planning systems designed for planning extracranial SRS employ three-dimensional pencil beam dose calculation methods.

Forward Versus Inverse Planning Forward planning is the most common planning approach for SRS. With forward planning, the treatment plan is developed through an iterative trial-and-error approach. From one iteration to the next, the planner attempts to determine a set of parameters (such as beam angle, beam weights, etc.) that give an acceptable plan. The iterative process is usually stopped when the planner is no longer able to make noticeable improvements in the plan quality. With inverse treatment planning, the user begins by outlining the target and any sensitive structures. A series of treatment goals are then deﬁned, and an optimization algorithm determines the plan parameters that provide a plan that best satisﬁes these goals. The quality of the plan is scored using an objective function and constraints. An objective function reduces an entire treatment plan into a single numerical value. The job of the optimizer is to either minimize or maximize the value of the objective function. Typically, an optimization will also include

one or more constraints. A constraint is a condition that must be satisﬁed in order for a solution to be considered feasible. The most basic constraint in any radiotherapy optimization is that the beam weights must be nonnegative. Compared with forward planning, inverse planning provides two potential advantages. First, the time required for planning can be reduced because much of the trial and error is removed. Second, inverse planning should lead to improved plan quality due to the ability of the optimizer to consider thousands of plan conﬁgurations in selecting the optimal plan. Unfortunately, the quality of inverse planning tools for SRS varies signiﬁcantly from one delivery technique to the next and one planning system to the next. Users of some systems will ﬁnd that an experienced planner using forward planning techniques produces the highest quality plans.

Evaluation of Plan Quality The most common tool for evaluating plan quality is the visualization of isodose curves superimposed on the patient’s anatomic images. Figure 7-8a shows a typical isodose plot for a Gamma Knife patient. In this case, the isodose curves are plotted as a percentage of the maximum target dose. Isodose curves can also be plotted as a percentage of the prescribed dose or as absolute dose lines. When evaluating plan quality, it is common practice to visualize isodose curves in the axial, sagittal, and coronal planes. Some systems can also display three-dimensional dose clouds that make it possible to quickly assess the quality of the dose coverage. DVHs also serve as an important tool for analyzing plan quality. A DVH plot from a Gamma Knife patient is shown in Figure 7-8b. For each structure, the DVH plots the fraction of the volume covered by each dose level. Both the dose and the volume can be expressed as either absolute or relative values. DVHs are particularly useful because they reduce a three-dimensional treatment plan into an easy-to-read twodimensional plot. The comparison of multiple plans is also simpliﬁed by overlying DVHs on the same plot. The disadvantage of DVHs is the lack spatial information. A DVH will indicate the presence of hot or cold spots, but it does not specify where

FIGURE 7-8. (a) A Gamma Knife isodose plot with the target outlined in red, the 50% isodose line in yellow, and the 30% isodose line in green. (b) Dose-volume histogram for target in this case.

7.

treatment planning for stereotactic radiosurgery

in the structure of interest this underdosage or underdosage is located. Consequently, slice-by-slice evaluation of a plan is critical in making a ﬁnal decision regarding a treatment plan. Additional parameters have been deﬁned to score both dose conformity and target coverage for SRS treatment plans [50–58]. The most commonly used parameter for scoring dose conformity is the conformity index deﬁned in the radiosurgery quality assurance guidelines of the Radiation Therapy Oncology Group (RTOG) [52]. The RTOG deﬁned the conformity index as the volume of the prescription isodose surface divided by the target volume. For a perfectly conformal plan where the prescription isodose line exactly matches the target volume, the conformity index would equal one. A case is considered to be per protocol if this ratio falls between 1.0 and 2.0. A shortcoming of this index is that it does not consider the degree of overlap between the prescription isodose curve and the target. A plan with a complete geometric miss of the target could still give a perfect conformity index. As an alternative, Lomax and Scheib have suggested a conformity index deﬁned as “the ratio of the volume within the target irradiated to at least the prescription isodose over the total volume enclosed by the prescription isodose” [55]. Consequently, this value ranges from 0 (no conformity) to 1.0 (for perfect conformation, where the prescription isodose is identical to the target volume). As a planning goal, Lomax and Scheib suggest a conformity index of 0.6 or higher. The RTOG has also deﬁned a homogeneity index that is equal to the maximum dose in the treatment volume divided by the prescription dose. A case is considered per protocol if this ratio is less than or equal to 2.0. In terms of target coverage, the RTOG considers a case to be per protocol if the isodose line equal to 90% of the prescribed dose completely encompasses the target. Lomax and Scheib suggest an alternative volumetric deﬁnition of coverage where target coverage is deﬁned as the percentage of the target volume covered by the prescription [55].

75

FIGURE 7-9. A patient positioned for treatment on a Gamma Knife. (Image Courtesy of Elekta, Inc.)

shaping of the dose distribution. The beneﬁts of plugging are illustrated in Figure 7-12. Figure 7-12a shows the isodose curve for a single 4-mm shot for a trigeminal neuralgia patient. In Figure 7-12b, a plugging pattern has been used to lower the dose to the brain stem. The plugging pattern is shown in Figure 7-12c. After the imaging is complete, the patient images are loaded into the Gamma Knife’s treatment planning system (GammaPlan). The images are registered with respect to the ﬁducial marks, and the treatment volume is outlined by a physician.

Gamma Knife Gamma Knife Unit The ﬁrst Gamma Knife was built in 1967 under the direction of Lars Leksell of the Karolinska Institute in Stockholm, Sweden. Currently, there are more than 200 units worldwide with more than 35,000 patients treated annually [59]. Inside of the shielded treatment unit of the Gamma Knife (Fig. 7-9), the beams from 201 radioactive sources are focused so that they intersect at a single location (Fig. 7-10). The result is an elliptical region of high dose with a rapid fall-off in dose outside of the boundaries of the ellipse. Each exposure to an elliptical region of high dose is referred to as a “shot” of radiation. For each exposure, the focusing helmet dictates the size of the high-dose region. A focusing helmet incorporates a separate collimator for each of the 201 cobalt-60 sources (Fig. 7-11a). The four focusing helmets provided with the Gamma Knife can be used to produce a shot of radiation that is 4 mm, 8 mm, 14 mm, or 18 mm in diameter (Fig. 7-11b). Within each helmet, individual collimators may be plugged to provide further

FIGURE 7-10. The 201 collimators focus the beam to a single intersection point. (Image Courtesy of Elekta, Inc.)

76

d.m. shepard et al.

are adjusted to place the center of the planned shot of radiation at the focal point of the 201 beams of radiation. After this, the personnel leave the room, the exposure time is programmed, the door of the shielding unit is opened, and the couch is advanced into the high-radiation ﬁeld. Next, the treatment couch is docked, and the shot of radiation is delivered. After the exposure is complete, the couch retracts and the door closes. If more than 1 shot of radiation is to be delivered, the process is repeated with the appropriate focusing helmet and x, y, and z positioning. The model C Gamma Knife provides an automated positioning system in which the shot positions and exposure times are directly transferred to a record-and-verify system, and the machine sets the coordinates in an automated fashion.

Treatment Planning

FIGURE 7-11. The four focusing helmets dictate the size of each shot of radiation. (Image Courtesy of Elekta, Inc.)

At the time of treatment, the patient lies on the couch of the treatment unit, and the appropriate focusing helmet is afﬁxed to the table. The patient’s stereotactic head frame is then attached to the focusing helmet, and the x, y, and z coordinates

For some cases, the treatment planning process is relatively straightforward. This is particularly true for small, spherical targets. For example, Figure 7-13 shows a case where the treatment volume is relatively spherical and approximately 6 mm in diameter. An 8-mm shot of radiation was placed so that it covers the entire tumor volume. The treatment planning process becomes much more complex when the tumor volume is large or irregularly shaped. These cases typically require several shots of radiation. Through an iterative process, the planner must determine the number of shots of radiation that are required along with the size, the location, and the weight that should be assigned to each. In Figure 7-14, a simple two-dimensional bean-shaped target is used to illustrate the general planning procedure [22]. Each frame illustrates the dosimetric impact of an added shot of radiation. In this case, 5 shots of radiation provide a conformal treatment plan. Note that three different shot sizes were used in creating this plan. Figure 7-15 shows a screen capture from the GammaPlan system. During the process of treatment planning, the plan quality is evaluated through the use of isodose plots and DVHs. In the GammaPlan system, users typically normalize to the maximum dose and seek to cover the target with the 50% isodose line. An imaging study with SRS ﬁducials is required for Gamma Knife planning. For cases where the proximity of a sensitive structure leads to concern over ability to produce an acceptable plan, a “mock” treatment planning session can be helpful. To achieve this, an imaging study is performed with the patient’s head positioned within the ﬁducial box without a frame.

Viewed from behind left

right

back FIGURE 7-12. The isodose curves for a trigeminal neuralgia patient (a) before and (b) after plugging. (c) The plugging pattern: 35 of the 201 sources have been blocked (blocked collimators are shown in black).

plan that is produced can depend upon both the experience and the patience of the user. Because of these factors, researchers have sought to develop an automated process for creating Gamma Knife treatment plans. A number of researchers have investigated techniques for automating the Gamma Knife treatment planning process [18, 19, 21–24, 60–64]. One approach incorporates the assumption that each shot of radiation can be modeled as a rigid sphere. The problem is then reduced to one of geometric coverage, and a ball-packing approach can be used to determine the shot locations and sizes [22, 60, 63, 64]. Researchers have also examined the optimization of plug patterns for Gamma Knife treatment plans [23]. By selectively plugging a subset of the 201 beams, one can provide further sparing of adjacent sensitive structures. In addition, Elekta has provided an automated planning solution as an add-on to the GammaPlan system, called the GammaPlan Wizard. In the research of Shepard et al. [22, 62], the dose distribution is modeled and a formal constrained optimization is used to determine the treatment plan. With this technique, the shot sizes, locations, and weights are optimized simultaneously in less than 10 minutes. The optimization does not require the user to provide initial shot locations, and the optimization model can include dose constraints applied to both the target and the sensitive structures. The treatment plan optimization is based on the use of migrating shot locations and a nonlinear programming approach. The clinical signiﬁcance of this automated system was assessed by comparing an optimized plan with a manual plan for 10 consecutive patients treated at our Gamma Knife facility. Each treatment plan was analyzed using DVHs in conjunction with the conformity index, the minimum target dose, and the integral normal tissue dose. The results are summarized in Table 7-1. The results demonstrate that the quality of treatment plan produced by the inverse planning tool consistently matches

FIGURE 7-13. A single-shot treatment plan for a Gamma Knife treatment. The target is outlined in blue, the 50% isodose line in yellow, and the 30% isodose line in green.

Cushioning around the patient’s head is used to maintain stable positioning. This imaging study is then used to analyze the quality plan that can be achieved.

Inverse Planning For some patients, the treatment planning process can become tedious and time consuming, and the quality of the treatment

100

0.9

80

0.8

0.8

70

0.7

0.7

50 40

2.5 centimeters

5

0.6 0.5

2.5

0.4

0.6 0.5

2.5

0.4

30

0.3

0.3

20

0.2

0.2

10

0.1

0.1

0

0

2.5 centimeters

a 5

1

5

0.9

60 2.5

1

5

90

centimeters

centimeters

5

77

treatment planning for stereotactic radiosurgery

centimeters

7.

5

2.5 centimeters

b

0

c

5

1.2

5

5 1.2

1.2

1.0

1.0

2.5

0.6

2.5 centimeters

0.8 2.5

0.6

centimeters

0.8

centimeters

centimeters

1.0

0.8 2.5 0.6

0.4

0.4

0.4

0.2

0.2

0.2

0

0 2.5 centimeters

5

2.5 centimeters

FIGURE 7-14. (a) A bean-shaped target. (b–f) The process of adding shots to create a treatment plan is illustrated.

5

0

78

d.m. shepard et al.

FIGURE 7-15. Screen capture from the GammaPlan system.

TABLE 7-1. Comparison of manual plans with those created using an inverse planning tool.

Patient number

1 2 3 4 5 6 7 8 9 10 Average

Number of shots

Conformity index

Volume of 30%

Percent coverage 90%

Minimum target dose

Target volume (cm3)

Manual

Opt

Manual

Opt

Manual

Opt

Manual

Opt

Manual

Opt

12.1 3.4 3.9 11.4 7.6 6.7 9.0 37.1 7.2 3.8 10.2

6 4 4 6 8 6 5 12 6 9 6.6

6 4 4 6 8 6 5 12 6 9 6.6

1.67 1.35 2.12 1.19 1.76 1.62 1.64 1.15 1.51 1.16 1.52

1.61 1.38 1.64 1.34 1.39 1.60 1.56 1.25 1.29 1.53 1.46

44.4 12.4 21.6 34.7 33.6 28.0 35.4 105.7 30.2 12.0 35.8

41.9 11.8 16.1 35.8 24.2 27.4 31.9 102.8 25.3 13.7 33.1

100 100 99 100 100 100 100 97 100 99 99.5

100 97 100 100 100 100 100 99 100 100 99.6

48% 46% 40% 45% 49% 52% 47% 39% 49% 43% 45.8

47% 44% 48% 49% 46% 52% 48% 43% 48% 48% 47.3

7.

treatment planning for stereotactic radiosurgery

79

or beats the corresponding plan developed by an experienced physician.

Linac-Based SRS with Circular Cones Initial developments in the use of linac-based SRS techniques were centered on the use of multiple converging arc treatments delivered with circular collimators. For each treatment plan, the user must select the number of arcs, the angular arrangement of each arc, the weight assigned to each arc, and the number of isocenters. When one isocenter is used, the high isodose levels at the isocenter are nearly spherical. A variety of arc arrangements have been reported [26]. Compared with the Gamma Knife, a greater number collimator sizes is available with circular collimators. By adjusting the weightings of the arcs or through an asymmetric arc arrangement, ellipsoidal instead of spherical isodose distributions can be created. For tumors that are small and convex in shape, it is often possible to treat with only one isocenter. When one isocenter is used, the dose uniformity in the target is high and the dose can be prescribed to a high isodose level. When the target volume is large or it deviates from a sphere or an ellipsoid, multiple isocenters are typically required to achieve adequate target dose conformity. The spherical high-dose regions must overlap in order to avoid leaving a cold spot in the target. Consequently, one must accept a decreased target dose uniformity compared with plans using a single isocenter. To date, no negative clinical consequences have been reported as a result of the lack of dose uniformity seen in Gamma Knife and linac-based circular collimator treatments. When multiple isocenters are used, the treatment planning problem is similar to the sphere-packing problem of the Gamma Knife [60], where spherical dose distributions are packed together to achieve a composite dose distribution conforming to the target volume. Because the setup and delivery time per isocenter is typically longer for linac-based radiosurgery compared with Gamma Knife–based radiosurgery, it is generally desirable to deliver a limited number of isocenters; however, the greater selection of collimator sizes and the availability of larger collimator sizes compared with the Gamma Knife reduces the need for the use of a large number of isocenters. Overall, the Gamma Knife centers and linac-based SRS centers have reported similar cures rates and complications levels.

Linac-Based SRS with Micro-Multileaf Collimators A multileaf collimator (MLC) is a ﬁeld-shaping device that uses movable leaves made out of a highly attenuating material such as tungsten in order to generate arbitrary ﬁeld shapes (see Fig. 7-16). MLCs used for routine external beam delivery typically have leaf widths that project to 1 cm at isocenter. These MLCs lack the geometric precision for shaping small irregularly shaped ﬁelds such as those in commonly encountered in SRS. With micro-multileaf collimators (mMLCs), each leaf projects to a width of between 2 and 5 mm at isocenter. mMLCs are suitable for SRS applications, because they are capable of

FIGURE 7-16. A photo of Varian’s Millennium mMLC. (Image courtesy of Varian Medical Systems. Copyright 2006, Varian Medical Systems. All rights reserved.)

shaping small, irregular ﬁelds with acceptable geometric error. Typical mMLCs have between 20 and 80 leaves, arranged in pairs. The maximum ﬁeld size of mMLCs generally varies from 8 to 20 cm, much greater than those available with traditional circular collimators. As a result, extracranial SRS and stereotactic body radiation therapy (SBRT) can be delivered. In addition, linear accelerator vendors now offer MLCs that incorporate leaves of varying widths. For example, Varian’s Millennium MLC has 120 leaves (60 leaves in each leaf bank). Over the central 20 cm of the ﬁeld, the leaves project to 0.5 cm at isocenter while at the edges of the ﬁeld the leaves project to 1 cm at isocenter. This type of MLC can be used for either traditional fractionated radiotherapy or SRS. Due to their ease of use and wide range of possible applications, the advent of mMLCs has led to a decline in popularity of SRS delivery using circular collimators. Compared with Gamma Knife and linac-based SRS with circular collimators, the use of a mMLC and a single isocenter can lead to plans that are less conformal with a less steep dose gradient at the target edge. The decreased dose gradient is in part the result of the use of larger beams and fewer unique beam paths. As mentioned previously, a routine Gamma Knife plan will involve as many as 201 beams, and a typical plan for a linac-based cone-beam treatment would include ﬁve arcs each covering approximately 100° of gantry rotation. By contrast, a typical micro-multileaf plan may have as few as six beams. Target size is a second issue that affects one’s ability to create a conformal treatment plan using a mMLC. For small targets, such as in the treatment of trigeminal neuralgia, high precision in target localization and positioning is required. mMLCs may not be suitable for target sizes signiﬁcantly less than 1 cm due to the undulating ﬁeld edges caused by the ﬁnite leaf width. For larger targets, mMLC can provide efﬁcient beam

80

d.m. shepard et al.

delivery and dose uniformity due to the use of a single isocenter [50]. However, the dose to surrounding structures increases as the size of the target increases [50]. As a result, the dose to normal structures may exceed the acceptable limits if one attempts to treat a large tumor with a single fraction. Generally, multileaf collimators (including mMLCs) can be used for three delivery approaches: (1) ﬁxed ﬁeld delivery; (2) conformal arc delivery where the leaves of the MLC are adjusted continuously to match the BEV projection of a target volume; and (3) intensity-modulated delivery.

Fixed Fields With mMLC, three-dimensional (3D) conformal therapy treatment planning techniques can be applied to SRS planning. If there are a sufﬁcient number of mMLC-shaped beams, the dose distribution quality can rival that obtained with multiple arcs using circular collimators and multiple isocenters. This is especially true when the targets are larger and nonspherical in shape. Because only one isocenter is needed, the dose uniformity created with mMLCs is generally better than that achieved with the use of circular collimators and multiple isocenters. With the exception of the need for stereotactic reconstruction from stereotactic frames and markers, the planning method for SRS essentially matches that for 3D conformal therapy. Each non-coplanar ﬁeld is shaped based on the BEV of the target with additional margins to account for the width of the beam penumbra, which is wider than the penumbra from circular collimators and increases with leaf width. The use of ﬁxed ﬁelds delivered with a mMLC is best suited for convex target shapes. With ﬁxed ﬁelds, there is only a limited ability to spare normal tissues in the center of a concave volume.

Dynamic Conformal Arcs Many of the commercially available mMLCs are capable of dynamic beam delivery where the ﬁeld shape changes continuously to conform to the BEV of the target during arced beam delivery. Dynamic conformal arc delivery combines the dosimetric advantages of arcs (reduced hot-streaks of dose through the patient) with the dose conformity that is possible with mMLC beam shaping. Additionally, this approach only requires a single isocenter. With the use of three or four non-coplanar arcs, a uniform high-dose volume can be created that conforms to the target. Treatment planning for dynamic conformal arcs requires the delineation of the target and critical structures. The planner must determine the number of arcs, their length and their arrangements. For planning and delivery control, each arc is approximated as a series of ﬁxed ﬁelds. The shapes are typically set based on the BEV of the target and critical structures. For example, one can set each ﬁeld shape contained within an arc to match the BEV of the target plus a 3-mm margin. Additionally, one may choose to use the BEV of a critical structure to design ﬁeld shapes that block that structure throughout the arc path. If the gantry of the accelerator rotates with a constant speed and maintains a constant dose rate during rotation, one must assign the same number of monitor units to each beam angle within a given arc.

The dose calculation for conformal arcs is more complicated than that for conventional arcs due to fact that the ﬁeld shapes changes while the beam is on. The planning system must calculate the dose contributions from a large number of irregularly shaped ﬁelds. In determining the angular spacing of the ﬁxed ﬁelds, one must balance the need for accuracy with the desire for a reasonably quick dose calculation time. When an arc is approximated with ﬁelds spaced more than 5° apart from one another, the lack of sufﬁcient ﬁeld overlap results in undulating features in the lower isodose lines away from the focal region. These features are not reﬂected in the actual delivery. Finer spacing of the beams will reduce such artifacts in the displayed isodose lines. In some treatment planning system, a long arc can also be broken into multiple sub-arcs. The weights of these sub-arcs can be optimized based on the user-deﬁned dose-volume constraints [66]. During delivery, each of these sub-arcs is treated as a separate beam. Generally, this technique works best on convex targets and less well on targets with concave surfaces.

Intensity-Modulated Fields With intensity-modulated radiation therapy (IMRT), a modulated intensity pattern is delivered from each beam direction. Consequently, radiation can be delivered to the target through preferred locations within each beam. From each beam direction, the dose delivered to the target is nonuniform. However, all of the beams in combination produce a highly conformal dose distribution. For both external beam radiation therapy and SRS, IMRT improves dose conformity compared with the use of conventional unmodulated beams [67]. It should be noted that for small intracranial targets, IMRT delivered with a mMLC may be unable to achieve the dose conformality obtained using sphere packing due to the fact the width of the leaves of the mMLC are of the order the size of the targets to be treated. The IMRT planning problem is modeled by subdividing each beam into a grid of beamlets. The weight, or intensity, of each of the beamlets is then determined. Because of the complexity of determining the appropriate beamlet weights, inverse (automated) planning techniques are employed. IMRT planning for SRS and IMRT planning for external beam radiation therapy share the same general procedures. The user deﬁnes a series of treatment goals, and an optimization algorithm determines the plan parameters that lead to a plan that best satisﬁes those goals.

CyberKnife The CyberKnife (Fig. 7-17) is a fully integrated radiosurgery system that uses dual kilovoltage imaging devices to locate the treatment site and direct the external treatment beam to it [68]. The treatment beam is provided by a linac mounted to a robotic arm that is capable of maneuvering and pointing the linac with nearly complete freedom within the treatment workspace (with the exception of the angles blocked by the integrated imaging system). During treatment, the imaging system repeatedly acquires and analyzes targeting radiographs, supplying updated target coordinates automatically through a control loop to the robotic arm. This enables the system to maintain alignment

7.

treatment planning for stereotactic radiosurgery

FIGURE 7-17. A CyberKnife.

of the beam with the target even while the patient moves [11, 69]. The maneuverability of the beam, unconstrained by an isocenter, combined with the capacity to adjust alignment during treatment, endow the CyberKnife with unique dose delivery capabilities. These capabilities enable (1) the delivery of highly conformal dose distributions to irregular target volumes, (2) the delivery of fractionated stereotactic radiotherapy treatments, and (3) the treatment of extracranial sites that are not amenable to localization and/or ﬁxation using conventional stereotactic frames. They are exploited in a treatment planning system that has been designed speciﬁcally for the CyberKnife.

Beam Characteristics The treatment beam is provided by a 6-MV X-band linear accelerator that does not employ a ﬂattening ﬁlter. The beam is collimated to a circular cross-section by one of a set of 12 interchangeable collimators. At a source-to-surface distance (SSD) of 80 cm, these collimators provide a beam diameter that ranges from 5 mm to 60 mm. The SSD of the beam can be varied from 60 cm to 100 cm. A continuous range of beam diameters can be achieved by varying the collimator size and SSD.

Treatment Sites, Planning Scenarios, and Imaging Requirements All treatment planning for the CyberKnife is based on a CT study used alone or in conjunction with supplemental diagnostic imaging such as MRI. As with most radiosurgical and stereo-

81

tactic radiotherapy systems, the CT study is utilized for both target delineation and dose calculation; however, with the CyberKnife, the CT also plays a critical role in the imageguidance aspect of the delivery process. If MRI images are to be used to enhance anatomic delineation, then the MRI study must be fused to the CT study prior to beginning the planning process. The integrated image-guidance system allows the CyberKnife to target any treatment site that can be located via radiographic landmarks. Targeting is achieved by registering the landmarks in the kilovoltagae (kV) alignment images with their counterparts in digitally reconstructed radiographs (DRRs) derived from the treatment planning CT study. There are two basic strategies for localization. Sites that maintain a rigid relationship with bony landmarks that are easy to image (such as intracranial lesions) can be located by measuring the position of the skeletal features in the treatment room. Sites in soft tissue or along the spine are located with the assistance of artiﬁcial ﬁducial markers implanted near the lesion in a manner similar to those being employed by other extracranial systems. This targeting capability enables the user of the CyberKnife to plan treatments for central nervous system (CNS) lesions in the cranium and anywhere along the spine, as well as soft-tissue sites in the thorax, abdomen, and pelvis. Although the general treatment planning procedure is not site-speciﬁc, each site introduces some distinctive elements to the planning process. The important role that the planning CT study plays in target localization and beam alignment during treatment places additional demands on its spatial resolution. To achieve optimal targeting accuracy [70], it is recommended that the CT slice thickness not exceed 1.25 mm [71]. Although this slice thickness is commonly used for intracranial lesions, it is a higher spatial resolution than is typically used for routine diagnostic studies of extracranial sites. Central nervous system lesions, including tumors and AVMs, are the most commonly treated sites. As examples of CNS applications, we take note of two specialized treatment planning problems: trigeminal neuralgia and spinal lesions. The trigeminal nerve is visualized for planning using CT cisternography [72]. The patient is scanned over the full head in the Trendelenberg position. Typically, 64 cGy is prescribed to the 80% isodose line, encompassing approximately 8 mm of nerve, and is delivered in one fraction. Beam alignment and tracking during treatment is based on the position of the cranium visualized in the treatment room radiographs. Treatments of lesions along the spine are delivered in 1 to 5 fractions. Targeting is based on ﬁducials implanted into the spine near the treatment site. The planning CT study (Fig. 7-18) is set up to visualize the region in and around the lesion, plus the targeting ﬁducials, with the patient lying supine in an alpha cradle. The typical treatment dose is 1200 to 2000 cGy prescribed to the 80% isodose line. Dose to the cord is limited to 800 cGy. Experience from numerous treatments has shown that patients resting supine in an alpha cradle move less than 3 mm over the course of a 30-minute treatment fraction [9]. The image-guidance system detects and corrects for intrafraction movement. In addition to CNS sites, the CyberKnife has been used to treat pancreatic, lung, prostate and other soft tissue tumors. These represent innovations in the application of radiosurgery

82

d.m. shepard et al.

FIGURE 7-18. Optimized plan for a spinal lesion.

and thus have not yet evolved standard protocols for treatment planning.

The Planning Process The CyberKnife treatment planning process combines a variety of user-managed and automatic planning operations to arrive at satisfactory deliverable plans. The planning system uses a CT study as the primary planning resource. The dose calculation process makes a coarse accommodation of anatomic inhomogeneities by distinguishing air, tissue, and bone based on the Hounsﬁeld number. The robotic manipulator has a great deal of ﬂexibility in positioning and aiming the treatment beam. This makes it possible to deliver both isocentric and non-isocentric dose distributions. To utilize the full beneﬁt of this ﬂexibility and at the same time make planning calculations tractable, the planning system works with a discrete set of linac positions (called nodes) and a discrete set of pointing directions at each node analogous to a multitude of isocenter-gantry-table positions in routine radiosurgery planning. During the delivery process, the robotic manipulator moves the linac from one node to another. At each node, the robot stops, aims the linac in the selected directions, and delivers dose increments in a “stop and shoot” sequence. The beam is not on while the robot is in motion.

A typical treatment plan has at its disposal a set of up to 110 nodes distributed approximately uniformly over about onehalf of a sphere centered on the treatment site. For each plan, these nodes are selected from a total of more than 300 possible linac positions. For non-isocentric delivery, the planning system deﬁnes up to 12 discrete pointing directions (vectors) at each node. These pointing vectors are aligned to designated points within the treatment volume to produce a set of overlapping beams designed to optimally cover the target volume while avoiding critical structures. Figure 7-19 illustrates this approach schematically. The combination of nodes and pointing vectors provides altogether a set of 1320 “beams” from which to construct a plan. The plan is developed by selecting beams (i.e., nodes and pointing vectors) from among the 1320 available and assigning them “weights” corresponding with the amount of dose each beam is to deliver. The planning process consists of a sequence of planning choices by the clinician combined with a set of constraints on the dose actually delivered by the plan. The ﬁrst step is to identify the treatment site. Because each anatomic location is optimally treated by a particular conﬁguration of nodes, the planning system has predetermined groups of nodes that are considered the best choices for each treatment site. These groups of nodes are called “paths.” There are 80 to 110 nodes in each path, chosen from among 300+ possible linac positions. When the

FIGURE 7-19. (a, b) Illustration of isocentric versus non-isocentric delivery.

treatment planning for stereotactic radiosurgery

clinician designates the treatment site, the planning system selects the appropriate path. In the second step, the clinician chooses to construct either a single-isocenter, a multiple-isocenter, or a non-isocentric plan. Although the isocentric option is reasonable for compact semispherical lesions, the non-isocentric technique is most commonly employed due to its greater ﬂexibility. The third step is to choose either forward or inverse planning. Non-isocentric plans are difﬁcult to design using forward planning and so inverse planning is typically used. It will be assumed from this point on that the clinician is interested in inverse planning. One or more collimators can be employed in a plan. Typically, for isocentric plans, the collimator diameter is matched to the diameter of the lesion. In non-isocentric delivery, the clinician has more freedom in selecting the collimator diameter(s). Typically, the collimator diameter is chosen to be approximately 60% of the lesion’s smallest projected cross-section. This makes it possible to obtain acceptable dose conformality, dose homogeneity, and dose fall-off at the boundary of the target without unduly lengthening the treatment time. As mentioned above, the beam diameter can be ﬁne-tuned by adjusting the SSD away from its nominal distance of 80 cm. Inverse planning works within a set of constraints to ﬁnd the best deliverable plan. The clinician speciﬁes a minimum target dose and a maximum dose to each critical structure that can potentially be transited by a beam. Additionally, the planner can create artiﬁcial dose constraint structures to further inﬂuence the plan. These structures can be used to attenuate or block beams so as to shield critical structures. They can also be used to modulate the dose reaching the lesion. The effective use of artiﬁcial constraints is a valuable acquired skill. In the contouring process, the clinician outlines the target volume, all relevant critical structures, and any artiﬁcial structures needed to help manage the dose distribution. Part of this step involves identifying up to 12 locations within the target lesion that serve as end points for the pointing vectors at each node. After the contouring is complete, the planning system has up to 1320 beams to choose from when optimizing the plan. The inverse calculation then selects from among the available beams to get optimal target coverage and critical structure avoidance. The optimizer then assigns beam weights that meet the dose constraints. If constraints applied in an inverse planning formulation are too strict, the system may be unable to ﬁnd a plan that simultaneously satisﬁes all of the constraints. Experienced CyberKnife planners have found it most effective to begin by deﬁning minimum target doses combined with fairly generous maximum doses to the critical structures. If the system can ﬁnd a solution for this ﬁrst approximation, the clinician then reduces the maximum doses to the critical structures until either the necessary constraints are met or a solution is no longer possible. If the obtainable solutions cannot quite meet the original critical structure constraints, then the clinician must decide if the best obtainable solution is good enough. After an optimized plan has been calculated, the clinician has an opportunity to ﬁne-tune it in a pseudo-forward planning step by selectively turning individual beams on and off. The ﬁnal decision made in the planning process is whether the original fractionation scheme should be modiﬁed in light of

83

the achievable dose limits to the target volume and critical structures. This is an important consideration for spine treatments because there is incomplete knowledge of the spinal cord’s tolerance of single-fraction doses.

Proton SRS There are currently 23 clinical proton radiotherapy facilities in the world (9 proposed over the next 4 years) [73]. The upfront costs of around $100 million as well as the operational costs are the primary reasons there are so few proton facilities. Of those centers offering proton therapy, fewer than half have an active radiosurgery program. By comparison, there are approximately 200 Gamma Knife centers worldwide [59] and a much larger number of linac-based SRS programs. The major advantage of high-energy protons and other heavy charged particles is the characteristic distribution of dose with depth (Fig. 7-20). As the beam passes through tissue, the dose initially remains approximately constant. Near the end of the range, however, the dose quickly increases to its peak value then rapidly falls off to near zero dose. The region of high dose at the end of the particle range is called the Bragg peak [74]. By adjusting the incident energy of the beam, the position of the Bragg peak can be controlled to match the depth of the lesions being treated. This is, however, a case of too much of a good thing. The Bragg peak is usually much narrower than the span of the average target. This in turn requires that the peak be “spread out” or modulated. The effect of this modulation is an increase in the surface dose of the entrance beam (Fig. 7-21). Proton delivery systems are not compact like conventional linacs or Gamma Knife units. Until recently, proton facilities that were originally designed for nuclear research used retroﬁtted equipment to bring the radiation to the patients. This meant using ﬁxed beamlines and innovative methods to deliver ﬁelds from various directions. Modern proton facilities designed and built for radiation therapy have incorporated gantry systems making their use similar to linacs [75, 76] (Fig. 7-22).

Dose Modeling Doses from proton beams are clinically described in cobalt Gray equivalent (CGE). The CGE represents the equivalent Photon & Proton Depth Doses 190 MeV Photons

100

170 MeV Photons 150 MeV Photons

Relative Dose (%)

7.

130 MeV Photons

75

6 MeV Photons

50 25 0 0

5

10 15 20 25 Depth in Water (cm) FIGURE 7-20. Dose distributions for 6-MV photons and protons of various energies.

84

d.m. shepard et al.

FIGURE 7-21. (a) SOBP dose distribution designed with a 16-cm range to the distal 90% and 3-cm modulation. The SOPB consists of ﬁve pristine peaks with respective ranges of 16.0, 15.4, 14.8, 14.2, and 13.6 cm, with relative weights of 0.488, 0.168, 0.137, 0.098, and 0.109. The proximal tail of the individual pristine peaks is omitted on all but two curves to improve the visual clarity of the graph. (b) Pristine Bragg peaks with the same energy but varying ﬁeld diameters from 1.0 to 4.8 cm. The smaller ﬁelds lack lateral dose equilibrium because more protons diverge from the central axis than converge to the central axis. Thus, for the same delivered monitor units, fewer protons reach the isocenter with small ﬁelds compared with large ﬁelds. The ﬁeld size effect is also energy dependent.

100

Dose

80

60

Relative Dose (%)

90

40 45 Ø 10 – 48 mm 20

20

5 16 Depth in Water (cm) 13 FIGURE 7-22. (a) The STAR (Stereotactic Alignment Radiosurgery) isocentric patient positioning device designed and built by Product Genesis Inc. and used by the Massachusetts General Hospital (Boston) group in conjunction with horizontal ﬁxed proton beamlines. The patient can be rotated enabling any portal combination aimed from the

top cranial hemisphere. (b) A 110-ton isocentric gantry and 6-axis robotic patient positioner designed and built as a collaborative effort between General Atomic Inc. and Ion Beam Application Inc. and used for proton SRT and SRS at the Northeast Proton Therapy Center, Massachusetts General Hospital, Boston.

7.

treatment planning for stereotactic radiosurgery

biological dose to achieve the same cell-kill as 60Co gamma rays. The ratio of the physical to biological dose is called the radiobiological effectiveness (RBE). Clinical applications of protons assume an RBE of 1.1 relative to 60Co [77, 78]. This means that 10% less physical dose is required when delivering protons beams compared with X-rays. The Bragg peak measured for a single energy beam is referred to as a “pristine” Bragg peak. Figure 7-20 shows pristine Bragg peaks for beams of various energies measured in a water phantom. The width of a pristine Bragg peak as measured from the proximal to distal 90% dose varies with energy. The width is narrower for shallow ﬁelds and broader for deeper ﬁelds, however, a typical width is around 0.5 cm. In order to treat larger lesions, multiple pristine Bragg peaks are combined to create a spread out Bragg peak (SOBP) (see Fig. 7-21a). The SOPB 90–90% width (also known as the modulation width) can be customized to the thickness of any lesion by varying the depth and weight of each pristine peak. It is important to note that the relative weights and depth pullbacks of the individual pristine peaks needed to generate a ﬂat SOBP are dependent not only on the 90–90% width but also on their entire depth dose proﬁle. The pristine peak shape is also affected by the ﬁeld diameter (see Fig. 7-21b). Therefore, for a speciﬁc beamline, both the energy and the ﬁeld diameter must be modeled in the treatment planning algorithm to properly generate the desired SOBP. A SOPB can be generated by delivering the dose from each pristine Bragg peak separately or by using beam spreading devices such as such as spinning absorber wheels or ridge ﬁlters [79]. The lateral dose penumbra of proton beams depend both on the inherent beam source geometry as well as individual treatment ﬁeld settings. Beamlines designed to treat large ﬁelds generally use a double scattering system, which can result in a 5-cm FWHM source size. When this source size is combined with a source-to-axis-distance (SAD) of 220 cm that is typical for a proton gantry system, the result is an 80–20% isocenter plane penumbra of 2.8 mm for a range of 8 cm, and 6.3 mm for a range of 16 cm. In a small-ﬁeld single-scattering system with an SAD of 450 cm, the source size is around 1 cm FWHM, and the penumbra is reduced to 2.3 mm for a range of 8 cm, and 5.3 mm for a range of 16 cm. Treatment planning systems with integrated proton beam algorithms, such as CMS Focus (CMS, Inc.), must be able to model both the lateral and depth dose proﬁles based on as many clinical variables as possible. This is achieved by using pencil beam algorithms that model multiple Coulomb scattering within beam-modifying devices as well as the patient [80, 81].

Preplanning Consideration Compared with linac and Gamma Knife delivery, protons result in a lower integral dose to the patient due to the sharp dose fall-off beyond the Bragg peak. Larger lesions beneﬁt the most from the reduction of dose to normal brain from protons compared with other modalities. Protons also make it possible to deliver a uniform target dose even with large lesions. This may be considered an advantage when treating AVM, where the target and normal brain are intertwined; however, additional clinical results are needed to demonstrate this conclusively.

85

The immobilization hardware and alignment techniques used for linac SRS can be applied to proton radiosurgery particularly when a gantry-based beamline is used. In ﬁxed beamlines, the need to rotate the patient can lead to torque on the patient’s head that can result in slight shifts of the head relative to an external reference frame. In such cases, it would be unwise to rely on conventional stereotactic external reference frames. The insertion of three 1/16-inch-diameter, 316LVM-grade stainless steel spheres in the outer table of the patient’s skull used in conjunction with a diagnostic-quality X-ray treatment alignment system provides a reliable reference coordinate system [82]. A beneﬁt of using an internal marker system is the ﬂexibility to image, plan, and treat on different days. This is an important consideration for proton facilities that have signiﬁcant overhead due to the need to fabricate custom beam shaping devices, which require pretreatment quality assurance/quality control and possibly dosimetric veriﬁcation. Patients also beneﬁt from not having to wear a stereotactic frame while waiting for their treatments as is necessary with same-day CT-plan-treat schedules. Using internal alignment markers does not necessarily preclude the use of stereotactic bony ﬁxation for CT imaging and/or the treatment. It does, however, provide the ﬂexibility to use stable noninvasive immobilization devices as assessed on an individual patient basis. Treatment planning for proton therapy requires CT imaging because the dose algorithms convert the pixel densities to proton stopping power. The stopping powers are used to calculate the penetration of the protons inside the patient. When required, secondary imaging studies such as positron emission tomography, MRI, and angiography can be fused or merged with the primary planning CT as is done with conventional treatments [83, 84]. Because of the proton penetration conversion from CT densities and the sharp dose fall-off beyond the Bragg peak, special precautions are necessary to ensure that the planning CT’s densities are not artiﬁcially altered. This means that if a CT requires contrast for target and normal structure delineation, it may be necessary to ﬁrst obtain a noncontrast CT to be used for treatment planning. A secondary scan performed with contrast is used for structure delineation and fused to the noncontrast planning CT. This is especially important when treating vascular lesions, where using the contrast CT scan to calculate the proton stopping powers would overestimate the required proton range and modulation and result in an increased treatment volume.

Clinical Treatment Planning Proton arc therapy is not feasible nor is it necessary to generate conformal plans. Three to ﬁve static ports are generally sufﬁcient to obtain satisfactory target dose conformity. Using BEV projections, the user determines the best combination of ports to optimize the dose conformality. When practical, the use of orthogonal beams minimizes overlap regions between ﬁelds. Brass apertures are custom milled to the shape of the lesion’s BEV projection (Fig. 7-23). The expansion of the aperture openings accounts for the penumbra of the speciﬁc portal. Blocking of speciﬁc critical structures is made easy using the BEV approach. Custom Lucite range compensators are fabricated to enable dose shaping to conform to the distal edge of the target.

86

d.m. shepard et al.

FIGURE 7-23. A custom brass aperture and Lucite compensator used to ensure a conformal proton beam.

Although the beam directions are determined on a case-bycase basis, an effort is made to use standard beam arrangements. Pituitary adenomas are located in and around sella and have similar dose constraints. A standard approach at the Massachusetts General Hospital (MGH) for pituitary lesions has been to use four ﬁelds (RL, LL, ASO, PSO) (Fig. 7-24). The lateral ﬁelds avoid the optic structures and brain stem while passing through the temporal lobes; the PSO avoids the optic structures and temporal lobes while passing through the brain stem; and the reduced-weight ASO avoids the temporal lobes and brain stem while passing through the optic structures. Another standardized approach is used to treat acoustic neuromas using three to four ﬁelds (L/RPO, L/RAO, L/RSO, L/RPSO). This ﬁeld combination limits the dose to the brain stem, cerebellum, and temporal lobe while avoiding bone heterogeneities as much as possible. In other cases such as AVMs, the variability in location, shape, and size make standardization difﬁcult (Fig. 7-25). In order to maximize dose uniformity across targets, it has been the MGH practice, for proton SRS, to avoid truncating a target so as to treat it using abutting or patch ﬁelds [85]. These techniques are regularly used when treating with fractionated proton SRT to improve dose conformality of very irregular targets; however, they create small, local high-dose points within the treatment volume.

FIGURE 7-24. A standard four-ﬁeld pituitary beam combination used for PSRS. The ﬁrst image shows a 3D view of the ports and the structures of interest. Typical doses are 18 to 20 CGE at 90% to the target with constraints of 8 and 12 CGE to the surface of the optic structures and brain stem, respectively.

7.

treatment planning for stereotactic radiosurgery

87

FIGURE 7-25. Six-ﬁeld PSRS radiosurgery isodose plan for a 9.6-cm3 right thalamic AVM in a 17-year-old patient. The treatment was delivered in two sessions, separated by 2 weeks, each delivering 8 CGE at

90% for a total of 16 CGE at 90%. Three ﬁelds (LL, RPO, ASO) were delivered at the ﬁrst session and three different ﬁelds (RL, LAO, LSO) were delivered at the second session.

Patient demographics are dependent on both the practice referral patterns as well as the facility’s ability to offer alternative treatments. Proton facilities that do not have an alternative modality option are likely to have a signiﬁcant percentage (~25%) of patients with metastatic disease compared with very few if an alternative such as linac SRS is available. At the MGH, 47% of PSRS patients have AVMs, 19% pituitary or cavernous sinus lesions, 13% acoustic neuromas, 11% menin-

giomas, 4% extracranial, and 6% have miscellaneous other disease.

Other Considerations When one is considering the overall advantage of protons for SRS, one should compare the dose gradient relative to standard Gamma Knife or linac-based systems. As with any

88

d.m. shepard et al.

such treatment device, there exist unique planning situations at which it will excel. When one considers, however, that the Gamma Knife and the linac-based system routinely produce plans that expose as little as 5% to 10% of normal nontarget tissues in the prescription isodose shell while providing dose gradients that drop the dose from the prescription dose to half of the prescription dose in 2.5 to 4 mm, the margin for improvement is relatively small. The image-guidance techniques available with linac-based systems such as cone-beam CT could also impact the degree of the advantage seen with proton therapy for SRS and SBRT.

Conclusion Technological advances in recent years have created a new array of options for SRS. The dosimetric advantage of SRS has also been gradually extended to extracranial sites. The introduction of micro-multileaf collimators, IMRT, and inverse planning into SRS gives us more control in shaping the highdose volume to conform to the target. Online or real-time image guidance also provides a new alternative to the use of traditional stereotactic frames. With these advances, fractionated treatments can now be reliably administered. Although the dosimetric goals and basic principles of SRS planning have not changed, the means of achieving the dosimetric goals has shifted more toward automation and computer optimization.

References 1. Ganz JC. Gamma Knife Surgery, 2nd ed. New York: SpringerVerlag, 1997:205. 2. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry 1983; 46(9):797–803. 3. Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two cases of intractable pain. Acta Chir Scand 1968; 134(8):585–595. 4. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102(4):316–319. 5. Wagner TH. Optimal delivery techniques for intracranial stereotactic radiosurgery using circular and multileaf collimators. In: Nuclear and Radiological Engineering. Gainesville, FL: University of Florida, 2000. 6. Takeuchi H, Yoshida M, Kubota T, et al. Frameless stereotactic radiosurgery with mobile CT, mask immobilization and micromultileaf collimators. Minim Invasive Neurosurg 2003; 46(2): 82–85. 7. Ryken TC, Meeks SL, Pennington EC, et al. Initial clinical experience with frameless stereotactic radiosurgery: analysis of accuracy and feasibility. Int J Radiat Oncol Biol Phys 2001; 51(4):1152– 1158. 8. Murphy MJ, Cox RS. The accuracy of dose localization for an image-guided frameless radiosurgery system. Med Phys 1996; 23(12):2043–2049. 9. 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(5):1400–1408. 10. Kamath R, Ryken TC, Meeks SL, et al. Initial clinical experience with frameless radiosurgery for patients with intracranial metastases. Int J Radiat Oncol Biol Phys 2005; 61(5):1467–1472. 11. Gerszten PC, Ozhasoglu C, Burton SA, et al. Evaluation of CyberKnife frameless real-time image-guided stereotactic radiosurgery for spinal lesions. Stereotact Funct Neurosurg 2003; 81(1–4):84–89.

12. Bova FJ, Buatti JM, Friedman WA, et al. The University of Florida frameless high-precision stereotactic radiotherapy system. Int J Radiat Oncol Biol Phys 1997; 38(4):875–882. 13. Meeks SL, Bova FJ, Wagner TH, et al. Image localization for frameless stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 2000; 46(5):1291–1299. 14. Buatti JM, Bova FJ, Friedman, et al. Preliminary experience with frameless stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 1998; 42(3):591–599. 15. Gehring MA, Mackie TR, Kubsad SS, et al. A three-dimensional volume visualization package applied to stereotactic radiosurgery treatment planning. Int J Radiat Oncol Biol Phys 1991; 21(2): 491–500. 16. Friedman WA, Bova FJ. The University of Florida radiosurgery system. Surg Neurol 1989; 32(5):334–342. 17. Coy SR, Houdek PV. Radiosurgery treatment planning. Semin Radiat Oncol 1995; 5(3):213–219. 18. Ferris MC, Lim J, Shepard DM. Radiosurgery treatment planning via nonlinear programming. Ann Operations Res 2003; 119:247– 260. 19. Ferris MC, Shepard DM. Optimization of Gamma Knife radiosurgery. In: Du Pardolas DZ, Wang J, eds. Discrete Mathematical Problems with Medical Applications. Providence, RI: AMS, 2000. 20. Flickinger JC, Lunsford LD, Wu A, et al. Treatment planning for gamma knife radiosurgery with multiple isocenters. Int J Radiat Oncol Biol Phys 1990; 18(6):1495–1501. 21. Luo L, Shu H, Yu W, et al. Optimizing computerized treatment planning for the Gamma Knife by source culling. Int J Radiat Oncol Biol Phys 1999; 45(5):1339–1346. 22. Shepard DM, Ferris MC, Ove R, Ma L. Inverse treatment planning for Gamma Knife radiosurgery. Med Phys 2000; 27(12): 2748–2756. 23. 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):420–427. 24. Zhang P, Dean D, Metzger A, Sibata C. Optimization of Gamma knife treatment planning via guided evolutionary simulated annealing. Med Phys 2001; 28(8):1746–1752. 25. Kubo HD, Pappas CT, Wilder RB. A comparison of arc-based and static mini-multileaf collimator-based radiosurgery treatment plans. Radiother Oncol 1997; 45(1):89–93. 26. Podgorsak EB, Pike GB, Olivier A, et al. Radiosurgery with high energy photon beams: a comparison among techniques. Int J Radiat Oncol Biol Phys 1989; 16(3):857–865. 27. Alheit H, Dornfeld S, Dawel M, et al. Patient position reproducibility in fractionated stereotactically guided conformal radiotherapy using the BrainLab mask system. Strahlenther Onkol 2001; 177(5):264–268. 28. Bourland JD, McCollough KP. Static ﬁeld conformal stereotactic radiosurgery: physical techniques. Int J Radiat Oncol Biol Phys 1994; 28(2):471–479. 29. Cardinale RM, Benedict SH, Wu Q, et al. A comparison of three stereotactic radiotherapy techniques; ARCS vs. noncoplanar ﬁxed ﬁelds vs. intensity modulation. Int J Radiat Oncol Biol Phys 1998; 42(2):431–436. 30. Grebe G, Pfaender M, Roll M, et al. Dynamic arc radiosurgery and radiotherapy: commissioning and veriﬁcation of dose distributions. Int J Radiat Oncol Biol Phys 2001; 49(5):1451–1460. 31. Hamilton RJ, Kuchnir FT, Sweeney SJ, et al. Comparison of static conformal ﬁeld with multiple noncoplanar arc techniques for stereotactic radiosurgery or stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 1995; 33(5):1221–1228. 32. Laing RW, Bentley RE, Nahum AE, et al. Stereotactic radiotherapy of irregular targets: a comparison between static conformal beams and non-coplanar arcs. Radiother Oncol 1993; 28(3):241– 246.

7.

treatment planning for stereotactic radiosurgery

33. Shiu AS, Kooy HM, Ewton JR, et al. Comparison of miniature multileaf collimation (MMLC) with circular collimation for stereotactic treatment. Int J Radiat Oncol Biol Phys 1997; 37(3): 679–688. 34. Soanes T, Hampshire A, Vaughan C, et al. The commissioning and quality assurance of the Automatic Positioning System on the Leksell gamma knife. J Neurosurg 2002; 97(5 Suppl):574– 578. 35. Tsai JS, Engler MJ, Ling MN, et al. A non-invasive immobilization system and related quality assurance for dynamic intensity modulated radiation therapy of intracranial and head and neck disease. Int J Radiat Oncol Biol Phys 1999; 43(2):455–467. 36. Tome WA, Meeks SL, Buatti JM, et al. A high-precision system for conformal intracranial radiotherapy. Int J Radiat Oncol Biol Phys 2000; 47(4):1137–1143. 37. 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. 38. Haedinger U, Krieger T, Flentje M, Wulf J. Inﬂuence of calculation model on dose distribution in stereotactic radiotherapy for pulmonary targets. Int J Radiat Oncol Biol Phys 2005; 61(1):239– 249. 39. Chaves A, Lopes MC, Alves CC, et al. A Monte Carlo multiple source model applied to radiosurgery narrow photon beams. Med Phys 2004; 31(8):2192–2204. 40. Pike B, Peters TM, Podgorsak E, et al. Stereotactic external beam calculations for radiosurgical treatment of brain lesions. Appl Neurophysiol 1987; 50(1–6):269–273. 41. Pike B, Podgorsak EB, Peters TM, Pla C. Dose distributions in dynamic stereotactic radiosurgery. Med Phys 1987; 14(5):780– 789. 42. 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):941–949. 43. Wu X, Ting JY, Markoe AM, et al. Stereotactic dose computation and plan optimization using the convolution theorem. I. Dose computation. Stereotact Funct Neurosurg 1996; 66(Suppl 1):302– 308. 44. Solberg TD, Holly FE, De Salles AA, et al. Implications of tissue heterogeneity for radiosurgery in head and neck tumors. Int J Radiat Oncol Biol Phys 1995; 32(1):235–239. 45. Kubsad SS, Mackie TR, Gehring MA, et al. Monte Carlo and convolution dosimetry for stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1990; 19(4):1027–1035. 46. Dong, L, Shiu A, Tung S, Hogstrom K. A pencil-beam photon dose algorithm for stereotactic radiosurgery using a miniature multileaf collimator. Med Phys 1998; 25(6):841–850. 47. Bardash M, Amols HI, Kohn S, et al. Rapid dose calculations for stereotactic radiosurgery. Med Phys 1992; 19(4):965–970. 48. Ayyangar KM, Jiang SB. Do we need Monte Carlo treatment planning for linac based radiosurgery? A case study. Med Dosim 1998; 23(3):161–168. 49. Verellen D, Linthout N, Bel A, et al. Assessment of the uncertainties in dose delivery of a commercial system for linac-based stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1999; 44(2):421–433. 50. Verhey LJ, Smith V, Serago CF. Comparison of radiosurgery treatment modalities based on physical dose distributions. Int J Radiat Oncol Biol Phys 1998; 40(2):497–505. 51. Smith V, Verhey L, Serago CF. Comparison of radiosurgery treatment modalities based on complication and control probabilities. Int J Radiat Oncol Biol Phys 1998; 40(2):507–513. 52. Shaw E, Kline R, Gillin M, et al. Radiation Therapy Oncology Group: radiosurgery quality assurance guidelines. Int J Radiat Oncol Biol Phys 1993; 27(5):1231–1239.

89

53. Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. Technical note. J Neurosurg 2000; 93(Suppl 3):219–222. 54. Nedzi LA, Kooy HM, Alexander E 3rd, et al. Dynamic ﬁeld shaping for stereotactic radiosurgery: a modeling study. Int J Radiat Oncol Biol Phys 1993; 25(5):859–869. 55. Lomax NJ, Scheib SG. Quantifying the degree of conformity in radiosurgery treatment planning. Int J Radiat Oncol Biol Phys 2003; 55(5):1409–1419. 56. Leung LH, Chua DT, Wu PM. A new tool for dose conformity evaluation of radiosurgery treatment plans. Int J Radiat Oncol Biol Phys 1999; 45(1):233–241. 57. Knoos T, Kristensen I, Nilsson P. Volumetric and dosimetric evaluation of radiation treatment plans: radiation conformity index. Int J Radiat Oncol Biol Phys 1998; 42(5):1169–1176. 58. 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– 232. 59. Elekta AB. 60. Cho PS, Kuterdem HG, Marks RJ 2nd. A spherical dose model for radiosurgery plan optimization. Phys Med Biol 1998; 43(10):3145– 3148. 61. Ferris MC, Lim J, Shepard DM. An optimization approach for radiosurgery treatment planning. SIAM J Optimiz 2003; 13(3):921– 937. 62. Shepard DM, Chin LS, DiBiase SJ, et al. Clinical implementation of an automated planning system for gamma knife radiosurgery. Int J Radiat Oncol Biol Phys 2003; 56(5):1488–1494. 63. Wu QJ, Bourland JD. Three-dimensional skeletonization for computer-assisted treatment planning in radiosurgery. Comput Med Imaging Graph 2000; 24(4):243–251. 64. Wu QJ, Bourland JD. Morphology-guided radiosurgery treatment planning and optimization for multiple isocenters. Med Phys 1999; 26(10):2151–2160. 65. Lutz W, Winston KR, Maleki N. A system for stereotactic radiosurgery with a linear accelerator. Int J Radiat Oncol Biol Phys 1988; 14(2):373–381. 66. Ma L, Yu CX, Earl M, et al. Optimized intensity-modulated arc therapy for prostate cancer treatment. Int J Cancer 2001; 96(6): 379–384. 67. Kramer BA, Wazer DE, Engler MJ, et al. Dosimetric comparison of stereotactic radiosurgery to intensity modulated radiotherapy. Radiat Oncol Invest 1998; 6(1):18–25. 68. Adler JR Jr, Murphy MJ, Chang SD, Hancock SL. Image-guided robotic radiosurgery. Neurosurgery 1999; 44(6):1299–1306; discussion 1306–1307. 69. Chang SD, Murphy M, Geis DP, et al. Clinical experience with image-guided robotic radiosurgery (the CyberKnife) in the treatment of brain and spinal cord tumors. Neurol Med Chir (Tokyo) 1998; 38(11):780–783. 70. Chang SD, Main W, Martin DP, et al. An analysis of the accuracy of the CyberKnife: a robotic frameless stereotactic radiosurgical system. Neurosurgery 2003; 52(1):140–146; discussion 146–147. 71. Heilbrun MP. CyberKnife Radiosurgery: A Practical Guide. The CyberKnife Society, 2003. 72. Mawad ME, Silver AJ, Hilal SK, Ganti SR. Computed tomography of the brain stem with intrathecal metrizamide. Part I: the normal brain stem. AJR Am J Roentgenol 1983; 140(3):553– 563. 73. Sisterson J. Sponsored by the Particle Therapy Co-Operative Group (PTCOG). Particles Newsletter 2005; 35. 74. Kahn FM. The Physics of Radiation Therapy, 3rd ed. Philadelphia: Lippincott Williams & Wilkins, 2003:560.

90

d.m. shepard et al.

75. Slater J, Miller D, Archambeau J. Development of a hospital based proton therapy center. Int J Radiat Oncol Biol Phys 1988; 14(4):761–775. 76. Flanz J, Bailey J, Bradley S, et al. Recent performance of the NPTC equipment compared with the clinical speciﬁcations. In: Proceedings of 15th Conference on the Application of Accelerators in Research and Industry. AIP Press, 1999. 77. Raju MR. Proton radiobiology, radiosurgery and radiotherapy. Int J Radiat Biol 1995; 67(3):237–259. 78. Paganetti H, Niemierko A, Ancukiewicz M, et al. Relative biological effectiveness (RBE) values for proton beam therapy. Int J Radiat Oncol Biol Phys 2002; 53(2):407–421. 79. Chu W, Ludewigt B, Renner T. Instrumentation for treatment of cancer using proton and light-ion beams. Rev Sci Instrum 1993; 64(8):2055–2122. 80. Russell KR, Isacsson U, Saxner M, et al. Implementation of pencil kernel and depth penetration algorithms for treatment planning of proton beams. Phys Med Biol 2000; 45(1):9–27. 81. Hong L, Goitein M, Bucciolini M, et al. A pencil beam algorithm for proton dose calculations. Phys Med Biol 1996; 41(8):1305– 1330. 82. Gall KP, Verhey LJ, Wagner M. Computer-assisted positioning of radiotherapy patients using implanted radiopaque ﬁducials. Med Phys 1993; 20(4):1153–1159. 83. Wells WM 3rd, Viola H, Atsumi S, et al. Multi-modal volume registration by maximization of mutual information. Med Image Anal 1996; 1(1):35–51. 84. van Herk M, Kooy HM. Automatic three-dimensional correlation of CT-CT, CT-MRI, and CT-SPECT using chamfer matching. Med Phys 1994; 21(7):1163–1178. 85. Bussiere MR, Adams JA. Treatment planning for conformal proton radiation therapy. Technol Cancer Res Treat 2003; 2(5):389–399.

Glossary conformity index deﬁned by the RTOG as the volume of the prescription isodose divided by the target volume constraint a condition that must be satisﬁed during inverse treatment planning for a plan to be considered feasible; the most basic constraint is that all beam weights must be nonnegative dose-volume histogram (DVH) a plot of the volume versus dose used to analyze the dose distribution on a structure by structure basis forward treatment planning an iterative approach to planning where the user manually changes each of the plan parameters until an acceptable plan is obtained homogeneity index deﬁned by the RTOG as the maximum dose in the treatment volume divided by the prescription dose inverse treatment planning an automated approach to planning where the user deﬁnes the treatment goals and an optimization algorithm is run that determines the parameters that best meet the goals micro-multileaf collimator (mMLC) a device attached to or incorporated into the head of a linear accelerator used to deﬁne ﬁeld shapes; compared with a conventional multileaf collimator (MLC), the leaves of a mMLC are less wide and project to 5 mm or less in width at the isocenter objective function a scoring function that reduces the entire treatment plan into a single numerical value that is to be either minimized or maximized

8

Designing, Building and Installing a Stereotactic Radiosurgery Unit Lijun Ma and Martin Murphy

Design Principles The ﬁrst stereotactic radiosurgery (SRS) unit was designed by Swedish neurosurgeon Dr. Lars Leksell in the 1950s [1]. The term stereotactic literally means “spatially ﬁxed.” In general, a SRS procedure involves delivering a single fraction of high-dose radiation, usually with the guidance of a rigid ﬁxation device (i.e., a stereotactic frame). The purpose of the frame is to map out the coordinate system of the target for accurate reference of the radiation beams [2–6]. Common types of radiation used for SRS are high-energy gamma rays (e.g., 60Co), high-energy x-rays, and charged particles such as protons. Building a photon-based (x-rays and gamma rays) or an ion-based (protons and heavy charged particles) SRS unit is different in principle. The goal of the photon-based units is to converge a large number of beams to the isocenter in order to produce rapid dose fall-off outside the focal area. In contrast, the ion-based units use a limited number of beams in order to spread the Bragg peak across the target area. The rapid dose fall-off is largely created via the distal fall-off of the Bragg peak [7–12]. Currently, the photon-based units employ radioactive sources (60Co) and electron linear accelerators. The ion-based units employ high-energy cyclotrons or synchrocyclotrons. Due to the high cost of these heavy-particle accelerators, the use of the ion-based SRS units are limited. In this chapter, we discuss the photon-based SRS units only. The major photon-based SRS units include Gamma Knife, conventional linear accelerator (linac)-based units, CyberKnife, and integrated x-ray systems such as Tomotherapy and Varian Trilogy units. Four design principles are used in building a SRS unit: 1. Fix the source of the radiation and ﬁx the patient position when beam is on. 2. Move the source of the radiation and ﬁx the patient position when beam is on. 3. Move the source of the radiation and also move the patient when beam is on. 4. Fix the source of the radiation and move the patient when beam is on. The ﬁrst principle is used in the Gamma Knife and early proton and heavy-ion systems; the second principle is used in

most linac-based SRS systems including the CyberKnife; the third principle is used in Tomotherapy and dynamic arc linacbased systems; and the fourth principle is uncommon but has been reported in the linac-based rotating-chair SRS system [13].

Major Components and Functions Gamma Knife Gamma Knife is a 60Co unit speciﬁcally designed for intracranial SRS treatments 14–17]. It is solely manufactured by the Elekta company (Stockholm, Sweden). The unit contains 201 60 Co sources ﬁxed on a hemispherical surface to deliver gamma rays of 1.25 MeV. The center of the hemisphere is the isocenter where all the beams from the sources intersect. The initial activity of a single source is about 3.0 Ci. Therefore, the total activity from all the sources exceeds 6000 Ci, which generates a starting dose rate of approximately 300 cGy/min at the isocenter. The source-to-isocenter distance is 40.1 cm for the Gamma Knife unit. This distance is signiﬁcantly shorter than the 80- to 100-cm source-to-axis distance (SAD) of the linear accelerators. Short source-to-focal distance is one of the distinct design characteristics of the Gamma Knife unit. This not only reduces the total source activity for high-dose delivery but also facilitates high mechanical precision near the isocenter. For example, if a small milling error occurs in the primary collimator, the error would be projected signiﬁcantly smaller at a short sourceto-isocenter distance than at a large source-to-isocenter distance. High source activity demands high-density materials for proper shielding of the Gamma Knife unit. With the shielding and the collimator assembly, a Gamma Knife unit weighs about 20 tons. Heavy and robust structure, ﬁxed source conﬁguration, and short source-to-focal distance are three major factors contributing to the high mechanical accuracy of the Gamma Knife unit. Historically, there are several models of Gamma Knife units (model U, model B, and model C). In the United States, the ﬁrst Gamma Knife (model U) was installed at the

91

92

l. ma and m. murphy

FIGURE 8-1. The initial model U Gamma Knife installed in the United States.

University of Pittsburgh in 1987. An illustration of the model U unit is given in Figure 8-1. Following U.S. FDA approval of later models (models B and C), the original model U is being gradually replaced. Major differences between the original and the new models include the couch driving mechanism (hydraulic vs. electrical), patient positioning assembly (manual vs. automatic), collimator design, and source patterns. Despite these differences, the structural layout of the Gamma Knife SRS remains identical for all of the models. Figure 8-2 illustrates the major components of the model C Gamma Knife unit. The key components of a Gamma Knife unit include a shielded source head housing 201 60Co sources, a movable couch, four helmets (18 mm, 14 mm, 8 mm, and 4 mm in beam aperture), removable plug collimators (201 for each helmet), console control units, patient positioning devices, audio/video monitoring devices, and treatment planning systems. The patient positioning device of the Gamma Knife consists of a Leksell frame (pinned to the patient skull) and its attachments to the helmets. The frame is locked onto the helmet through a pair of bars (i.e., trunnions) or the automatic positioning system (APS). The details of the APS device attached to a helmet are illustrated in Figure 8-3. The APS device uses

FIGURE 8-3. Detailed illustration of the patient APS with the mounted helmet: a, Leksell frame; b, shielding door to the source; c, APS device (right side); d, helmet (4-mm shown). (Courtesy of Elekta AB, Stockholm, Sweden.)

a micro step motor to position the Leksell frame to precise localization coordinates. The positional accuracy of the APS system is speciﬁed to be less than 0.2 mm. Figure 8-4 shows four plug collimators on each helmet of the Gamma Knife unit (i.e., 18 mm, 14 mm, 8 mm, and 4 mm). The size of these plug collimators approximately equals the full width at half maximum (FWHM) of the beam proﬁles measured along the major axes at the isocenter. Precise alignment of the plug collimator directly affects the precision of the beam. Manufacturing >800 individual plug collimators comprised a major part of the effort in building the Gamma Knife unit. The treatment planning system is a separate component of the Gamma Knife unit. The approved treatment parameters from the treatment planning system are uploaded directly onto the console control computer via a serial cable. The salient features of the system as well as the three-dimensional dose calculation algorithms are discussed in a separate chapter.

Linear Accelerator–Based SRS Units Historically, the linac-based SRS units were built based on the existing linear accelerators using an add-on collimation system and couch stabilization devices [18–29]. The majority of the

FIGURE 8-2. Illustration of the key components of the model C Gamma Knife unit: a, source head unit; b, helmet changer; c, position control pendant; d, docking veriﬁcation pendant; e, helmet, f, patient automatic positioning system (APS) device. (Courtesy of Elekta AB, Stockholm, Sweden.)

FIGURE 8-4. Plug collimators of the Gamma Knife unit. From left to right: 4-mm, 8-mm, 14-mm, and 18-mm collimators.

8.

designing, building and installing a srs unit

93

FIGURE 8-5. A set of tertiary cones for traditional linac-based SRS deliveries. From left to right, the cone diameters are 2 cm, 3 cm, 4 cm, and 5 cm.

linac SRS units still use a set of tertiary cones or add-on micromultileaf collimators to deliver non-coplanar converging arc beams at the isocenter. The arrangement of these non-coplanar arc beams is to spread the beam to form a spherical isodose distribution similar to that of a single-shot delivery of Gamma Knife SRS. Figure 8-5 shows a set of tertiary collimator cones for the linac-based SRS units. The comparison of a 2-cm cone with an 18-mm Gamma Knife plug collimator is shown in Figure 8-6. As illustrated in Figures 8-5 and 8-6, the linac-based SRS cones are signiﬁcantly larger in size than the Gamma Knife collimator for similar beam aperture sizes. The large size of the linac cones is intended for extended source-to-diaphragm distance in order to sharpen the beam penumbra at the isocenter. Unlike Gamma Knife units and early linac systems where the frame isocenter is mechanically ﬁxed at the center of the beam collimators, most modern linac-based SRS units rely on wall-mount lasers and a special localizer box attached to the stereotactic frame for aligning the target coordinates. Therefore, the accuracy of the laser system is critical to the overall treatment accuracy of the linac-based SRS units. As a general rule, a pretreatment isocenter alignment check is carried out for each patient treatment.

FIGURE 8-6. The 18-mm plug collimator of the Gamma Knife unit compared with the 2-cm cone collimator for a linac-based SRS unit.

Figure 8-7 shows the isocenter alignment devices for the linac-based unit and the Gamma Knife unit. In the Gamma Knife alignment tool (Fig. 8-7b), a pin pricks the ﬁlm at the isocenter set by the mechanical coordinates. The ﬁlm is then exposed and the pin-pricked mark is then compared with the center of the beam proﬁle to determine mechanical and radiologic isocenter alignment [4, 21, 26, 30, 31]. For linac-based units, several methods have been reported. The most common method is the Winston-Lutz test method [24, 29]. To carry out Winston-Lutz test, a metallic spherical ball ﬁxed at the tip of a wire is typically used (Fig. 8-7a). The ball is ﬁrst aligned with the isocenter using wall-mount lasers. A series of port ﬁlms are then exposed downstream for a “dry-run” of actual beam deliveries. Typically, the ﬁlm is placed on a special holder attached to the gantry port extended underneath the ball thus allowing its shallow projection onto the ﬁlm. If the linac is equipped with a high-resolution electronic portal imager (EPID) [32], the image can then be taken with the device in

FIGURE 8-7. Isocenter test tools for (a) linac-based SRS units and (b) for the Gamma Knife unit.

94

l. ma and m. murphy

lieu of the ﬁlm. Once exposed, the projected ball center is compared with the ﬁeld edge to detect any deviations from the center of the ﬁeld. The Winston-Lutz test is especially useful for linac-based SRS because it checks the isocenter alignment and also checks potential collisions among the gantry, couch, and tertiary collimator setups during the patient deliveries. Once the alignment between the laser and mechanical isocenter is veriﬁed, the target coordinates for the linac-based SRS treatments are set via a laser-based localizer. A picture of the localizer is given in Figure 8-8. To facilitate alignment of the isocenter, the localizer can be adjusted with 6 degrees of freedom (x, y, z translation and pitch, yaw, and roll) to correct for non-ﬂat couch top and small rotational errors. The coordinates on the localizer are commonly set in positive numbers in order to avoid misreading of positive or negative signs in setting up the coordinates. Common techniques for linac-based SRS use either noncoplanar converging arc beams or dynamic rotation beams. In dynamic rotation delivery, the gantry rotates while the couch rotates (∼150o) at the same time. The advantage of the dynamic arc-beam delivery is that only a single treatment setup is required and no parallel-opposed beam occurs during the entire delivery [26, 27]. With the advent of precise and high-output linear accelerators, the ﬁxed-cone collimator is gradually being replaced by micro-multileaf collimators (mMLCs). In practice, these highresolution multileaf collimators (MLCs) shape the beam aperture conformally in the beam’s-eye view while the gantry rotates for the linac-based SRS delivery. The use of the shaped beams rather than the ﬁxed cones represents a signiﬁcant paradigm shift in the linac-based SRS deliveries as the dose distributions from superposing arcs are no longer elliptical in shape analogous to the Gamma Knife shot delivery. Many investigators classiﬁed the MLC-based SRS as the shaped-beam SRS delivery to distinguish it from the traditional ﬁxed-cone linac-based SRS deliveries [23, 25, 28, 33–35].

Mini-MLC

FIGURE 8-9. A dedicated linear accelerator unit with built-in multileaf collimator for shaped-beam SRS delivery.

For most large tumor treatments, the shaped-beam SRS is capable of delivering the treatments with single isocenter rather than packing multiple spherical shot-like dose distributions inside the target volume. This improves the delivery efﬁciency in terms of total monitor units (MU) as well as the total setup time. With optimized beam weights, the shaped-ﬁeld SRS is expected to deliver more uniform and more conformal dose distribution for large and irregular targets in comparison with the conventional ﬁxed-ﬁeld approach [28]. A picture of a dedicated shaped-beam SRS unit is shown in Figure 8-9. In addition to beam-shaping capability using a built-in MLC or mMLC, a dedicated SRS unit also incorporates the capability of performing online imaging guidance of the delivery. The goal is to register and verify the target locations and to carry out real-time adjustment of the treatment setup. The mechanical accuracy of the unit has been reported to be of the order 0.5 mm [36]. There have been preliminary reports of the use of such dedicated SRS units for functional SRS deliveries such as trigeminal neuralgia treatments [35].

CyberKnife

FIGURE 8-8. A laser-based stereotactic localizer for linac-based SRS treatments.

The CyberKnife [37] is a type of linac-based stereotactic radiosurgery system; however, it differs in two fundamental ways from most conventional linear accelerator systems. First, it does not rely on a stereotactic frame for target localization. Instead, the system employs an integrated X-ray image-guidance system to locate and monitor radiographic landmarks such as the cranium that indicate the target position. Second, the linac is supported and maneuvered by a robotic arm instead of a rotating gantry. This allows the treatment beam to be positioned and aimed with 6 degrees of freedom, which in turn enables the system to adapt to arbitrary patient positions. With this capability, it is not necessary to immobilize the patient at a predeﬁned treatment isocenter. By combining image guidance with complete beam maneuverability, the CyberKnife was the ﬁrst radio-

8.

designing, building and installing a srs unit

surgical system to achieve stereotactic targeting precision without using externally attached devices for either targeting or immobilization. Figure 8-10 shows a typical CyberKnife system. The robotic arm supports an X-band 6-MV linear accelerator. The treatment beam is shaped by cylindrical collimators similar to those used in conventional linac SRS systems. Two orthogonally aligned diagnostic X-ray sources and ﬂat-panel imaging detectors are positioned on either side of the patient. Prior to treatment, the patient undergoes a computed tomography (CT) exam for treatment planning purposes. This CT study also serves as the reference image for the X-ray guidance system. At the time of treatment, the patient lies on the couch in a position that closely but not exactly duplicates his position in the treatment planning CT study. The imaging system acquires a pair of X-ray images of the patient’s anatomy that are compared with a corresponding pair of digitally reconstructed radiographs (DRRs) that have been calculated from the planning CT study. The comparison process establishes the difference between the patient’s pose (position and orientation) in the CT study and his pose at the moment of treatment by matching either bony anatomy or internal radiopaque ﬁducials in the images. Then, rather than use this information to correct the patient position so as to reproduce the planning pose, the position coordinates are automatically sent to the robotic arm, which adjusts the alignment of the treatment beam instead. Thus, the planned beam positions are adapted to the patient position, rather than vice versa. This strategy avoids the need to put the patient at a precisely deﬁned treatment position. Furthermore, it allows the system to adapt to patient movement by continuing to take positioning images throughout treatment. This eliminates the need to rigidly immobilize the patient. The CyberKnife’s frameless alignment capability makes the concept of a ﬁxed mechanical isocenter irrelevant. This carries over to the planning of dose distributions. The treatment beams can be arranged in complex non-isocentric patterns of overlapping rays rather than in the spherically symmetric pattern required by isocentric delivery systems with ﬁxed collimators. This improves the uniformity and conformality of the dose for irregular target volumes.

FIGURE 8-10. The CyberKnife system: 1, diagnostic X-ray sources; 2, 6-MV X-band linac; 3, ﬂat-panel X-ray imaging detectors.

95

Treatment delivery proceeds in a stop, align, and shoot sequence. The total dose for the treatment plan is divided among 80 to 110 linac positions distributed spherically around the patient to obtain the maximum geometric distribution of beams. The robotic arm moves the linac to a designated position (called a node), where it stops while X-ray images are acquired. After automatically processing the images to determine the target position, the linac beam alignment is adjusted and the beam is turned on to deliver a dose increment. The system then advances to the next node and repeats the process. The entire sequence of steps at each node (positioning, image acquisition and analysis, beam realignment, and dose delivery) typically takes 10 to 20 seconds, resulting in a total fraction duration of 20 to 30 minutes. The design, installation, and commissioning of a CyberKnife facility differ in several ways from conventional linear accelerator facilities. For example, the robot can point the beam anywhere in the treatment room. This must be accommodated in the shielding design. Unlike the vault for a gantry-mounted linac, which can concentrate shielding in the plane of gantry rotation, a CyberKnife vault requires a more uniform conﬁguration of shielding. At the same time, though, the beam dwell time in any particular direction is typically less than for a gantry system delivering a plan with the same number of monitor units. This can reduce the required thickness of shielding. Because the robot can move freely in three dimensions, the positions of all physical objects within its workspace (e.g., couch, cameras, etc.) must be programmed into the robot’s workspace model to avoid collisions. If any new object is moved into the workspace without including it in the workspace model, or if an existing object is moved away from its prescribed position, there is risk of a collision. The complete CyberKnife system (robot, linac, imaging system, and couch) is installed as an integrated unit. Because the imaging system is solely responsible for determining beam alignment with the patient, its position calibration with respect to the linac delivery system is critically important. The coalignment of the imaging and robot coordinate frames proceeds in three steps. First, the robot is calibrated to identify a point in its own coordinate frame that is near the center of the imaging coordinate frame. When this step is done, the robot can align the linac beam to within 0.7 mm (RMS variation) of a ﬁxed point from anywhere in its workspace [38]. The second calibration step measures the intrinsic and extrinsic camera models describing the magniﬁcation, distortion, image plane alignment, and other characteristics of the imaging system. This measurement is made with a phantom consisting of a planar grid of radiopaque balls placed in the ﬁeld of view of the two imaging systems. The resulting camera model is used to create the DRRs used for image-guided alignment. Using a well-calibrated camera model, the DRRs can reproduce actual X-ray images with 0.1-mm precision. The last calibration step establishes the position of the imaging coordinate origin within the coordinate frame of the robotic arm. The imaging origin is referred to as a virtual isocenter to distinguish it from the ﬁxed mechanical isocenter of the Gamma Knife or a gantry-based linac SRS system. The virtual isocenter position is ﬁrst estimated using laser and image alignment tools. Then it is measured precisely using a combination dosimetric/imaging phantom. This phantom consists of an

96

l. ma and m. murphy

isocenter within the robot workspace. The mean offset in each direction is used to reﬁne the position of the isocenter in the robot coordinate frame. The random scatter around the mean is then identiﬁed as the residual overall dose alignment error of the system. Through successive reﬁnements in the calibration and analysis process, this residual error has been reduced from 1.7 mm RMS variation [38] to 1.1 mm [39]. Once the system alignment is calibrated, any residual offset (as well as any later perturbation) of the imaging and robot frames appears as a systematic dose offset detectable via the end-to-end phantom calibration process just described. Therefore, it is important to continue these measurements on a regular quality assurance schedule after installation, to detect irregularities in system alignment, and to further reduce any residual systematic offset.

Integrated Units FIGURE 8-11. An imaging/dosimetry phantom used to calibrate and verify the placement of doses within the robotic delivery system frame of reference.

anthropomorphic skull with a cube of radiochromic ﬁlm layers ﬁxed inside (Fig. 8-11). The ﬁlm cube records a threedimensional image of a dose distribution relative to internal imaging landmarks. The dosimetric phantom is scanned in a CT exam and then planned to receive a spherical dose distribution. The phantom is placed on the treatment couch near the center of the imaging system, where it is located via the image-guidance system before receiving the planned dose exposure. Then the phantom is disassembled to measure the center of the actual delivered dose distribution for comparison with the planned position. The deviation is recorded as an offset in (x, y, z). This procedure exactly emulates a treatment from end-to-end and therefore detects all the imprecision in the planning, alignment, and delivery processes. The end-to-end calibration test is repeated a number of times, resulting in a scattered distribution of dose offsets, usually with a nonzero mean. The offset of the mean from zero corresponds with a systematic error in locating the virtual imaging

Several integrated units have recently emerged as SRS delivery units. The purpose of the integrated units is to deliver SRS as well as modern treatments such as adaptive imaging-guided radiation therapy [40, 41]. Two distinct units are the Tomotherapy unit (Tomotherapy, Madison, WI) and the Varian Trilogy Unit (Varian Oncology, Palo Alto, CA). The Tomotherapy unit employs a compact 6-MeV S-band linac mounted on a rotating ring gantry (Fig. 8-12). There are 64 individual binary (either open or shut completely) collimators that shape a fan-beam X-ray across the axial plane. During the delivery, the patient is transported through the bore as the beam rotates around to form a spiral trajectory around the patient. The unit is equipped with a large array of xenon detectors and a kilovoltage (kV) X-ray tube that are capable of performing online CT of the patient. The Trilogy unit is built upon the high-end performance of the Varian Clinac series. In addition to conventional linac ﬂattening ﬁlters, the unit has a separate ﬁlter assembly for delivering enhanced dose rates (>1000 MU/min) to accommodate SRS treatments. The unit is also equipped with either tertiary cones or a built-in multileaf collimator. In addition, the system has a side-mounted kV X-ray tube and a ﬂat-panel imager to provide cone-beam CT imaging capabilities (Fig. 8-12b).

Slip Ring Gantry

Gantry

kV Source Flat panel detector Couch CT Couch

FIGURE 8-12. Two integrated SRS units with online CT capabilities: (a) the Tomotherapy unit; (b) the Trilogy unit. (Part A courtesy of Tomotherapy. Part B courtesy of Varian.)

8.

97

designing, building and installing a srs unit

TABLE 8-1. Estimated overall uncertainties in intracranial SRS deliveries. Source of errors

Imaging studies (resolution and distortions) Mechanical plus setup errors Tissue/target motion Treatment planning Total

Estimated error (mm)

1.0–3.0 0.3–2.0 0.5–1.0 0.5 1.3–3.8

Physically, both the Tomotherapy and the Trilogy systems are suitable for SRS treatments. The Tomotherapy unit uses a fan beam to irradiate the target in the axial plane via slice-byslice fashion; while the Trilogy unit makes use of a cone-beam to irradiate the target from coplanar or nonplanar directions. The mechanical accuracy of these units is estimated to be of the order 0.5 to 1.0 mm. More experiences with these new units for SRS treatments are expected in the future.

Accuracy The treatment accuracy is the distinguishing characteristic of all SRS units. The mechanical accuracy is reported to be less than 0.3 mm for Gamma Knife, 0.5 mm for dedicated SRS units, 0.5 to 0.7 mm for CyberKnife and Tomotherapy units, and 1.0 mm for standard linac-based SRS units. To evaluate the overall physical accuracy of the delivery, other sources of errors also need to be considered. These include (1) uncertainties in the diagnostic imaging studies, (2) dose calculation uncertainties, and (3) target motion and setup errors. In general, the imaging resolutions of CT and magnetic resonance (MR) studies and the spatial distortions in the imaging acquisition process are the common source of uncertainties for all SRS units. This is largely caused by the irregularities in the gradient ﬁelds and the susceptibility from the stereotactic frame in the MR studies for SRS treatments. Table 8-1 summarizes the estimated error levels for the SRS deliveries. The total error shown in Table 8-1 was calculated via quadrature sum of the errors from the individual sources. Overall, the imaging studies and setup variations are the major sources of uncertainties for most SRS deliveries.

Measurements have been carried out to document such dose for both Gamma Knife and linac-based SRS units in the treatment of intracranial lesions [4, 42]. As percentage of the target dose, the Gamma Knife and linac-based units produced ∼0.4% to blood forming organs, ∼0.5% to thyroid, ∼0.04% to gonads, and ∼0.05% to breast or thorax region. Because dmax for the 60Co unit is 0.5 cm and for 4 to 6 MV is in the range of 1.0 to 1.5 cm, the lens dose is expected to be slightly higher for the 60Co units. The lens dose for both units is estimated to ∼ 2.5% of the target dose. Such a percentage could be important for treatments involving large dose deliveries such as trigeminal neuralgia where 70 to 90 Gy may be used. For room shielding purpose, the bunker design depends on numerous factors that include the machine load, how frequently a wall is irradiated, the distance between the wall and the isocenter, and so forth. To ameliorate the requirements for primary irradiations, it is often advantageous to construct a bunker in the basement and direct the primary beam toward the underground earth. Figure 8-13 shows a sample ﬂoor plan hosting a SRS unit. As an illustration, we here give an example calculation for the shielding design of the ﬂoor plan in Figure 8-13. The equation for shielding design calculations is as follows: R = WUT × (d0/d)2 × 10(−t/TVL) × S, where R is the exposure per week, W is the work load (i.e., maximum dose per week), U is the use factor (i.e., fraction of the time that the beam is directed or scattered toward the barrier; U = 1 if scattered radiation is considered), T is the occupancy factor (i.e., fraction of the time a person is present at the point), d0 is the reference dose calibration distance (e.g., d0 = 100 cm for the linac-based SRS unit), t is the thickness of the barrier materials, TVL is the tenth-value layer of the barrier material, and S is the scatter factor (e.g., S = 1 if primary is used, S = 0.001 if scatter radiation is considered). If we need to calculate the required wall thickness near the console operator at the point A, the calculation parameters are determined as follows: W = maximum 5 patients per day × 5 days per week × 5000 cGy per treatment = 125,000 cGy/week; TVL = 13.5 inches for the concrete of the wall; d0 = 1 m and d = 4 m; U = 1; T = 1, and S = 1 for the case; and the exposure limit is 0.002 rad/week. We have

C storage

B

Computer Room

Treatment Room

a

A

CONSOLE

One of the major tasks in building and installing a SRS unit is to design a bunker room and to enforce radiation protection rules. There are two aspects of source shielding and radiation protection for the SRS delivery: (1) reduce extrafocal radiation that contributes to the dose deposition outside of the target volume and (2) reduce radiation exposure to the operators and the general public outside of the treatment room. The radiation to be shielded is generated either from the primary beam directly hitting outside of the target or from the scattered radiation and the leakage radiation from the head of the SRS unit. When considering extrafocal radiation internal to the patient, the concern is the secondary malignancy via carcinogenesis effects, particularly in the treatment of young patients.

SRS Unit D

EARTH BELOW GRADE

Shielding Design

Alley E

FIGURE 8-13. A sample ﬂoor plan for installing a dedicated SRS unit. The designated points (A, B, C, D, E) are selected calculation points for barrier transmissions.

98

l. ma and m. murphy

R = 125,000 × 1 × 1 × (1/4)2 × 10(−t/13.5) × 1 ≤ 0.002. Solving for t of the above equation, we have t ≥ 89 inches. This means that the minimum concrete thickness at the point A should be about 7.5 feet in order to meet the exposure limit of 2 mR/week. It is clear that increasing the source-to-barrier distance or adding high-density shielding materials such as lead in the wall (reducing TVL value) can signiﬁcantly shorten the wall thickness therefore increase the room size. In case of need, the workload can be restricted to maintain the exposure level to a satisfactory level.

Radiation Safety Program Before building or installing a SRS unit, an institutional radiation safety committee should be formed to implement a radiation safety program. The goal of the program is to maintain the radiation exposures to all employees and individual members of the public to be “As Low as Reasonably Achievable” (ALARA). This is commonly referred to as the ALARA principle. To implement the program, a general rule is to monitor the radiation level that is signiﬁcantly below the regulated legal dose limits (e.g., a factor of 10 lower) In general, ﬁlm badges are used to monitor the radiation exposure to all SRS operators and visitors. A radiation safety ofﬁcer (RSO) should be designated for the SRS treatments. The key responsibilities of RSOs are to oversee the radiation exposure reports (e.g., monthly) and to provide regular (e.g., annual) refresher radiation safety training to the authorized users. The RSO needs to be responsive to emergency situations and serve as the contact person between the institution and the federal and the state agencies. For some states in the United States, the RSO is also required to be a qualiﬁed operator of the SRS unit. In case of pregnancy of an authorized user, the RSO needs to determine whether it is appropriate for her to continue to operate the unit. In general, a declared worker may work as long as the total exposure to the embryo or fetus is maintained within the ALARA dose limits.

FIGURE 8-14. Example of calibration phantoms for SRS units: (a) for linac-based SRS units; (b) for Gamma Knife. These phantoms are also used for regular quality assurance measurements or serve as the scat-

Installation and Acceptance Installation and acceptance of a SRS unit requires extensive tests of the functionalities of the unit. Common to all SRS units, three categories of tests should be performed: (1) dose rate calibration and radiation survey; (2) radiation and mechanical isocenter alignment; and (3) imaging acquisition and treatment planning process. For example, the absolute dose rate for a newly installed Gamma Knife should exceed 300 cGy/min with the 18-mm helmet. This is important because this dose rate not only affects the delivery efﬁciency but it also ensures adequate source life and extends the use time before the next reloading is needed. The dose calibration geometry should use tissue-equivalent phantoms of standard shapes as illustrated in Figure 8-14. Another important test is the radiation and the mechanical isocenter alignment. This is performed using the special mechanical alignment tool for the Gamma Knife unit, the Winston-Lutz test method for the linac-based units, and the dosimetry phantom for the CyberKnife, as described above. In general, the smallest ﬁeld size should be measured to ensure maximum accuracy. Example measurement results for a Gamma Knife unit and a linac-based SRS unit are given in Figures 8-15 and 8-16. The most important test in installing a SRS unit is to ensure that the delivered dose proﬁles matched the prescribed ones. As a general practice, the beam proﬁles along all major axes of a single-isocenter delivery should compare with the calculated proﬁles. The agreement should satisfy minimum manufacturer speciﬁcations such as 2% at the central axis and 2 mm in the dose gradient region. For accurate delivery using small collimators (e.g., 4 to 5 mm in diameter), the requirements can be more stringent such as 1% or 1 mm in the dose proﬁles. Such dose proﬁles can be measured using radiochromic ﬁlms with the insert pieces ﬁt into the standard calibration phantoms as illustrated in Figure 8-14. The exposed radiochromic ﬁlms need to be scanned using laser densitometers with resolution of 100 μm or better. An example result for the measured dose proﬁle is given in Figure 8-17.

tering medium for radiation survey tasks. (Part A courtesy of Lucy phantom.)

8.

99

designing, building and installing a srs unit

FIGURE 8-15. Results of radiation and mechanical alignment tests for the new Gamma Knife unit (model C): (a) 4-mm shot; (b) 8-mm shot. The agreement between the pin-pricked center and ﬁeld edge outline was found within 0.25 mm.

FIGURE 8-16. Results of Winston-Lutz test for isocenter alignment checks for a linac-based SRS unit: (a) acquired using electronic portal imager (EPID) with MLC-shaped square ﬁeld; (b) acquired using radiographic ﬁlm for a circular cone. Both ﬁeld shapes can be used for the test.

110 Measurements Reference

RELATIVE DOSE (%)

100 90 80 70 60 50 40 30 20 10

FIGURE 8-17. Measured dose proﬁles compared with the calculated proﬁle in accepting a GKSRS unit. The agreement was found to be within 1% and 1-m requirements for the 4-mm helmet shown here.

0 85

90

95

100 Z-AXIS

105

110

115

100

l. ma and m. murphy TABLE 8-2. Servicing period for major components of the SRS units. Recommended check and servicing frequency

Major components:

Couch motion mechanism (all units) APS (GK) Microswitches (GK) Source activity (GK) Stereotactic frames and ﬁducial boxes (all units) Alarm and interlocks (all) Audio/video system (all) Laser systems (LB, TU, CK) Gantry rotation (LB, TU) Localizing camera systems (LB, CK) Robotic arms (CK)

Annually Annually Monthly Every 7–10 years Weekly Daily Annually Daily Annually Weekly Quarterly

GK, Gamma Knife; LB, Trilogy/linac-based units; CK, CyberKnife; TU, Tomotherapy unit.

FIGURE 8-18. The grid phantom for detecting spatial distortion of the SRS imaging studies.

Besides these major tests, additional acceptance tests may be carried out to satisfy federal and state regulations and institutional protocols for installing a SRS unit. Most of these tests are also part of a quality assurance program in operating the SRS unit. Detailed tasks are recommended in the American Association of Physicists in Medicine (AAPM) task group TG42 report [4]. Example additional tests include timer or monitor chamber linearity tests, radiation survey, wipe tests (for Gamma Knife units), emergency beam-off switch, door interlocks, patient audio-video monitoring and communication circuits, and so forth. One important but often neglected test in accepting a SRS unit is to check the quality and the accuracy of imaging modalities for the treatment planning. To carry out such a task, a MR or CT compatible phantom with precalibrated grid pattern should be scanned (preferably with the ﬁducial localizer box) and imported into the treatment planning system. The procedure aims to validate the image resolution and detect any spatial distortion for the imaging studies. One standard test phantom is shown in Figure 8-18. In general, it is recommended that additional contrast anthropomorphic head phantoms should be scanned for all imaging modalities including CT, MR, angiogram, and so forth, with the acceptance of a SRS unit.

ment tools such as laser systems or patient positioning devices. Because the skull surface measurement can be visually checked against the patient imaging studies, the chance of operation error and malfunction is small. If an error occurs, it can be reversed and corrected without signiﬁcant risk of injuring the patient. However, malfunctions in the laser system and patient positioning device will directly affect the treatment locations and may produce unrecoverable errors. As general practice, regular testing and servicing of high-risk components of SRS units is required (Table 8-2). Most service and maintenance jobs listed above are performed in the treatment room except the source loading and reloading task of the Gamma Knife, because that unit requires manipulation of >1200 Ci of 60Co sources. The radiation exposure is generally higher than the design limit of a treatment room. Typically, a special hot-cell bunker is constructed for carrying out such tasks. The hot-cell should be large enough to accommodate the source head and the source loading unit. An illustration of the source loading unit with an opened unit head is shown in Figure 8-19. In particular, the door to the hot cell should be shielded GK Source Ball

Loading Unit

Manipulators

Service and Maintenance Proper functioning of a SRS unit is only as good as its weakest link. To maintain high-performance quality, routine service and maintenance is important. Before building and installing a SRS unit, it is recommended that a comprehensive probable risk analysis (PRA) be carried out. The goal of PRA is to assign priorities or possible risk factors to individual components, steps, or sequences of SRS unit operations. This aims to eliminate potential causes leading to unrecoverable or disastrous events. For example, a relatively low priority can be assigned to functionality of the skull surface measurement device while a high priority can be given to the functionality of isocenter align-

FIGURE 8-19. Source loading unit with opened unit head for handling high-activity 60Co sources. (Courtesy of Elekta AB, Stockholm, Sweden.)

8.

designing, building and installing a srs unit

TABLE 8-3. Minimum requirements of a qualiﬁed user for performing SRS treatments. Member of the team

Qualiﬁcations

Radiation oncologist

Certiﬁcation (e.g., American Board of Radiology or equivalent) SRS training during or after residency Certiﬁcation (e.g., American Board of Neurological Surgery) SRS training during or after residency Certiﬁcation (e.g., American Board of Radiology or equivalent) SRS physics training and unit operations and quality assurance State licensing and certiﬁcation (e.g., American Registry of Radiological Technologists, ARRT) Unit operation training

Neurosurgeon

Qualiﬁed physicist

Radiation therapist

but allow operation of the robotic manipulators by an operator from the outside. Lead bricks are typically used for this purpose.

Personnel Training and Qualiﬁcations Operation of the SRS units requires qualiﬁed users. The qualiﬁcations and the training of the personnel should follow the practice guideline from the American College of Radiology (ACR). The minimum requirements for the key members of the team are summarized in Table 8-3. All team members should receive basic operation instructions and radiation safety training. The team may also include nursing staff, neuroradiologists, and anesthesiologists in case a young child needs to be treated.

101

with the Winston-Lutz tests, and so forth. The routine QA for SRS deliveries generally include unit functionality, safety interlocks, radiation monitor, and imaging modality functionality and accuracy. In addition to general procedures, qualiﬁed operators must carry out unit-speciﬁc QA tests. For example, Gamma Knife uses radioactive 60Co sources. Therefore, Nuclear Regulatory Commission (NRC) and state regulations mandate speciﬁc tests for Gamma Knife units. One of these tests often includes annual wipe testing of all the helmets to ensure no source leakage occurs during routine operations. A QMP is also required for most SRS deliveries. Major components of such a program include documentation of patient identiﬁcation by at least two methods (name, photo, etc.), as well as records showing that the treatments followed written directives and were completed without deviations. In case of any deviations, the reporting level for misadministration and variance is very strict for SRS. The misadministration of a single-fraction SRS delivery is normally deﬁned as the incidence when the treatment dose differs from the prescribed dose by 10% or more. Because of such stringent criteria, small deviations from the standard protocol are likely to result in reportable misadministrations. Therefore, most SRS programs require qualiﬁed physicists to supervise all QA tasks for the procedure. In-treatment checks such as validation that the target coordinates match the treatment plan must be carried out. Redundant QA checks by other team members are recommended. All qualiﬁed SRS users should receive regular refreshing training. The training should cover radiation safety instructions and reporting requirements for radiation workers as well as QA chart reviews. One important part of training is to refresh emergency procedures such as when to and how to act in case of major equipment failures or medical emergencies.

Cost and Budget Quality Control, Quality Assurance, and Quality Management Quality control is the process to validate whether a SRS unit complies with the design requirements. Once a SRS unit has been built and installed, a series of tests needs to be performed to ensure initial quality of the unit. Quality control processes govern all the manufacturing tests involved in testing the quality of the unit. Because SRS units are medical devices, FDA regulates the manufacturing quality control process. For the unit to be used in human treatments, manufacturers of the SRS units need to submit proper documentation showing good manufacturing practice (GMP) in designing and producing the device. Gamma Knife and linac-based SRS units involve moving components such as gantry, couch, and patient positioning devices. Strict quality assurance (QA) programs should be enforced. The QA program can be separated into routine QA and patient-speciﬁc QA procedures. The routine QA procedures include daily, monthly, and semi-annual/annual QA procedures, user refreshing training, and emergency procedures. Patient-speciﬁc QA procedures include quality management program (QMP) checks, pretreatment checks such as dry-runs

Budgeting SRS units is a complex issue. On appearance, purchasing add-on options to retroﬁt a conventional linear accelerator is appealing because the accelerator was already in use; however, it is a misconception that any conventional linear accelerator can be retroﬁtted for SRS delivery. Along with the fact that isocenter precision is a factor of 2 or 3 higher for the SRS units, the SRS deliveries demand high dose rate and stable output of a large number of monitor units. Wear and tear on the major linear accelerator components such as the klystron, magnetron, and wave guide can be high for such treatments. For gantry-based SRS delivery, the speed of the gantry should be over 5 MU/degree and preferably 10 MU/degree or higher to allow efﬁcient treatment deliveries. Extra cost considerations also include pretreatment QA effort, machine scheduling, patient ﬂow logistics and room design, and so forth. Today, many centers opt to purchase a dedicated SRS unit or integrated SRS unit in order to maximize SRS treatment capabilities. A list of estimated startup costs for various SRS units is summarized in Table 8-4. In general, the acquisition and the ownership cost are higher for the dedicated SRS units than for standard linear accelerators. Aside from operation logistics, patient care and patient

102

l. ma and m. murphy

TABLE 8-4. Total estimated acquisition cost of different SRS units. Units

Estimated cost (US$)

Gamma Knife

$3 million to $4 million (includes source units, planning software, bunker, service contract and training); additional $500,000 to $1.0 million for source reloading with 1–2 weeks down time $2 million to $4 million (includes dedicated accelerator, treatment planning system, bunker, add-on imaging components, service contract and training) $3 million (includes accelerator units, online-CT, treatment planning system, service and training) $3 million (includes X-band accelerator units, bunker, imaging devices, service contract and training)

Dedicated linac-based unit (e.g., Novalis unit)

Integrated unit (e.g. Tomotherapy or Trilogy units) CyberKnife

quality of life should be the driving factor for purchasing an SRS unit. It is evident that the patients are more likely to pursue an SRS option that yields the best possible results while providing the most agreeable treatment experiences.

References 1. Leksell L. The stereotaxic method and radiosurgery of the brain, Acta Chir Scand 1951; 102 (4):316–9. 2. 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): 1665–1672. 3. Corn BW, Curran WJ Jr, Shrieve DC, et al. Stereotactic radiosurgery and radiotherapy: new developments and new directions. Semin Oncol 1997; 24 (6):707–714. 4. Schell MC, Bova FJ, Larson DA, et al. Stereotactic Radiosurgery, Report of the American Association of Physicists in Medicine Task Group No. 42. College Park: American Institute of Physics, 1995. 5. Phillips MH, Stelzer KJ, Grifﬁn TW, et al. Stereotactic radiosurgery: a review and comparison of methods. J Clin Oncol 1994; 12(5):1085–1099. 6. Loefﬂer JS, Shrieve DC, Wen PY, et al. Radiosurgery for intracranial malignancies. Semin Radiat Oncol 1995; 5(3):225–234. 7. Harsh GR, Thornton AF, Chapman PH, et al. Proton beam stereotactic radiosurgery of vestibular schwannomas. Int J Radiat Oncol Biol Phys 2002; 54(1):35–44. 8. Larsson B, Leksell L, Rexed B, et al. The high-energy proton beam as a neurosurgical tool. Nature 1958; 182(4644):1222–1223. 9. Larsson B, Sarby B. Equipment for radiation surgery using narrow 185 MeV proton beams. Dosimetry and design. Acta Oncol 1987; 26(2):143–158. 10. Lawrence JH, Tobias CA, Linfoot JA, et al. Heavy particles and the Bragg peak in therapy. Ann Intern Med 1965; 62:400–407. 11. Levy RP, Fabrikant JI, Frankel KA, et al. Heavy-charged-particle radiosurgery of the pituitary gland: clinical results of 840 patients. Stereotact Funct Neurosurg 1991; 57(1–2):22–35. 12. Weber DC, Chan AW, Bussiere MR, et al. Proton beam radiosurgery for vestibular schwannoma: tumor control and cranial nerve toxicity. Neurosurgery 2003; 53(3):577–586; discussion 586– 588.

13. McGinley PH, Butker EK, Crocker IR, et al. A patient rotator for stereotactic radiosurgery. Phys Med Biol 1990; 35(5):649–657. 14. Leksell DG. Stereotactic radiosurgery. Present status and future trends. Neurol Res 1987; 9(2):60–68. 15. Lindquist C. Gamma Knife radiosurgery. Semin Radiat Oncol 1995; 5:197–202. 16. Maitz AH, Wu A, Lunsford LD, et al. Quality assurance for gamma knife stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1995; 32(5):1465–1471. 17. Wu A, Lindner G, Maitz H, et al. Physics of gamma knife approach on convergent beams in stereotactic radiosurgery. Int J Radiati Oncol Biol Phys 1990; 18:941–949. 18. Bourland JD, McCollough KP. Static ﬁeld conformal stereotactic radiosurgery: physical techniques. Int J Radiat Oncol Biol Phys 1994; 28(2):471–479. 19. Bova FJ, Friedman WA, Mendenhall WM. Stereotactic radiosurgery. Med Prog Technol 1992; 18(4):239–251. 20. Colombo F, Benedetti A, Pozza F, et al. Stereotactic radiosurgery utilizing a linear accelerator. Appl Neurophysiol 1985; 48(1–6):133– 145. 21. Falco T, Lachaine M, Poffenbarger B, et al. Setup veriﬁcation in linac-based radiosurgery. Med Phys 1999; 26(9):1972–1978. 22. Friedman WA, Bova FJ, Spiegelmann R. Linear accelerator radiosurgery at the University of Florida. Neurosurg Clin N Am 1992; 3(1):141–166. 23. Leavitt DD, Watson G, Tobler M, et al. Intensity-modulated radiosurgery/radiotherapy using a micromultileaf collimator. Med Dosim 2001; 26(2):143–150. 24. Lutz W, Winston KR, Maleki N. A system for stereotactic radiosurgery with a linear accelerator. Int J Radiat Oncol Biol Phys 1988; 14(2):373–381. 25. Nedzi LA, Kooy HM, Alexander E 3rd, et al. Dynamic ﬁeld shaping for stereotactic radiosurgery: a modeling study. Int J Radiat Oncol Biol Phys 1993; 25(5):859–869. 26. Podgorsak EB, Olivier A, Pla M, J. Hazel, et al. Physical aspects of dynamic stereotactic radiosurgery. Appl Neurophysiol 1987; 50(1–6):263–268. 27. Podgorsak EB, Olivier A, Pla M, et al. Dynamic stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1988; 14(1):115–126. 28. Solberg TD, Boedeker KL, Fogg R, et al. Dynamic arc radiosurgery ﬁeld shaping: a comparison with static ﬁeld conformal and noncoplanar circular arcs. Int J Radiat Oncol Biol Phys 2001; 49(5):1481–1491. 29. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for stereotactic radiosurgery. Neurosurgery 1988; 22(3):454–464. 30. Colombo F, Francescon P, Cora S, et al. A simple method to verify in vivo the accuracy of target coordinates in linear accelerator radiosurgery. Int J Radiat Oncol Biol Phys 1998; 41(4):951–954. 31. Gibbs FA Jr, Buechler D, Leavitt DD, et al. Measurement of mechanical accuracy of isocenter in conventional linearaccelerator-based radiosurgery. Int J Radiat Oncol Biol Phys 1993; 25(1):117–122. 32. Boyer AL, Antonuk L, Fenster A, et al. A review of electronic portal imaging devices (EPIDs). Med Phys 1992; 19(1):1–16. 33. Leavitt DD, Gibbs FA Jr, Heilbrun MP. Dynamic ﬁeld shaping to optimize stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1991; 21(5):1247–1255. 34. Pedroso AG, De Salles AA, Tajik K. Novalis shaped beam radiosurgery of arteriovenous malformations. J Neurosurg 2004; 101(Suppl 3):425–434. 35. Smith ZA, De Salles AA, Frighetto L. Dedicated linear accelerator radiosurgery for the treatment of trigeminal neuralgia. J Neurosurg 2003; 99(3):511–516. 36. Rahimian J, Chen JC, Rao AA. Geometrical accuracy of the Novalis stereotactic radiosurgery system for trigeminal neuralgia. J Neurosurg 2004; 101(Suppl 3):351–355.

8.

designing, building and installing a srs unit

37. Adler JR, Murphy MJ, Chang SD, et al. Image-guided robotic radiosurgery. Neurosurgery 1999; 44:299–306. 38. Murphy MJ, Cox RS. Dose localization accuracy for an imageguided frameless radiosurgery system. Med Phys 1996; 23(12):2043– 2049. 39. Chang SD, Main W, Martin DP, et al. An analysis of the accuracy of the CyberKnife: a robotic frameless stereotactic radiosurgery system. Neurosurgery 2003; 52:140–147.

103

40. Mackie TR, Balog J, Ruchala K, et al. Tomotherapy. Semin Radiat Oncol 1999; 9(1):108–117. 41. Mackie TR, Holmes T, Swerdloff S, et al. Tomotherapy: a new concept for the delivery of dynamic conformal radiotherapy. Med Phys 1993; 20(6):1709–1719. 42. Berk HW, Larner JM, Spaulding C, et al. Extracranial absorbed doses with Gamma Knife radiosurgery. Stereotact Funct Neurosurg 1993; 61(Suppl 1):164–172.

PA R T III

Stereotactic Radiosurgery Techniques

105

9

Gamma Knife Radiosurgery Ajay Niranjan, Sait Sirin, John C. Flickinger, Ann Maitz, Douglas Kondziolka, and L. Dade Lunsford

Historical Review Professor Lars Leksell ﬁrst coupled an orthovoltage X-ray tube with his ﬁrst-generation guiding device to focus radiation on the Gasserian ganglion to treat facial pain. He subsequently investigated cross-ﬁred protons as well as X-rays from an early-generation linear accelerator (linac) for radiosurgery. In the 1960s, he became dissatisﬁed with the cumbersome nature of crossﬁred proton beams and the poor reliability and wobble of thenexisting linear accelerators. Leksell and Larsson ﬁnally selected cobalt-60 as the ideal photon radiation source and developed the Gamma Knife [1, 2]. They placed 179 60Co sources in a hemispherical array so that all gamma rays (radiation from the decay of 60Co) focused on a single point thereby creating cumulative radiation isocenters of variable volume depending on the beam diameter. The ﬁrst Gamma Knife created a discoidshaped lesion suitable for neurosurgical treatment of movement disorders and intractable pain management. Clinical work with the Gamma Knife began in 1967 at the manufacturing site, the Motala AB workshop near Linköping, Sweden. The ﬁrst patient had a craniopharyngioma. The patient’s head was immobilized using a plaster-molded headpiece. In 1975, a series of surgical pioneers at the Karolinska Hospital, Stockholm, began to use a reengineered Gamma Knife (spheroidal lesion) 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). Lunsford ﬁrst performed Gamma Knife radiosurgery in August 1987 at University of Pittsburgh Medical Center. In the United States, based on the available published literature, arteriovenous malformations (AVMs) and skull base tumors that failed other treatments were considered the initial indications for radiosurgery. A cautious approach was adopted while waiting for increased scientiﬁc documentation. The encouraging results of radiosurgery for benign tumors and vascular malformations led to an exponential rise of radiosurgery cases and sales of radiosurgical units (Tables 9-1, 9-2, and 9-3). In recent years, metastatic brain tumors have become the most common indica-

tion for radiosurgery. Brain metastases now comprise 30% to 50% of radiosurgery cases at busy centers.

The Evolution of Gamma Knife: Models A, B, and C The Gamma Knife has evolved steadily since 1967. Three commercially produced models are now used worldwide. In the ﬁrst models (model U or A), 201 cobalt sources were arranged in a hemispherical array. These units present challenging 60Co loading and reloading issues. To eliminate this problem, the unit was redesigned so that sources were arranged in a circular (O-ring) conﬁguration (models B, C, and 4-C) (Fig. 9-1). Gamma Knife radiosurgery usually involves multiple isocenters of different beam diameters to achieve a treatment plan that conforms to the irregular three-dimensional volumes of most lesions. The total number of isocenters may vary depending upon the size, shape, and location of the target. Each isocenter has a set of three x, y, z stereotactic coordinates corresponding with its location in three-dimensional space as deﬁned using a rigidly ﬁxed skull stereotactic frame. In terms of actual dose delivery, this means several changes in the patient’s head position within the helmet. In 1999, the model C Gamma Knife was introduced. The ﬁrst model C in the United States was installed at the University of Pittsburgh Medical Center in March 2000. This technology combines advances in dose planning with robotic engineering and uses a submillimeter accuracy automatic positioning system (APS). This technology obviates the need to manually adjust each set of coordinates in a multiple-isocenter plan. The robotic positioning system (Fig. 9-2) moves the patient’s head to the target coordinates deﬁned in the treatment 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 treatment. Because the treatment time is shortened, a precise three-dimensional (3D) plan can be generated using multiple smaller beams achieving volumetric conformality. Such an

107

108

a. niranjan et al.

TABLE 9-1. Numbers of active Gamma Knife units worldwide. Continent

Country

Asia

China Japan India Korea Philippines Singapore Taiwan Thailand Canada Mexico United States Austria Belgium Croatia Czech Republic Egypt France Germany Greece Iran Italy Jordan The Netherlands Norway Romania Spain Sweden Switzerland Turkey United Kingdom Argentina Brazil

North America

Europe and Middle East

Latin America

Total units

Active units

17 47 3 7 1 1 6 1 3 2 90 2 1 1 1 1 2 5 1 1 4 1 1 1 1 1 2 1 2 3 1 1

TABLE 9-2. Brain disorders treated worldwide using Gamma Knife radiosurgery by December 2004. Brain disorder

Indications

Vascular disorders

AVM

Benign tumors

Malignant tumors

Functional targets

Ocular disorders

212

approach results in a steeper dose fall-off extending beyond the target (higher selectivity). 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 4-C, was introduced. The ﬁrst unit was installed at the University of

Number of patients treated

39,847

Aneurysm Cavernous malformation Other vascular Vestibular schwannoma Meningioma Pituitary adenoma Pineal region tumor Craniopharyngioma Hemangioblastoma Trigeminal schwannoma Chordoma Other benign tumors Glial tumors (grade I–II) Glial tumors (grade III–IV) Metastatic tumor Chondrosarcoma Nasopharyngeal carcinoma Other malignant tumors Trigeminal neuralgia Parkinson disease Pain Epilepsy Obsessive compulsive disorder Other functional targets Uveal melanoma Glaucoma Other ocular disorders

177 437 3,328 28,306 36,602 24,604 2,619 2,748 1,296 1,781 1,336 4,408 505 20,614 100,098 273 1,087 4,070 17,799 1,208 491 1,879 117 646 1,062 158 33

Total indications

297,529

Pittsburgh in January 2005. The model 4-C is equipped with enhancements designed to improve workﬂow, increase accuracy, and provide integrated imaging capabilities. The integrated imaging, powered by Leksell GammaPlan, offers the ability to fuse images from multiple sources. These images can also be exported to a CD-ROM, so the referring physician can

TABLE 9-3. Peer-reviewed publications on the outcome of Gamma Knife radiosurgery. Disease category

Diagnosis

Vascular malformation

Arteriovenous malformation Cavernous malformation Acoustic neuroma Meningioma Pituitary adenoma Metastases Glial tumors Craniopharyngioma Non–acoustic schwannoma Glomus tumor Pineal tumor Hemangioma Hemangioblastoma Trigeminal neuralgia Movement disorders Epilepsy Obsessive compulsive disorder

Benign tumors

Malignant tumors Other tumors

Functional disorders

Number of Gamma Knife publications

Years of experience

Multi-institutional trials

Randomized controlled trial

85 14 124 60 49 130 46 19 22 13 7 11 13 75 37 48 5

1989–2005 1995–2005 1969–2005 1991–2005 1993–2005 1990–2005 1992–2005 1994–2004 1993–2004 1997–2005 1990–2005 1999–2005 1996–2004 1991–2005 1991–2005 1991–2005 1991–2005

1 0 0 1 0 1 0 0 0 0 0 0 0 2 0 4 0

1 0 0 0 0 3 1 0 0 0 0 0 0 1 0 0 0

9.

109

gamma knife radiosurgery

Intergrated Shielding Plastic Automatic Beam LCD screen doors cover Positioning SystemTM channel

Cobalt-60 Shielding sources

Helmet in treatment position

Protection panels

FIGURE 9-1. Schematic diagram of model 4-C Gamma Knife unit. (Courtesy of Elekta AB, Stockholm, Sweden.)

receive pre- or postoperative images for reference and followup. The planning information can be viewed on both sides of the treatment couch. The helmet changer and robotic APS are faster and reduce total treatment time.

The Radiosurgery Procedure Following are the basic steps of Gamma Knife radiosurgery: 1. Daily quality assurance of the radiosurgery system. 2. Application of the stereotactic guiding device to the patient’s head. 3. Stereotactic brain imaging using magnetic resonance imaging (MRI), computed tomography (CT), and/or an angiogram. 4. Quality assurance of images. 5. Determination of target volume(s). 6. Conformal radiosurgery dose planning by the radiosurgery team. 7. Stereotactic delivery of radiation to the target volume by positioning the patient’s head inside a collimator helmet (Gamma Knife), or on treatment couch (linac-based systems). 8. Removal of stereotactic guiding device.

Daily Quality Assurance of the Radiosurgery System FIGURE 9-2. Automated positioning system (APS). APS is a robotic device that positions the patient’s head at planned target coordinates.

Gamma Knife quality assurance testing is performed by an authorized medical physicist every morning. The medical

110

a. niranjan et al.

physicist ensures that the system tests required by U.S. Nuclear Regulatory Commission (NRC) regulations are performed and functioning properly. These tests include the permanently mounted radiation monitor and its remote indicator, handheld radiation monitor, patient viewing and communication systems, door interlock, timer termination of exposure, emergency stops, beam status indicators, availability of the release rods for the emergency removal of a patient, test run of the automatic patient positioning unit, microswitch test, and function of the helmet hoist. Apart from daily quality assurance tests, monthly and annual quality assurance is also performed in addition to preventative maintenance of the Gamma Knife unit.

Application of the Stereotactic Guiding Device For Gamma Knife radiosurgery, appropriate stereotactic frame placement is the initial critical aspects of the procedure. Prior to frame placement, the radiosurgery team should review the preoperative images and discuss optimal frame placement strategy. The preoperative images should be kept in plain view while applying the head frame. An effort should be made to keep the lesion as close to the center of the frame as possible. The possibility of collision by the frame base ring, the posts/pins assembly, or the patient’s head with the collimator helmet during treatment should also be considered prior to the frame application. Steps to avoid the possible collision should be taken during frame placement. The authors use the following strategies for optimal frame placement. To target lower lesions (skull base tumors, acoustic tumors, cerebellar lesions), the frame is positioned lower by placing the ear bars in the top holes of the earpieces on the Leksell G frame. For higher lesions, (sagittal sinus meningioma, metastases high in the frontal or parietal lobe), the frame is positioned higher by placing the ear bars in the bottom holes of the earpieces. For anterior targets, (anterior frontal lobe tumors, cavernous sinus tumors, sellar lesion and lesion anterior to sella, and anterior temporal lesions), the frame is shifted forward by placing earpieces posteriorly on the base ring of the frame. The posterior edge of the earpiece is kept at 90 to 75 mm on the ydimension of the head frame (instead of 95 to 100 mm) depending upon the shift that is needed to bring the lesion closest to the center of the frame. For anterior target, short posterior posts are preferred to avoid collision of the posterior post/pin assembly with the collimator helmet. To target posterior lesions (occipital lobe tumors, transverse sinus tumors, cerebellar lesions), the frame is shifted backward by positioning the earpieces forward. The posterior edge of the earpiece is kept at 110 to 125 mm instead of 95 to 100 mm. The anterior posts are positioned as low as possible on the supraorbital region to avoid collision of the frontal post/pin assembly with the collimator helmet. For radiosurgery planning, a higher gamma angle (120° to 140°) is used if a potential collision is detected at the default angle of 90°. To reach lateral targets (lateral metastases, convexity tumors, far lateral tumors), the frame is shifted laterally (right or left) toward the lesion. While shifting the frame laterally, it is important to make sure that there is enough space on the contralateral side to allow positioning of the ﬁducial box on the base ring of the frame. The MRI or CT ﬁducial should be tried on the frame prior to sending the patient to the MRI unit. If the ﬁducial box does not ﬁt on to the frame due to excessive shifting of the frame, the frame will have to be repositioned.

Stereotactic Brain Imaging Techniques Aside from the frame application, the next most important aspect of radiosurgery is accurate imaging of the target. MRI is the preferred imaging modality. CT is used when MRI is not possible. Angiograms are used in conjunction with MRI for AVM radiosurgery.

Stereotactic MRI The highlights of stereotactic imaging include optimal contrast between normal and abnormal tissues in addition to high spatial resolution, short scan time, and thin slices so that accurate target localization can be achieved. The use of MRI in stereotactic planning has enhanced accurate targeting of lesions that are usually not adequately deﬁned by any other imaging modality. Some physicians prefer the fusion of magnetic resonance and CT images for stereotactic guidance, as they believe that in certain type of scanners, distortion may affect the accuracy of target localization in MRI. For the initial 2 years, the authors used both MRI and CT for stereotactic planning. Signiﬁcant target coordinate differences were not observed using the Leksell stereotactic system. Since 1993, MRI has been used for stereotactic radiosurgery planning in almost all eligible cases using a 1.5-tesla unit. In addition, arteriovenous malformations are imaged also by biplane angiography. At our institution, high-resolution, gadolinium-enhanced 3D localizer (T2* images) image sequence is used ﬁrst to localize the tumor in axial, sagittal, and coronal images. This sequence (3-mm-thick slices 2-mm apart) only takes 45 seconds for 11 axial, 11 sagittal, and 11 coronal slices. Using the axial images, the ﬁducials can be measured and compared with the opposite side to exclude the possibility of magnetic resonance (MR) artifacts and to conﬁrm that there is no angulation or head tilt. Alternatively, T1-weighted sagittal scout images (3mm-thick slices with 1 mm) using spin echo pulse sequence can be obtained for lesion localization. The average time for this sequence is approximately 1.5 minutes. For stereotactic imaging of most lesions, a 3D-volume acquisition using fast spoiledGRASS (gradient recalled acquisition in steady state) sequence at 512 × 256 matrix and 2 NEX (number of excitations) covering the entire lesion and surrounding critical structures is preferred. To deﬁne the radiosurgery target, this volume is displayed as 1- or 1.5-mm-thick axial slices. The ﬁeld of view (FOV) is kept at 25 cm × 25 cm in order to visualize all ﬁducials. The approximate imaging time for this sequence is 6 to 8 minutes. We generally prefer 3D spoiled-GRASS sequence for most lesions. Additional sequences are performed when more information is needed. Pituitary lesions are particularly difﬁcult to image especially if there has been prior surgery. A half dose of paramagnetic contrast is usually given to image pituitary adenomas. For residual pituitary tumors, after trans-sphenoid resection, a fat-suppression SPGR sequence is recommended in order to differentiate tumor from the fat packed in the resection cavity. For cavernous malformations, an additional variable echo multi planar (VEMP) imaging is obtained to deﬁne the hemosiderin rim. For thalamotomy planning, an additional fast inversion recovery sequence is performed to differentiate basal ganglia from white-matter tracts. Brain metastases patients receive a double dose of contrast agent, and the entire brain is imaged by 2 mm slices to identify all of the lesions. Before

9.

gamma knife radiosurgery

removing the patients from the MR scanner, the images must be checked for accuracy.

Stereotactic CT Imaging When using CT imaging instead of MRI, it is advisable to use short posterior posts to avoid artifacts from the posts and pins. Care should also be taken in deciding the optimal place for the pins because they cause artifacts on CT. An effort should be made to keep the lesion away from the pin artifacts. With modern CT scanners, 1- or 2-mm-thick slices (depending upon the size of the lesion) without any gap can be obtained in 4 to 5 minutes. Before removing the patients from the CT table, accuracy checks are performed to make sure that images would be accepted by the planning system and the lesions have coordinates that are achievable in the Gamma Knife unit.

Stereotactic Angiography Angiography is the gold standard for AVM radiosurgery planning. It should be used in conjunction with MR or CT imaging to provide the third imaging dimension. The technique of angiography differs slightly from the conventional digital angiography as the stereotactic angiographic images are used not only for AVM nidus deﬁnition but also to guide radiation to the target. The orthogonal images (instead of oblique or rotated) are preferred but are not necessary. For AVM nidi that are not properly visualized in orthogonal planes, a rotation of up to 10° in two dimensions or aspects can be used without compromising the accuracy of radiation delivery. Before removing the angiography catheter, the images should be reviewed to make sure that all the ﬁducials are seen on the images. Digital subtraction techniques, despite a potential radial distortion error, have proved satisfactory.

Quality Assurance of Images Regular quality control checks of the MR unit are performed in order to maintain accuracy of images. With a properly shimmed magnet, regular servicing, and strict quality assurance on the unit as well as on the images, MRI provides high-resolution images with accurate target localization. A special frame holder is used in order to avoid head movement during MRI. Accuracy of images is checked for each image sequence by comparing the known frame measurement with image measurements in addition to the distance from the posterior ﬁducials to the middle ﬁducials. The images are exported from Imaging Suite via the Ethernet. Images are deﬁned using Leksell GammaPlan software (LGP) after the images are transferred to the LGP computer. The measurements are again checked and compared with the known frame measurement and also the distance to the middle ﬁducial in order to check for any distortion during image transfer.

111

and outlining by surgeon or oncologist becomes an important step. The surgeon’s input is required to deﬁne radiosurgery targets for patients with AVMs, tumors, and functional neurosurgery as used by some centers. Deﬁning the target volume also helps in calculating conformality index.

Techniques of Conformal Dose Planning In the process of treatment planning, several strategies can be used. The model C allows treatment using robotic automatic patient positioning system (APS mode), manual positioning (trunnion mode), or mixed treatment (some isocenters in APS mode and some in trunnion mode). A different approach would be used if only a trunnion treatment plan was possible versus an APS treatment plan. Universally, in LGP one can start planning from the middle of the target and then move to the top and bottom of the target. Another approach is to start at the bottom or top and build from the starting point. Some surgeons prefer to outline the target volume before planning the treatment volume. Beginners can also use the inverse dose-planning algorithm (Wizard) to create a plan and then optimize it manually. When planning a treatment using trunnions only, one might tend to use larger collimators (especially for larger lesions) to reduce the time and maximize coverage of the target. For example, for a medium-size acoustic tumor, in trunnion mode, one might use a few 14-mm collimators for the majority of the tumor and a few 4-mm collimators for the intracanalicular portion of the tumor. In APS mode, however, one would most likely use multiple 8-mm isocenters for the majority of the tumor and 4-mm isocenters for the intracanalicular portion because the total time spent would be less. There would be no need to go into the treatment room to set each isocenter. As long as the isocenters are in close proximity to one another, the software would automatically put them into the same treatment run and the patient would move from one set of coordinates to the next until all isocenters of one collimator size were treated. The conformal dose planning is enhanced by the use of multiple small collimators. There are other tricks to use in planning; for example, using a steep (125°) gamma angle for posterior lesions (cerebellar or occipital) to avoid frame collisions. Another technique available for single isocenters lesions is to match the gamma angle to the angle of the target. In APS mode, during the set-up phase of planning, the idea is to try to group as many isocenters in the same run as possible, even if it means changing all of the isocenters to high docking or a different gamma angle than the default of 90°. If a different gamma angle is used, the plan must be rechecked for accuracy and adjusted if necessary. In the current version of LGP, multiple targets (multiple tumors) can be treated using different isodose prescription and different central doses with the use of multiple matrices.

Determination of Target Volume(s)

Techniques of Stereotactic Radiation Delivery Using Gamma Knife

Target determination is an important step in order to make a conformal plan. Target volume can be outlined using the LGP software (manual or semiautomatic mode). Experienced surgeons can create conformal dose plans without outlining the target; however, for new centers especially where physicists assume the initial responsibility for planning, target deﬁnition

The model C Gamma Knife allows radiation delivery using trunnion mode (manual patient positioning) or APS mode (robotic positioning) or a combination of the two (mixed treatment). In trunnion treatment, the x, y, and z of each isocenter are set manually and double-checked to avoid errors. The same coordinates and the time obtained from LGP are entered into

112

a. niranjan et al.

the control console, and radiation is administered. The APS plan is transferred directly from planning computer to control computer. The operator selects the run (a combination of isocenters of same beam diameter) that matches the collimator helmet on the Gamma Knife unit. The APS is moved to the dock position, and the patient’s head frame is ﬁxed into the APS. The accuracy of the docking position is checked. The system prompts the user to perform clearance checks ﬁrst for all those planned isocenters in which the pins, posts, frame, or patient’s head would be less than 12 mm away from the inner surface of the collimator helmet (even though they may not match with the collimator size that is being used for ﬁrst run). The clearance check is performed by moving the patient to those positions under APS manual control and by checking the risk of collision with the collimator helmet. After the clearance check, the system prompts the surgeon to carry out position checks. In the position checks, all the isocenters using the same helmet are checked, one by one, by moving the patient’s head to these positions using APS manual control. The team moves out after the position checks are completed, and the radiosurgical dose is administered. The APS moves the patient to all the planned positions, one by one, until the isocenters using that size of collimator helmet are completed. The team monitors the patient and the coordinates of different isocenters on the control computer. If other runs using a different gamma angle but using the same helmet are planned, then the patient is taken out, the next run is selected, APS is moved to the dock position, and the patient’s head is again ﬁxed in the APS using the planned angle (72°, 90°, 110°, or 125°). Position checks are performed, and procedure commences is begun. Similarly, if additional runs using different helmets are planned, the helmet is changed, the patient’s head is positioned in the APS, and the position checks for all the isocenters for that helmet and gamma angle are performed.

Long-Term Outcome After Gamma Knife Radiosurgery Radiosurgery for Brain Vascular Malformations AVM Radiosurgery Untreated, patients with AVMs have a 2% to 4% annual risk of hemorrhage and a signiﬁcant lifetime risk of death if hemorrhage occurs. Approximately 40% to 50% of patients with untreated AVMs will experience some physical deterioration in their working capacity during a 20- to 40-year period after presentation. The goals of radiosurgery are to achieve complete AVM obliteration, to improve symptoms, and to preserve existing neurologic function. Obliteration is a process resulting from endothelial proliferation within the AVM blood vessel walls, supplemented by myoﬁbroblast proliferation. This leads to contraction and eventual obliteration of the AVM blood vessel lumens. The process is cumulative, with earliest obliterations noted within 2 to 3 months, 50% of the effect often seen within 1 year, 80% within 2 years, and 90% within 3 years. Radiosurgery is an effective primary management strategy for patients with small to moderate size (less than 10 cm3)

AVMs (Fig. 9-3). Current AVM radiosurgical studies report obliteration rates of 50% to 95% after radiosurgery (Table 9-4) [3–12]. Dose-volume guidelines for AVM management have been published [5]. AVM outcomes are best for those patients with small-volume AVMs located in noncritical locations. Children may respond faster than adults in terms of the obliteration rate. Within 3 years, radiosurgery offers the potential advantage of complete AVM obliteration in 80% to 95% of small AVMs. Small size and noncritical location predict good outcomes with either AVM resection or radiosurgery. Pollock et al. reported clinical and angiographic variables that affect the results of AVM radiosurgery [13]. When 220 patients were subjected to a multivariate analysis with patient outcomes as the dependent variable, four factors were associated with successful AVM radiosurgery: smaller AVM volume, fewer draining veins, younger patient age, and hemispheric AVM location. Preradiosurgical embolization was a negative predictor of successful AVM radiosurgery. These investigators concluded that AVM obliteration without new neurologic deﬁcits can be achieved in at least 80% of patients with small-volume, hemispheric AVMs after single-session AVM radiosurgery [13]. Multimodality strategies employing combinations of embolization, microsurgery, and radiosurgery may allow greater numbers of patients with AVMs to be successfully treated. Multimodality management is especially needed for large AVMs. Embolization is used as an adjunct to radiosurgery at some centers. The purpose of embolizing large AVMs prior to radiosurgery is to permanently decrease the volume of the AVM. With a smaller AVM, a higher and more effective radiation dose can be administered. We have not found embolization overly effective for radiosurgery. There are several reasons for this. Embolization can only be an effective adjunct to radiosurgery if it results in permanent reduction of a deﬁnitive nidus volume. Reduction in ﬂow or obliteration of a small part within the overall AVM does not help reduce the subsequent radiosurgery volume. Partial embolization is often helpful prior to microsurgery and facilitates bloodless surgery. For large AVMs (more than 15 cm3) where an effective radiosurgery dose cannot be given for fear of causing radiation-related complications, we recommend prospectively staged (volumetric) radiosurgery instead of adjuvant embolization [14]. In staged radiosurgery, a component of AVM nidus is treated using an effective radiosurgery dose, which cannot be prescribed to a larger volume. The remaining volume undergoes second-stage radiosurgery at 3 to 6 months. Staged radiosurgery may allow for the elimination of larger AVMs over a period of 4 to 5 years. In general, most reports indicate that patients remain at risk for hemorrhage during the latency interval until AVM obliteration is complete [15, 16]; however, the long-term results of radiosurgery (5- to 14-year results after Gamma Knife radiosurgery) suggest that the majority of AVM patients (73%) have reduced risk of future hemorrhage after 2 years. If at the end of 3 years residual AVM is identiﬁed by imaging, repeat radiosurgery may be considered (as may other management strategies designed to complete obliteration of the AVM) [17].

Cavernous Malformation Radiosurgery Cavernous malformation radiosurgery has provided a therapeutic option for patients with symptomatic, hemorrhagic

9.

113

gamma knife radiosurgery

FIGURE 9-3. (a) Posterior-anterior (top left) and lateral (top right) carotid angiograms showing large left parietal AVM at the time of Gamma Knife radiosurgery. Axial MR poster (bottom) showing the GammaPlan with 50% isodose line covering AVM nidus. (b) Axial contrast-enhanced T1-weighted MR image (top left) and T2-weighted

image (top right) of the same patient 3 years after radiosurgery showing some radiation-related changes and absence of ﬂow voids signal suggesting nidus obliteration. Posterior-anterior (bottom left) and lateral (bottom right) carotid angiograms showing complete nidus obliteration after 3 years.

malformations in high-risk brain locations not amenable to microsurgery [18, 19]. Radiosurgery is performed for patients with symptomatic, imaging-conﬁrmed hemorrhages for which resection is believed to be associated with high risk. Hasegawa et al. studied the long-term hemorrhage rate in 82 patients after cavernous malformation radiosurgery [19]. Most patients had multiple hemorrhages from brain stem or diencephalic cavernous malformations. During an average observation period of 4.33 years (for a total of 354 patient-years) before treatment, 202 hemorrhages were noted, for an annual hemorrhage rate of 33.9%, excluding the ﬁrst hemorrhage. After radiosurgery, 19 hemorrhages were identiﬁed during an average of 5 years (for

a total of 401 patient-years). The annual hemorrhage rate was 12.3% per year for the ﬁrst 2 years after radiosurgery, followed by 0.76% per year from years 2 to 12. Eleven patients had new neurologic symptoms without hemorrhage after radiosurgery (13.4%). A signiﬁcant decrease in the symptomatic hemorrhage rate after stereotactic radiosurgery of cerebral cavernous malformations indicates radiosurgery is an effective management strategy for patients with hemorrhagic malformations in high-risk brain locations. When patients serve as their own control, the pre-radiosurgery hemorrhage rates fall dramatically 2 to 5 years after radiosurgery.

TABLE 9-4. Results of Gamma Knife radiosurgery for intracranial AVMs. First author

Year

No. of patients

Volume (cm3)

Margin dose (Gy)

Follow-up (months)

Obliteration rate (%)

Morbidity (%)

Re-bleed rate (%)

Chang [3] Regis [9]

2000 2001

254 45

12.1 0.55

16.2 23

24.7 18

78.2 (4 year) 82

5 4.4

Kurita [6] Coffey [4] Pollock [8] Steiner [10] Flickinger [5] Maruyama [7] Shin [12]

2000 1995 2003 1992 2002 2004

30 121 144 247 351 50

1.35 — 5.7 — 5.7 1.5

18.4 — 18 — 20 20

52.2 — 86 — 51 72

52.2 (3 year) 72 64 81 75 66 (6 year)

6.9 2.5 5 3.2 — 16

17.2 4 8 1.9/year — 1.7/year

— — 4 — — 0

2004

400

1.9

20

65

72 (3 year)

6.9

1.9/year

—

6.8 4.4

Cyst formation (%)

0.4 —

114

a. niranjan et al.

TABLE 9-5. Results of Gamma Knife radiosurgery for acoustic tumors.

First author

Year

No. of patients

Median volume (cm3)

Margin dose (Gy)

Follow-up (months)

Tumor control (%)

Flickinger [29] Delbrouck [30] Unger [21] Lunsford [27] Kondziolka [28] Noren [23] Niranjan [24]

2004 2003 2002 2005 1998 1998 1999

313 95 100 829 162 254 29

1.1 — 3.4 2.5 — — 0.4

13 — 13 13 16 13.6 14

24 12 76 120 60–120 33

98.6 — 96 97 98 93 100

Nakamura [25] Prasad [22] van Eck [20] Muacevic [26]

2000 2000 2005 2004

78 153 78 219

— — 2.28 —

— — 13 —

— — 22 72

81 92 97.5 97

Radiosurgery for Brain Tumors Vestibular Schwannoma Radiosurgery The management options for vestibular schwannomas include observation, microsurgery, and radiosurgery. Vestibular schwannoma radiosurgery using the Gamma Knife unit was ﬁrst performed by Leksell in 1969. Acoustic tumor radiosurgery has evolved steadily in the past decade [20–30]. Advanced dose planning software, MR-guided dose planning, and dose optimization reﬂect the evolution of this technology. The goal of radiosurgery is tumor growth control with preservation of cochlear, facial, and trigeminal nerve functions. The results after radiosurgery of acoustic tumors have established it as the effective minimally invasive alternative to microsurgery. In a review of his 17 years’ experience with radiosurgery in more than 1000 acoustic neuromas patients, Lunsford et al. reported that more than 97% of patients had tumor growth control [27]. Patients with acoustic tumors are evaluated with high-resolution MRI scan and undergo clinical evaluation as well as audiologic tests that include pure tone average (PTA) and speech discrimination score (SDS). Hearing is graded using the Gardner-Robertson’s modiﬁcation of the Silverstein and Norell classiﬁcation [31] and facial nerve function is assessed according to the House-Brackmann grading system [32]. “Serviceable” hearing (class I and II) is deﬁned as a PTA or speech reception threshold (SRT) lower than 50 dB and a speech discrimination score better than 50%. After radiosurgery, all patients are followed up with serial gadolinium-enhanced MRI scans, which are requested at 6-month intervals for 2 years. If there is no appreciable change in tumor size, subsequent MRIs are requested at 2-year intervals. All patients who have some preserved hearing are advised to obtain audiologic evaluation (PTA and SDS) near the time of their MRI follow-up. TUMOR GROWTH CONTROL Recent reports suggest a tumor control rate of 93% to 100% after radiosurgery (Table 9-5). Kondziolka et al. studied 5- to 10-year outcomes in 162 acoustic tumor patients who had radiosurgery at the University of Pittsburgh [28]. In this study, a long-term 98% tumor control rate was reported (Fig. 9-4).

Cranial nerve preservation (%) VIII

78.6 67 55 50–77 47 60 73 (<14 Gy, 100) — 65 83.40 90

VII

100 — 94.00 99 79 86 100 — — 98.70 99.50

V

95 — 95 97 73 92 100 — — 97.40 95

Sixty-two percent of tumors became smaller, 33% remained unchanged, and 6% became slightly larger. Some tumors initially enlarged 1 to 2 mm during the ﬁrst 6 to 12 months after radiosurgery as they lost their central contrast enhancement. Such tumors generally regressed in volume compared with their size before radiosurgery. Only 2% of patients required tumor resection after radiosurgery. Niranjan et al. analyzed outcome of intracanalicular tumor radiosurgery performed at the University of Pittsburgh. All patients (100%) had imaging documented tumor growth control [24]. Flickinger et al. performed an outcome analysis of acoustic neuroma patients treated between August 1992 and August 1997 at the University of Pittsburgh [29]. The actuarial 5-year clinical tumor control rate (no requirement for surgical intervention) was 99.4 + 0.6% [29, 33]. The long-term (10- to 15-year) outcome of benign tumor radiosurgery has been evaluated. In a study that included 157 patients with acoustic tumors, 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 (73%) patients, no change in 40 (25.5%) patients, and an increase in 3 (1.9%) patients who had later resection [34]. No patient developed a radiation-associated malignant or benign tumor (deﬁned as a histologically

FIGURE 9-4. (Left) Axial contrast-enhanced T1-weighted MR image showing a right-sided acoustic tumor. Gamma Knife radiosurgery was performed using 12.5 Gy to 50% isodose line. (Right) Two-year followup axial contrast-enhanced T1-weighted MR image showing tumor shrinkage.

9.

gamma knife radiosurgery

conﬁrmed and distinct neoplasm arising in the initial radiation ﬁeld after at least 2 years had passed). HEARING PRESERVATION Pre-radiosurgery hearing now can be preserved in 60% to 70% of patients. In a long-term (5- to 10-year follow-up) study conducted at the University of Pittsburgh, 51% of the patients had no change in hearing ability [28, 33]. All patients (100%) who were treated with a margin dose of 14 Gy or less maintained serviceable level of hearing after intracanalicular tumor radiosurgery [24]. 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 13-Gy tumor margin dose. The 5-year actuarial rates of hearing level preservation and speech preservation were 68.8% and 86.3.2%, respectively, for patients (n = 103) treated with >14 Gy as tumor margin dose [29]. Unlike microsurgery, immediate hearing loss is uncommon after radiosurgery. If hearing impairment is noted, it is gradual over 6 to 24 months. Early hearing loss after radiosurgery (within 3 months) is rare and may result from neural edema, inﬂammation, or demyelination. The exact mechanism of delayed hearing loss after radiosurgery is still unclear. Perhaps gradual obliteration of microvessels, inﬂammatory response, or even direct radiation axonal or cochlear injury is implicated. The effect of radiation on normal microvessels supplying the cochlear nerve or cochlea itself is not known; however, with doses as low as 12 to 13 Gy (which are sufﬁcient to halt the tumor growth), vascular obliteration of normal vessels seems less likely. This dose probably does not adversely affect the vessels as well as the axons. Although with current imaging techniques the cochlear nerve cannot be well visualized, effort should be made to achieve high conformality at the anterior and inferior margins of the tumor. Conformal dose planning using 4-mm collimator for the intracanalicular tumor may prevent further injury to the cochlear nerve. FACIAL NERVE AND TRIGEMINAL NERVE PRESERVATION Facial and trigeminal nerve function can now be preserved in the majority of patients (>99%). In the early experience at the University of Pittsburgh, normal facial function was preserved in 79% of patients after 5 years and normal trigeminal nerve function was preserved in 73% [28]. These facial and trigeminal nerve preservation rates reﬂected the higher tumor margin dose of 18 to 20 Gy used during the CT-based planning era before 1991. In a recent study using MR-based dose planning, 13-Gy tumor margin dose was associated with no 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 2.5% risk of new facial weakness and 3.9% risk of facial numbness (5-year actuarial rates) [29]. None of the patients who had radiosurgery for intracanalicular tumor developed new facial or trigeminal neuropathy. NEUROFIBROMATOSIS TYPE 2 Patients with acoustic neuroma associated with neuroﬁbromatosis type 2 (NF 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 NF 2 tend to form nodular clusters that engulf or even inﬁltrate

115

the cochlear nerve. Complete resection may not be always possible. Radiosurgery has been performed for patients with NF 2. Subach et al. studied 40 patients (with 45 tumors) who were treated with radiosurgery for NF 2 [35]. Serviceable hearing was preserved in 6 of 14 (43%) patients, and this rate improved to 67% after modiﬁcations made in technique in 1992. The overall tumor control rate was 98% [36]. Only one patient showed imaging-documented growth. Normal facial nerve function and trigeminal nerve function was preserved in 81% and 94% of patients, respectively. It now appears that preservation of serviceable hearing in patients with NF 2 is an attainable goal with modern radiosurgery technique.

Meningioma Radiosurgery The optimal treatment for meningioma 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. Multimodality management should then be considered. Recurrence rates are higher for meningiomas in critical locations where only subtotal resections are possible due to limited access and involvement of the 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. Table 9-6 shows recent results of meningioma radiosurgery from various institutions [37–58]. Tumor control rates ranged from 98% (at 2 years) to 75% (at 8 years). Excellent clinical outcomes after skull base meningioma radiosurgery have been reported (Fig. 9-5). Meningiomas attached to 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 neurologic preservation and patient satisfaction. Surgical excision is the preferred ﬁrst-line approach for convexity, anterior fossa, or lateral sphenoid ridge meningiomas, which can be easily approached. For meningiomas at all other intracranial locations, radiosurgery can be offered as the ﬁrst management approach unless the tumor needs debulking because of mass effect. Larger tumors involving critical locations such as optic chiasm may require combined approaches. Malignant meningiomas especially require multimodality management that includes resection, radiosurgery, and radiation therapy.

Pituitary Adenoma Radiosurgery Multimodality management is needed for patients with pituitary tumors. The primary aim of treatment for clinically nonfunctioning pituitary macroadenomas is tumor removal and preservation of visual function. Transsphenoidal surgery is the preferred approach for managing most pituitary adenomas. Radiosurgery is often indicated as an adjuvant management after partial resection or later recurrence of nonfunctioning pituitary adenomas; however, radiosurgery can be performed as the primary management of nonfunctioning adenomas in carefully selected patients such as those who have major surgical risks or for patients who decline microsurgery. Cavernous sinus invasion can occur de novo in patients with large pituitary macroadenomas but is more commonly seen as a residual tumor

116

a. niranjan et al.

TABLE 9-6. Results of Gamma Knife radiosurgery for meningiomas. No. of patients

Author

Year

Liscak [50]

1999

53

Aichholzer [37] Roche [58] Muthukumar [52] Huffmann [43] Kim [44] Feigl [41] Kondziolka [48] Pan [55] DiBiase [39] Flickinger [42] Iwai [44]

2000 2003 1998 2005 2005 2005 1999 1998 2004 2003 2003

46 32 41 15 23 127 99 80 121 219 42

Chang [38] Nicolato [53] Eustacchio [40] Lee [49]

2003 2002 2002 2002

187 156 121 159

Kobayashi [46] Morita [51]

2001 1999

87 88

Tumor location

Cavernous sinus Skull base Petroclival Tentorial Atypical Superﬁcial Mixed Mixed Mixed Mixed Mixed Cavernous sinus Mixed Mixed Skull base Cavernous sinus Mixed Skull base

Mean follow-up (months)

19 48 56 36 35 32 29.3 42 21 54 29 49.4 37.3 48.9 60–118 35 24.2 35

after attempted microsurgical resection. The cranial nerve complication risks and cerebrovascular risks of cavernous sinus microsurgery warrant consideration of radiosurgery. In many cases, the cavernous sinus mass can be treated while selectively sparing not only the optic apparatus but also the pituitary stalk and residual pituitary gland within the sella. Tumor growth control rates of 90% to 100% have now been conﬁrmed by multiple centers after pituitary adenoma radiosurgery [59]. The antiproliferative effect of radiosurgery has been reported in nearly all patients who underwent Gamma Knife radiosurgery. Relatively few patients (who usually had received lower margin doses) eventually required additional treatment. For secretory adenomas, medical management is extremely useful as either ﬁrst-line therapy or as an adjunct in a combined multimodality approach to overall patient management. Tumor

FIGURE 9-5. (Left) Axial contrast-enhanced T1-weighted MR image showing a skull base meningioma at the time of radiosurgery. (Right) Three-year follow-up axial contrast-enhanced T1-weighted MR image showing complete tumor disappearance.

Temporary

Persistent

100.0

3.8

0.0

15.9 13 15.3 16 16 13.8 16 12–20 14 14 11

96.0 100.0 97.5 86.6 95.6 96.4 95.0 91.0 91.7 93.2 90.5

2.0 6.2 — 6.6 43.0 2.5 14 5.0 — — 4.7

9.0 6.2 2.5 0.0 0.0 1.2 5 2.5 8.3 5.3 0.0

10.1 8.3 6.8 6.5

15.1 14.6 13 13

97.1 96.0 99.2 93.1

10.7 4.0 3.3 1.8

0.0 1.0 1.7 5.0

Diameter 25.8 10

14.5 16

93.1 95.0

10.3 2.2

3.4 12.5

Mean volume (cm )

7.8 Diameter 23.5 mm — Diameter 20 mm — 4.7 5.9 4.7 — 4.5 5 14.7

Mean margin dose (Gy)

Complications (%)

Tumor control (%)

3

12

resection is the preferred management strategy when medical management fails to normalize pituitary function. Radiosurgery is often indicated as an adjunct to control residual or recurrent secretory adenoma. The initial ﬁrst stage extracavernous microsurgery is often optimal in order to reduce the subsequent tumor volume, create space between the tumor and the optic apparatus, and thus allow safe delivery of the highest dose of radiosurgery possible. The goal of radiosurgery for functional adenomas is pituitary hormone normalization. Biochemical remission for growth hormone (GH)-secreting adenomas is deﬁned as GH level suppressed to below 1 μg/L on oral glucose tolerance test (OGTT) and normal age-related serum insulin-like growth factor-1 (IGF-1) levels. OGTT remains the gold standard for deﬁning a cure of acromegaly. The IGF-1, however, is far more practical. Decrease of random GH to less than 2.5 μg/L is achieved more frequently than the normalization of IGF-1 but it is necessary to obtain the fulﬁllment of both criteria. Hormonal normalization after radiosurgery was achieved in 23% to 82% of cases in the published series [60–64]. The suppression of hormonal hyperactivity is more effective when higher doses of radiation are used. In a study at the University of Pittsburgh, 38% of recurrent tumor patients were cured (GH ≤1 μg/L), and overall, 66% had growth hormone levels ≤5 μg/L 3 to 5 years after radiosurgery [65]. The impact of radiosurgery has a latency of about 20 to 28 months [66, 67]. During this interval, hormone-suppressive medications may be beneﬁcial. Because hormone-suppressive medication during radiosurgery may act as a radioprotective agent, this medication should be discontinued at least 6 to 8 weeks prior to the radiosurgery and may be resumed after a week. Patients with Cushing disease (adrenocorticotropic hormone [ACTH]-secreting adenomas) respond to radiosurgery, but

9.

gamma knife radiosurgery

more than one procedure may be needed. Often, tumor cannot be well deﬁned during the initial imaging. In addition, there is a latency of about 14 to 18 months for maximal therapeutic response [66, 67]. In various published series, 38% to 83% hormone normalization after radiosurgery has been observed [68–71]. Most prolactinomas can be controlled successfully by dopaminergic suppressive therapy. Surgery is indicated for cases of intolerance to medical treatment, in cases where women desire to have children, or when patients are dopamine agonist resistant (5% to 10% of patients). Some patients prefer microsurgery or radiosurgery to the need for highly expensive lifelong medical treatment. In published studies of patients treated with radiosurgery, 25% to 29% showed normalization [60, 61]. The possible radioprotective effect of dopaminergic drugs should be taken into account. Because patients treated on dopamine agonist during radiosurgery had lower remission rate, it is therefore recommended that prolactinoma radiosurgery be performed during a period of drug withdrawal. New pituitary hormone deﬁciency has been reported in 0 to 30% of patients after radiosurgery for functional pituitary adenomas [60, 61]. The most important factor inﬂuencing hypopituitarism after radiosurgery seems to be the mean dose to the hypophysis (pituitary stalk). Vladyka et al. observed some worsening of gonadotropic, corticotropic, or thyrotropic functions 12 to 87 months after radiosurgery, usually within 4 to 5 years after radiosurgery [72]. Deterioration in pituitary functions was observed when pituitary stalk received higher doses (>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, and can be excluded from higher dose, the risk of hypopituitarism is extremely small. 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 predominately 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, prolactinomas 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 to 5 years but can be avoided by minimizing radiation to pituitary stalk and hypothalamus. Somatostatin analogues and dopamine agonists may have a radioprotective effect [60, 73]. Although the radioprotective effect of these drugs was not conﬁrmed in subsequent studies [62–64], it is advisable to stop these drugs prior to radiosurgery. Short-acting form of somatostatin analogues can be given until 2 weeks prior to GK. Long-acting somatostatin analogues should be discontinued 4 months prior to GK. Dopamine agonists should be discontinued 2 months prior to radiosurgery. After radiosurgery, once hormone levels are normal on medical therapy, somatostatin analogues should be stopped for 4 months each year to assess for biochemical cure. Similarly, dopamine agonists should be stopped for 2

117

months. A panel of tests to detect hypopituitarism should be done at 6-month intervals for the ﬁrst 5 years and then yearly.

Radiosurgery of 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 widemargin fractionated radiotherapy. It has been used mainly for patients with tumors >3.5 cm in diameter as part of a multimodality approach to malignant gliomas. Early radiosurgery reports widely varied in the outcomes for malignant gliomas with a median survival for GBM patients ranging from 9.5 months to 26 months. These variations could result from patient selection biases and other prognostic factors [74]. 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 patients was 16 months after radiosurgery and 26 months after diagnosis. A 2-year survival rate was 51%. For patients with anaplastic astrocytomas, median survival after radiosurgery was 21 months and after diagnosis was 32 months. A 2-year survival rate after diagnosis was 67%. Other centers have recently reported survival rates that seem signiﬁcantly improved compared with 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) [75]. 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 to 70.2 Gy), and the median GK-SRS dose to the prescription volume was 17.1 Gy (range, 10 to 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 vs. 25 months in EBRT plus GK-SRS). Age, Karnofsky performance status, and the addition of GKSRS were all found to be signiﬁcant predictors of overall survival. No acute grade 3 or grade 4 toxicity was encountered. There is 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 survival beneﬁt from radiosurgical boost for malignant gliomas. LOWER-GRADE GLIOMAS Low-grade gliomas have been treated with radiosurgery. Simonova et al. treated 68 patients with low-grade gliomas using Gamma Knife surgery [76]. The median patient age was 17 years and median target volume was 4200 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

118

a. niranjan et al.

time to response of 18 months. In this series, the progressionfree 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 low-grade gliomas (12 grade I astrocytomas, 39 grade II astrocytomas) using Gamma Knife [77]. The mean margin dose was 12.5 Gy for grade I and 15.7 Gy for grade II tumors. In the mean follow-up 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 histologic 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 [78]. The median radiosurgical dose to the tumor margin was 15 Gy (range, 9.6 to 22.5 Gy). After radiosurgery, serial imaging demonstrated complete tumor resolution in 10 patients, reduced tumor volume in 8, stable tumor volume in 7, and delayed tumor progression in 12 patients. 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 [78–80]. Acute complications after 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%. Reoperation 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 The best initial management for brain metastases patients remains to be deﬁned. Current options include fractionated radiation therapy alone, surgery alone, radiosurgery alone, surgery plus radiation therapy, or radiosurgery plus radiation therapy. There are several features that make brain metastases the most common indication for radiosurgery. Most brain metastases are roughly spherical and therefore can be easily targeted by radiosurgery. Brain metastases are compact targets. Although peritumoral microscopic spread is likely, conformal radiosurgery provides additional therapeutic beneﬁt because of the fall-off zone of radiation outside the imaging-deﬁned margin. Advances in neuroimaging have led to early diagnosis of metastases while these are still small and without signiﬁcant mass effect and symptoms. Whereas single brain metastasis without mass effect is the ideal indication of radiosurgery, multiple metastases are treated when the total target volume allows for safe and effective dose delivery. Radiosurgery is not recommended for patients with large metastatic tumors causing signiﬁcant mass effect. Such patients should undergo surgical excision.

FIGURE 9-6. (Left) Axial contrast-enhanced T1-weighted MR image showing left frontal brain metastases with signiﬁcant surrounding edema. (Right) Six-month follow-up axial contrast-enhanced T1weighted MR image showing signiﬁcant tumor shrinkage and no edema in surrounding brain parenchyma.

A large number of scientiﬁc publications deﬁne the effectiveness of radiosurgery for brain metastases (Fig. 9-6). Table 9-7 lists several large representative series of patients [81–92]. These reports include patients with various primary histologies. The local tumor control ranges from 25% to 97%, and median survival ranges from 6 to 27 months. The Gamma Knife Users Group studied the outcomes of radiosurgery in 116 patients with solitary brain metastases. Radiosurgery was part of the initial management of 71 patients, and 45 patients had recurrent tumors after prior whole-brain fractionated radiation therapy. In this study, actuarial local control rate of 67% at 2 years was reported. Shiau et al. recently reported their radiosurgery experience in 219 brain metastases in 100 patients [93]. The actuarial tumor control, deﬁned as freedom from progression, was 82% and 77% at 6 and 12 months, respectively. These data substantially validate the clinical observation of improved local control after radiosurgery (Fig. 9-7). Although local tumor control rates have improved, mortality is usually related to the uncontrolled primary tumor or metastatic spread to other organs. In general, survival and morbidity results of radiosurgery are superior to those reported for surgical resection followed by whole-brain radiation therapy. The results show that radiosurgery is associated with high local control and low morbidity in comparison with surgical resection. When interpreting radiosurgery results, one should also take into account the facts that patients in surgical series are selected for their suitable locations and good general condition, whereas no such selection is performed for radiosurgery. On the contrary, those who are not suitable candidates for surgery either due to the eloquent brain locations or poor medical condition are included in the radiosurgery series. Rutigliano et al. performed a cost-beneﬁt comparison of Gamma Knife radiosurgery and surgical resection for solitary brain metastases and concluded that radiosurgery had a lower uncomplicated procedure cost, a lower average complication per procedure, was more effective, and had a better incremental cost effectiveness per life-year [93].

Radiosurgery for Pineal Gland Tumors Management of pineal region tumors remains a signiﬁcant challenge because of the anatomic complexity of the area and the

9.

119

gamma knife radiosurgery

TABLE 9-7. Results of Gamma Knife radiosurgery for metastatic brain tumors. Author

Year

No. of patients

Serizawa [90] Lippitz [86]

2005 2004

521 215

Nam [87] Pan [88] Gerosa [82] Gerosa [83] Petrovich [89] Sheehan [91]

2005 2005 2002 2005 2002 2002

130 191 804 504 458 273

Amendola [81]

2000

68

Hasegawa [84] Hoffman [85]

2003 2001

121 113

Simonova [92]

2000

237

Median survival (months)

9 Multiple 7.8 Single 13.7 9 14 13.5 14.5 9 7 7.8 8 12

6–12

Systemic control

Margin dose (Gy)

TPFSR 1-year

70% lung —

13% —

Median 20 Median 22

95.70% 93.9%

23% 20%

73% 74% — — 25% 78%

45% lung All lung 88% lung All NSLC 50% melanoma All NSLC

62% — 29% 31% 43% —

Mean 17.9 Mean 18 Mean 20.6 Mean 21.4 Median 18 Median 16

63.90% 91% 94% 95% 87% 86%

65% 46% 55% 60% 58%

53%

All breast

97%

15–24

0% 17%

35% lung All lung

36% 61%

Mean 18.5 Median 18

0%

43% lung

100%

84% local control 79% 81% GK alone 86% with XRT 95.3%

WBRT

Primary cancer

No No

Median 21.5

Single tumor

>50% 22% 80% 45%

100%

WBRT, whole-brain radiation therapy; TPFSR 1-year, one-year tumor progression free survival; NSLC, non–small cell lung cancer.

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 [95, 96]. At the authors’ institution, 14 patients with parenchymal pineal tumors were treated between 1989 and 1997. Local tumor control was achieved in 13 patients while one died of tumor progressions despite chemotherapy and craniospinal irradiation prior to radiosurgery. Neuroimaging followup showed complete disappearance of tumor in 3 patients, decrease in tumor size in 7, no change in tumor size in 3, and tumor growth in 1 patient.

Radiosurgery for Skull Base Tumors Radiosurgery is a primary and adjuvant management for tumors of skull base [97–103] (Table 9-8). From September 1987 through December 2004, 238 miscellaneous skull base tumors were treated with Gamma Knife radiosurgery at the University

FIGURE 9-7. (Left) Axial contrast-enhanced T1-weighted MR image of an 80-year-old man showing a large hemangioblastoma after attempted tumor resection. Gamma Knife radiosurgery was performed using 15 Gy prescribed to the tumor margin (tumor volume, 16.6 cm3). (Right) Follow-up axial MR image shows signiﬁcant regression of the tumor 4 years after Gamma Knife radiosurgery.

of Pittsburgh Medical Center. These tumors and their subsequent management are described below in more detail.

Non–Acoustic Schwannomas Thirty-ﬁve patients underwent radiosurgery for trigeminal 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 [99]. The records of 23 patients were reviewed with a median follow-up of 40 months. Twenty of 23 (91%) patients had tumor growth control, with regression noted in 15 and no further tumor growth in 5. Patients who had subsequent tumor enlargement underwent a second radiosurgical procedure. Twelve of 23 (52%) trigeminal nerve sheath tumor patients reported systemic improvement. Nine (39%) patients had no change in their symptoms. Only two patients noted new neurologic complaints such as facial weakness (one patient) and 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 the 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 concomitant neurologic symptoms. Most such symptoms will resolve as the tumors regress during the next 3 to 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 [100]. Three patients underwent Gamma Knife radiosurgery for facial schwannomas, all identiﬁed at the time of prior

120

a. niranjan et al.

TABLE 9-8. Results of Gamma Knife radiosurgery for skull base tumors. First author

Year

Diagnosis

Saringer [102] Zhang [103] Nettel [99] Pollock [101] Pan [100] Muthukumar [98] Miller [97]

2001 2002 2004 2002 2005 1998 1997

Glomus tumor Jugular foramen schwannomas Trigeminal schwannomas Non–vestibular schwannomas Trigeminal schwannomas Chordomas, chondromas Carcinomas, sarcomas

No. of patients

Mean follow-up (months)

Tumor control (%)

Complications (%)

13 27 23 23 46 15 32

60 38.7 40 43 68 40 27

100 92.5 91 96 93 73 91

0 0 8 17 8 0 3

microsurgery and associated with recurrence or subtotal prior resection. Tumors of the ninth and tenth cranial nerves pose special challenges. Twenty-six patients with jugular bulb schwannomas underwent radiosurgery 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 4 patients 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% (8 decreased and 8 were stable in size) after jugular foramen schwannoma radiosurgery [104]. Zhang et al. reported 96% (26/27) tumor growth control with a follow-up period of 38.7 months [103]. In the series of non–vestibular schwannomas, Pollock et al. reported 96% (22/23) tumor growth control after Gamma Knife radiosurgery [101].

Craniopharyngioma Multimodality therapy is often necessary for craniopharyngioma patients because of the development of refractory cystic components of their tumors. Radiosurgery is usually part of a multimodality management when prior therapies have failed [105, 106]. Forty-three 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 to 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 to 20 Gy), and the maximum dose was 25 Gy (range, 21.8 to 40 Gy). 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 to 150) from radiosurgery. Overall, 14 of 29 tumors regressed or vanished, and 10 remained stable after radiosurgery. Further tumor growth was noted in ﬁve patients, three of who underwent surgical resection and one who 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 multimodality management including microsurgery, radiosurgery, and intracavitary radiation rather than stereotactic or fractionated radiation therapy. The goal has been to maintain endocrinologic 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 [107–109].

Glomus Tumors Radiosurgery using the Gamma Knife has been performed in 16 patients in a 17-year interval. This sparse number of patients (of 7200 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. 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 brain stem 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 months [110]. Neurologic improvement or stability was observed in the majority of patients in published series. Centers using linac-based radiosurgery 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 seven 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. Because they may hemorrhage dramatically at the time of attempted removal, it is prudent for

9.

gamma knife radiosurgery

121

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 to 19 Gy at the margin [111]. All patients had symptomatic improvement, and all had shown a dramatic reduction in the overall volume of the 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 to 5 cases in each also achieved reduction in tumor volume after radiosurgery [112–114].

Radiosurgery appears to be a safe and effective management for small-volume tumors, but over the course of many years, especially from years 5 to 10 after initial surgery and radiosurgery, recurrence rates continue to increase [98, 118]. 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 for tumor cytoreduction. More recently, we have embarked on the usage of endoscopic transsphenoidal resection followed by radiosurgery. In our 1998 report, with an average of 4 years of experience, 15 patients were evaluated. In 13 cases, it was used as an adjunctive treatment and in two patients as an alternative to microsurgical resection. Eight patients had clinical improvement, three remained stable, and four died. Two of the four patients who died had tumor progression outside of the radiosurgical volumes, but two patients died of unrelated disorders. Tumor reduction was noted in 5 of 11 patients. Five patients had deﬁned additional growth and underwent repeat resection [98].

Hemangioblastoma

Invasive Skull Base Cancers

Thirty-six patients with intracranial hemangioblastomas, usually in conjunction with the syndrome of von Hippel–Lindau disease (VHL), have been treated by radiosurgery at our center. Early experience from several centers indicated that radiosurgery could lead to tumor control or regression [115, 116] (Fig. 9-7). For the most part, we have treated tumors with documented tumor growth, which are usually solid, and almost exclusively located in the posterior fossa, cerebellum, and brain stem. Such tumors are generally treated when they have shown evidence of objective growth and neurologic symptoms develop. Prophylactic radiosurgery for hemangioblastomas in the case of VHL is not performed unless tumor growth or new symptoms are documented. Multifocality is often a characteristic of the 20% of hemangioblastomas that are associated with VHL. Radiosurgery is a potential therapeutic option for these patients where resection of multiple tumors might be precluded because of brain location. For those patients with cystic hemangioblastomas, we have less optimism related to the overall role of radiosurgery at least as a single option. Cyst-associated tumors with nonenhancing cyst cavities were controlled by including only the enhancing nodule in the target volume; 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 [117].

After combined otolaryngological and neurosurgical procedures, we have used adjuvant radiosurgery for invasive skull base cancers (28 patients over the past 17 years). Fourteen patients had adenocarcinomas, 13 squamous cell carcinomas, and 1 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 [119–121].

Chordoma and Chondrosarcoma During our 17-year experience, 26 patients with chordoma and 17 patients with chondrosarcomas have undergone management with radiosurgery. We continue to regard these tumors as difﬁcult tumors to manage. Almost invariably, they require multimodality 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 multimodality treatment. Radiosurgery has been used both as a primary and adjuvant management strategy.

Radiosurgery for Functional Brain Disorders Trigeminal Neuralgia Radiosurgery Our current experience included 513 patients, managed since 1992. There were 305 (60%) women and 208 men. The mean age was 68 years (range, 16 to 92 years), and the mean duration of symptoms was 8 years. Our last detailed review studied 220 consecutive radiosurgery procedures for typical trigeminal neuralgia, all performed between 1992 and 1998 [122]. All 220 patients had trigeminal neuralgia that was idiopathic, longstanding, 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 to 564 months). Pain was predominately 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 (61.4%) patients, including microvascular decompression, glycerol rhizotomy, radiofrequency rhizotomy, balloon microcompression, peripheral neurectomy, or ethanol injections. Thus, the majority of patients represented both medical and surgical failures. In the remaining 85 (38.6%) patients, radiosurgery was the ﬁrst surgical procedure performed. The median central dose at trigeminal nerve was 80 Gy (Fig. 9-8). The pain relief after radiosurgery was graded into four categories: excellent, good, fair, and poor. Complete pain relief without the use of any

122

a. niranjan et al.

FIGURE 9-8. (Left) Axial contrast-enhanced T1-weighted MR image showing trigeminal neuralgia radiosurgery dose plan. (Right) Axial contrast-enhanced T1-weighted MR image of the same patient 6 months later shows contrast enhancement at the site of radiosurgery.

medication was deﬁned as an excellent outcome. We recommended all patients with complete pain relief to taper off their medications, and some patients were in the process of tapering at the time of evaluation (or refused to taper off because of the fear of a recurrence). Those patients with complete pain relief but who were still using some medication were considered as good outcomes. Patients with partial pain relief (more than 50% pain relief) were considered to have a fair outcome [11]. No pain relief or less than 50% pain relief were considered as poor. Placement within a category was decided by the patient rather than by the physician. Criteria for improvement included a reduction in both the frequency and severity of pain attacks. Of the 220 patients, 47 (25.1%) required further additional surgical procedures because of poor 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 responded to radiosurgery within 6 months of the procedure (median, 2 months). The ﬁrst evaluation was performed for all patients within 6 months after radiosurgery. At the initial follow-up assessment, excellent results were obtained in 105 (47.7%) patients, and excellent plus good results were found in 139 (63.2%) patients. More than 50% pain relief (excellent, good, or fair) was noted in 181 (82.3%) patients. At the last follow-up evaluation, 88 (40%) patients had excellent outcomes, 121 (55.9%) patients had excellent plus good outcomes, and 152 (69.1%) patients were fair or better. Thirty (13.6%) patients had recurrence of pain after the initial achievement of pain relief (25 patients after complete relief, 5 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 more than 50% pain relief (excellent, good, or fair) was 2 months (2.0 ± 0.05), and median time to achieving complete pain relief (good or excellent) was also 2 months (2.0 ± 5.1). At 6 months after treatment, 81.4 ± 2.6% patients had achieved more than 50% pain relief, and by 12 months, 85.6 ± 2.47% (actuarial statistics). Complete pain relief (good or excellent) was achieved in 64.9 ± 3.2% of the patients at 6 months, 70.3 ± 3.16% by 1 year, and in 75.4 ± 3.49% of patients by 33 months.

Prior authors, including our group, noted a latency interval to pain relief of approximately 1 to 2 months; however, approximately 15% of patients had no improvement in their pain even after 12 months. The duration of pain relief after initial response in all patients was also analyzed. Patients who never responded to radiosurgery were recorded as having a relief duration of zero months. More than 50% pain relief (excellent, good, or fair) was achieved and maintained in 76% of patients at 1 year, 71% of patients at 2 years, 67% of patients at 3 years, 65% of patients at 3.5 years, and 56% of patients at 5 years. Complete pain relief (excellent or good) was achieved and maintained in 63.6 ± 3.3% of patients at 1 year, 59.2 ± 3.5% of patients at 2 years, and 56.6 ± 3.8% of patients at 3 years. A history of no prior surgery 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. The median time to developing paresthesia was 8 months (range, 1 to 19 months). After 19 months, no patient developed any new sensory symptoms. 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 with all other surgical options. In this study, less than 10% of patients developed increased facial paresthesia and/or facial sensory loss. The majority of our patients described their numbness or paresthesia as minor and not bothersome. 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 [123]. We advocate a maximum dose of 50 to 60 Gy at a second procedure, and usually target a volume anterior to the prior target. Doing so has led to a pain response similar to that after primary radiosurgery in properly selected patients.

Movement Disorder Radiosurgery There is a small subset of movement disorder patients who have conditions that may make them unacceptable candidates for invasive stereotactic neurosurgical intervention. Such conditions are chronic use of anticoagulants and severe cardiac or respiratory disease. In addition, very elderly and noncompliant patients are usually considered poor surgical candidates. Finally, some patients voluntarily choose a less-invasive alternative to open stereotactic technique. Stereotactic radiosurgery is an option for this subset of patients with movement disorders. Gamma Knife is the preferred radiosurgical tool for treatment of movement disorders.

Radiosurgical Targets for Functional Disorders VIM nucleus of the thalamus is targeted for tremor patients. In our series, we determined the VIM coordinates based on the position of the nucleus relative to the AC-PC line and the anatomic information gathered from very-high-resolution MRI.

9.

gamma knife radiosurgery

Current planning software incorporates computerized atlases, which can be registered with the images and can be projected on MR images. The x, y, and z coordinates can be determined for the target using either SurgiPlan (Elekta Corp, Georgia), GammaPlan (Elekta Corp) or Multiview (Elekta Corp). Duma et al. reported results of radiosurgical pallidotomy for Parkinson disease. The target localization of globus pallidus interna (GPi) was determined by coordinates based on anatomic information gathered from very-high-resolution MRI and subjective surgeon correlation with the Schaltenbrand atlas. The 50% isodose line of a single or double isocenter 4-mm collimator plan was placed at the center of GPi. Keep et al. performed radiosurgical subthalamotomy using a primary target of 13 mm lateral to, 2 mm posterior to the midpoint of, and 5 mm inferior to the AC-PC line. This was optimized using atlas reference and experience. We are hesitant to perform radiosurgical pallidotomy because of the nearby optic tract. Friehs et al. reported targeting the center of the heads of the caudate nuclei bilaterally to treat the bradykinesia and rigidity of parkinsonism, and Pan et al. targeted the anterior portion of the VL nucleus for dystonia.

Gamma Knife Thalamotomy We have previously reported our experience with the treatment of essential tremor (ET) and Parkinson 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 [124]. 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 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 (range, 1 week to 8 months) [125]. No change in tremor occurred in four Gamma Knife thalamotomies (8.6%), “mild” improvement was seen in 4 (8.6%), “good” improvement was seen in 13 (28%), and “excellent” improvement in 13 (28%). In 12 thalamotomies (26%), the tremor was eliminated completely. The high-dose (160 Gy mean maximum dose) thalamotomy lesion was more effective at reducing tremor than the low dose (120 Gy mean maximum dose). One patient, after bilateral treatment, suffered a mild acute dysarthria 1 week after GK thalamotomy. Ohye et al. reported 36 Gamma Knife thalamotomies in 31 patients. Maximum dose was 150 Gy in the ﬁrst 6 cases, which was subsequently reduced to 130 Gy [126, 127]. 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. The authors conﬁrmed two different patterns of post-radiosurgical lesions on follow-up imaging. One was a round punchedout lesion with enhancing borders with good symmetry, 7 to 8 mm in diameter to the enhancing edges. The second type of lesion seen extended to surrounding areas including the capsule with “rail-like” high signal along the border of the thalamus and

123

GP. Young et al. in a large series of patients reviewed their use of GK thalamotomy for the treatment of tremor [128]. Their series included 102 patients with parkinsonian tremor, 52 patients with essential tremor, and 4 patients with tremor of other etiology. The single 4-mm collimator was used with doses varying from 110 to 160 Gy. At a median follow-up of 52.5 months (range, 11 to 93 months), 76% were tremor free, and 12% were “nearly free of tremor.” Thus, there was failure in 12%. In 52 patients with disabling ET, median follow-up was 26 months. At 1 year, 92% were completely or nearly tremor free; after 4 years of follow-up this percentage decreased to 88%. One patient experienced a transient complication of contralateral balance disturbance; one patient had mild contralateral paresthesia in the face and upper extremity without detectable sensory deﬁcit and no impairment of function. A third patient had a mild weakness and dysphasia. All complications were believed to be due to lesions that became larger than expected. The overall complication rate was 1.3%.

Gamma Knife Pallidotomy Duma et al. performed Gamma Knife pallidotomy on 18 patients with medically recalcitrant and disabling symptoms of PD. Fifteen patients were treated using a single 4-mm collimator with a median central dose of 160 Gy (range, 90 to 165 Gy) [125]. Three patients were treated using a combination of two 4-mm shots with a dose of 160 Gy. Only 6 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 patient had numbness in the contralateral hemibody. Nine patients (50%) had one or more complications related to treatment. Okun et al. reported similar complications of GK pallidotomy in a report describing eight patients seen in an 8-month period referred for complications of GK radiosurgery [129]. Complications included hemiplegia, homonymous ﬁeld cut, weakness, dysarthria, hypophonia, aphasia, hemihypesthesia, and pseudobulbar laughter. Friedman et al. had similar experience [130]. They described their results in four patients using Gamma Knife pallidotomy in advanced disease. No patient improved in a signiﬁcant manner within the follow-up 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.

Other Functional Targets Friehs et al. reported the efﬁcacy of GKRS caudatotomy for the treatment of the bradykinesia and rigidity of parkinsonism [131]. One month after treatment, 6 of 10 patients showed clear beneﬁt from bilateral 4-mm head of caudate lesions without any treatment-related complications. Keep et al. reported radiosurgical subthalamotomy using the GK in a single case report [132]. The 73-year-old patient received 120 Gy central dose using the 4-mm collimator helmet. At 2 weeks, she was able to reduce her Sinemet dose. At 5 weeks, she had no tremor, rigidity, or dyskinesia and walked easily with improved balance while using only a one-point cane for support. At 3

124

a. niranjan et al.

months, she had partial return of increased motor tone and cogwheel rigidity. At 1 year, after medication adjustments, she was able to move with ease and had no tremor. Imaging at 42 months showed a well-demarcated signal focus corresponding with the subthalamic nucleus. Pan et al. reported two patients who underwent radiosurgery for torsion spasm to evaluate the efﬁcacy of Gamma Knife radiosurgery as an alternative treatment. The target was located at the anterior portion of the ventrolateral nucleus. The maximum doses were 150 Gy and 145 Gy, respectively. Double isocenters with a 4-mm collimator were used. Follow-up lasted for 18 months and 8 months, respectively. Both patients had excellent clinical improvement 2 to 3 months after Gamma Knife radiosurgery, respectively. The authors concluded that Gamma Knife radiosurgery might be a safe and efﬁcient treatment for torsion spasm. Gamma Knife radiosurgical thalamotomy is a safe and effective alternative to invasive radiofrequency or DBS. This is not the case with radiosurgical pallidotomy. The paucity of radiosurgical pallidotomy reports in the literature reﬂects a lack of enthusiasm in the procedure. Subtle differences in lesion targeting have the potential to affect outcome. Without physiologic feedback, differentiation of internal and external globus pallidus is impossible during gamma pallidotomy. The lack of clinical improvement may therefore have been attributable to inaccurate physiologic lesioning within the GP without physiologic monitoring. The high complication rate of 50% in pallidotomy series is likely due to the variability and unpredictability of the lesion size when the globus pallidus serves as the target. This unpredictability and variability is not seen in the VIM thalamotomy series. It seems that there is a differential sensitivity to radiation between these two locations. Historically, the pallidum has exhibited a “supersensitivity” to hypoxia, and this may be the reason for higher complication rate. The pallidum is known to contain high levels of iron, which typically rises with age. It has been hypothesized that the presence of iron within this structure may catalyze free-radical reactions causing toxicity to the aging brain.

Conclusion In the past 15 years, we have witnessed dramatic improvements in the stereotactic radiosurgery technologies. Gamma Knife radiosurgery now offers better image-handling features including image fusion; faster, more compact platforms that make the calculations almost real-time; automated patient positioning reducing the potential for human error; and the inverse treatment planning. In the future, more accurate imaging techniques and improved software to handle those images as well as advanced inverse planning software will provide better treatment resulting in better patient outcomes.

References 1. Lunsford LD. Lars Leksell. Notes at the side of a raconteur. Stereotact Funct Neurosurg 1996; 67(3–4):153–168. 2. Ammar A. Lars Leksell’s vision—radiosurgery. Acta Neurochir Suppl 1994; 62:1–4. 3. Chang JH, Chang JW, Park YG, Chung SS. Factors related to complete occlusion of arteriovenous malformations after gamma knife radiosurgery. J Neurosurg 2000; 93(Suppl 3):96–101.

4. Coffey RJ, Nichols DA, Shaw EG. Stereotactic radiosurgical treatment of cerebral arteriovenous malformations. Gamma Unit Radiosurgery Study Group. Mayo Clin Proc 1995; 70(3):214–222. 5. 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(3):347–354. 6. Kurita H, Kawamoto S, Sasaki T, et al. Results of radiosurgery for brain stem arteriovenous malformations. J Neurol Neurosurg Psychiatry 2000; 68(5):563–570. 7. Maruyama K, Kondziolka D, Niranjan A, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for brainstem arteriovenous malformations: factors affecting outcome. J Neurosurg 2004; 100(3):407– 413. 8. 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(6):1291–1296. 9. Regis J, Massager N, Levrier O, et al. (Gamma-knife radiosurgery for brainstem arteriovenous malformations. Preliminary results]. Neurochirurgie 2001; 47(2–3 Pt 2):291–297. 10. 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):1–8. 11. Szeifert GT, Major O, Kemeny AA. Ultrastructural changes in arteriovenous malformations after gamma knife surgery: an electron microscopic study. J Neurosurg 2005; 102(Suppl):289–292. 12. Shin M, Maruyama K, Kurita H, et al. Analysis of nidus obliteration rates after gamma knife surgery for arteriovenous malformations based on long-term follow-up data: the University of Tokyo experience. J Neurosurg 2004; 101(1):18–24. 13. Pollock BE, Flickinger JC, Lunsford LD, Maitz A, Kondziolka D. Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 1998; 42(6):1239–1244. 14. Pollock BE, Kline RW, Stafford SL, Foote RL, Schomberg PJ. The rationale and technique of staged-volume arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 2000; 48(3):817–824. 15. Pollock BE, Flickinger JC, Lunsford LD, Bissonette DJ, Kondziolka D. Hemorrhage risk after stereotactic radiosurgery of cerebral arteriovenous malformations. Neurosurgery 1996; 38(4): 652–659. 16. Izawa M, Hayashi M, Chernov M, et al. Long-term complications after gamma knife surgery for arteriovenous malformations. J Neurosurg 2005; 102(Suppl):34–37. 17. Maesawa S, Flickinger JC, Kondziolka D, Lunsford LD. Repeated radiosurgery for incompletely obliterated arteriovenous malformations. J Neurosurg 2000; 92(6):961–970. 18. Liscak R, Vladyka V, Simonova G, Vymazal J, Novotny J Jr. Gamma knife surgery of brain cavernous hemangiomas. J Neurosurg 2005; 102(Suppl):207–213. 19. 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(6):1190–1197. 20. van Eck AT, Horstmann GA. Increased preservation of functional hearing after gamma knife surgery for vestibular schwannoma. J Neurosurg 2005; 102(Suppl):204–206. 21. Unger F, Walch C, Schrottner O, Eustacchio S, Sutter B, Pendl G. Cranial nerve preservation after radiosurgery of vestibular schwannomas. Acta Neurochir Suppl 2002; 84:77–83. 22. Prasad D, Steiner M, Steiner L. Gamma surgery for vestibular schwannoma.(see comment]. J Neurosurg 2000; 92(5):745–759. 23. Noren G. Long-term complications following gamma knife radiosurgery of vestibular schwannomas. Stereotact Funct Neurosurg 1998; 70(Suppl 1):65–73.

9.

gamma knife radiosurgery

24. Niranjan A, Lunsford LD, Flickinger JC, Maitz A, Kondziolka D. Dose reduction improves hearing preservation rates after intracanalicular acoustic tumor radiosurgery. Neurosurgery 1999; 45(4):753–762. 25. Nakamura H, Jokura H, Takahashi K, Boku N, Akabane A, Yoshimoto T. Serial follow-up MR imaging after gamma knife radiosurgery for vestibular schwannoma. Ajnr: Am J Neuroradiol 2000; 21(8):1540–1546. 26. Muacevic A, Jess-Hempen A, Tonn JC, Wowra B. Results of outpatient gamma knife radiosurgery for primary therapy of acoustic neuromas. Acta Neurochir Suppl 2004; 91:75–78. 27. Lunsford LD, Niranjan A, Flickinger JC, Maitz A, Kondziolka D. Radiosurgery of vestibular schwannomas: summary of experience in 829 cases. J Neurosurg 2005; 102(Suppl):195–199. 28. Kondziolka D, Lunsford LD, McLaughlin MR, Flickinger JC. Long-term outcomes after radiosurgery for acoustic neuromas.[see comment]. N Engl J Med 1998; 339(20):1426–1433. 29. 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(1):225–230. 30. Delbrouck C, Hassid S, Massager N, et al. Preservation of hearing in vestibular schwannomas treated by radiosurgery using Leksell Gamma Knife: preliminary report of a prospective Belgian clinical study. Acta Otorhinolaryngol Belg 2003; 57(3):197–204. 31. Gardner G, Robertson JH. Hearing preservation in unilateral acoustic neuroma surgery. Ann Otol Rhinol Laryngol 1988; 97(1):55–66. 32. House JW, Brackmann DE. Facial nerve grading system. Otolaryngol Head Neck Surg 1985; 93(2):146–147. 33. 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):1–6. 34. Kondziolka D, Nathoo N, Flickinger JC, Niranjan A, Maitz AH, Lunsford LD. Long-term results after radiosurgery for benign intracranial tumors. Neurosurgery 2003; 53(4):815–821. 35. 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. J Neurosurg 1999; 90(5):815–822. 36. 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(5):815–822. 37. Aichholzer M, Bertalanffy A, Dietrich W, et al. Gamma knife radiosurgery of skull base meningiomas. Acta Neurochir (Wien) 2000; 142(6):647–652. 38. Chang JH, Chang JW, Choi JY, Park YG, Chung SS. Complications after gamma knife radiosurgery for benign meningiomas. J Neurol Neurosurg Psychiatry 2003; 74(2):226–230. 39. DiBiase SJ, Kwok Y, Yovino S, et al. Factors predicting local tumor control after gamma knife stereotactic radiosurgery for benign intracranial meningiomas. Int J Radiat Oncol Biol Phys 2004; 60(5):1515–1519. 40. Eustacchio S, Trummer M, Fuchs I, Schrottner O, Sutter B, Pendl G. Preservation of cranial nerve function following Gamma Knife radiosurgery for benign skull base meningiomas: experience in 121 patients with follow-up of 5 to 9.8 years. Acta Neurochir Suppl 2002; 84:71–76. 41. Feigl GC, Bundschuh O, Gharabaghi A, Samii M, Horstmann GA. Volume reduction in meningiomas after gamma knife surgery. J Neurosurg 2005; 102(Suppl):189–194. 42. Flickinger JC, Kondziolka D, Maitz AH, Lunsford LD. Gamma knife radiosurgery of imaging-diagnosed intracranial meningioma. Int J Radiat Oncol Biol Phys 2003; 56(3):801–806.

125

43. Huffmann BC, Reinacher PC, Gilsbach JM. Gamma knife surgery for atypical meningiomas. J Neurosurg 2005; 102(Suppl):283–286. 44. Iwai Y, Yamanaka K, Ishiguro T. Gamma knife radiosurgery for the treatment of cavernous sinus meningiomas. Neurosurgery 2003; 52(3):517–524. 45. Kim DG, Kim Ch H, Chung HT, et al. Gamma knife surgery of superﬁcially located meningioma. J Neurosurg 2005; 102(Suppl): 255–258. 46. Kobayashi T, Kida Y, Mori Y. Long-term results of stereotactic gamma radiosurgery of meningiomas. Surg Neurol 2001; 55(6): 325–331. 47. 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):405–413. 48. Kondziolka D, Levy EI, Niranjan A, Flickinger JC, Lunsford LD. Long-term outcomes after meningioma radiosurgery: physician and patient perspectives. J Neurosurg 1999; 91(1):44–50. 49. Lee JY, Niranjan A, McInerney J, Kondziolka D, Flickinger JC, Lunsford LD. Stereotactic radiosurgery providing long-term tumor control of cavernous sinus meningiomas. J Neurosurg 2002; 97(1):65–72. 50. 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(5):473–480. 51. Morita A, Coffey RJ, Foote RL, Schiff D, Gorman D. Risk of injury to cranial nerves after gamma knife radiosurgery for skull base meningiomas: experience in 88 patients. J Neurosurg 1999; 90(1):42–49. 52. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for tentorial meningiomas. Acta Neurochir (Wien) 1998; 140(4):315–320; discussion 20–21. 53. Nicolato A, Foroni R, Alessandrini F, Maluta S, Bricolo A, Gerosa M. The role of Gamma Knife radiosurgery in the management of cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys 2002; 53(4):992–1000. 54. Nicolato A, Foroni R, Alessandrini F, Bricolo A, Gerosa M. Radiosurgical treatment of cavernous sinus meningiomas: experience with 122 treated patients. Neurosurgery 2002; 51(5):1153– 1159. 55. Pan DH, Guo WY, Chang YC, et al. The effectiveness and factors related to treatment results of gamma knife radiosurgery for meningiomas. Stereotact Funct Neurosurg 1998; 70(Suppl 1):19–32. 56. Pendl G, Schrottner O, Eustacchio S, Feichtinger K, Ganz J. Stereotactic radiosurgery of skull base meningiomas. Minim Invasive Neurosurg 1997; 40(3):87–90. 57. Pendl G, Eustacchio S, Unger F. Radiosurgery as alternative treatment for skull base meningiomas. J Clin Neurosci 2001; 8(Suppl 1):12–14. 58. Roche PH, Pellet W, Fuentes S, Thomassin JM, Regis J. Gamma knife radiosurgical management of petroclival meningiomas results and indications. Acta Neurochir (Wien) 2003; 145(10):883– 888. 59. Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD. Radiosurgery for nonfunctioning pituitary adenoma. Neurosurg Focus 2003; 14(5):1–9. 60. Pollock BE, Nippoldt TB, Stafford SL, Foote RL, Abboud CF. Results of stereotactic radiosurgery in patients with hormoneproducing pituitary adenomas: factors associated with endocrine normalization. J Neurosurg 2002; 97(3):525–530. 61. Landolt AM, Lomax N. Gamma knife radiosurgery for prolactinomas. J Neurosurg 2000; 93(Suppl 3):14–18. 62. Attanasio R, Epaminonda P, Motti E, et al. Gamma-knife radiosurgery in acromegaly: a 4-year follow-up study. J Clin Endocrinol Metab 2003; 88(7):3105–3112.

126

a. niranjan et al.

63. Castinetti F, Taieb D, Kuhn JM, et al. Outcome of gamma knife radiosurgery in 82 patients with acromegaly: correlation with initial hypersecretion. J Clin Endocrinol Metab 2005; 90(8):4483– 4488. 64. Gutt B, Wowra B, Alexandrov R, et al. Gamma-knife surgery is effective in normalising plasma insulin-like growth factor I in patients with acromegaly. Exp Clin Endocrinol Diabetes 2005; 113(4):219–224. 65. Niranjan A, Szeifert G, Kondziolka D, Flickinger JC, Maitz A, Lunsford LD, eds. Gamma Knife Radiosurgery for Growth Hormone-Secreting Pituitary Adenoma. Basal: Karger, 2002. 66. Jane JA Jr, Vance ML, Woodburn CJ, Laws ER Jr. Stereotactic radiosurgery for Hypersecreting pituitary tumors: part of multimodality approach. Neurosurgery Focus 2003; 14(14):1–5. 67. Laws ER Jr, Vance ML. Radiosurgery for pituitary tumors and craniopharyngiomas. Neurosurg Clin N Am 1999; 10(2):327–336. 68. Rahn T, Thoren M, Hall K, Backlund EO. Stereotactic radiosurgery in Cushing’s syndrome: acute radiation effects. Surg Neurol 1980; 14(2):85–92. 69. Sheehan JM, Vance ML, Sheehan JP, Ellegala DB, Laws ER Jr. Radiosurgery for Cushing’s disease after failed transsphenoidal surgery. J Neurosurg 2000; 93(5):738–742. 70. Kobayashi T, Kida Y, Mori Y. Gamma knife radiosurgery in the treatment of Cushing disease: long-term results. J Neurosurg 2002; 97(5 Suppl):422–428. 71. 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(3):839–841. 72. 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(2):309–316. 73. Landolt AM, Haller D, Lomax N, et al. Octreotide may act as a radioprotective agent in acromegaly. J Clin Endocrinol Metab 2000; 85(3):1287–1289. 74. Larson DA, Gutin PH, McDermott M, et al. Gamma knife for glioma: selection factors and survival. Int J Radiat Oncol Biol Phys 1996; 36(5):1045–1053. 75. Nwokedi EC, DiBiase SJ, Jabbour S, Herman J, Amin P, Chin LS. Gamma knife stereotactic radiosurgery for patients with glioblastoma multiforme. Neurosurgery 2002; 50(1):41–46. 76. Simonova G, Novotny J Jr, Liscak R. Low-grade gliomas treated by fractionated gamma knife surgery. J Neurosurg 2005; 102(Suppl):19–24. 77. 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):42–46. 78. Hadjipanayis CG, Kondziolka D, Gardner P, et al. Stereotactic radiosurgery for pilocytic astrocytomas when multimodal therapy is necessary. J Neurosurg 2002; 97(1):56–64. 79. Hadjipanayis CG, Kondziolka D, Flickinger JC, Lunsford LD. The role of stereotactic radiosurgery for low-grade astrocytomas. Neurosurg Focus 2003; 14(5):1–15. 80. Hadjipanayis CG, Niranjan A, Tyler-Kabara E, Kondziolka D, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for wellcircumscribed ﬁbrillary grade II astrocytomas: an initial experience. Stereotact Funct Neurosurg 2002; 79(1):13–24. 81. 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(2):88–92. 82. Gerosa M, Nicolato A, Foroni R, et al. Gamma knife radiosurgery for brain metastases: a primary therapeutic option. J Neurosurg 2002; 97(5 Suppl):515–524. 83. Gerosa M, Nicolato A, Foroni R, Tomazzoli L, Bricolo A. Analysis of long-term outcomes and prognostic factors in patients with

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96. 97.

98.

99.

100.

101.

102.

non-small cell lung cancer brain metastases treated by gamma knife radiosurgery. J Neurosurg 2005; 102(Suppl):75–80. 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):1318–1326. Hoffman R, Sneed PK, McDermott MW, et al. Radiosurgery for brain metastases from primary lung carcinoma. Cancer J 2001; 7(2):121–131. Lippitz BE, Kraepelien T, Hautanen K, et al. Gamma knife radiosurgery for patients with multiple cerebral metastases. Acta Neurochir Suppl 2004; 91:79–87. Nam TK, Lee JI, Jung YJ, et al. Gamma knife surgery for brain metastases in patients harboring four or more lesions: survival and prognostic factors. J Neurosurg 2005; 102(Suppl): 147–150. Pan HC, Sheehan J, Stroila M, Steiner M, Steiner L. Gamma knife surgery for brain metastases from lung cancer. J Neurosurg 2005; 102(Suppl):128–133. 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(5 Suppl):499– 506. Serizawa T, Saeki N, Higuchi Y, et al. Gamma knife surgery for brain metastases: indications for and limitations of a local treatment protocol. Acta Neurochir (Wien) 2005; 147(7):721–726. Sheehan JP, Sun MH, Kondziolka D, Flickinger J, Lunsford LD. Radiosurgery for non-small cell lung carcinoma metastatic to the brain: long-term outcomes and prognostic factors inﬂuencing patient survival time and local tumor control. J Neurosurg 2002; 97(6):1276–1281. Simonova G, Liscak R, Novotny J Jr, Novotny J. Solitary brain metastases treated with the Leksell gamma knife: prognostic factors for patients. Radiother Oncol 2000; 57(2):207–213. Shiau CY, Sneed PK, Shu HK, et al. Radiosurgery for brain metastases: relationship of dose and pattern of enhancement to local control. Int J Radiat Oncol Biol Phys 1997; 37(2):375–383. Rutigliano MJ, Lunsford LD, Kondziolka D, Strauss MJ, Khanna V, Green M. The cost effectiveness of stereotactic radiosurgery versus surgical resection in the treatment of solitary metastatic brain tumors. Neurosurgery 1995; 37(3):445–453; discussion 53– 55. Reyns N, Hayashi M, Chinot O, et al. The role of Gamma Knife radiosurgery in the treatment of pineal parenchymal tumours. Acta Neurochir (Wien) 2006; 148(1):5–11. Kobayashi T, Kida Y, Mori Y. Stereotactic gamma radiosurgery for pineal and related tumors. J Neurooncol 2001; 54(3):301–309. Miller RC, Foote RL, Coffey RJ, et al. The role of stereotactic radiosurgery in the treatment of malignant skull base tumors. Int J Radiat Oncol Biol Phys 1997; 39(5):977–981. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for chordoma and chondrosarcoma: further experiences. Int J Radiat Oncol Biol Phys 1998; 41(2):387–392. Nettel B, Niranjan A, Martin JJ, et al. Gamma knife radiosurgery for trigeminal schwannomas. Surg Neurol 2004; 62(5):435–444; discussion 44–46. Pan L, Wang EM, Zhang N, et al. Long-term results of Leksell gamma knife surgery for trigeminal schwannomas. J Neurosurg 2005; 102(Suppl):220–224. Pollock BE, Foote RL, Stafford SL. Stereotactic radiosurgery: the preferred management for patients with nonvestibular schwannomas? Int J Radiat Oncol Biol Phys 2002; 52(4):1002–1007. 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(3):141–146.

9.

gamma knife radiosurgery

103. Zhang N, Pan L, Dai JZ, Wang BJ, Wang EM, Cai PW. Gamma knife radiosurgery for jugular foramen schwannomas. J Neurosurg 2002; 97(5 Suppl):456–458. 104. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for jugular foramen schwannomas. Surg Neurol 1999; 52(2):172–179. 105. Chiou SM, Lunsford LD, Niranjan A, Kondziolka D, Flickinger JC. Stereotactic radiosurgery of residual or recurrent craniopharyngioma, after surgery, with or without radiation therapy. NeuroOncology 2001; 3(3):159–166. 106. Hasegawa T, Kondziolka D, Hadjipanayis CG, Lunsford LD. Management of cystic craniopharyngiomas with phosphorus-32 intracavitary irradiation. Neurosurgery 2004; 54(4):813–820. 107. Chung WY, Pan DH, Shiau CY, Guo WY, Wang LW. Gamma knife radiosurgery for craniopharyngiomas. J Neurosurg 2000; 93(Suppl 3):47–56. 108. Amendola BE, Wolf A, Coy SR, Amendola MA. Role of radiosurgery in craniopharyngiomas: a preliminary report. Med Pediatr Oncol 2003; 41(2):123–127. 109. Ulfarsson E, Lindquist C, Roberts M, et al. Gamma knife radiosurgery for craniopharyngiomas: long-term results in the ﬁrst Swedish patients. J Neurosurg 2002; 97(5 Suppl):613–622. 110. Pollock BE. Stereotactic radiosurgery in patients with glomus jugulare tumors. Neurosurg Focus 2004; 17(1–3):E10. 111. Thompson TP, Lunsford LD, Flickinger JC. Radiosurgery for hemangiomas of the cavernous sinus and orbit: technical case report. Neurosurgery 2000; 47(3):778–783. 112. Peker S, Kilic T, Sengoz M, Pamir MN. Radiosurgical treatment of cavernous sinus cavernous haemangiomas. Acta Neurochir 2004; 146(4):337–341. 113. Nakamura N, Shin M, Tago M, et al. Gamma knife radiosurgery for cavernous hemangiomas in the cavernous sinus. Report of three cases. J Neurosurg 2002; 97(5 Suppl):477–480. 114. Kida Y, Kobayashi T, Mori Y. Radiosurgery of cavernous hemangiomas in the cavernous sinus. Surg Neurol 2001; 56(2):117–122. 115. Georg AE, Lunsford LD, Kondziolka D, Flickinger JC, Maitz A. Hemangioblastoma of the posterior fossa. The role of multimodality treatment. Arquivos de Neuro-Psiquiatria 1997; 55(2): 278–286. 116. Pan L, Wang EM, Wang BJ, et al. Gamma knife radiosurgery for hemangioblastomas. Stereotact Funct Neurosurg 1998; 70(Suppl 1):179–186. 117. Jawahar A, Kondziolka D, Garces YI, Flickinger JC, Pollock BE, Lunsford LD. Stereotactic radiosurgery for hemangioblastomas of the brain. Acta Neurochir 2000; 142(6):641–644. 118. Kondziolka D, Lunsford LD, Flickinger JC. The role of radiosurgery in the management of chordoma and chondrosarcoma of the cranial base. Neurosurgery 1991; 29(1):38–45.

127

119. Kondziolka D, Lunsford LD. Stereotactic radiosurgery for squamous cell carcinoma of the nasopharynx. Laryngoscope 1991; 101(5):519–522. 120. 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(2):160–165. 121. Lee N, Millender LE, Larson DA, et al. Gamma knife radiosurgery for recurrent salivary gland malignancies involving the base of skull. Head & Neck 2003; 25(3):210–216. 122. 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(1):14– 20. 123. Hasegawa T, Kondziolka D, Spiro R, Flickinger JC, Lunsford LD. Repeat radiosurgery for refractory trigeminal neuralgia. Neurosurgery 2002; 50(3):494–500. 124. 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(3):443– 446. 125. Duma CM. Movement Disorder Radiosurgery. Philadelphia: Lippincott, Williams & Wilkins, Inc., 2003. 126. 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):120–127. 127. Ohye C, Shibazaki T, Zhang J, Andou Y. Thalamic lesions produced by gamma thalamotomy for movement disorders. J Neurosurg 2002; 97(5 Suppl):600–606. 128. Young RF. Gamma knife radiosurgery as an alternative form of therapy for movement disorders. Arch Neurol 2002; 59(10):1660– 1662; author reply 2–4. 129. Okun MS, Stover NP, Subramanian T, et al. Complications of gamma knife surgery for Parkinson disease. Arch Neurol 2001; 58(12):1995–2002. 130. 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(1): 143–150. 131. Friehs GM, Ojakangas CL, Schrottner O, Ott E, Pendl G. Radiosurgical lesioning of the caudate nucleus as a treatment for parkinsonism: a preliminary report. Neurol Res 1997; 19(1): 97–103. 132. Keep MF, Mastrofrancesco L, Erdman D, Murphy B, Ashby LS. Gamma knife subthalamotomy for Parkinson disease: the subthalamic nucleus as a new radiosurgical target. Case report. J Neurosurg 2002; 97(5 Suppl):592–599.

1 0

Linear Accelerator Radiosurgery William A. Friedman

Introduction Stereotactic radiosurgery (SRS) is a minimally invasive treatment modality that delivers a large, single dose of radiation to a speciﬁc intracranial target while sparing surrounding tissue. Unlike conventional fractionated radiotherapy, SRS does not maximally exploit the higher radiosensitivity of brain lesions relative to normal brain (therapeutic ratio). Its selective destruction is dependent mainly on sharply focused, high-dose radiation and a steep dose gradient away from the deﬁned target. The biological effect is irreparable cellular damage (probably via DNA strand breaks) and delayed vascular occlusion within the high-dose target volume. Because a therapeutic ratio is not required, traditionally radioresistant lesions can be treated. Because destructive doses are used, however, any normal structure included in the target volume is subject to damage. The basis for SRS was conceived more than 40 years ago by Lars Leksell [1]. He proposed the technique of focusing multiple beams of external radiation on a stereotactically deﬁned intracranial target. The averaging of these intersecting beams results in very high doses of radiation to the target volume but innocuously low doses to non–target tissues along the path of any given beam. His team’s implementation of this concept culminated in the development of the Gamma Knife. The modern Gamma Knife employs 201 ﬁxed cobalt radiation sources in a ﬁxed hemispherical array, such that all 201 photon beams are focused on a single point. The patient is stereotactically positioned in the Gamma Knife so that the intracranial target coincides with the isocenter of radiation. Using variable collimation, beam blocking, and multiple isocenters, the radiation target volume is shaped to conform to the intracranial target. An alternate radiosurgical solution using a linear accelerator (linac) was ﬁrst described in 1984 by Betti et al. [2]. Colombo et al. described such a system in 1985 [3], and linacs have subsequently been modiﬁed in various ways to achieve the precision and accuracy required for radiosurgical applications [4–7]. In 1986, a team composed of neurosurgeons, radiation oncologists, radiation physicists, and computer programmers began development of the University of Florida linac-based radiosurgery system [8]. This system has been used to treat more than 2000 patients at the University of Florida since May 1988 and is in use at multiple sites worldwide. Many other commercial versions of radiosurgical systems are currently available, includ-

ing the BrainLAB system, the Radionics (X-knife) system, the Accuray (CyberKnife) system, and others. Most linac radiosurgical systems rely on the same basic paradigm: A collimated X-ray beam is focused on a stereotactically identiﬁed intracranial target. The gantry of the linac rotates around the patient, producing an arc of radiation focused on the target (Figs. 10-1 and 10-2). The patient couch is then rotated in the horizontal plane and another arc performed. In this manner, multiple non-coplanar arcs of radiation intersect at the target volume and produce a high target dose, with minimal radiation to surrounding brain. This dose concentration method is exactly analogous to the multiple intersecting beams of cobalt radiation in the Gamma Knife. The target dose distribution can be tailored by varying collimator sizes, eliminating undesirable arcs, manipulating arc angles, using multiple isocenters, and differentially weighting the isocenters [9]. In our center, multiple isocenters are used to achieve highly conformal dose distributions, exactly analogous to the Gamma Knife technique (Fig. 10-3). Some linear accelerator systems use an alternative approach that relies upon a computer-driven multileaf collimator to generate nonspherical beam shapes that are conformal to the beam’s-eye view of the tumor. The multileaf collimator can be adjusted statically or dynamically as the linear accelerator rotates. Intensity modulation can be used to achieve dose distributions that are close to those seen with multiple isocenters, and treatment time can be reduced. Achievable dose distributions are similar for linac-based and Gamma Knife systems. With both systems, it is possible to achieve dose distributions that conform closely to the shape of the intracranial target, thus sparing the maximum amount of normal brain. Recent advances in stereotactic imaging and computer technology for dose planning, as well as reﬁnements in radiation delivery systems, have led to improved efﬁcacy, fewer complications, and a remarkable amount of interest in the various applications of SRS. Perhaps of equal importance is the fact that increasing amounts of scientiﬁc evidence have persuaded the majority of the international neurosurgical community that radiosurgery is a viable treatment option for selected patients suffering from a variety of challenging neurosurgical disorders. This chapter will present a brief description of linac radiosurgical technique, followed by a review of the more common applications of stereotactic radiosurgery in the treatment of

129

130

w.a. friedman

FIGURE 10-1. Linear accelerators are the preferred device, worldwide, for conventional radiotherapy. They accelerate electrons to near light speed, then collide them with a heavy metal (like tungsten) in the head of the machine. The collision mainly produces heat, but a small percentage of the energy is converted into highly energetic photons. These photons, because they are electronically produced, are called Xrays. The X-radiation is collimated and focused on the target.

FIGURE 10-3. This choroidal ﬁssure AVM required four 1-cm isocenters to produce a conformal plan. The inner line (70% isodose) is the prescription dose line. The outer line (35% isodose) is half of the prescription dose.

FIGURE 10-2. This diagram shows an add-on device, designed to improve the accuracy of the linear accelerator, in place. The linac arcs around the patient, with its beam always focused on the stereotactically

positioned target. The patient is then moved to a new horizontal (table) position and another arc performed. The result is multiple, noncoplanar arcs of radiation, all converging on the target point.

10.

linear accelerator radiosurgery

intracranial disease (benign tumors, malignant tumors, and arteriovenous malformations).

131

with these difﬁcult tumors continue to be less than optimal [10–12]. A signiﬁcant amount of experience has been accumulated using SRS in the treatment of schwannomas and meningiomas. We will focus on each of these tumor types in turn.

Linac Radiosurgery Technique Although the details of radiosurgical treatment techniques differ somewhat from system to system, the basic paradigm is quite similar everywhere. Following is a detailed description of a typical radiosurgical treatment at the University of Florida. Almost all radiosurgical procedures in adults are performed on an outpatient basis. The patient reports to the neurosurgical clinic the day before treatment for a detailed history and physical, as well as an in-depth review of the treatment options. If radiosurgery is deemed appropriate, the patient is sent to the radiology department for a volumetric magnetic resonance imaging (MRI) scan. A radiosurgical plan can be generated, in advance, using this MRI study. The next morning, the patient arrives at 7:00 a.m. A stereotactic head ring is applied under local anesthesia. No skin shaving or preparation is required. Subsequently, stereotactic computed tomography (CT) scanning is performed. One-millimeter slices are obtained throughout the entire head. The patient is then transported to an outpatient holding area where he and his family have breakfast and relax until the treatment planning process is complete. The stereotactic CT scan and the nonstereotactic volumetric MRI scan are transferred via Ethernet to the treatmentplanning computer. The CT images are quickly processed so that each pixel has an anteroposterior, lateral, and vertical stereotactic coordinate, all related to the head ring previously applied to the patient’s head. Using image fusion software, the nonstereotactic MRI is fused, pixel for pixel, with the stereotactic CT. The “pre-plan” performed the day before is, likewise, fused to the stereotactic CT. Final dosimetry then begins and continues until the neurosurgeon, radiation oncologist, and radiation physicist are satisﬁed that an optimal dose plan has been developed. A variety of options are available for optimizing the dosimetry. The fundamental goal is to deliver a radiation ﬁeld that is precisely conformal to the lesion shape (see Fig. 10-3) while delivering a minimal dose of radiation to all surrounding neural structures. A detailed discussion of dosimetric options is available in Chapter 7. When dose planning is complete, the radiosurgical device is attached to the linac. The patient then is attached to the device and treated. The head ring is removed and, after a short observation period, the patient is discharged. The radiosurgical device is disconnected from the linac, which is then ready for conventional usage. Close clinical and radiologic follow-up is arranged at appropriate intervals depending on the pathology treated and the condition of the patient.

Radiosurgery for Benign Tumors SRS has proved useful for the treatment of a variety of benign intracranial neoplasms. These tumors commonly arise from the skull base, where their dramatic impact on quality of life belies their benign histology and small size. Despite progressive improvement in microsurgical techniques, outcomes for patients

Vestibular Schwannomas Among benign intracranial tumors, vestibular schwannoma (acoustic neuroma) has to date been the most frequent target for stereotactic radiosurgery. This common tumor (representing approximately 10% of all primary brain tumors) is a benign proliferation of Schwann cells arising from the myelin sheath of the vestibular branches of the eighth cranial nerve. These tumors are slightly more common in women, present at an average age of 50 years, and occur bilaterally in patients with neuroﬁbromatosis type 2. Leksell ﬁrst used stereotactic radiosurgery to treat a vestibular schwannoma in 1969 [13]. SRS is a logical alternative treatment modality for this tumor for several reasons. A vestibular schwannoma is typically well demarcated from surrounding tissues on neuroimaging studies. The sharp borders of this noninvasive tumor make it a convenient match for the characteristically steep dose gradient produced at the boundary of a radiosurgical target. This allows the radiosurgeon to minimize radiation of normal tissue. Excellent spatial resolution on gadolinium-enhanced MRI facilitates radiosurgical dose planning. These tumors typically occur in an older population that may be less ﬁt for microsurgical resection under general anesthesia. Finally, the location of these tumors at the skull base in close proximity to multiple critical neurologic structures (i.e., cranial nerves, brain stem) leads to appreciable surgical morbidity and rare mortality even in expert hands. This makes the concept of an effective, less invasive, less morbid alternative treatment that can be performed in a single day under local anesthesia quite attractive. Whether or not radiosurgery ﬁts this description has been extensively debated. Certainly, the role of radiosurgery is limited by its inability to expeditiously relieve mass effect in patients for whom this is necessary. The radiobiology of SRS also requires lower, potentially less effective doses for higher target volumes in order to avoid complications. This limits the use of SRS to the treatment of smaller tumors. Despite these limitations, there is a growing body of literature that substantiates the claim that radiosurgery is a safe and effective alternative therapy for acoustic schwannomas. The published experience using linac-based radiosurgery for the treatment of vestibular schwannomas is relatively limited compared with the Gamma Knife literature. Foote et al. [14] performed an analysis of risk factors associated with radiosurgery for vestibular schwannoma at University of Florida (UF). The aim of this study was to identify factors associated with delayed cranial neuropathy after radiosurgery for vestibular schwannoma (VS) and to determine how such factors may be manipulated to minimize the incidence of radiosurgical complications while maintaining high rates of tumor control. From July 1988 to June 1998, 149 cases of VS were treated using linear accelerator radiosurgery at the University of Florida. In each of these cases, the patient’s tumor and brain stem were contoured in 1-mm slices on the original radiosurgical targeting images. Resulting tumor and brain-stem volumes were coupled

132

w.a. friedman

with the original radiosurgery plans to generate dose-volume histograms. Various tumor dimensions were also measured to estimate the length of cranial nerve that would be irradiated. Patient follow-up data, including evidence of cranial neuropathy and radiographic tumor control, were obtained from a prospectively maintained, computerized database. The authors performed statistical analyses to compare the incidence of posttreatment cranial neuropathies or tumor growth between patient strata deﬁned by risk factors of interest. One hundred thirty-nine of the 149 patients were included in the analysis of complications. The median duration of clinical follow-up for this group was 36 months (range, 18 to 94 months). The tumor control analysis included 133 patients. The median duration of radiology follow-up in this group was 34 months (range, 6 to 94 months). The overall 2-year actuarial incidences of facial and trigeminal neuropathies were 11.8% and 9.5%, respectively. In 41 patients treated before 1994, the incidences of facial and trigeminal neuropathies were both 29%, but in the 108 patients treated since January 1994, these rates declined to 5% and 2%, respectively. An evaluation of multiple risk factor models showed that maximum radiation dose to the brain stem, treatment era (pre-1994 compared with 1994 or later), and prior surgical resection were all simultaneously informative predictors of cranial neuropathy risk. The radiation dose prescribed to the tumor margin could be substituted for the maximum dose to the brain stem with a small loss in predictive strength. The overall radiologic tumor control rate was 93% (59% tumors regressed, 34% remained stable, and 7.5% enlarged), and the 5-year actuarial tumor control rate was 87% (95% conﬁdence interval [CI], 76% to 98%). Based on this study, the authors currently recommend a peripheral dose of 12.5 Gy for almost all acoustics as that dose most likely to yield long-term tumor control without causing cranial neuropathy. Spiegelmann et al. [15, 16] have reported their experience. They reviewed the methods and results of linac radiosurgery in 44 patients with acoustic neuromas who were treated between 1993 and 1997. CT 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 to 20 Gy, whereas in the last 20 patients the dose was reduced to 11 to 14 Gy. After a mean follow-up period of 32 months (range, 12 to 60 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 linac-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 [17]. A mean marginal dose of 19.4 Gy (range, 16 to 20 Gy) was delivered to the 70% isodose line with a single isocenter. Mean follow-up duration was 19 months (range, 12 to 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 Rome, Italy [18]. Five of these had neuroﬁbromatosis and seven (30%) had undergone previous surgery. Total radiation dose to the tumor margin ranged from 12 to 45 Gy (median, 30 Gy) and was delivered in one to ﬁve sessions. One or two isocenters were used, and mean duration of follow-up was 40 months (range, 24 to 46 months). Results using this less conventional method of multisession radiosurgery were comparable with 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.” The use of linac radiosurgery for acoustics is brieﬂy discussed in reports by Delaney [19] and Barcia [20]. In addition, fractionated stereotactic radiation therapy (SRT) has been used as an alternative management for vestibular schwannomas [1, 5]. This method is proposed as a way of exploiting the precision of stereotactic radiation delivery to minimize dose to normal brain while employing lower fractionated doses in an effort to minimize complications. Thus far, most radiosurgeons feel that optimal results can be achieved with highly conformal singlefraction radiosurgery while sparing the patient the inconvenience of a prolonged treatment course. As of April 2005, the University of Florida experience with vestibular schwannomas comprised 386 patients. The indications for radiosurgery were age >60 (180 cases), failed surgery (81), preference (118), medical inﬁrmity (6). The median treatment volume was 2 cm3. With a median follow-up of 32 months for the entire group, 108 tumors are unchanged, 154 are smaller (Figs. 10-4 and 10-5), and 11 (4%) tumors are larger. Only four

FIGURE 10-4. Pretreatment MRI scan shows left-sided vestibular schwannoma.

10.

linear accelerator radiosurgery

FIGURE 10-5. Four years after treatment, the MRI scan shows the schwannoma of Fig. 10-4 to be much smaller.

(1%) patients have required surgery because of tumor growth after radiosurgery.

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 [21, 22]. 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 [23]. A grade I resection, which is complete tumor removal with 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 grades 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

133

surgical resection or Gamma Knife radiosurgery [24]. 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-free survival rates with radiosurgery were superior to Simpson grades 2, 3, and 4 resections. Complications were lower in the radiosurgery group. Multiple linear accelerator radiosurgical series have been published [25–28]. Hakim and colleagues described the largest such series, and the only one to report actuarial statistics [29]. One hundred twenty-seven patients with 155 meningiomas were treated. Actuarial tumor control for patients with benign tumors was 89.3% at 5 years. Six (4.7%) patients had permanent radiation-induced complications. The University of Florida report on linear accelerator radiosurgery treatment of meningiomas is the largest yet published [30]. 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 (Figs. 10-6 and 10-7). 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 that was surgically excised when it enlarged. The other had a heman-

FIGURE 10-6. MRI scan shows right cavernous sinus meningioma. The patient presented with a sixth nerve paresis.

134

w.a. friedman

FIGURE 10-7. Three years later, the meningioma of Fig. 10-6 is barely visible. The sixth nerve paresis is completely resolved. We believe that radiosurgery is the treatment of choice for many cavernous sinus meningiomas.

giopericytoma of the lateral cavernous sinus that was surgically excised when it enlarged.

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 proved 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 10 times more common than primary brain tumors with an annual incidence of between 80,000 and 150,000 new cases each year [31]. Fifteen percent to 40% 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 neurologic manifestations. Debate exists concerning the optimum treatment for metastatic brain disease. In autopsy series, brain metastases occur in up to 50% of cancer patients [32]. Approximately 30% to 40% present with a solitary metastasis. Brain metastases frequently cause debilitating symptoms that can seriously impact the patient’s quality

of life. With no treatment or steroid therapy alone, survival is limited (1 to 2 months). Whole-brain radiotherapy (WBRT) extends median survival, but the duration of survival is typically low (3 to 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 [33]. They found a signiﬁcant improvement in survival (40 weeks vs. 15 weeks) and local recurrences in the CNS (20% vs. 52%) for patients in the surgery plus WBRT 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 [34]. In contrast, Mintz et al. studied a group of 84 patients and did not show an advantage of surgery plus radiotherapy over radiotherapy alone [35]. 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 [36]. Surrogate end points, 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 end points 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 intraarterial 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 [37–39], Black [40, 41], and Adler [42–44] published early reports on linear accelerator radiosurgery for brain metastases. Alexander [41] reported on 248 patients. Median tumor volume was 3 cm3 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 multi-institutional study of 122 patients [45]. Actuarial 1- and 2-year survivals were 53% and 30%, respectively. Local control was 86%. Many other linac series have been reported [39, 46–53]. 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

10.

linear accelerator radiosurgery

radiosurgical treatment [54–56]. 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 [57, 58], the Radiation Therapy Oncology Group (RTOG) recursive partitioning categories [59], and radiation dose. Recently, the University of Florida published their experience with radiosurgery for brain metastases [60]. 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% (Figs. 10-8 and 10-9). 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.

Malignant Gliomas Current conventional treatment for malignant gliomas involves a combination of surgery, radiation, and, often, chemotherapy. The prognosis in these patients remains poor [61]. The majority of recurrences occur within 2 cm of the enhancing lesion as seen on initial imaging. Gross total excision may be associated with prolonged median survival in patients with malignant gliomas. Some studies have shown that other aggressive local therapies, such as interstitial brachytherapy, may favorably impact survival [62–64]. Radiosurgery is another attempt at forestalling local recurrence by aggressive local therapy.

FIGURE 10-8. The patient with known breast carcinoma presented with symptomatic pontine lesions. She was treated with radiosurgery (15 Gy to the 80% isodose line).

135

FIGURE 10-9. Three years later, the site of the lesion of Fig. 10-8 was barely visible.

Malignant gliomas account for approximately 40% of the 17,000 primary brain tumors diagnosed annually in the United States. The prognosis for long-term survival remains poor. More than 80% of recurrences are found within 2 cm of the original tumor site. Many attempts have been made to improve long-term survival by improving local control [65, 66]. Such therapies include aggressive surgical removal, brachytherapy, chemotherapy wafers, and radiosurgery. In this retrospective study, we have attempted to ascertain whether the use of radiosurgical boost, whether given as part of initial tumor therapy or at the time of recurrence, increases survival compared to historical controls. A number of linear accelerator radiosurgery series have been published. Shrieve and colleagues reported on 32 patients receiving interstitial brachytherapy and 86 patients receiving radiosurgical boost [67]. They found similar survival rates between the two groups and recommended radiosurgery because of its outpatient, noninvasive nature. Hall and colleagues reported 35 patients and believed that radiosurgery did confer a survival advantage, with fewer complications than brachytherapy [68]. Buatti et al., at the University of Florida, reported on 11 patients treated with radiosurgical boost [65]. No signiﬁcant survival advantage was found. Likewise, Masciopinto and colleagues [69] reported on 31 patients so treated and found that the “curative value of radiosurgery is signiﬁcantly limited by peripheral recurrence.” Other studies include those of Regine [70], Prisco [71], and Gannett [72]. A recurring theme in all retrospective studies of brain tumor therapies is the question of selection bias inﬂuencing the results of therapy more than the therapy itself. In an attempt to control for selection bias in retrospective treatment trials for malignant gliomas [73], Curran [74] developed the recursive partitioning analysis categories, and Sarkaria and colleagues used this methodology to analyze 115 patients from three institutions treated with linear accelerator radiosurgery [75]. They found that patients treated with radiosurgery had a signiﬁcantly improved 2-year and median survival compared with RTOG

136

w.a. friedman

historical controls. The improvement was seen predominately in the worst prognostic classes (3 to 6 classes). Kondziolka performed a similar analysis on 65 patients who underwent upfront radiosurgery [76]. He also found that patients in RTOG classes 3 to 5 appeared to beneﬁt. At the University of Florida, we have retrospectively reviewed 100 patients with WHO grade III and IV malignant gliomas who received SRS boost therapy for residual or recurrent enhancing disease [77]. The patients in our study were divided into recursive partitioning analysis (RPA) classiﬁcations for comparison with historical controls. Class III and IV patients had median survival times very similar to the historical controls. Class V patients demonstrated an increase in median survival (15.6 vs. 8.9 months) and 2-year survival rate (12.5% vs. 6%) compared with historical controls. Eloquent location correlated with poorer survival. This may be due to the selection of less aggressive therapies for this group of patients. Recurrence at time of radiosurgery was associated with longer survival. Very probably, this reﬂects the fact that patients judged “eligible” for radiosurgery at time of recurrence are already selected for longer survival than the average patient treated up front. However, it remains possible that radiosurgery at time of recurrence is truly more effective than upfront radiosurgery. What about drawbacks of the recursive partitioning technique? The RTOG classes used are broad and do not include all known prognostic variables, most notably tumor size. In addition, important linear variables like age, mental status, and KPS are converted into binary ones. This approach, therefore, is ﬂawed, as are all attempts at retrospective analysis. Irish and colleagues, in an analysis of 101 consecutive malignant glioma patients, have shown that those “eligible” for radiosurgery have a median survival of 23.4 months, compared with 8.6 months for “ineligible” patients [78]. Likewise, Curran found a marked survival advantage in radiosurgery “eligible” versus “ineligible” patients [79]. The only complete solution to the issue of selection bias affecting outcome is a prospective randomized study. Such a study has been performed and the results recently published. RTOG Study 93-05 randomized patients with glioblastoma into two treatment arms [80]. One received postoperative radiosurgery, followed by conventional radiotherapy and BCNU chemotherapy. The other arm received radiotherapy and chemotherapy without radiosurgery. At a median follow-up time of 61 months, the median survival in the radiosurgery group was 13.5 months compared with 13.6 months in the standard treatment arm. There were no signiﬁcant differences in 2or 3-year survival, patterns of failure, or quality of life between the two groups. Notably, RTOG 93-05 did not address the use of radiosurgery for recurrent malignant gliomas.

Arteriovenous Malformations Patient Selection Open surgery is generally favored if an arteriovenous malformation (AVM) is amenable to low-risk resection (e.g., low Spetzler-Martin grade, young healthy patient) or is believed to be at high risk for hemorrhage during the latency period

between radiosurgical treatment and AVM obliteration (e.g., associated aneurysm, prior hemorrhage, large AVM with diffuse morphology, venous outﬂow obstruction). Radiosurgery is favored when the AVM nidus is small (<3 cm) and compact, when surgery is judged to carry a high risk or is refused by the patient, and when the risk of hemorrhage is not believed to be extraordinarily high. Endovascular treatment, although rarely curative alone, may be useful as a preoperative adjunct to either microsurgery or radiosurgery. The history, physical examination, and diagnostic imaging of each patient are evaluated and the various factors outlined above are weighed in combination to determine the best treatment approach for a given case. The decision about optimal AVM treatment is best made by a multidisciplinary team composed of experts in operative, endovascular, and radiosurgical treatment.

Stereotactic Image Acquisition The most problematic aspect of AVM radiosurgery is target identiﬁcation. In some series, targeting error is listed as the most frequent cause of radiosurgical failure [81, 82]. The problem lies with imaging. Although angiography very effectively deﬁnes blood ﬂow (feeding arteries, nidus, and draining veins), it does so in only two dimensions. Using the two-dimensional data from stereotactic angiography to represent the three-dimensional target results in signiﬁcant errors of both overestimation and underestimation of AVM nidus dimensions [83–85]. Underestimation of the nidus size may result in treatment failure, and overestimation results in the inclusion of normal brain within the treatment volume. This can cause radiation damage to normal brain, which—when affecting an eloquent area—may result in a neurologic deﬁcit. To avoid such targeting errors, a true three-dimensional image database is required. Both contrast-enhanced CT and MRI are commonly used for this purpose. Diagnostic (nonstereotactic) angiography is used to characterize the AVM, but because of its inherent inadequacies as a treatment planning database, stereotactic angiography has been largely abandoned at our institution. We use contrast-enhanced, stereotactic CT as a targeting image database for the vast majority of AVMs. Our CT technique employs rapid infusion (1 mL/s) of contrast while scanning through the AVM nidus with 1-mm slices. The head ring is bolted to a bracket at the head of the CT table, ensuring that the head/ring/localizer complex remains immobile during the scan. This technique yields a very clear three-dimensional picture of the nidus. Alternative approaches use MRI/MRA as opposed to CT. Attention to optimal image sequences in both CT and MRI is essential for effective AVM radiosurgical targeting.

Dose Selection Various analyses of AVM radiosurgery outcomes have elucidated an appropriate range of doses for the treatment of AVMs [86–89]. We prefer to deliver a dose of 20 Gy to the periphery of the AVM nidus whenever possible. Larger AVMs, or those in critical locations, may require a lower dose—but this will reduce the chances of complete obliteration.

10.

linear accelerator radiosurgery

FIGURE 10-10. Pretreatment angiogram shows a left parietal AVM. It was treated with radiosurgery (17.5 Gy to the 70% line using three isocenters).

Follow-up Standard follow-up after AVM radiosurgery typically consists of annual clinic visits with MRI/MRA to evaluate the effect of the procedure and monitor for neurologic complications (Figs. 10-10 and 10-11). If the patient’s clinical status changes, he is followed more closely at clinically appropriate intervals. Each patient is scheduled to undergo cerebral angiography at 3 years after radiosurgery, and a 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 followup. 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.

137

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” that 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 cm3 (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 AVMs [89]. 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 SpetzlerMartin 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 occurrence of transient radiation-induced complications. Better conformality correlated with a reduced incidence of transient complications. Lower Spetzler-Martin scores, higher doses, and steeper dose gradients correlated with radiographic success. 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 [90]. 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 University of Florida Experience From May 18, 1988 to March 22, 2005, 544 patients with AVMs were treated at the University of Florida. The mean age was 40 (range, 4 to 78 years). The median treatment volume was 7 cm3 (range, 2 to 45.3 cm3). Many patients early in the series were treated with single isocenters (259), 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 1750 cGy and the mean follow-up duration was 31 months. Presenting symptoms included the following: headache/ incidental (188), seizure (227), hemorrhage (179), progressive neurological deﬁcit (23). Spetzler-Martin scores were as follows: I, 29; II, 188; III, 228; IV, 98. AVMs were further delineated into four nidus volume categories: A, <1 cm3; B, 1 to 4 cm3; C, 4 to 10 cm3; D, >10 cm3. Angiographic/MRI cure rates were as follows: A, 92%; B, 79%; C, 64%; and D, 36%. Ellis et al. [81] performed a detailed

FIGURE 10-11. Two years later, the angiogram of the AVM patient of Fig. 10-10 is normal.

138

w.a. friedman

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 end points included angiographic cure, radiosurgical failure (documented persistence of AVM ﬂow 3 years after retreatment), and death. The mean original lesion volume was 13.8 cm3 and the mean volume at retreatment was 4.7 cm3, for an average volume reduction of 66% after the initial “failed” treatment. Only two (3.8%) AVMs failed to demonstrate size reduction after primary treatment. The median doses on initial and repeat treatment were 12.5 and 15 Gy, respectively. To date, 25 retreated patients have reached a deﬁnitive end point. These include 15 (60%) angiographically documented cures, 9 (36%) angiographically documented failures, and 1 fatal hemorrhage. A single permanent radiation-induced complication occurred among 52 (1.9%) patients, and 1 patient experienced a transient deﬁcit that resolved with steroid therapy. Two hemorrhages (one fatal) occurred during a total of 130 patient-years at risk, resulting in a 1.5% annual incidence of posttreatment hemorrhage. If one includes retreatments in the analysis of radiosurgical success, the results are as follows: A, 100%; B, 92%; C, 85%; D, 82%. Ten (1.8%) patients sustained a permanent radiationinduced complication. Seventeen (3.1%) had a transient radiation-induced complication. These problems usually resolved within several months of steroid therapy. Most importantly, 42 patients suffered hemorrhages after radiosurgical treatment, and 8 were fatal. Hemorrhage during the latent period after radiosurgery is the major drawback of this procedure. Only surgery at this point can immediately eliminate the risk of hemorrhage in patients with AVMs.

References 1. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316–319. 2. Betti OO, Derechinsky VE. Hyperselective encephalic irradiation with a linear accelerator. Acta Neurochir Suppl 1984; 33:385–390. 3. Colombo F, Benedetti A, Pozza F, et al. External stereotactic irradiation by linear accelerator. Neurosurgery 1985; 16:154–160. 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; 11:1185–1192. 5. McGinley PH, Butker EK, Crocker IR, Landry JC. A patient rotator for stereotactic radiosurgery. Phys Med Biol 1990; 35:649–657. 6. Podgorsak EB, Olivier A, Pla M, et al. Dynamic stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1988; 14:115–126. 7. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for stereotactic radiosurgery. Neurosurgery 1988; 22:454–464. 8. Friedman WA, Bova FJ. The University of Florida radiosurgery system. Surg Neurol 1989; 32:334–342. 9. Friedman WA, Buatti JM, Bova FJ, Mendenhall WM. LINAC Radiosurgery: A Practical Guide. Berlin: Springer-Verlag, 1998. 10. Samii M, Matthies C. Management of 1000 vestibular schwannomas (acoustic neuromas): surgical management and results with an emphasis on complications and how to avoid them. Neurosurgery 1997; 40:11–23. 11. Samii M, Matthies C. Management of 1000 vestibular schwannomas (acoustic neuromas): The facial nerve–preservation and restitution of function. Neurosurgery 1997; 40:684–695.

12. Samii M, Matthies C. Management of 1000 vestibular schwannomas (acoustic neuromas): Hearing function in 1000 tumor resections. Neurosurgery 1997; 40:248–262. 13. Leksell L. A note on the treatment of acoustic tumors. Acta Chir Scand 1971; 137:763–765. 14. Foote KD, Friedman WA, Buatti JM, et al. Analysis of risk factors associated with radiosurgery for vestibular schwannoma. J Neurosurg 2001; 95(3):440–449. 15. Spiegelmann R, Gofman J, Alezra D, Pfeffer R. Radiosurgery for acoustic neurinomas (vestibular schwannomas). Isr Med Assoc J 1999; 1(1):8–13. 16. Spiegelmann R, Lidar Z, Gofman J, et al. Linear accelerator radiosurgery for vestibular schwannoma. J Neurosurg 2001; 94(1): 7–13. 17. Martens F, Verbeke L, Piessens M, Van Vyve M. Stereotactic radiosurgery of •vestibular schwannomas with a linear accelerator. Acta Neurochir 1994; 62(Suppl):88–92. 18. Valentino V, Raimondi AJ. Tumour response and morphological changes of acoustic neurinomas after radiosurgery. Acta Neurochir 1995; 133:157–163. 19. Delaney G, Matheson J, Smee R. Stereotactic radiosurgery: an alternative approach to the management of acoustic neuromas. Med J Austral 1992; 156:440. 20. Barcia Salorio JL, Hernandez G, Ciudad J, et al. Stereotactic radiosurgery in acoustic neurinoma. Acta Neurochir Suppl 1984; 33:373–376. 21. Sekhar LN, Jannetta PJ, Burkhart LE, Janosky JE. Meningiomas involving the clivus: a six-year experience with 41 patients. Neurosurgery 1990; 27:764–781. 22. Sekhar LN, Altschuler EM. Meningiomas of the cavernous sinus. In: Al-Mefty O, ed. Meningiomas. New York: Raven Press, 1991:445–460. 23. Simpson D. The recurrence of intracranial meningiomas after surgical treatment. J Neurol Neurosurg Psychiatry 1957; 20:22–39. 24. Pollock BE, Stafford SL, Utter A, et al. 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(4):1000–1005. 25. Valentino V, Schinaia G, Raimondi AJ. The results of radiosurgical management of 72 middle fossa meningiomas. Acta Neurochir 1993; 122:60–70. 26. Villavicencio AT, Black PM, Shrieve DC, et al. Linac radiosurgery for skull base meningiomas. Acta Neurochir (Wien) 2001; 143(11):1141–1152. 27. 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; 19:1021–1026. 28. Spiegelmann R, Nissim O, Menhel J, et al. Linear accelerator radiosurgery for meningiomas in and around the cavernous sinus. Neurosurgery 2002; 51(6):1373–1380. 29. Hakim R, Alexander III E, Loefﬂer JS, et al. Results of linear accelerator-based radiosurgery for inracranial meningiomas. Neurosurgery 1998; 42:446–454. 30. Friedman WA, Murad G, Bradshaw P, et al. Linear accelerator radiosurgery for meningiomas. J Neurosurg 2005; 103:206–209. 31. Lohr F, Pirzkall A, Hof H, et al. Adjuvant treatment of brain metastases. Semin Surg Oncol 2001; 20:50–56. 32. DeAngelis LM. Brain tumors. N Engl J Med 1990 2001; 344: 114–123. 33. 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. 34. Noordijk EM, Vecht CJ, Haaxma-Reiche H, et al. The choice of treatment of single brain metastasis should be made based on extracranial tumor activity and age. Int J Radiat Oncol Biol Phys 1994; 29:711–717.

10.

linear accelerator radiosurgery

35. 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; 78:1470–1476. 36. Haines SJ. Moving targets and ghosts of the past: outcome measurement in brain tumour therapy. J Clin Neurosci 2002; 9:109–112. 37. 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; 13:279–282. 38. Sturm V, Kimmig B, Engenhardt R, et al. Radiosurgical treatment of cerebral metastases. J Stereo Func Neurosurg 1991; 57:7–10. 39. Voges J, Treuer H, Erdmann J, et al. LINAC radiosurgery in brain metastases. Acta Neurochir 1994; 62(Suppl):72–76. 40. Black PM. Solitary brain metastases. Radiation, resection, or radiosurgery? Ch 1993; 103:367S–369S. 41. 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. 42. Adler JR, Cox RS, Kaplan I, Martin DP. Stereotactic radiosurgical treatment of brain metastases. J Neurosurg 1992; 76:444–449. 43. Fuller BG, Kaplan ID, Adler J, et al. Stereotaxic radiosurgery for brain metastases: The importance of adjuvant whole brain irradiation. Int J Radiat Oncol Biol Phys 1992; 23:413–418. 44. Joseph J, Adler JR, Cox RS, Hancock SL. Linear acceleratorbased stereotaxic radisourgery for brain metastases: the inﬂuence of number of lesions on survival. J Clin Oncol 1996; 14:1085–1092. 45. 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. 46. Becker G, Jeremic B, Engel C, et al. Radiosurgery for brain metastases: the Tuebingen experience. Radiother Oncol 2002; 62: 233–237. 47. Breneman JC, Warnick RE, Albright RE, et al. Stereotactic radiosurgery for the treatment of brain metastases. Cancer 1997; 79: 551–557. 48. Buatti JM, Friedman WA, Bova FJ, Mendenhall WM. Treatment selection factors for stereotactic radiosurgery of intracranial metastases. Int J Radiat Oncol Biol Phys 1995; 32:1161–1166. 49. Caron J-L, Souhami L, Podgordak EB. Dynamic stereotactic radiosurgery in the palliative treatment of cerebral metastatic tumors. J Neuro-Oncol 1992; 12:173–179. 50. Gutin PH, Wilson CB. Radiosurgery for malignant brain tumors. J Clin Oncol 1990; 8:571–573. 51. 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; 24:619–625. 52. Mehta M, Noyes W, Craig B, et al. A cost-effectiveness and costutility analysis of radiosurgery vs. resection for single-brain metastases. Int J Radiat Oncol Biol Phys 1997; 39:445–454. 53. Valentino V. The results of radiosurgical management of 139 single cerebral metastases. Acta Neurochir Suppl 1995; 63:95– 100. 54. Cho KH, Hall WA, Gerbi BJ, et al. Patient selection criteria for the treatment of brain metastases with stereotactic radiosurgery. J Neurooncol 1998; 40:73–86. 55. Fernandez-Vicioso E, Suh JH, Kupelian PA, et al. Analysis of prognostic factors for patients with single brain metastasis treated with stereotactic radiosurgery. Radiat Oncol Invest 1997; 5:31– 37. 56. Maor MH, Dubey P, Tucker SL, et al. Stereotactic radiosurgery for brain metastases: results and prognostic factors. Int J Cancer 2000; 90:157–162. 57. Goodman KA, Sneed PK, McDermott MW, et al. Relationship between pattern of enhancement and local control of brain metas-

58.

59.

60.

61. 62.

63.

64.

65.

66. 67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

139

tases after radiosurgery. Int J Radiat Oncol Biol Phys 2001; 50:139–146. 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; 37:375–383. 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; 37:745–751. Ulm AJ, Friedman WA, Bova FJ, et al. Linear accelerator radiosurgery in the treatment of brain metastases. Neurosurgery 2004; 55:1076–1085. DeAngelis LM. Brain tumors. N Engl J Med 2001; 344(2):114– 123. 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):1213–1217. Chang CN, Chen WC, Wei KC, et al. High-dose-rate stereotactic brachytherapy for patients with newly diagnosed glioblastoma multiformes. J Neurooncol 2003; 61(1):45–55. Prados MD, Gutin PH, Phillips TL, et al. Interstitial brachytherapy for newly diagnosed patients with malignant gliomas: the UCSF experience. Int J Radiat Oncol Biol Phys 1992; 24(4): 593–597. 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):205–210. Friedman WA, Foote KD. Linear accelerator radiosurgery in the management of brain tumours. Ann Med 2000; 32(1):64–80. 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): 275–282; discussion 82–84. Hall WA, Djalilian HR, Sperduto PW, et al. Stereotactic radiosurgery for recurrent malignant gliomas. J Clin Oncol 1995; 13(7):1642–1648. Masciopinto JE, Levin AB, Mehta MP, Rhode BS. Stereotactic radiosurgery for glioblastoma: a ﬁnal report of 31 patients. J Neurosurg 1995; 82(4):530–535. Regine WF, Patchell RA, Strottmann JM, et al. 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): 421–426. Prisco FE, Weltman E, de Hanriot RM, Brandt RA. Radiosurgical boost for primary high-grade gliomas. J Neurooncol 2002; 57(2): 151–160. Gannett D, Stea B, Lulu B, et al. 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):461–468. Roberge D, Souhami L. Stereotactic radiosurgery in the management of intracranial gliomas. Technol Cancer Res Treat 2003; 2(2):117–125. 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):704–710. Sarkaria JN, Mehta MP, Loefﬂer JS, et al. Radiosurgery in the initial management of malignant gliomas: survival comparison with the RTOG recursive partitioning analysis. Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1995; 32(4): 931–941. Kondziolka D, Flickinger JC, Bissonette DJ, et al. Survival beneﬁt of stereotactic radiosurgery for patients with malignant glial neoplasms. Neurosurgery 1997; 41(4):776–783; discussion 83–85.

140

w.a. friedman

77. Ulm AJ, Friedman WA, Bova FJ, et al. Radiosurgery for malignant gliomas: the University of Florida experience. Neurosurgery 2005; 57:512–517. 78. 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):307–312. 79. Curran WJ Jr, Scott CB, Weinstein AS, et al. Survival comparison of radiosurgery-eligible and -ineligible 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):857–862. 80. 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):853–860. 81. Ellis TL, Friedman WA, Bova FJ, et al. Analysis of treatment failure after radiosurgery for arteriovenous malformations. J Neurosurg 1998; 89(1):104–110. 82. Pollock BE, Flickinger JC, Lunsford LD, et al. Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 1998; 42(6):1239–1244; discussion 44–47. 83. Bova FJ, Friedman WA. Stereotactic angiography: an inadequate database for radiosurgery? Int J Radiat Oncol Biol Phys 1991; 20:891–895.

84. Blatt DL, Friedman WA, Bova FJ. Modiﬁcations in radiosurgical treatment planning of arteriovenous malformations based on CT imaging. Neurosurgery 1993; 33:588–596. 85. Spiegelmann R, Friedman WA, Bova FJ. Limitations of angiographic target localization in planning radiosurgical treatment. Neurosurgery 1992; 30:619–624. 86. 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: 873–879. 87. Karlsson B, Lindquist C, Steiner L. Prediction of obliteration after gamma knife surgery for cerebral arteriovenous malformations. Neurosurgery 1997; 40(3):425–430; discussion 30–31. 88. Pollock BE, Kondziolka D, Lunsford LD, et al. Repeat stereotactic radiosurgery of arteriovenous malformations: factors associated with incomplete obliteration. Neurosurgery 1996; 38(2): 318–324. 89. 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. 90. Foote KD, Friedman WA, Ellis TL, et al. Salvage retreatment after failure of radiosurgery in patients with arteriovenous malformations. J Neurosurg 2003; 98:337–341.

1 1

Proton Beam Radiosurgery: Physical Bases and Clinical Experience Georges Noel, Markus Fitzek, Loïc Feuvret, and Jean Louis Habrand

Introduction The introduction of photon therapy into the armamentarium of cancer treatment represented a breakthrough in the early 20th century. Since then, major technical innovations in radiation oncology have been continuously associated with major improvements in local tumor control and patient quality of life. The development of computer technology two decades ago allowed an unprecedented step forward with the design of computer-driven linear accelerators, multileaf collimation, threedimensional treatment planning, and intensity modulation of the beam. In parallel, biological concepts on tumor control probability have been modeled and studied in relation to tumor volume, clonogens sensitivity, biological environment, and so forth. A basic concept remained, however; that is, the probability of local and to some extent distant control are dose-correlated to the primary tumor. This gave the impetus to the exploration of new forms of particles, especially the “biologically” effective ones, like neutrons. In our opinion, disappointing results in this ﬁeld should not discourage, but rather pave avenues to new investigations, dealing with heavy ions for example. An alternative approach is based on purely “ballistically” effective particles, with protons as a paradigm. They have been investigated successfully in multiple dose-escalation trials concerning so-called radioresistant tumors, like ocular melanomas and low-grade sarcomas at the skull base. These results remind us that the simpler the concept, the better! In less-challenging clinical situations, special types of radiation have also been of interest: for many years, radiation oncologists have learned to treat superﬁcial or semideep tumor sites routinely with an electron beam, the optimal and simplest way to spare surrounding organs. Similarly, the management of deep-sited

conditions can be elegantly approached using high-energy protons. Although the technical complexity and cost associated with their production and delivery has frequently been put forward by detractors, it is generally (from a ballistic standpoint) the simplest way to deliver the dose with optimal conformation and with a superior dose-scattering limitation around the target. Thus, the development of proton therapy seems ultimately associated with the need for ionizing radiation that minimizes the risk of long-term side-effects, including carcinogenicity. This might be one of the great public health challenges for the coming decade.

Early History of Proton Radiation Protons are part of the hadron family. This term derives from the old Greek αδρος, which means “strength.” They are actually high-energy particles made up of quarks (elementary particles of the atomic nucleus). In radiology, it has become synonymous with protons or neutrons, as they are basic components of the atomic nucleus, and even with the atomic nuclei themselves. The latter are also known as “light ions”(especially helium, oxygen, and carbon ions) or “heavy ions” (especially neon and argon) [1]. The history of proton therapy began in December 1904 when William Henry Bragg described the peak absorption in the air of alpha particles. In their founding paper (published 1905), Bragg and Kleeman reported the deﬁnite path of the particles, correlated with their initial energy and sharp increase of the ionization density as they approached their range-end [2]. Unlike the rapid development of X-rays in diagnosis and treatment, the medical applications of charged particles did

141

142

g. noel et al.

TABLE 11-1. Hadron therapy centers (except neutrons) and numbers of treated patients (data of January 2005). Center

Country

Particle

Closed centers Uppsala (1) PSI (SIN) Louvain-la-Neuve Dubna (1) Los Alamos Berkeley 184 Berkeley Berkeley Harvard MPRI (1) PMRC (1) TRIUMF

Dates of creation

No. treated patients

Sweden Switzerland Belgium Russia USA USA USA USA USA USA Japan Canada

Protons Pions Protons Protons Pions Protons Helium Heavy ions Protons Protons Protons Pions

1957–1976 1980–1993 1991–1993 1967–1996 1974–1982 1954–1957 1957–1992 1975–1992 1961–2002 1993–1999 1983–2000 1979–1994

73 503 21 124 230 30 2,054 433 9,116 34 700 367

Open centers Western Europe Uppsala (2) PSI (72 MeV) PSI (200 MeV) Nice Orsay Clatterbridge INFN-LNS GSI HMI

Sweden Switzerland Switzerland France France England Italy Germany Germany

Protons Protons Protons Protons Protons Protons Protons Heavy ions Protons

1989 1984 1996 1991 1991 1989 2002 1997 1998

418 4,182 209 2,555 2,805 1,372 82 198 546

Eastern Europe ITEP St. Petersburg Dubna (2)

Russia Russia Russia

Protons Protons Protons

1969 1975 1987

3,785 1,145 296

Asia Chiba PMRC (2) HIMAC NCC HIBMC HIBMC WERC Shizuoka Wanjie, Zibo

Japan Japan Japan Japan Japan Japan Japan Japan China

Protons Protons Heavy ions Protons Protons Heavy ions Protons Protons Protons

1979 2001 1994 1998 2001 2002 2002 2003 2004

145 492 1,796 300 483 30 19 100 1

USA USA USA USA Canada

Protons Protons Protons Protons Protons

1990 1993 2001 1994 1995

9,585 21 973 632 89

South Africa Pions Ions Protons All particles

Protons

1993

468 1,100 4,511 40,801 46,412

North America Loma Linda MPRI (2) NPTC, MGH UCSF-CNL TRIUMF Africa iThemba LABS Total

Source: From Particles. A newsletter of the Particle Therapy Cooperative Group. 2005; 35:10. Available at http://ptcog.web.psi.ch/ptles35.pdf. Used with permission.

11.

143

proton beam radiosurgery

not start until after World War II. Robert Wilson, who had been chief of the cyclotron team during the Manhattan Project, elected to devote (like many other physicists at that time) his future research activities to the beneﬁt of mankind. Soon thereafter (1946) was brought out “radiological use of fastprotons” [3]. This was the ﬁrst proposal for use of proton beam in radiation treatment. He challenged scientists to achieve a uniform dose-distribution whatever the tumor thickness and proposed a solution still valid (i.e., the superimposition of “native” Bragg peaks of different energies and penetration generated by a rotating “ridge ﬁlter” later on described as the “spread out” Bragg peak). This prophetic announcement did not come true until the pioneering experiments in a variety of neurologic disorders by Dr. William Sweet, head of the Neurosurgery Department at Massachusetts General Hospital (MGH), in the early 1960s, then joined by Dr. Raymond Kjellberg, a young Swedish surgeon trained at the Berkeley cyclotron. All in all, it had taken 15 years between the Wilson paper and the ﬁrst patient treated on May 25, 1961! This was a 2year-old girl with a suprasellar tumor (reported in [4]). This group concentrated thereafter on pituitary conditions such as acromegaly, diabetic retinopathy, and eventually in the 1980s on arteriovenous malformations (AVMs) [5]. However, between 1967 and 1973, the future of proton therapy remained unclear with the projected cyclotron shut down. Fortunately, the project was deﬁnitely strengthened upon the arrival of Dr. Herman Suit, recently appointed chief of radiation medicine at MGH. He was joined by Michael Goitein, PhD, in charge of the physics project, and addressed radiobiological issues, in relation to the protons’ relative biologic effectiveness (RBE): this was assessed in 1972, to 1.1, a ﬁgure still in use in most centers worldwide [6]. In February 1974, the ﬁrst patient with a pelvic sarcoma was treated with a fractionated schedule. In parallel, an active ophthalmologic program was initiated in ocular melanomas under the auspices of the Massachusetts Eye and Ear Inﬁrmary (MEEI), which became for many years the cyclotron spearhead. The history has retained the ﬁrst patient’s name, Mr. Mc Kelvey [4]. By January 2005, approximately 40,000 patients worldwide had been treated with protons, almost half at the HCL (Table 11-1) [7].

matter and presents a major straggling effect (deﬁned as “lateral penumbra”) on depth-dose proﬁles. In contrast with indirectly ionizing radiation, charged particles are directly ionizing and exhibit a deﬁnite range in matter. Because of their low masses, electron beams interact at rapid velocity (close to speed of light), with relatively limited and uniform interactions with matter. They also come rapidly to their range-end at the usual energy domain in radiotherapy. When they come to their end, they are strongly submitted to a strong lateral scattering (i.e., large penumbra). Mono energetic proton beams generated in large accelerators such as synchrotrons produce particles of highly uniform range (of the order 1%). This limited but actual heterogeneity is due in part to the nuclear interactions that can affect some of the particles. With current generators, deep beam penetrations (of the order 15 to 30 cm in water equivalent material) can be achieved. As the mass of protons are about 2000 times that of electrons, they evidence considerably higher kinetic energy (at equal velocity) and so the capability for releasing this energy much more considerably as they are slowing down. Furthermore, the projectiles are barely deﬂected during the process (that translates into a sharp lateral penumbra on dose-proﬁles). These properties translate into the so-called Bragg curve (Fig. 11-1): relatively limited interactions in their initial path and limited dose-absorption and a steep rise close to their range-end, followed by a sharp fall-off of the dose. There is an increased linear energy transfer (LET) in the Bragg peak that is in turn linked with increased cell killing and RBE (see below). This phenomenon could potentially have a deleterious effect on normal tissues located at the Bragg peak, although the concomitant abrupt fall-off of the dose seems to minimize considerably its impact. Because of these uncertainties, when planning a proton treatment, it is our practice to avoid abutment of the distal end region to a critical organ surface (especially brain stem).

1.0 proton Bragg peak

0.8

Facilities 0.6

SOBP

DOSE

Approximately 30 centers worldwide have implemented a clinical program dealing with heavy particles. Proton therapy itself approaches 20 centers (Table 11-1).

0.4

Physical Properties Photons as they pass through matter interact with atomic electrons and induce ionization when the energy transfer is superior to the energy that binds them with the nucleus. This corresponds with their main energy transfer process. Because of the statistical nature of the events, and the major deﬂections of interacting particles, the photon beam has no deﬁnite path in

0.2

0.0 0

10

20

30

DEPTH (cm) FIGURE 11-1. Native Bragg peak (in red) and spread-out Bragg peak result of the addition of several native Bragg peaks.

144

g. noel et al.

FIGURE 11-2. (A) Plastic and (B) metallic modulators to modulate, respectively, eye and base of the skull tumors.

When the Bragg peak is commonly “spread out” (in order to cover targets thicker than the “native” Bragg peak width) using a modulator (Fig. 11–2), by the emission of protons of various energies, the narrow, high-LET zone is repositioned many times within the “plateau” and “diluted” among longrange low-LET protons. This dilution does not affect obviously the end of the deeper peak. Another drawback of spreading out the peak is an entrance-dose increase (Fig. 11-1). It should be pointed out that the stopping power of different tissues to protons differs substantially as their densities decrease: corrections need to be done when the beam passes from bones to soft tissues (water equivalent), and above all to air cavities. Technical solutions call for the design of appropriate “compensators,” or the production of “pencil beams” of predeﬁned energies, made possible by the most recent accelerators. We will also mention that the rare nuclear fragmentations that affect the beam at its very end are positron emitters and could in theory be detected using a positron scanning. This property is not as prominent as with heavy ions where it has come to clinical application.

Protons Production and Delivery Systems: Conventional Versus Innovative Heavy charged particles are accelerated in large circular cyclotrons, synchrocyclotrons, and synchrotrons (Fig. 11-3). Although synchrotrons seem the most appropriate to produce protons of high and variable energies, there has recently been a renowned interest for modiﬁed cyclotrons that can achieve excellent performance at relatively low cost. The basic principle of proton production is based on the acceleration in an electromagnetic ﬁeld of a hydrogen plasma (i.e., stripped H nuclei ionized by an electric arc). As mentioned above, the beam can be spread out by changing its energy by (1) combining multiple energies by steps of a few MeV; and (2) interposing upstream of a ﬁxed energy beam speciﬁc rotating “ridge” ﬁlters that alter the beam’s path (Fig. 11-2) [8]. Large accelerators were initially developed in a nuclear physics environment, and so early proton facilities were found at university campuses or nuclear plants. This came

FIGURE 11-3. (A) Magnets of the synchrocyclotron at Orsay (2 yellow rings) and (B) magnet to change the beam direction to serve different rooms.

11.

proton beam radiosurgery

along with simple technology, that is, ﬁxed beam, passive scattering, hand-made compensators, and so forth. Recently, hospital-based facilities have been developed, especially in the United States. This has been made possible because modern accelerators are somewhat more compact than the former ones. Nonetheless, recent equipment goes along with sophisticated beam delivery systems including isocentric gantries that are considerably space-consuming. Thus, the main reason for this geographic transfer is likely elsewhere and lies in the need for patients and staff to beneﬁt from immediate medical support, multidisciplinary management, imaging panel, social support, and so forth. Moreover, part of the indications can be approached using a combination of protons and of photons (which lowers substantially the treatment cost), and younger patients can call for general anesthesia. These services are provided by regular departments of radiation oncology. Two sets of energy beams should be made available to fully cover indications: 50 to 60 MeV for ophthalmologic programs, and 200 to 250 MeV for deep sites. Multiple technical innovations have improved clinical indications and treatment accuracy for the past decade. Isocentric gantry technology is still in progress, but prototypes of remarkable accuracy have been developed despite considerable dimensions (i.e., 6 to 11 m) [9]. Gantries, like in photon therapy, allow isocentric planning and treatment and expand arrangement capabilities of beams (including non-coplanar if combined with couch rotations). In turn, this opens up new potential indications, especially intrathoracic and abdominal malignancies. They also affect positively set-up complexity, duration, and accuracy as well as patient comfort. Proton intensity modulation of the beam is a recent concept that is somewhat different from that developed with photons [10]; although the goal is similar (optimal conformation to the target), the approach is different and based on the beam’s penetration modulation rather than the beam’s ﬂuence modulation. Proton beams in the past have been exclusively generated by passive scattering through lead foils (passive diffusion) and then adapted to the tumor distal shape by compensating depth variations. This was achieved using individualized “boluses” for each beam (compensating also for tissue heterogeneities). Unfortunately, this approach does not compensate for proximal tumor shape and is responsible for unnecessary irradiation of tissues located upstream of the target. This “intrinsic” drawback can only be minimized using a multiple-beams arrangement. Recent innovations have aimed to conform dose tightly in three dimensions: The variable modulation method based on a dynamic multileaf collimator acting in conjunction with stepwise Bragg peaks penetration, in order to trim the target volume in subelements from depth to surface; pencil beam scanning eliminates the need for individualized collimators and compensators [11], but requires tight control processes of the dose. Any altered proton ﬂuence, or treatment interruption, or patient’s displacement can seriously affect dose delivery. This is true for raster scanning, an approach in which the beam is deﬂected by two dipole magnets, and more importantly in spot (i.e., point by point) scanning, probably the “ultimate” conformation in radiation oncology.

145

Radiobiological Considerations RBE measurements constitute an integral part of the calibration of clinical hadron beams and contribute, together with dosimetry and, possibly, microdosimetry, to specify the radiation quality. Today, the worldwide reference biological system is the “intestinal crypt regeneration in mice.” RBE multiple biological investigations were conducted at the Harvard cyclotron from the 1960s to the early 1980s that explored mainly brain tolerance in mammals to single fractions administered with different beam apertures. Most of the current knowledge on single-dose radiosurgery (both with photons and protons) has emerged from these data. The rodent intestinal crypt regeneration tests allowed early assessment of the RBE value for rapidly proliferating tissues. It was assessed to a mean of 1.1 [12], although it was likely higher in the peak region (or in the deeper peak as far as the spread-out Bragg peak): 1.4 to 1.5 as mentioned previously. This led to the deﬁnition of the CGE (cobalt Gray equivalent) as the proton dose “unit” by the Boston group. One CGE corresponds with the physical dose times the estimated 1.1 mean RBE. Some experiments conducted elsewhere suggested RBE values ranging between 0.9 and 1.25 [12, 13]. Some lower values can be attributed to the use of orthovoltage equipment as reference radiation instead of cobalt-60. Microdosimetric investigations have recently been conducted that conﬁrm and reﬁne animal models. If RBE estimates can substantially affect dose-reports, physical dose measurements take their own toll: this was evidenced in a recent multicentric intercomparison in which doseunderestimates up to 17% were found for those dealing with Faraday cups against ionization chambers [14]. This led, for example, the Boston group to move ongoing prescribed tumordoses and tolerance-doses to the critical organs up and down by 10%, respectively.

Patient Setup Patient setup is largely conditioned by the treatment room conﬁguration. The procedure implemented at the Orsay center will be brieﬂy discussed in this section and in the following one, which is based on a ﬁxed horizontal beamline. The patient can be either lying on his back or seated. For immobilization purposes, a thick custom-made thermoplastic mask is used. This rather “conventional” approach is offset by a stereotactic alignment that includes the implantation under local anesthesia of four to ﬁve radiopaque ﬁducial markers in the outer skull (made of gold or titanium). Orthogonal X-ray ﬁlms are compared with digitally reconstructed radiographs (DRRs) before each daily session to check alignment and repeated for each ﬁeld. The ﬁducial markers allow appropriate translation and angular corrections using an original computer program (Rotaplus) developed at the Orsay Proton Therapy Center (CPO). Proton treatment time averages 25 minutes with 20 minutes for patient setup and 5 minutes for irradiation itself. Appropriate corrections are made with high precision, using a robotic chair or a couch. Setup accuracy has been checked to less than 1 mm in the x, y, and z directions and less than 1° rotation (Fig. 11-4).

146

g. noel et al.

< 2 mm x 0.8

FIGURE 11-4. Patient setup. Gold ﬁducial marker (A). Fiducial makers in outer bone (B1) or in skull (B2). Patient postioning in two different robotic chairs (C).

Patient Treatment Planning and Simulation The process starts with the acquisition of 3-mm-thick slices from a contrast-enhanced computed tomography (CT) scan and 1.5-mm-thick slices from a contrast-enhanced magnetic res-

onance imaging (MRI) scan, both performed with the patient in a supine position. Target volume and organs at risk beneﬁt from matched or fused imagings, combining both modalities. Threedimensional treatment plans, including dose-volume histograms (DVHs), are generated using a “homemade” software (ISIS 3D

11.

proton beam radiosurgery

147

π patchesεσ

FIGURE 11-5. Beams in patch technique. Example: Distal penumbra of left posterior (A) and a right lateral-posterior (B) beams are in contact with the lateral penumbra of a right lateral beam (C) allowing delivery of a tumor dose (blue) without irradiating the brain stem– spinal cord junction (D).

Treatment Planning System; Technology Diffusion Company). This treatment planning system (TPS) has the advantage of combining photon, electron, and proton dose-distributions. A unique advantage of proton beams arrangement should be mentioned here: the ability to design “patched” ﬁelds, in which the lateral penumbra of a beam is matched with the distal one of a

second (and sometimes a third) beam. This leads to remarkable “bended” isodoses around critical structures such as brain stem or spinal cord (Fig. 11-5). Individualized modulators and compensators are eventually designed by a computer-driven milling machine for each proton beam. Beam shaping is performed using Cerrobend blocks or electro-cut brass blocks (Fig. 11-6).

Without compensator

FIGURE 11-6. Collimator (A) to limit irradiation in two dimensions (2D irradiation) and compensator (B) to compensate indepth the irradiation (third dimension). Collimator was used for a superior ﬁeld and shielded optic nerves and chiasm.

With compensator

148

g. noel et al.

Clinical Applications of Proton Beam Therapy

Eye Tumors

Proton therapy has proved highly valuable in the management of slow-growing, radioresistant tumors that abut radiosensitive structures. In adults, ocular melanomas and base of the skull chordomas and chondrosarcomas typically represent such a challenge and have enjoyed remarkable improvements both in their outcome and in radiation-related toxicity. Proton therapy should also be considered seriously in the therapeutic armamentarium of patients with exquisite sensitivity to radiation (i.e., primarily young patients). Promising experiences have been brought out in pediatric conditions located at the skull base and brain [15–24]. Past and ongoing phase I and II studies could pave new avenues in the treatment of adult gliomas [15–17, 25, 26], intracranial AVMs [27–30], medulloblastomas [31, 32], head and neck malignancies [28, 33, 34], especially nasopharyngeal carcinomas [35–37], esophagus [38, 39], prostate [40, 41], rectum [42, 43], gynecological carcinomas [44–46], soft tissues sarcomas [47], lung carcinomas [48], and hepatocarcinomas [49].

Uveal melanomas have been treated with protons since 1975 at HCL, followed by a remarkable program at the Paul Scherrer Institute (PSI) in Switzerland. Total dose ranged between 60 and 70 CGE in four to ﬁve consecutive fractions [50]. The ﬁrst retrospective analyses evidenced an increased 3-year overall survival for the groups treated with protons over those with enucleation [51–58] (Table 11-2). A deﬁnitive retrospective analysis of MGH-MEEI patients showed in 1488 cases a 5-year local control rate of 96%, along with useful vision preservation in 65% and 39% whether the tumor was small or large, respectively [52]. Egger et al. reported a 10-year local control rate of 94.8% in the Swiss series. In multivariate analysis, prognostic factors for local control were gender, tumor size, margin close to the ciliary body, and extent of safety margin around the GTV [53]. Courdi et al. reported similar ﬁgures in his series of 538 patients treated in Nice. Five-year speciﬁc local control and survival rates were 89% and 86.3%, respectively. The metastatic rate was 8%. Fifteen enucleations were performed for tumor progression, 12 for

TABLE 11-2. Ocular melanoma (relevant series only). Authors

Location

No. of cases

Type of study

Dose/fractionation

Results

Munzenrider et al. [52]

Choroid

1488

P

70 CGE in 4 Fr

Choroid

2435

P

60–70 CGE in 4 Fr

Courdi et al. [54]

Choroid

538

P

57.2 CGE in 4 Fr

Lumbroso et al. [55]

Choroid

1062

P

60 CGE in 4 Fr

Damato et al. [56]

Choroid

349

P

58.4 CGE in 4 Fr

Damato et al. [57]

Iris

88

P

58.4 CGE in 4 Fr

Gragoudas et al. [58]

Choroid

Rand.

70 CGE vs. 50 CGE To decrease radiation-induced complication

Median follow-up 59 months Large tumors Relapse: 2.4% 5- and 10-year FMS: 68%, 60% Useful vision: 39% Small tumors Relapse: 0.5% 5- and 10-year FMS: 86%, 79% Useful vision: 65% Median follow-up 40 months 5- and 10-year LC: 95.8%, 94.8% 10-year OS: 72.6% Useful vision: 50% LC: 89% 5-year DFS: 86% Mets: 8% Useful vision: 50% Median follow-up 38 months LC: 97% 5-year DFS: 78% Mets: 15% Useful vision: 47% Median follow-up 37.2 months 5-year LC: 96.5 % 5- and 8-year FMS: 90%, 83.9% 5-year useful vision: 79.1% Median follow-up 32.4 months 4-year LC: 96.7 % 4-year cataract rate: 63% Dose reduction did not limit loss of visual acuity. Local tumor recurrence and mets death rates = both groups

Egger et al. [53]

188

DFS, disease-free survival rate; FMS, free-metastasis survival; Fr, fraction(s); LC, local control rate(s); OS, overall survival rate; P, prospective-retrospective series; Rand, randomized.

11.

proton beam radiosurgery

major complications, and 6 for both [54]. The Orsay group with 1062 patients and a median 32 months follow-up reported a 5-year local control and overall survival rate of 97% and 78%, respectively. Metastases were detected in 15%. Prognostic factors for survival were location of the tumor, diameter, thickness, and tumor volume [55]. Approximately 50% of patients, with a pretreatment useful vision, retained it after proton beam. A total of 7.9% of enucleations were performed for tumor relapse or for complications. All published series evidence neovascular glaucoma (45% of the enucleation) as the leading cause for enucleation when tumor is controlled [52–55]. In order to decrease the complication rate, a randomized trial comparing two radiation dose levels (i.e., 50 vs. 70 CGE) in ﬁve fractions was investigated in 188 MEEI patients. There was no difference in terms of local control, overall and disease-free survival, complication rate, and preservation of a useful vision. It has also been suggested to lower dose per fraction while keeping total dose constant in order to decrease the complication rate without affecting the excellent local control [58]. Recently, Liverpool-Clatterbridge published results of the ﬁrst large series of iris melanoma. Local 4-year control rate was high, 86.7%, and one of the most remarkable ﬁndings of this study was the minimal nature of the complications after treatment (mainly cataract), which was eminently treatable [57]. Because of the development of stereotactic fractionated irradiation, a comparative theoretical dosimetric study was performed at Paul Scherrer Institute. A ﬁxed proton horizontal beam and intensity-modulated spot scanning proton therapy (IMPT), with multiple non-coplanar beam arrangements, was compared with linear accelerator–based stereotactic radiotherapy using a static or a dynamic microleaf collimator and intensity-modulated radiotherapy. From imaging of a patient without eye melanoma (with a brain metastasis) were deﬁned several cases of tumor location. The results suggested that use of fractionated stereotactic radiotherapy compared with protons provided similar levels of dose conformation. Tumor dose inhomogeneity was always increased with photon planning. Furthermore, to dose all the contralateral organs at risk was completely eliminated with proton planning [59]. We will also mention ocular hemangiomas, a benign condition located at the posterior eye pole that can affect vision seriously, as a potential indication for low-dose proton beam [60]. Interestingly, recently Tsina et al. reported results of treatment of 76 eyes with ocular metastasis in 63 patients. Treatment delivered 28 CGE in two fractions. This treatment did not need implantation of ﬁducial markers. Proton beam irradiation is a useful therapeutic approach for uveal metastases; it allows retention of the globe, achieves a high probability of local tumor control, and helps to avoid pain and visual loss [61]. Proton beam has become the gold standard for primary ocular malignancies in adults. Its current indications cover posterior and equatorial sites and also large anterior ones (i.e., >5 mm thickness). Few indications remain for radioactive plaques (mainly small anterior tumors) and enucleation (massive ocular involvement) [62].

149

CNS Malignancies Base of the Skull and Upper Cervical Spine Chordomas and Chondrosarcomas Chordomas and low-grade chondrosarcomas of the skull base and upper cervical spine (see Table 11-3) [63–70] represent a paradigm for three-dimensional conformal proton therapy for many reasons: (1) both are radioresistant, that is, poorly controlled with the conventional doses administered in the literature. Surgery plus conventional radiotherapy (i.e., with doses below 60 Gy) yield a 5-year local control rate of 17% to 33% [71–73] and an overall survival rate of 33% to 90% [63, 74–77]. (2) Both are rarely totally resectable, because they are located in sites of difﬁcult surgical access: sphenoid body, petrous bone, clivus, spinal canal. This makes in turn escalated doses suitable. (3) Both lie along and even abut sensitive anatomic structures whose injury could be lethal or highly damaging such as optic nerves, chiasm, brain stem, and spinal cord. (4) Both are located in bony areas that are clearly visible on imaging (especially CT scan) and so easily delineated on three-dimensional simulation and accurately positioned for treatment using bony landmarks. The limited number of new cases per year was also compatible with the time-consuming process that was made necessary for patient setup. The Bragg peak with its sharp lateral and distal dose fall-off proves theoretically of major interest in such challenging situations, and this led to early clinical experiments in Boston in the late 1970s. The impact of dose escalation was clearly suggested through nonrandomized studies that ranged between 55.8 CGE and 83 CGE [23, 63–65, 78–82] delivered with protons (alone or in combination with photons). Local control was switched to 59% to 100%, according to various prognosticators: the strongest one was the pathologic type with chondrosarcomas faring better than chordomas: 59% to 78% for chordomas versus 78% to 100% for chondrosarcomas [23, 64, 65, 78, 80, 83]. Also of prognostic value (but less convincingly) were evidenced: tumor volume, local extension, and ﬁnally gender in chordomas alone (females having a less favorable outcome) [84]. However, this last prognostic factor remains unclear [85]. Interestingly, in the Boston historical series of 141 patients, on 26 local relapses (18%), only 6 of 26 were found after doses ≥70 CGE and 15 of 26 after doses below that level. The main reason for underdosage was the compliance with a dose-constraint due to the proximity of an organ at risk [83]. These ﬁndings led the Proton Radiation Oncology Group (PROG) to explore in a randomized fashion two dose-levels (PROG 25– 86): (1) 66.6 versus 72 CGE in the low-risk group (i.e., all chondrosarcomas and male chordomas), and (2) 72 versus 79 CGE in the high-risk group (i.e., all cervical sites and female chordomas). Results are still pending. In a similar population, Noel et al. showed the importance of quality of radiation, in terms of dose-uniformity within the GTV, a ﬁnding also related to underdosed areas [23]. Two other ﬁndings from our experience should also be mentioned: the delay between treatment and failure that rarely exceeds 3 years. The possibility in few cases is of regional failures that can take the form of tumor seeding within nodal drainage, and/or surgical route, which is highly typical of chordomas.

150

g. noel et al.

TABLE 11-3. Chordoma (CH) and chondrosarcoma (CS) of the base of the skull and upper cervical spine. Type of study

Dose/fractionation/aim

Results

68

P

69 CGE, 1.8 CGE, 80% protons

CH CS

58

P

71 CGE (65–79)

Munzenrider and Liebsch [65]

CH CS

621

P

67 CGE (66–83)

Noel et al. [22, 23, 66]

CH CS

67

P

67 CGE (60–70), 1/3 protons

Igaki et al. [67]

CH

13

P

72 CGE (63–95)

Weber et al. [68]

CH CS

29

P

68–74 CGE

Baumert et al. [69]

Brain

7

TP

Lomax et al. [70]

Miscellaneous

9

TP

Large range of total dose but same dose for photon and proton dosimetry. Study of the distribution of dose of three types of irradiation Ph, IMRT, Pr

Median follow-up 34 months 5-year LC: 82% 10-year LC: 58% Median follow-up 33 months CH 5-year LC: 59% CH5-year OS: 79% CS 5-year LC: 75% CS 5-year OS: 100% Median follow-up 41 months CH 5-, 10-year LC: 73%, 54% CH 5-, 10-year OS: 80%, 54% CS 5-, 10-year LC: 98%, 94% CS 5-, 10-year LC: 91%, 88% Median follow-up 29–31 months CH 4-year LC: 53.8% CH 5-year OS: 80.5% CS 3-year LC: 85% CS 4-year OS: 75% Median follow-up 69.3 months CH 5-year LC: 46% CH 5-year OS: 66;7% Median follow-up 29 months CH 3-year LC: 87.5% CH 3-year OS: 90% CS 3-year LC: 100% CS 3-year OS: 93.8% Mean CI Photons Protons 1.5 (1.15–2.03) 1.2 (1.05–1.38)

Authors

Tumors

Austin-Seymour et al. [78]

CH CS

Hug et al. [104]

No. of cases

healthy tissues irradiated with Pr vs. Ph/ IMRT. Tumor coverage with Pr vs. Ph and equal to IMRT.

CI, conformity index; IMRT, intensity-modulated radiotherapy (using photons); LC, local control rate; OS, overall survival rate; P, prospective-retrospective series; Ph, photons; Pr, protons; TP, theoretical publication.

Lower spinal/paraspinal conditions proved extraordinarily difﬁcult to manage using the ﬁxed beamlines only available at that time and are generally excluded from these studies [86]. Nonetheless, Hug et al. reported on a limited series of 20 patients an outcome close to the rest of the population: 56% and 100% 5-year local control in chordomas and chondrosarcomas, respectively. Total dose ranged from 55 to 82 CGE. Five of 16 chordomas and 0 of 4 chondrosarcomas failed. There was a tendency for improved local control as far as patients treated upfront with radiation rather than at the time of relapse, patients who underwent total/subtotal removal, and those who received a dose above 77 CGE [87]. Salvage therapy after proton therapy has been considered as purely symptomatic in the absence of demonstrated chemosensitivity and of possibility of reirradiation within 5 years. The introduction of recent biological agents (Glivec) could pave new avenues in these situations as well as less advanced presentations [88].

Meningiomas The indication for proton therapy is deﬁnitely more controversial in this tumor type. The main reason is that no clear-cut

dose-response relationship has been evidenced so far (Table 11-4) [89–94]. The initial study by Austin-Seymour et al. in 13 patients was a mixture of benign, malignant, and atypical variants. Median dose was 59.4 CGE (range, 54 to 71.6) and followup 26 months. Local control was 100% [90]. Gudjonsson et al. described results of a hypofractionated regime in 19 patients, delivering 26 CGE in 6 fractions. With a median follow-up of 36 months, local control was again 100% [89]. Miralbell et al. reported on 11 patients treated with photons and protons after incomplete surgery. With a median follow-up of 53 months, no relapse was observed against 6 of 25 patients with comparable tumors irradiated with photons alone [91]. Later on, the HCL published the results of 46 patients with benign meningiomas. Median dose delivered in tumor volume was 59 CGE. Median follow-up was 53 months. Five- and 10-year local control rates were 100% and 88%, respectively, and overall survival rates were 93% and 77%, respectively [24]. The same group published the results in 31 malignant or atypical meningiomas irradiated either conventionally with photons alone or with an escalated dose by a combination of photons and protons. Fiveyear local control rates were 17% and 80%, respectively [93]. Seventeen patients treated at the CPO with a combination of photons and protons to a slightly escalated dose were recently

11.

151

proton beam radiosurgery

TABLE 11-4. Meningiomas. Authors

Tumors

Gudjonsson et al. [89]

Benign, atypical, malignant meningiomas Benign, atypical, malignant meningiomas Benign, atypical, malignant meningiomas Benign, atypical, malignant meningiomas Benign meningiomas

Austin-Seymour et al. [90] Miralbell et al. [91]

Noel et al. [92]

Wenkel et al. [24]

Hug et al. [93]

Noel et al. [94]

No. of cases

Benign, atypical, malignant meningiomas Benign meningiomas

Type of study

Dose/fractionation/aim

Results

19

P

36 CGE in 4 Fr

Median follow-up 36 months No tumoral progression

13

P

59.4 CGE, 1.8 CGE/Fr

Follow-up ≥26 months No tumoral progression

11

P

Combination of photon-protons 54–69 CGE 1.8 CGE/Fr

Median follow-up 53 months No tumoral progression

17

P

Combination of photon-protons 61 CGE 1.8 CGE/Fr

46

P

59 CGE (53–74)

31

P

50–72 CGE

51

P

60.6 CGE Clinical improvement

Median follow-up 37 months 4-year LC: 82.4% 5-year OS: 88.9% Median follow-up 53 months 5- and 10-year LC: 100% and 88% 5- and 10-year OS: 93% and 77% 5-year LC: Photons: 17% Protons: 80% ocular symptoms: 68.8% other symptoms: 67%

Fr, fraction(s); LC, local control rate; OS, overall survival rate; P, prospective/retrospective.

analyzed. Patients were irradiated after surgery or at the time of relapse after surgery. Median follow-up was 37 months. Four-year local control and 5-year overall survival were 87.5% and 88.9%, respectively. The median disease-free interval was increased by 24 months for the patients treated after relapse, a ﬁgure similar to the disease-free interval between surgery and relapse [92, 95]. Our group also brought out the outcome of 51 patients with purely benign meningiomas. The main ﬁnding was the improvement in functional outcome (67% cases) and especially ocular preservation (68.8%) [94].

Gliomas The survival of patients diagnosed with a glioblastoma multiform is dismal (Table 11-5). It has been demonstrated that the majority of patients will fail locally (i.e., within 2 cm around the macroscopic tumor extension). The positive impact of dose-

escalation has been shown using brachytherapy or radiosurgical boosts with photons. Blomquist and Carlsson suggested a proton-based strategy in grades III to IV gliomas: (1) surgical removal of the bulky tumor, (2) high-precision, high-dose proton beam fractionated irradiation in a limited volume encompassing the area at risk plus minimal margin around [96]. The 90-Gy dose level has been investigated through multiple studies: Tatsuzaki et al. conducted dosimetric investigations showing that at this level using protons, none of the brain stem received 60 Gy/CGE compared with 5 cm3 with photons, and that the volume of “nontarget” brain receiving >70 CGE was almost doubled by photons (175 cm3 and 94 cm3, respectively). The reverse side of the coin was a superior dose to the skin delivered with protons alone (63 vs. 45 CGE) [25]. Baumert et al. compared the dose-distribution between modulated-intensity protons and photons using multileaf collimator in seven cases of brain tumors. The conformity index (CI) was better

TABLE 11-5. Brain gliomas. Authors

Tumors

Tatsuzaki et al. [25]

Glioblastoma

No. of cases

1

Type of study

Dose/fractionation/aim

Results

TP

60 Gy/CGE + 30 Gy/CGE (boost) Comparison of normal irradiated volume

20

Phase II

68,2 CGE, 1.8/Fr, grade II 79.7 CGE, 1.8/Fr, grade III Dose escalation

23

Phase II

90 CGE, 1.8/Fr, 2 Fr/day, at least 33% of the total dose with Pr Dose escalation

Photons Protons Normal brain 175 cm3 94 cm3 Brain stem 5 cm3 0 cm3 Max. dose chiasm 60 Gy 60 CGE Max. dose skin 45 Gy 63 CGE Grade III: median survival 29 Grade II not reached 5-year OS grade II: 71% 5-year OS grade III: 23% No increase in LC rate or time of relapse Median survival: 20 months 1-, 2-, 3-year OS: 78%, 34%, 18% Survival: 5–11 months compared with photons

Fitzek et al. [17]

Grade II–III gliomas

Fitzek et al. [16]

Glioblastoma

LC, local control rate; Fr, fraction(s); Pr, protons; OS, overall survival rate; TP, theoretical publication.

152

g. noel et al.

for protons than for photons for complex or concave lesions, or when the target volume was close to critical structures. In other cases, the CI was comparable for both modalities. The number of beams was lower with protons than with photons, which suggests a decrease in the integral dose to the brain [69]. Fitzek et al. reported clinical outcome of a phase II trial that enrolled 23 patients irradiated for a glioblastoma. This institutional study explored a dose of 90 CGE delivered predominately with protons and accelerated fractionation. The median survival time was 20 months, with 4 patients being alive 22 to 60 months after diagnosis. Analysis according to the RTOG prognostic criteria or MRC indices showed a 5- to 11-month increase in median survival time compared with patients treated conventionally in the literature. Radiation necrosis was evidenced in 7 of 29 patients, and survival was signiﬁcantly shortened in patients treated at the time of failure (p = 0.01). Tumor relapse occurred most commonly in areas that received doses >70 CGE; conversely, only one case failed in the 90-CGE volume. The authors concluded that future investigations should aim to cover more widely areas at risk to 90 CGE, although in practice this is rarely achievable due to the risk of radiation necrosis [16]. Another phase II trial was reported by the same group in 20 patients with less aggressive gliomas (grades II to III of the Daumas-Duport classiﬁcation). Prescribed dose were 68.2 CGE in grade II and 79.7 CGE in grade III. Five-year overall survival (OS) rates were 71% and 23%, respectively. The authors concluded that tumor recurrence was neither prevented nor noticeably delayed in these patients compared with the published series on photons [17].

Arteriovenous Malformations Single-dose proton therapy at MGH in the early 1960s gave the impetus to radiosurgical programs worldwide (Table 11-6). A single dose of 10 to 50 CGE was delivered according to the tumor size and radiobiological data sets as mentioned above. Obliteration rate was 20% and complication rate 3% [5]. In an updated series of 95 patients, Amin-Hanjani et al. reported a 17% (before irradiation) to 9% (after irradiation) reduction of the annual hemorrhage rate, but the complication rate was substantial: 26.5%, including 16.3% permanent neurologic deﬁcits and 3% death [27]. Seifert et al. reported a 16% obliteration rate for German patients treated with protons in the United States. They observed clinical improvement in 44%, stability in 27%, and worsening in 29% cases. An unexpected ﬁnding was that therapy was less effective (and so not recommended) for

lesions >3 cm [97], which contradicts dosimetric considerations that plead for the superiority of protons in lesions >4 cm [98]. Recently, Vernimmen et al. reported the experience of a South Africa proton center regarding 64 patients treated for predominately large intracranial AVMs. Irradiation was delivered according to a hypofractionated schedule and dose ranged between 18.4 and 22 single-fraction equivalent CGE. Obliterations were observed in 67% of the lesions with a volume inferior to 14 cm3 and 43% in those with volume superior to 14 cm3. Grade IV complications were reported in 3% of the patients [30].

Vestibular Schwannoma This is another typical case for monofractionated radiosurgery. Weber et al. reported the results of 88 patients with vestibular schwannomas that were treated at HCL with proton beam stereotactic radiosurgery. Two to four convergent ﬁxed beams of 160-MeV protons were applied. The median cross section and target volume were 16 mm and 1.4 cm3, respectively. Previous surgical resection had been made possible in 15 (17%) patients. Facial and trigeminal nerves functions were normal in 79 (89.8%) patients. Eight (9%) patients had good or excellent hearing, and 13 (15%) patients a useful hearing. A median dose of 12 CGE was prescribed to the 70% to 108% isodose lines. Median follow-up was 38.7 months. The actuarial 2- and 5-year tumor control rates were 95.3% and 93.6%, respectively. The actuarial 5-year cumulative radiologic reduction rate was 94.7%. Of the 21 patients (24%) with functional hearing, 7 retained “serviceable” hearing ability. Actuarial 5-year normal facial and trigeminal nerve function preservation rates were 91.1% and 89.4%, respectively. Univariate analysis revealed that prescribed dose (p = 0.005), maximum dose (p = 0.006), and the inhomogeneity index (p = 0.03) were associated with a signiﬁcant risk of long-term facial neuropathy [99, 100]. These results compare favorably with other published radiosurgical series. In summary, current knowledge on proton therapy in skull base and probably spinal canal low-grade sarcomas make proton therapy highly suitable in a majority of patients both in terms of oncologic outcome and quality of life. The place of surgical resection remains crucial because it can greatly improve ballistics to critical structures, and possibly by reducing tumor burden. Randomized studies (and possibly meta-analyses) might conﬁrm these ﬁndings as the number of patients submitted to this approach is expanding. The introduction of modern technologies especially isocentric gantries should make possible

TABLE 11-6. Arteriovenous malformations. Authors

Tumors

No. of cases

Type of study

Dose/fractionation/aim

Results

Kjellberg et al. [5]

AVM

75

P

10.5–50 CGE, 1 Fr

Amin-Hanjani et al. [27] Vernimmen et al. [30]

AVM

95

P

AVM

64

P

Median max. dose: 18.3 CGE (8–36.7), 1 Fr 18.38–22.05 SFEGyE

Complete occlusion rate: 20% Partial occlusion rate: 56% Death: 2 cases Reduction hemorrhage risk: 17% to 9% Complication: 26.6% Median follow-up: 62 months Obliteration rate for lesion >14 cm3: 67% Obliteration rate for lesion >14 cm3: 43% Complication: 3%

Fr, fraction(s); P, prospective; SFEGyE, single-fraction equivalent Gy equivalent.

11.

proton beam radiosurgery

the management of a variety of spinal/paraspinal malignancies in the future. PT remains controversial in meningiomas although it represents an elegant option for sparing normal tissues. In high-grade gliomas, PT has been disappointing despite substantial tumor dose increase. Future directions should explore combined chemoradiation and light ions.

Pediatric Tumors Despite the prominent role held by chemotherapy in this age group, radiotherapy still plays a crucial role in most solid tumors especially those located in the brain (approximately 20% of cases). There is a considerable amount of data indicating that radiotherapy even at low dose can induce severe sideeffects especially in young children (Table 11-7) [101–106]. For example in the CNS, it can affect cognitive function and pituitary-driven growth at >20 Gy in whole brain, sellar area, auditory function at >30 Gy in cochlea, and so forth. Another major concern comes from irradiation of bones in the prepubertal age, with impairment of growth plates at >15 Gy. Carcinogenicity has also represented a serious threat, especially in the usual combined chemo-radiation approaches. A salient feature of proton dose-distribution that can be of considerable interest in this age group is the absolute reduction of the integral dose (mainly due to the lack of “exit” beam) that minimizes areas receiving low and moderate doses around the target. On the other hand, there are practical serious limitations that have made this approach still almost conﬁdential in this indication, to mention a few: (1) The rarity of pediatric oncology centers (correlated with the rarity of the disease itself) that are restricted to a few places not necessarily close to a particle center. This comes along with the unique expertise required from radiation oncologists, physicists, and technologists involved. (2) The difﬁculty for performing extensive and uncomfortable daily setups without the help of deep sedation and even general anesthesia in the youngest patients. There is obviously a need for careful patient selection through preliminary dosimetric investigations in order for example to deﬁne the merits of different technological approaches like conformal photons or IMRT. In a comparative study between protons and conformal photons, in optical pathway gliomas, Fuss et al. showed that the CI (ratio of GTV to non-GTV encompassed in the 95% isodose) was better with protons than with photons. They also showed that doses delivered with protons in normal optic nerve, chiasm, pituitary gland, and temporal lobes were respectively reduced by 47%, 11%, 13%, and 39% compared with those delivered with photons [101]. For tumors located in the posterior fossa, Lin et al. showed that cochlea received 25% of the proton dose versus 75% of photons. Furthermore, 40% of temporal lobes were fully spared from protons, whereas 90% were exposed to photons to a minimum 30% of the total dose [31]. Based on theoretical models, Miralbell et al. reported a potential 10% drop in the predictable risk of IQ decline in medulloblastoma treated with CNS irradiation by the age of 4 years, using protons compared with photons. NTCP (normal tissue complication probabilities) values were also lowered by protons but to a modest extent when compared with highly conformal photons [107]. The same authors compared both techniques in cervical irradiation to 27 Gy at the age of 2 to 3

153

years. As far as cervical spine, the volume receiving ≥50% dose was ﬁvefold using photons versus of protons (i.e., 100% vs. 20%). As far as heart, the difference was even more striking (≥45% dose, photons; 100%, protons; 0% volume). Low doses to thyroid, liver, and gonads were similarly reduced. The authors’ estimates of “height sparing effect” by protons was of the order 10 cm [32]. It is interesting to mention that most authors acknowledge that the beneﬁt of protons increases with tumor size, a datum of special interest in brain malignancies [26, 70, 98]. Similarly, reduction of the carcinogenic risks is of primary concern in this age group and can be expected from protons (see above) [102]. The Swiss group developed mathematical algorithms based on radioprotection estimates to quantify these risks [108]. Carcinogenic risk between “rival” techniques (i.e., conformal photons/protons, and photons/ protons IMRT) were appraised: in parameningeal rhabdomyosarcomas, the “protective” effect of protons was ≥2-fold, and in medulloblastoma, 8- to 15-fold [109]. Elegant beam arrangements are made possible with protons as mentioned previously. Sophisticated patch techniques make sparing of abutting critical organs feasible. Impressive sparing of highly sensitive structures such as growing plates has been brought out in retinoblastomas by Krengli [102], orbital rhabdomyosarcomas, and lumbar neuroblastomas by Hug [104]. Recent theoretical study including cases of retinoblastoma, medulloblastoma, and pelvic sarcoma cases concluded that protons delivered superior target dose coverage and sparing of normal structure. In pelvic sarcoma studied in this series, none of the ovaries received dose superior to 2 Gy; furthermore, as expected, proton lowdose volume is greatly inferior to that obtained with IMRT. These volumes receiving low dose of irradiation have been suspected to be the site of radiation-induced secondary cancer [103]. Clinical series on the use of protons in children are still scarce. The Boston group reported on 18 children aged 4 to 18 years with a skull base chordoma/chondrosarcoma. With a median 72 months follow-up, 5-year OS and RFS were 68% and 63%, respectively. The complication rate was limited to one temporal lobe necrosis [64]. The preliminary Orsay experience was reported by Noel et al. on 17 children (median age, 12 years), with skull base sarcomas as the main indication, who received combined photon-proton irradiation (approximately 50%–50% of the dose). With a median 27 months follow-up (range, 3 to 81), 3-year local control was 91.7% and 1-, 2-, and 3-year OS 93.3%, 83%, and 83%, respectively. One child failed in-ﬁeld and one at the margin. No late side-effect was reported although follow-up is deﬁnitely short [18, 106]. McAllister et al. at Loma Linda University Medical Center (LLUMC) reported on 28 children treated for grades 2 to 4 glioma. At the time of analysis, three were dead from disease, one was alive with tumor progression, and the others were with NED. Complication rate was low [105]. Hug et al. reported on 27 children treated by protons for low-grade malignancies. Six relapsed, four died, and the others were with NED and complicationfree. A subgroup of six children with optic glioma had visual preservation or improvement [21]. Giant cell tumors seen in the pediatric age are regarded as benign conditions, although they can behave aggressively. In this situation, high-dose proton therapy is warranted (just as in the adult sarcomas) and can induce prolonged remissions [110].

154

g. noel et al.

TABLE 11-7. Pediatric cancers. No. of cases

Type of study

Authors

Tumors

Dose/fractionation/aim

Results

Lin et al. [31]

Posterior fossa

9

TP

54 Gy/CGE Comparison 3D Ph and Pr

Fuss et al. [101]

Optic glioma

7

TP

50.4–50 Gy/CGE Dose distribution comparison for Pr, 3D conformal Ph, standard pH (std)

Krengli et al. [102]

Retinoblastoma

1

TP

Lee et al. [103]

Retinoblastoma

3

TP

46 CGE in GTV, 40 CGE in CTV Proton beam arrangements for various intraocular tumor locations Comparison of different technique of Ph including IMRT and proton

% prescribed Pr PH dose in cochlea 25% 75% Pr: 40% temporal lobes excluded irradiated volume PH: 90% temporal lobes received ≥31% of prescribed dose CI 3D Ph Std Ph Pr 2.9 7.3 2.3 Reduction of dose/Pr 3D Ph Std Ph Contralateral optic 47% 77% nerve 11% 16% Chiasm 13% 16% Pituitary 39% 54% Temporal lobes PT risks K2 + cosmetic / functional sequelae

Mean % volume/ technique 5 Gy orbit 20 Gy optic nerve 25 Gy cochlea 10 Gy pituitary 10 Gy thyroid 10 Gy lung 10 Gy kidney 15 Gy heart 5 Gy ovary 20 Gy vertebra 30 Gy bowel

3D RT 25 53 64 91 24 15 18 2 100 20 11

IMRT 69 55 33 81 100 14 15 59 29 29 12

Medulloblastoma

3

Pelvic sarcoma

3

Hug et al. [104]

Neuroblastoma

1

P

34.2 CGE Dose distribution

50% ipsilateral kidney <16 CGE 50% contralateral kidney <1 CGE 50% liver <2.6 CGE 20% liver ≥10 CGE Spinal cord <3 CGE

Benk et al. [64]

CH-CS

18

P

Follow-up ≥72 months 5-year LC: 78% 5-year OS: 68%

Krengli et al. [102]

Various ocular tumor

69 CGE, 1.8 CGE /Fr, 80% protons Combination photons and protons (60–80 CGE) 46 CGE in GTV, 40 CGE in CTV

Hug et al. [21]

Low-grade glioma

27

P

Mean dose 55.2 CGE (50.4–63) 1.8 CGE/Fr

McAllister et al. [105]

Miscellaneous

28

P

Noel et al. [18, 106]

Miscellaneous

17

P

Pr only: median: 54 CGE (40–70) Mixed Pr and pH: Ph: median 36 Gy (18–45) Pr: median 18 CGE (13–32) Mixed Pr and pH: Ph: median 40 Gy (24–54) Pr: median 20 CGE (9–31)

Protons 10 29 6 21 7 2 2 0 0 9 5

Pr: K2 + cosmetic outcome and functional sequelae Median follow-up 39 months OS: 85% LC: 78% 4 hypopituitarisms 1 asymptomatic brain necrosis 4 relapses Complications: 2 seizures 1 hormonal deﬁcit 1 cataract Median follow-up: 27 months (3–81) 1 in site + 1 marginal relapse 3-year CL: 91.7% 1-, 2-, 3-year OS: 93.3%, 83%, 83%

LC, local control rate; Fr, fraction(s); CI, conformity index; CTV, clinical target volume; GTV, gross tumor volume; K2, secondary cancer; Ph, photons; Pr, protons; OS, overall survival rate; std, standard; P, prospective/retrospective series; TP, theoretical publication.

11.

155

proton beam radiosurgery

TABLE 11-8. Pelvic cancers. Authors

Tumors

Type of study

Dose/fractionation/aim

Results

Isacsson et al. [42]

Inoperable rectum

No. of cases

6

TP

46 Gy/CGE Photons 4 b; protons 3 b Photons 4 b + protons 3 b

Tatsuzaki et al. [43]

Rectum

1

TP

50 Gy/CGE ± 5 Gy/CGE (boost) Comparison photons 2, 3, 4 b Protons 1 b, conformal or not

Cella et al. [111]

Prostate

1

TP

81 CGE • Ph 6 MeV 18 conformal (6 b) • Pr 214 MeV (2 b) • IMRT 15 MeV (5 b) • IMPT 177–200 MeV (5 b)

Schulte et al. [41, 112]

Localized prostate stage T1-T2b

911

P155

Pr alone or combining with Ph 74–75 CGE

Shipley et al. [113]

Prostate, T3–4, Nx, N0–2

202

Rand.

Ph: 50.4 Gy + 16.8 Gy Pr: 50.4 Gy + 25.2 CGE, 1.8 Gy/CGE/Fr Comparison dose escalation

Reduction of the irradiated digestive tract, femoral head and of bladder volume with proton, increased TCP from 8 to 14% if NTCP = 5% Proton treatment allows higher doses to be delivered to the tumor, with a probable increase in TCP, or a reduction in NTCP • Better homogeneity with proton • Low or mean dose volume decreased with proton • Improvement of NTCP of the rectum (<5%, grade 3) IMPT or IMRT 5-year biological survival: 82% No grade 3–4 complication Grade 2 rectal complication rate: 3.5% Grade 2 bladder complication rate: 5.4%. Median follow-up: 61 months Improvement of local control for the most undifferentiated tumors with Pr

b, beams; Fr, fraction(s); IMPT, intensity-modulated protontherapy; IMRT, intensity-modulated radiotherapy; LC, local control rate; NTCP, normal tissue complication probability; OS, overall survival rate; Ph, photons; Pr, protons; Rand., randomized; TCP, tumor control probability; TP, theoretical publication.

In summary, pediatric conditions represent a major challenge to radiation oncologists, mainly due to the risk for serious long-term side-effects. This challenge could possibly be overcome by a more systematic use of proton therapy especially in young children as suggested by preliminary studies. The pediatric community is certainly waiting anxiously for deﬁnitive clinical evaluations.

Pelvic Cancers Protons are attractive in this location because there is a number of relatively sensitive structures such as bladder, rectum, and femoral heads to be spared. On the other hand, tumors are generally located deeply and require beam energy >160 MeV, which is not widely available. (Table 11-8) [111–113]. If so, protons can be offered with good chances to achieve improved conformation, tumor homogeneity, and decreased integral dose as evidenced by the dosimetric intercomparisons by Cella et al. To a dose of 81 Gy, only intensity-modulation photons and protons could comply with acute rectal toxicity [111]. It has also been estimated that doses could be escalated by 20% this way, with parallel normal organs volume reduction up to 60%, if an isocentric gantry was available [44–46]. In case of mixed beams, Tatsuzaki et al. evidenced a possible relationship between beneﬁt and proportion of photons and protons [43]. Similarly in inoperable rectal carcinoma, Isacsson et al. estimated that tumor control probability was increased by 8% with mixed photon-protons and 14% with protons alone, for a ﬁxed 5% complication risk [42]. Most clinical studies have dealt with prostate carcinomas managed both at late and early stages: in Boston, between 1982

and 1992, T3-T4, Nx, N0-2, and M0 were elected to receive 50.4 Gy with photons in a box technique followed in a random fashion by a conformal boost of 25.2 CGE with protons or 16.8 Gy with photons. Total dose was slightly superior using protons rather than photons (75.6 CGE vs. 67.2 Gy). Among 202 patients, 8-year local control after proton and photon treatments were 77% and 60%, respectively (p = 0.089) with similar OS and DFS. A subgroup analysis in 57 poorly differentiated (Gleason 4 or 5) tumors evidenced an 8-year local control rate of 84% with protons versus 19% with photons (p = 0.0014). Grades 1 to 2 postirradiation rectal bleeding (i.e., not requiring surgery or hospitalization), and correlated with telangiectasia, affected 34% and 16% patients treated with protons and photons, respectively (p = 0.013) [113, 114]. This increased toxicity in the proton arm was related to a superior total dose but also with a relatively simple technique for boosting based on a single perineal ﬁeld. Oppositely, LLUMC explored early stages. On 911 patients managed with protons between 1991 and 1996, 5-year “biological” (i.e., normal PSA) local control was 85% [41, 112]. In summary, there is a considerable potential for pelvic tumor sites (and related ones like sacral malignancies) for centers with the most advanced equipment. The place of proton therapy in prostate carcinoma remains highly controversial especially in early stages, until comparative studies with alternative techniques (Brachytherapy, surgery, etc.) are contemplated.

Head and Neck Tumors There is again a substantial body of data based on dosimetric intercomparisons but few clinical reports. (Table 11-9) [115, 116]. In a comparative study testing current photon and proton

156

g. noel et al.

TABLE 11-9. Head and neck tumors. Authors

Tumors

Type of study

Dose/fractionation/aim

Results

Miralbell et al. [28]

Cancer of the maxillary sinus

1

TP

64 Gy/CGE • Classic Ph • Conformal 3D Ph • Ph-Pr

Slater et al. [34]

Cancer tonsillar T2 and T3

2

TP

T2: 70 Gy/CGE + 5 Gy/CGE boost; Ph 2 b; Pr 1 b T3: 70 Gy + 5 Gy boost; Ph 2 b; 70 CGE + 15 CGE boost Pr 1 b

• homogeneity with mixed and 3D technique • Dose escalation in tumoral volume without dose increase in organs risk T2: dose in contralateral parotid gland + mandible T3 same result with conformity

Slater et al. [33]

Oropharyngeal carcinoma stages II–IV Advanced head and neck cancer Orbital or periorbital tumor

29

P

50.4 Gy Ph 28 Fr of 1.8 Gy + 25.5 CGE Pr as boost in 17 Fr

6

TP

4

TP

54 Gy/CGE PH-electrons; photons 3D: IMRT Passive Pr or spot scanning 30–50.4 Gy/CGE Modulated intensity with photons or protons

Brown et al. [35] Noel et al. [37]

Nasopharyngeal carcinoma Nasopharyngeal carcinoma

2

TP

70 Gy/CGE; 3D Ph and Pr

5

TP

Lin et al. [36]

Relapsed nasopharyngeal carcinoma

70 Gy/CGE; 3D Ph; Ph + Pr boost Dose distribution in tumoral volume and organs at risk 62.8 CGE (59.4–70.2); 1.8 to 2 CGE/Fr

Cozzi et al. [115] Miralbell et al. [116]

No. of cases

12

P

Median follow-up: 28 months 2- and 5-year LC: 96% and 88% 2- and 5-year DFS: 81% and 65% • homogeneity with protons • dose spinal cord and parotid with protons NTCP (mean) Pr Ph Ipsilateral lens 0.30% 1.63% Ipsilateral choroid 1.03% 1.10% Ipsilateral parotid 0% 0.10% Contralateral lens 0.30% 1.40% Mean dose 5 Gy in the tumor with protons and organs at risk CI: 3.1 for mixed Ph-Pr and 5.7 for Ph only 78% of the 68 criteria studied in favor of Ph-Pr combination Median follow-up duration: 23.7 months 2-year LC: 50% 2-year DFS: 50% 2-year OS: 50%

b, beams; Fr, fraction(s); CI, conformity index; DFS, disease-free survival; IMRT, intensity-modulated radiotherapy; LC, local control rate; CL, rate of local control; NTCP, normal tissue complication probability; OS, overall survival; Ph, photons; Pr, protons; TCP, tumor control probability; TP, theoretical publication.

techniques (including passive diffusion, raster and spot scanning; see above) on ﬁve different treatment plans by Cozzi et al., main organs (i.e., spinal cord and parotid gland) shielding was always superior using protons. Dose homogeneity in the target was also improved but to a modest extent in comparison with conformal photons [117]. Miralbell et al. made a similar evaluation in maxillary sinus with similar conclusions but pointed out the difﬁculty evaluating the actual proton dose within air cavities [28]. There is also a study by Slater et al. in tonsillar primary, evidencing the potential improved sparing of salivary glands and mandible [34]. Recently, the same group presented results about 29 patients treated with accelerated irradiation including proton boost for an oropharyngeal tumor. Two- and 5-year disease-free survival were 81% and 65%, respectively. The dose increase, up to 75.9 CGE, was particularly well tolerated with only three cases of grade 3 complication [33]. Nasopharyngeal carcinoma represents a paradigm as there is evidence in the literature on the positive role of doseescalation. Brown et al. [35] exempliﬁed two cases. As far as clinical studies, protons have been investigated in recurrent nasopharyngeal carcinomas. It has been acknowledged since the initial reports by C.C.Wang and others that reirradiation was feasible but at the price of increased toxicity in extensive failures at the skull base. Lin et al. reported on 16 such cases. At initiation of therapy, 20 patients had symptoms consistent with intracranial involvement. Median dose at reirradiation was

62.8 CGE (range, 59.4 to 70.2), and cumulated with the previous one, 134.6 CGE (range, 110 to 148). With a median 23.7 months follow-up (range, 4 to 47), 2-year local/regional DFS and OS were both 50%. The authors pointed out that chances for prolonged remission were correlated with target dose homogeneity. Despite the high cumulative dose, only two severe complications were observed, due to a good sparing of the organs at risk the second time [36]. In summary, proton therapy has not gained wide acceptance in head and neck carcinomas for multiple reasons, such as the current interest for alternative techniques that are more readily accessible (especially photon IMRT), the difﬁculty of modeling with current biophysical algorithms, and considerable tissue heterogeneities, especially in the paranasal area. These limitations could be overcome in the future with expected technological and biophysical improvements. It will certainly remain debatable to put protons forward for an entire treatment program, including lymphatic drainage coverage.

Bronchial Cancer and Esophagus Bush et al. explored stages I to III lung cancers in 37 patients (Table 11-10). The group was divided in two according to whether patients had good or poor cardiopulmonary function. The ﬁrst group received 73.8 Gy by combined photons and

11.

157

proton beam radiosurgery

TABLE 11-10. Thoracic and esophagus cancers. Authors

Tumors

Type of study

Dose/fractionation/aim

Results

Bush et al. [48]

Stage I and III lung cancer

37

P

Tsunemoto et al. [39]

Stage T1–4 esophagus cancer Stage T1–4 esophagus cancer

16

P

Good cardiac function Ph 45 Gy + Pr 28.8 CGE; 1.8 Gy/CGE/Fr Bad cardiac function Pr 51 CGE, 5.1 CGE/Fr 80–88 CGE

46

P

Median duration of 14-month monitoring 2-year LC: 87% 2-year DFS: 63% 2-year OS: 31% Complete response in 13 patients LC T1: 83% T2–4: 29% OS Overall: 34% T1: 55%, T2–4: 13% DFSS Overall: 67% T1: 95% T2–4: 33%

Sugahara et al. [38]

No. of cases

Ph-Pr combination* Median dose Ph: 48 Gy Median dose Ph: 31.7 CGE Total dose range: 69.1–87.4 CGE Pr only Total dose range: 75–89.5 Gy

DFS, disease-free survival rate; DFSS, disease-free speciﬁc survival rate; Fr, fraction(s); LC, local control rate; OS, overall survival rate; Ph, photons; Pr, protons; P, prospective. *RBE = 1 in this study.

protons (45 Gy + 28.8 CGE) to the mediastinum and GTV + cone down to the primary only. The second group received exclusive protons to the GTV only, up to 51 CGE with accelerated regime (10 weekly fractions over 2 weeks). With a median follow-up of 14 months, local control was 87%. Two-year OS and DFS were 63% and 31%, respectively [48]. These results compared favorably with those of photons in the literature. Tsunemoto et al. managed 16 esophageal carcinomas with doses escalated up to 88 CGE. Thirteen enjoyed complete response [39]. Saguhara reported on 46 cases treated with mixed beams (40) or protons alone (6). Respective median doses were 76 and 82 CGE. Eighteen patients failed either locoregionally (16) or distantly (2). Five-year local control was 83% in T1 and 29% in T2-4. Five-year OS and DFS in T1-T2-T4 were 34%, 55%, 13% and 67%, 95%, 33%, respectively. These results suggest that proton beam therapy can be a reasonable option at least in early presentations [38]. In summary, no ﬁrm conclusion can be drawn from such limited and heterogeneous series, and comparative studies with alternative therapies are missing.

Toxicity Despite the relatively few long-term side-effects reported, a precious database has been accumulated especially by the Boston group with remarkable accuracy on dose, volume, and risk estimates. It provides highly valuable pieces of information on tolerance of normal organs to radiation (especially in the brain) that might expand our knowledge beyond the “particles” community and feed, for example, NTCP models in threedimensional conformal therapy. As far as temporal lobes, Santoni et al. conducted a study in 96 chordoma/chondrosarcoma patients. The 2- and 5-year complication rates (i.e., clinical and/or radiologic symptoms of radionecrosis) were 8% and 13%, respectively, after doses of

66 to 72 CGE. Surprisingly, male gender showed up as an independent risk factor [117] (we mentioned above the adverse role of female gender in failures in the same population). Glosser et al. focused on neuropsychological outcome in 17 patients who received radiation up to 66 CGE for skull base malignancies. They did not observe any early or late cognitive side-effect related to temporal lobe necrosis among NED patients; however half had psychomotor speed impairments and few developed minor transient symptoms like depression or anxiety [118]. Brain-stem tolerance was studied by Debus et al. in the same population. They reported 17 of 367 (4.5%) injuries. The 10-year complication rate was 12% and the mean free interval 10 months (90% within 3 years). Risk of injury was correlated with the number of previous surgical procedures, dose >60 CGE, and association with diabetes mellitus [119]. In benign meningiomas, Wenkel et al. found 1 of 46 (2%) BS necrosis, which is consistent with the lower dose administered (i.e., 59 CGE) [24]. The authors mentioned noncompliance to recommended doseconstraints as a major risk factor. Spinal cord tolerance was studied by Marucci et al. Thirteen minor and four severe injuries out of 85 patients with cervical spine malignancies were found (15% and 4.4%, respectively). There was no dose-effect relationship in the range explored (≤55 to 58 CGE cord center and ≤67 to 70 CGE cord surface). The only predictor of toxicity was the number of previous surgeries [120]. Radiation-related hypothalamic-pituitary endocrine damage was reported in the Boston series by Munzenrider et al. in 40% of cases [65]. From Slater et al., patients deteriorate between 14 and 45 months postirradiation and the risk is dosedependent: 50% at the 67.6 CGE level [121]. In the Orsay population, risk of symptomatic toxicity has been set to 8 of 64 (12.5%) but with a somewhat shorter follow-up [92]. Neurovisual impairments represent undoubtedly a major concern in escalated-dose skull base protocols (along with BS toxicity). This is due to their dramatic impact on quality of life

158

g. noel et al.

for patients with generally extended life expectancy. It strongly inﬂuences the approach of our group in the management of skull base sarcomas, as patients with tumor abutting the visual pathway are generally offered a limited reoperation in order to remove this extension by at least 1 or 2 mm. Given the sharp proton lateral penumbra (15%/mm), this means a 5- to 10-CGE drop toward the critical structure. Risk estimate has been appraised by Habrand et al. to ∼10% at 55 CGE and ≥20% at >60 CGE [122]. In Wenkel’s series of meningiomas, 4 of 46 (9%) experienced neurovisual symptoms [24]. Kim et al. demonstrated that the “patch” technique (see above) could be at risk just as usual photon beam matching if ﬁeld abutments concerned nerves or chiasm [123]. Several series point out the predisposing role of underlying vasculopathies related to diabetes mellitus or HBP [92, 122]. Permanent hearing loss is a symptom frequently related to cochlear injury [65, 93, 110, 124]. Schoenthaler et al. showed that in two of three affected patients, dose had been >62.7 CGE [110]. The rest of cranial nerves has been found relatively “radioresistant” as evidenced by Munzenrider and Liebsch [65]. In the detailed Urie et al. analysis in 27 patients, based on anatomic nerve reconstructions in three dimensions, 17 symptomatic injuries (in 5 patients) of 594 structures were found (3%), with a mean 74 months interval. Using logistic regression, they estimated the risk for “common” cranial nerves to 1% at 60 CGE and 5% at 70 CGE [124]. Using the model based on radioprotection data mentioned above, Schneider et al. came up with a twofold decrease in the carcinogenic risk for adults irradiated for Hodgkin disease when protons replaced photons [109]. Carcinogenicity in the pediatric age has been discussed above. All previous data concern adult patients treated with conventional fractionation. We are still missing age-based information in children. In summary, a remarkably low complication rate has been reported despite the considerable dose generally administered. This makes proton therapy highly attractive in most tumor sites, especially in children.

Socioeconomic Aspects Initiating a proton program is not as painful as expected when a “physics machine” becomes available for therapy. For example, the Orsay project has been conducted so far on an existing synchrocyclotron, kindly offered by the French Scientiﬁc Research Authority (CNRS). Adaptations for two treatment rooms were approximately *600,000, and yearly running cost (including beam’s hour, and staff salaries) approximately *2 million. With approximately 250 ocular and 50 intracranial malignancies managed per year, patients have been charged approximately *1000 per session, a ﬁgure still 10-fold that of conventional radiation (but less compared with sophisticated three-dimensional photons). The treatment cost has been actually substantially reduced since generally one-half to two-thirds of treatment is performed using conformal photons. Fully operational new equipment is more costly by 2 orders of magnitude. Goitein et al. have simulated the relative cost of state-of-the-art proton and photon facilities (i.e., with isocentric gantries and

intensity modulation capabilities) [8]. The following costs were attributed to protons: two-gantry facility, *62.5 million; and operating cost per fraction, *1025. As far as photons, they were respectively *16.8 million and *425. The authors estimated that relative treatment costs would be of the order 2.4 initially and down to 1.7 in the long run. Cost-beneﬁt evaluations are also emerging that suggest a ﬁnancial beneﬁt in using protons: this has been shown in pediatric medulloblastoma, when rehabilitation of children disabled by photon irradiation was taken into account [125]. New technologies that aim to produce protons at a lower cost are also in progress, like particle beams generated by ultraintense laser pulses [126].

Conclusion There is a respectable body of evidence showing the ballistic superiority of protons over conformal photons currently. This might be conﬁrmed in future investigations comparing intensity modulation with both particles. From a clinical standpoint, current successes in intracranial and ocular malignancies do not preclude the need for future controlled trials. In this context, the place of new particles, especially light ions, is emerging and should also be deﬁned.

References 1. Orecchia R, Zurlo A, Loasses A, et al. Particle beam therapy (hadron therapy): basis for interest and clinical experience. Eur J Cancer 1998; 34:459–468. 2. Bragg WH, Kleeman R. On the alpha particles of radium, and their loss of range in passing through various atoms and molecules. Philos Mag 1905; S6:318–340. 3. Wilson RR. Radiological use of fast proton. Radiology 1946; 47: 487–491. 4. Wilson R. A Brief History of the Harvard University Cyclotrons. Cambridge, MA: Harvard University Department of Physics, 2004. 5. Kjellberg RN, Hanamura T, Davis KR, et al. Bragg-peak protonbeam therapy for arteriovenous malformations of the brain. N Engl J Med 1983; 309:269–274. 6. Urano M, Verhey LJ, Goitein M, et al. Relative biological effectiveness of modulated proton beams in various murine tissues. Int J Radiat Oncol Biol Phys 1984; 10:509–514. 7. Brown A, Suit H. The centery of the discovery of the Bragg peak. Int J Radiat Oncol Biol Phys 2004; 73:268. 8. Goitein G, Lomax A, Pedroni ES. Treating cancer with protons. Physics Today 2002; 45–55. 9. Moyers MF, Lesyna W. Isocenter characteristics of an external ring proton gantry. Int J Radiat Oncol Biol Phys 2004; 60:1622– 1630. 10. Tourovsky A, Lomax AJ, Schneider U, Pedroni E. Monte Carlo dose calculations for spot scanned proton therapy. Phys Med.Biol 2005; 50:971–981. 11. Chu WT, Ludewigt BA, Renner TR. Instrumentation for treatment of cancer using proton and light-ion beams. Rev Sci Instrum 1993; 64:2055–2122. 12. Paganetti H, Niemierko A, Ancukiewicz M, et al. Relative biological effectiveness (RBE) values for proton beam therapy. Int J Radiat Oncol Biol Phys 2002; 53:407–421. 13. Molokanov AG, Spurny F. Biological weighted effective dose in 205 MeV clinical proton beam. Phys Med Biol 2005; 50:281– 287.

11.

proton beam radiosurgery

14. Delacroix S, Bridier A, Mazal A, et al. Proton dosimetry comparison involving ionometry and calorimetry. Int J Radiat Oncol Biol Phys 1997; 37:711–718. 15. Archambeau JO, Slater JD, Slater JM, Tangeman R. Role for proton beam irradiation in treatment of pediatric CNS malignancies. Int J Radiat Oncol Biol Phys 1992; 22:287–294. 16. Fitzek MM, Thornton AF, Rabinov JD, et al. Accelerated fractionated proton/photon irradiation to 90 cobalt gray equivalent for glioblastoma multiforme: results of a phase II prospective trial. J Neurosurg 1999; 91:251–260. 17. Fitzek MM, Thornton AF, Harsh G, et al. Dose-escalation with proton/photon irradiation for Daumas-Duport lower-grade glioma: results of an institutional phase I/II trial. Int J Radiat Oncol Biol Phys 2001; 51:131–137. 18. Habrand JL, Mammar H, Ferrand R, et al. Proton beam therapy (PT) in the management of CNS tumors in childhood. Strahlenther Onkol 1999; 175(Suppl 2):91–94. 19. Hug EB, Slater JD. Proton radiation therapy for pediatric malignancies: status report. Strahlenther Onkol 1999; 175(Suppl 2): 89–91. 20. Hug EB, Muenter MW, Adams JA, et al. 3-D-conformal radiation therapy for pediatric giant cell tumors of the skull base. Strahlenther Onkol 2002; 178:239–244. 21. Hug EB, Muenter MW, Archambeau JO, et al. Conformal proton radiation therapy for pediatric low-grade astrocytomas. Strahlenther Onkol 2002; 178:10–17. 22. Noel G, Habrand IL, Jauffret, de Crevoisier. Radiation therapy for chordoma and chondrosarcoma of the base of the skull and the cervical spine: prognostic factors and patterns of failure. Strahlenther Onkol 2003; 179:241–248. 23. Noel G, Feuvret L, Ferrand R, et al. Radiotherapeutic factors in the management of cervical-basal chordomas and chondrosarcomas. Neurosurgery 2004; 55:1252–1260. 24. Wenkel E, Thornton AF, Finkelstein D, et al. Benign meningioma: partially resected, biopsied, and recurrent intracranial tumors treated with combined proton and photon radiotherapy. Int J Radiat Oncol Biol Phys 2000; 48:1363–1370. 25. Tatsuzaki H, Urie MM, Linggood R. Comparative treatment planning: proton vs. x-ray beams against glioblastoma multiforme. Int J Radiat Oncol Biol Phys 1992; 22:265–273. 26. Wambersie A, Gregroire V, Brucher JM. Potential clinical gain of proton (and heavy ion) beams for brain tumors in children. Int J Radiat Oncol Biol Phys 1992; 22:275–286. 27. Amin-Hanjani S, Ogilvy CS, Candia GJ, et al. Stereotactic radiosurgery for cavernous malformations: Kjellberg’s experience with proton beam therapy in 98 cases at the Harvard Cyclotron. Neurosurgery 1998; 42:1229–1236. 28. Miralbell R, Crowell C, Suit HD. Potential improvement of three dimension treatment planning and proton therapy in the outcome of maxillary sinus cancer. Int J Radiat Oncol Biol Phys 1992; 22: 305–310. 29. Steinberg GK, Fabrikant JI, Marks MP, et al. Stereotactic heavycharged-particle Bragg-peak radiation for intracranial arteriovenous malformations. N Engl J Med 1990; 323:96–101. 30. Vernimmen FJ, Slabbert JP, Wilson JA, et al. Stereotactic proton beam therapy for intracranial arteriovenous malformations. Int J Radiat Oncol Biol Phys 2005; 62:44–52. 31. Lin R, Hug EB, Schaefer RA, et al. Conformal proton radiation therapy of the posterior fossa: a study comparing protons with three-dimensional planned photons in limiting dose to auditory structures. Int J Radiat Oncol Biol Phys 2000; 48:1219– 1226. 32. Miralbell R, Lomax A, Russo M. Potential role of proton therapy in the treatment of pediatric medulloblastoma/primitive neuroectodermal tumors: spinal theca irradiation. Int J Radiat Oncol Biol Phys 1997; 38:805–811.

159

33. Slater JD, Yonemoto LT, Mantik DW, et al. Proton radiation for treatment of cancer of the oropharynx: Early experience at Loma Linda University Medical Center using a concomitant boost technique. Int J Radiat Oncol Biol Phys 2005; 62:494–500. 34. Slater JM, Slater JD, Archambeau JO. Carcinoma of the tonsillar region: potential for use of proton beam therapy. Int J Radiat Oncol Biol Phys 1992; 22:311–319. 35. Brown AP, Urie MM, Chisin R, Suit HD. Proton therapy for carcinoma of the nasopharynx: a study in comparative treatment planning. Int J Radiat Oncol Biol Phys 1989; 16:1607–1614. 36. Lin R, Slater JD, Yonemoto LT, et al. Nasopharyngeal carcinoma: repeat treatment with conformal proton therapy–dose-volume histogram analysis. Radiology 1999; 213:489–494. 37. Noel G, Boisserie G, Dessard-Diana B, et al. Comparaison par les histogrammes dose-volume de deux techniques d’irradiation conformationnelle des cancers du nasopharynx classés T4N0M0: l’une par photons et protons et l’autre par photons seuls. Cancer Radiother 2002; 6:337–348. 38. Sugahara S, Tokuuye K, Okumura T, et al. Clinical results of proton beam therapy for cancer of the esophagus. Int J Radiat Oncol Biol Phys 2005; 61:76–84. 39. Tsunemoto H, Ishikawa T, Morita S, et al. Indications of particle radiation therapy in the treatment of carcinoma of the esophagus. Int J Radiat Oncol Biol Phys 1992; 22:321–324. 40. Sandler HM, McShan DL, Lichter AS. Potential improvement in the results of irradiation for prostate carcinoma using improved dose distribution. Int J Radiat Oncol Biol Phys 1992; 22:361– 367. 41. Schulte RW, Slater JD, Rossi CJ Jr, Slater JM. Value and perspectives of proton radiation therapy for limited stage prostate cancer. Strahlenther Onkol 2000; 176:3–8. 42. Isacsson U, Montelius A, Jung B, Glimelius B. Comparative treatment planning between proton and X-ray therapy in locally advanced rectal cancer. Radiother Oncol 1996; 41:263–272. 43. Tatsuzaki H, Urie MM, Willett CG. 3-D comparative study of proton vs. x-ray radiation therapy for rectal cancer. Int J Radiat Oncol Biol Phys 1992; 22:369–374. 44. Levin CV. Potential for gain in the use of proton beam boost to the para-aortic lymph nodes in carcinoma of the cervix. Int J Radiat Oncol Biol Phys 1992; 22:355–359. 45. Slater JD, Slater JM, Wahlen S. The potential for proton beam therapy in locally advanced carcinoma of the cervix. Int J Radiat Oncol Biol Phys 1992; 22:343–347. 46. Smit BM. Prospects for proton therapy in carcinoma of the cervix. Int J Radiat Oncol Biol Phys 1992; 22:349–353. 47. Gademann G, Wannenmacher M. Charged particle therapy to pediatric tumors of the retroperitoneal region: a possible indication. Int J Radiat Oncol Biol Phys 1992; 22:375–381. 48. Bush DA, Slater JD, Bonnet R, et al. Proton-beam radiotherapy for early-stage lung cancer. Chest 1999; 116:1313–1319. 49. Kawashima M, Furuse J, Nishio T, et al. Phase II study of radiotherapy employing proton beam for hepatocellular carcinoma. J Clin Oncol 2005; 23:1839–1846. 50. Gragoudas ES, Marie LA. Uveal melanoma: proton beam irradiation. Ophthalmol Clin North Am 2005; 18:111–118. 51. Seddon JM, Gragoudas ES, Albert DM, et al. Comparison of survival rates for patients with uveal melanoma after treatment with proton beam irradiation or enucleation. Am J Ophthalmol 1985; 99:282–290. 52. Munzenrider J, Gragoudas E, Walsh S. conservative treatment of uveal melanoma: survival, local control, eye retention, and vision in patients with intermediate and large tumor after proton beam therapy. Int J Radiat Oncol Biol Phys 1994; 30(Suppl 1):233. 53. Egger E, Schalenbourg A, Zografos L, et al. Maximizing local tumor control and survival after proton beam radiotherapy of uveal melanoma. Int J Radiat Oncol Biol Phys 2001;51:138–147.

160

g. noel et al.

54. Courdi A, Caujolle JP, Grange JD, et al. Results of proton therapy of uveal melanomas treated in Nice. Int J Radiat Oncol Biol Phys 1999; 45:5–11. 55. Lumbroso L, Levy C, Plancher C, et al. [Results of proton beam irradiation for treatment of choroidal melanoma]. J Fr Ophtalmol 2002; 25:290–297. 56. Damato B, Kacperek A, Chopra M, et al. Proton beam radiotherapy of choroidal melanoma: the Liverpool-Clatterbridge experience. Int J Radiat Oncol Biol Phys 2005; 62:1405–1411. 57. Damato B, Kacperek A, Chopra M, et al. Proton beam radiotherapy of iris melanoma. Int J Radiat Oncol Biol Phys 2005; 63: 109–115. 58. Gragoudas ES, Lane AM, Regan S, et al. A randomized controlled trial of varying radiation doses in the treatment of choroidal melanoma. Arch Ophthalmol 2000; 118:773–778. 59. Weber DC, Bogner J, Verwey J, et al. Proton beam radiotherapy versus fractionated stereotactic radiotherapy for uveal melanomas: a comparative study. Int J Radiat Oncol Biol Phys 2005; 63: 373–384. 60. Frau E, Rumen F, Noel G,. Low-dose proton beam therapy for circumscribed choroidal hemangiomas. Arch Ophthalmol 2004; 122:1471–1475. 61. Tsina EK, Lane AM, Zacks DN, et al. Treatment of metastatic tumors of the choroid with proton beam irradiation. Ophthalmology 2005; 112:337–343. 62. Desjardins L, Levy C, D’hermies F, et al. [Initial results of proton therapy in choroidal melanoma at the d’Orsey Center for Proton Therapy; the ﬁrst 464 cases]. Cancer Radiother 1997; 1:222–226. 63. Saito K, Unni KK, Wollan PC, Lund BA. Chondrosarcoma of the jaw and facial bones. Cancer 1995; 76:1550–1558. 64. Benk V, Liebsch NJ, Munzenrider JE, et al. Base of skull and cervical spine chordomas in children treated by high- dose irradiation. Int J Radiat Oncol Biol Phys 1995; 31:577–581. 65. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlenther Onkol 1999; 175(Suppl 2):57–63. 66. Noel G, Feuvret L, Calugaru V, et al. Chordomas of the base of the skull and upper cervical spine. One hundred patients irradiated by a 3D conformal technique combining photon and proton beams. Acta Oncol 2005; 44(7):700–708. 67. Igaki H, Tokuuye K, Okumura T, et al. Clinical results of proton beam therapy for skull base chordoma. Int J Radiat Oncol Biol Phys 2004; 60:1120–1126. 68. Weber DC, Rutz HP, Pedroni E, et al. Results of spot scanning proton radiation therapy for chordoma and chondrosarcoma of the skull base: the Paul Scherrer institut experience. Int J Radiat Oncol Biol Phys 2005; 63:401–409. 69. Baumert BG, Lomax AJ, Miltchev V, Davis JB. A comparison of dose distributions of proton and photon beams in stereotactic conformal radiotherapy of brain lesions. Int J Radiat Oncol Biol Phys 2001; 49:1439–1449. 70. Lomax AJ, Bortfeld T, Goitein G, et al. A treatment planning inter-comparison of proton and intensity modulated photon radiotherapy. Radiother Oncol 1999; 51:257–271. 71. Fuller DB, Bloom JG. Radiotherapy for chordoma. Int J Radiat Oncol Biol Phys 1988; 15:331–339. 72. Keisch ME, Garcia DM, Shibuya RB. Retrospective long-term follow-up analysis in 21 patients with chordomas of various sites treated at a single institution. J Neurosurg 1991; 75:374–377. 73. Romero J, Cardenes H, la Torre A, et al. Chordoma: results of radiation therapy in eighteen patients. Radiother Oncol 1993; 29:27–32. 74. Forsyth PA, Cascino TL, Shaw EG, et al. Intracranial chordomas: a clinicopathological and prognostic study of 51 cases. J Neurosurg 1993; 78:741–747. 75. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the cranial base: results and follow- up of 60 patients. Neurosurgery 1995; 36:887–896.

76. Mark RJ, Tran LM, Sercarz J, et al. Chondrosarcoma of the head and neck. The UCLA experience, 1955–1988. Am J Clin Oncol 1993; 16:232–237. 77. Ruark DS, Schlehaider UK, Shah JP. Chondrosarcomas of the head and neck. World J Surg 1992; 16:1010–1015. 78. Austin-Seymour M, Munzenrider J, Goitein M, et al. Fractionated proton radiation therapy of chordoma and low-grade chondrosarcoma of the base of the skull. J Neurosurg 1989; 70:13–17. 79. Hug EB, Loredo LN, Slater JD, et al. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg 1999; 91:432–439. 80. Hug EB, Slater JD. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. Neurosurg Clin N Am 2000; 11:627–638. 81. Noel G, Habrand JL, Mammar H, et al. Combination of photon and proton radiation therapy for chordomas and chondrosarcomas of the skull base: the Centre de Protontherapie D’Orsay experience. Int J Radiat Oncol Biol Phys 2001; 51:392–398. 82. Suit HD, Goitein M, Munzenrider J, et al. Deﬁnitive radiation therapy for chordoma and chondrosarcoma of base of skull and cervical spine. J Neurosurg 1982; 56:377–385. 83. Austin JP, Urie MM, Cardenosa G, Munzenrider JE. Probable causes of recurrence in patients with chordoma and chondrosarcoma of the base of skull and cervical spine. Int J Radiat Oncol Biol Phys 1993; 25:439–444. 84. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for chordoma and chondrosarcoma: further experiences. Int J Radiat Oncol Biol Phys 1998; 41:387–392. 85. Halperin EC. Why is female sex an independent predictor of shortened overall survival after proton/photon radiation therapy for skull base chordomas? Int J Radiat Oncol Biol Phys 1997; 38:225–230. 86. Terahara A, Niemierko A, Goitein M, et al. Analysis of the relationship between tumor dose inhomogeneity and local control in patients with skull base chordoma. Int J Radiat Oncol Biol Phys 1999; 45:351–358. 87. Hug EB, Fitzek MM, Liebsch NJ, Munzenrider JE. Locally challenging osteo- and chondrogenic tumors of the axial skeleton: results of combined proton and photon radiation therapy using three-dimensional treatment planning. Int J Radiat Oncol Biol Phys 1995; 31:467–476. 88. Casali PG, Messina A, Stacchiotti S, et al. Imatinib mesylate in chordoma. Cancer 2004; 101:2086–2097. 89. Gudjonsson O, Blomquist E, Nyberg G, et al. Stereotactic irradiation of skull base meningiomas with high energy protons. Acta Neurochir (Wien) 1999; 141:933–940. 90. Austin-Seymour M, Munzenrider J, Linggood R, et al. Fractionated proton radiation therapy of cranial and intracranial tumors. Am J Clin Oncol 1990; 13:327–330. 91. Miralbell R, Linggood RM, De La MS, et al. The role of radiotherapy in the treatment of subtotally resected benign meningiomas. J Neurooncol 1992; 13:157–164. 92. Noel G, Mammar H, Ferrand R, et al. Résultats de l’irradiation par association de photons et de protons de 47 chordomes et chondrosarcomes et de 17 méningiomes. Expérience du Centre de Protonthérapie d’Orsay. Rev Neurol (Paris) 2002; 158:497–509. 93. Hug EB, DeVries A, Thornton AF, et al. Management of atypical and malignant meningiomas: role of high-dose, 3D-conformal radiation therapy. J Neurooncol 2000; 48:151–160. 94. Noel G, Bollet MA, Calugaru V, et al. Functional outcome of patients with benign meningioma treated by 3D conformal irradiation with a combination of photons and protons. Int J Radiat Oncol Biol Phys 2005; 62:1412–1422. 95. Noel G, Habrand IL, Mammar H, et al. Highly conformal therapy using proton component in the management of meningiomas: preliminary experience of the centre de protonthérapie d’Orsay. Strahlenther Onkol 2002; 178:480–485.

11.

proton beam radiosurgery

96. Blomquist E, Carlsson J. Strategy for planned radiotherapy of malignant gliomas: postoperative treatment with combinations of high dose proton irradiation and tumor seeking radionuclides. Int J Radiat Oncol Biol Phys 1992; 22:259–263. 97. Seifert V, Stolke D, Mehdorn HM, Hoffmann B. Clinical and radiological evaluation of long-term results of stereotactic proton beam radiosurgery in patients with cerebral arteriovenous malformations. J Neurosurg 1994; 81:683–689. 98. Verhey LJ, Smith V, Serago CF. Comparison of radiosurgery treatment modalities based on physical dose distributions. Int J Radiat Oncol Biol Phys 1998; 40:497–505. 99. Harsh GR, Thornton AF, Chapman PH, et al. Proton beam stereotactic radiosurgery of vestibular schwannomas. Int J Radiat Oncol Biol Phys 2002; 54:35–44. 100. Weber DC, Chan AW, Bussiere MR, et al. Proton beam radiosurgery for vestibular schwannoma: tumor control and cranial nerve toxicity. Neurosurgery 2003; 53:577–586. 101. Fuss M, Hug EB, Schaefer RA, et al. Proton radiation therapy (PRT) for pediatric optic pathway gliomas: comparison with 3D planned conventional photons and a standard photon technique. Int J Radiat Oncol Biol Phys 1999; 45:1117–1126. 102. Krengli M, Hug EB, Adams JA, et al. Proton radiation therapy for retinoblastoma: Comparison of various intraocular tumor locations and beam arrangements. Int J Radiat Oncol Biol Phys 2005; 61:583–593. 103. Lee CT, Bilton SD, Famiglietti RM, et al. Treatment planning with protons for pediatric retinoblastoma, and pelvic sarcoma: how do protons compare with other conformal techniques? Int J Radiat Oncol Biol Phys 2005; 63:362–372. 104. Hug EB, Nevinny-Stickel M, Fuss M, et al. Conformal proton radiation treatment for retroperitoneal neuroblastoma: introduction of a novel technique. Med Pediatr Oncol 2001; 37:36–41. 105. McAllister B, Archambeau JO, Nguyen MC, et al. Proton therapy for pediatric cranial tumors: preliminary report on treatment and disease-related morbidities. Int J Radiat Oncol Biol Phys 1997; 39:455–460. 106. Noel G, Habrand JL, Helfre S, et al. Proton beam therapy in the management of central nervous tumors in childhood: the preliminary experience of the Centre de Protonthérapie d’Orsay. Med Pediatr Oncol 2003; 40:309–315. 107. Miralbell R, Lomax A, Bortfeld T, et al. Potential role of proton therapy in the treatment of pediatric medulloblastoma/primitive neuroectodermal tumors: reduction of the supratentorial target volume. Int J Radiat Oncol Biol Phys 1997; 38:477–484. 108. ICRP. Recommendations of the International Commission on Radiobiological Protection. Publication 60. Annals of the ICRP. Oxford: Pergamon Press 1991; 21:1–3. 109. Miralbell R, Lomax A, Cella L, Schneider U. 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; 54:824–829. 110. Schoenthaler R, Fullerton B, Maas A, et al. Relationship between dose to auditory pathways and audiological outcomes in skull-base tumor patients receiving high-dose proton-photon radiotherapy. Int J Radiat Oncol Biol Phys 1996; 36:291.

161

111. Cella L, Lomax A, Miralbell R. Potential role of intensity modulated proton beams in prostate cancer radiotherapy. Int J Radiat Oncol Biol Phys 2001; 49:217–223. 112. Slater JD, Rossi CJ Jr, Yonemoto LT, et al. Conformal proton therapy for early-stage prostate cancer. Urology 1999; 53:978–984. 113. Shipley WU, Verhey LJ, Munzenrider JE, et al. Advanced prostate cancer: the results of a randomized comparative trial of high dose irradiation boosting with conformal protons compared with conventional dose irradiation using photons alone. Int J Radiat Oncol Biol Phys 1995; 32:3–12. 114. Benk VA, Adams JA, Shipley WU, et al. Late rectal bleeding following combined X-ray and proton high dose irradiation for patients with stages T3-T4 prostate carcinoma. Int J Radiat Oncol Biol Phys 1993; 26:551–557. 115. Cozzi L, Fogliata A, Lomax A, Bolsi A. A treatment planning comparison of 3D conformal therapy, intensity modulated photon therapy and proton therapy for treatment of advanced head and neck tumours. Radiother Oncol 2001; 61:287–297. 116. Miralbell R, Cella L, Weber D, Lomax A. Optimizing radiotherapy of orbital and paraorbital tumors: intensity- modulated X-ray beams vs. intensity-modulated proton beams. Int J Radiat Oncol Biol Phys 2000; 47:1111–1119. 117. 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 the skull. Int J Radiat Oncol Biol Phys 1998; 41: 59–68. 118. Glosser G, McManus P, Munzenrider J, et al. Neuropsychological function in adults after high dose fractionated radiation therapy of skull base tumors. Int J Radiat Oncol Biol Phys 1997; 38:231– 239. 119. Debus J, Hug EB, Liebsch NJ, et al. Brainstem tolerance to conformal radiotherapy of skull base tumors. Int J Radiat Oncol Biol Phys 1997; 39:967–975. 120. Marucci L, Niemierko A, Liebsch N, et al. Spinal cord tolerance to high dose fractionated 3D conformal proton-photon irradiation as evaluated by equivalent uniform dose and dose volume histogram analysis. Radiother Oncol 2002; 64(Suppl 1):S55. 121. Slater JD, Austin-Seymour M, Munzenrider J, et al. Endocrine function following high dose proton therapy for tumors of the upper clivus. Int J Radiat Oncol Biol Phys 1988; 15:607–611. 122. Habrand JL, Austin-Seymour M, Birnbaum S, et al. Neurovisual outcome following proton radiation therapy. Int J Radiat Oncol Biol Phys 1989; 16:1601–1606. 123. Kim J, Munzenrider JE, Maas A, et al. Optic neuropathy following combined proton and photon radiotherapy for base of skull tumors. Int J Radiat Oncol Biol Phys 1997; 39:272. 124. Urie MM, Fullerton B, Tatsuzaki H, et al. A dose response analysis of injury to cranial nerves and/or nuclei following proton beam radiation therapy. Int J Radiat Oncol Biol Phys 1992; 23:27–39. 125. Lundkvist J, Ekman M, Ericsson SR, et al. Cost-effectiveness of proton radiation in the treatment of childhood medulloblastoma. Cancer 2005; 103:793–801. 126. Malka V, Fritzler S, Lefebvre E, et al. Practicability of protontherapy using compact laser systems. Med Phys 2004; 31:1587–1592.

1 2

Robotics and Radiosurgery Cesare Giorgi and Antonio Cossu

Robotic Basics and History of Use in Medicine and Surgery Similar to surgery, radiosurgery is rapidly becoming a ﬁeld of application of robotics. Surgeons are confronted with the limits of their dexterity, endurance, and ability to process large amounts of data while operating. Radiation oncologists are able to plan treatments using large amounts of morphologic and functional data and start using robots to deliver the dose with higher precision. The introduction of automation in these ﬁelds of medicine has different origins and has occurred at a different pace. Surgery has entered the Information Age with the advent of modern diagnostic capabilities and computer power. The parallel development of endoscope, operative microscope, microinstruments, and navigators has downsized the operative ﬁeld and minimized surgical exposure. This has resulted in reduction of morbidity-mortality and shortening of hospital stay. The challenge now consists of the development of humaninstrument interfaces, both motor and sensory, to expand surgical possibilities beyond human capabilities. Radiosurgery faces less demanding tasks consisting of the development of devices that achieve high dose conformity and homogeneity in more complex disease geometries and in possibly moving organs.

History The term robot is less than a century old and was introduced by playwright Karel Capeck in R.U.R. (Rossum’s Universal Robots). Robota is a Czech word for “slave”; in Capeck’s play, living and intelligent working machines built to free humans from work [1]. The concept is at least as old as our culture; Aristotle, two dozen centuries ago, wrote in The Politics, “There is only one condition in which we can imagine managers not needing subordinates, and masters not needing slaves. This condition would be that each [inanimate] instrument could do its own work. . . . as if a shuttle should weave of itself, and a plectrum should do its own harp playing” [2]. In the following centuries, skill and imagination have left increasingly complex testimony of the concept of an automaton, not only in Western but also in Islamic culture, where in the 13th century Ibn Ismail Ibn al Razzaz al-Jazari published his

Al-jami bain al-lim wal-amal al-naﬁ ﬁ sinat’at al-hiyal (“Treatise on the theory and practice of the mechanical arts”) [3]. Western countries expressed mechanical “creatures” working with increasingly complicated clockwork mechanisms from the 14th century throughout the 17th century, to reach maximal expression with the mechanical wonders that ﬂourished at the end the 18th century, when the Swiss inventors Pierre and Henri-Louis Jacquet-Droz created their Automatic Scribe, which could write messages up to 40 characters long, and a robotic woman playing the piano [2]. At the turning of the following century, automation exited the role of exotic curiosity and entered the realm of useful devices. Joseph Jacquard invented a programmable loom, operated by punch cards, and went to mass production [4]. Toward the end of that century, Seward Babbitt created the ﬁrst robot, consisting of a crane with a gripper to remove ingots from a furnace, the ﬁrst machine designed to substitute for human work in a hostile environment [5]. Contemporarily, Nikola Tesla manufactured wireless controlled vehicles, and coined the term teleautomatics for his study of robotics. [6]. In the 20th century, the term robotics became popular after publication of Isaac Asimov’s “Runaround” story, which introduced his Laws of Robotics. These laws express the concepts that robots must obey human instructions and protect themselves but never cause harm to human beings directly or through inaction [1]. These fundamentals, more than six decades old, will always stand at the base of any design, particularly of machinery interacting with human life. The introduction of computers had an astounding impact on robot technology in the following decades, starting in 1948 with Norbert Wiener’s concept of cybernetics; communication and control in electronic, mechanical, and biological systems [7]. Programming of robots was the ﬁrst step, accomplished by George Davol in 1946 by means of a magnetic process recorder, and in 1954 with a computer [8]. Robots made their appearance on a production line of an automobile factory in 1963. The next leap was made by Shakey (so-called because of its jerky motion), the ﬁrst robot with vision, bump detectors at the base, a TV camera, and triangulating range ﬁnder capable of interaction with the surroundings. It was the ﬁrst mobile robot that could claim to reason about its actions [9]. Soon this achievement allowed for eye-hand coordination in assembly

163

164

c. giorgi and a. cossu

robots. In 1974, touch and pressure sensors were added to perform small parts assembly. The past three decades saw robots becoming widely represented components in the ﬁelds of industry, research, and even entertainment.

Robots in Medicine and Surgery: Neurosurgery Robots made their appearance in medicine in the mid-1980s, and the ﬁrst application was neurosurgical [10]. Stereotaxy was a ﬁeld where computational ability overcame manual skills. It was the time when neurophysiology was the dominant guideline to surgery. The advent of computers and the introduction of digital three-dimensional (3D) neuroanatomic images prompted pioneer researchers to use robots to help surgeons calculate the trajectory of stereotactic probes and hold instruments along that trajectory. Targets and trajectories were calculated from a multimodal set of data fast and error-free and transferred to the surgical stereotactic space to assist in performing functional procedures or biopsies. Neuromate, a contemporary 6 degrees of freedom (6DOF) articulated arm, evolved from original work published by Benabid et al. in Grenoble in 1987 [11]. The Neuromate performs all calculations involved and is able to direct the probe to target, by registering itself with respect to the stereotactic space. The Minerva, a Cartesian robot, was developed in Switzerland in 1994 [12] to perform biopsies under computed tomography (CT) control. Soon after, a similar device, completely nonmagnetic, was manufactured in Japan to perform biopsies of cerebral lesions within the gantry of a magnetic resonance imaging (MRI) instrument [13]. The concept of “controlling a probe position in stereotactic space” was extended to the robotic operative microscope, where the “probe” was the optical axis. The Zeiss MKM, Elekta Surgiscope, and the Möller-Wedel scope mounted on a Staubli Rx90 robot [14–16] were all devices that allowed the surgeon to “navigate” within the brain with the conﬁdence provided by the display of the position of his line of sight within the 3D reconstruction of the brain anatomy, based on CT or MR data. The movement of those robots was passive, controlled by the surgeon’s hand. Diffusion of these instruments within the neurosurgical community after the curiosity aroused at the presentation of each of them has not followed the expected trend: computer-assisted surgery (CAS), or image-guided surgery, has won the leadership. The obvious advantage of extracting accurate 3D morphologic data from neuroanatomic images and displaying the position of surgical instruments as in a virtual rendering of the surgical ﬁeld has rapidly gained broad consensus even among the most conservative neurosurgeons. With CAS, operative skills remain human, but the surgeon’s conﬁdence is enhanced by the added perception of complete control of the operative scenario. Minimally invasive procedures are facilitated, even in the absence of obvious anatomic landmarks. Distortions of preoperative imaging (swelling and brain shift) during surgery have to be taken into account. Intraoperative imaging techniques (US, CT, or MRI) are being evaluated in order to extend the beneﬁts of image-guided methods throughout the procedure. Following this road, with continuous perfection of CAS and consequent further miniaturization of

surgery, even in neurosurgery eventually the limits of human dexterity and stamina will be unveiled.

Robot Development in Other Fields of Surgery Other ﬁelds of medicine witnessed robots being developed to the clinical stage in the early 1990s. Davies described a device for soft tissue removal that developed into Prorobot, a system for transurethral removal of prostatic tissue, used in clinical practice [17]. Shortly after, Robodoc became the ﬁrst industrial production robot for orthopedics (hip replacement). It was an example of transfer of technology from industry to medicine. Bony structures were handled with a precision far exceeding that obtained with traditional surgery [18]. The most successful ﬁeld for robot development in surgery has been that of telemanipulators, used in endoscopic surgery. The Automated Robotic System for Optimal Position (AESOP; Computer Motion) was the ﬁrst robotic arm to position and hold an endoscopic camera to reach commercial diffusion and to clear the U.S. FDA as a surgical robot, in 1996 [19]. Two similar products appeared in the same years, the ZEUS (Computer Motion) and the da Vinci (Intuitive Surgical) telemanipulators, which have been used since then worldwide in a number of thoracic and abdominal endoscopic procedures with the largest bulk of literature regarding cardiac surgery for coronary artery bypass, mitral valve repair, and atrial septal surgery [20, 21]. Both systems consist of a surgical interface and a controller that relay action performed by the surgeon to the robotic arms, equipped with custom-designed endoscopic instruments. Visual perception of the operative ﬁeld is achieved through an endoscope mounted on a third arm, which is also controlled by the surgeon. In spite of all technical efforts though, natural visual 3D perception cannot be matched by camera pictures, and true force feedback equal to that obtained by natural human contact with tissue is still far from being effectively simulated. Still, the advantages of surgery, performed with these robots, have not been clearly demonstrated in experimental and clinical settings. It is true that hand tremor is ﬁltered and movements are scaled down. Compared with non–robotic laparoscopy, hand-eye coordination is maintained and hand movements are natural. Shear forces are neutralized, and operator’s postural distress is minimized. These facts lead us to say that the level of perfection reached by existing robotic devices allows them to be deﬁned as dexterity enhancers, far from being able to replace human operators. The deﬁnition of robots, “a mechatronic device interacting with its environment under remote or programmed control” [22], applies to the case of surgical robots, but the “interaction with the environment” is their true limit.

Robot Types Different robot types are currently used in the medical ﬁeld (Table 12-1). Selective compliance arm for robotic assembly (SCARA) and articulated robots, usually derived from industrial applications, are typically used in surgery. The reason for this choice is evident in the requirements of such procedures, where the robot is used to precisely handle low-weight

12.

robotics and radiosurgery

TABLE 12-1 Robot types in medical applications. Name

Type

Application

Freedom

Parallel

CyberKnife

Articulated

Robodoc Caspar da Vinci

SCARA Articulated Four articulated arms

Radiotherapy/radiosurgery: patient positioning Radiotherapy/radiosurgery: linac positioning Orthopedic surgery Orthopedic surgery General purpose: heart, thoracic, urologic

instrumentations. In terms of kinematics, they are serial robots, the end-effector is moved thanks to a series of joints, each responsible for a degree of freedom. SCARA and articulated robots characteristically handle large working volume. On the contrary, compared with other robots, they have poor performance in terms of payload. In radiotherapy, robots can be used to move the linear accelerator (linac), move the patient, or to conform the radiation beam. In the existing scenario, the robot-moving linac (CyberKnife) uses the 6 degrees of freedom both to shape the beam and to perform patient position tracking. Other devices tend to use a combination of the multileaf collimators to shape the beam and the robotic couch to position the patient and to track organ movement. Theoretically, the multileaf collimator alone could shape the beam and track the position, but the ﬁeld of view is ﬁxed and there are too few degrees of freedom. Speed of the blade is also too high for the current technology.

Moving Radiation Source This philosophy has been followed in developing the CyberKnife device, where a 6DOF robotic arm allows total freedom in positioning the radiation source. The choice of an articulated robot, combined with the necessary restrictions in terms of robot weight and size, results in a limitation of the source energy.

Robotic Couches with Existing Linacs The chosen technology in terms of patient positioning is the parallel robot, where the motion results from the simultaneous

165

movement of several actuators (typically six, one for each DOF). Compared with the articulated robot, this kind of kinematics allows high payload and high degree of accuracy at the cost of a reduced working volume.

Conformal Micro-Multileaf This kind of robotic device is mainly made by several actuators moving each leaf during the treatment in order to precisely conform the radiation to the shape of the lesion. Advantages and disadvantages of the different solutions will be examined in the next chapters. Table 12-2 contains a brief classiﬁcation of different robot kinematics.

Robotics and Image Guidance Radiation therapy, similar to surgery, has evolved with the progress of imaging technologies and computers. Conformal planning and delivery of maximal dose to the tumor, while sparing surrounding structures, has become more and more efﬁcient. The introduction of intensity-modulated radiation therapy (IMRT) in 1992 has made it possible to tailor the distribution of the dose according to the often non-homogeneous lesion grading, complex geometry, and the presence of surrounding critical structures. MRI, including perfusion-diffusion, spectroscopy and diffusion tensor imaging, SPECT and PET all contribute to the deﬁnition of planning target volume (PTV). Patient positioning has become extremely critical; methods and devices currently in use may soon be considered inadequate to deliver such sophisticated treatment plans with acceptable precision. Body respiratory movements and natural internal organ motion, particularly for thoracic and upper abdominal lesions, represent an additional challenge. They need to be accurately described and tracked in order to extend to moving targets the advantages of contemporary treatment planning techniques. Image-guided radiation therapy (IGRT) captures images of the body immediately prior to treatment delivery. Images can be produced by ultrasound, ﬂuoroscopy, X-rays, or cone beam tomography; they can be matched with treatment plan images, in order to detect and correct movements of the organs, or compensate for displacement of the target due to respiratory

TABLE 12-2. Kinematics of most common robot types. Type of robot

Kinematics

Cartesian

Positioning is done in the workspace with prismatic joints. This conﬁguration is useful when a large workspace must be covered or when consistent accuracy is expected from the robot. It has a revolute motion about a base, a prismatic joint for height, and a prismatic joint for radius. This robot is well suited to round workspaces. A robot whose axes form a polar coordinate system. A robot that has two parallel rotary joints to provide compliance in a plane. This robot conforms to cylindrical coordinates, but the radius and rotation is obtained by a two planar links with revolute joints. A robot whose arm has at least three rotary joints. The robot uses three revolute joints for positioning. Generally, the work volume is spherical. This robot most resembles the human arm, with a waist, shoulder, elbow, and wrist. A parallel robot is one whose arms (primary axes) have at least three concurrent prismatic joints or both prismatic and rotary joints.

Cylindrical Polar SCARA Articulated Parallel

166

c. giorgi and a. cossu

movements or other physiologic conditions. This increases the conﬁdence of radiotherapists to give a higher dose to the tumor and reduce the side effects due to irradiation of surrounding critical structures.

Ultrasound Imaging Devices like the MOMOS Radiation Oncology Batcam, an echographic probe used to locate precisely the prostate, use ultrasound to image the target precisely to improve treatment accuracy. Prior to daily treatment section, the physician obtains ultrasound images that are registered in the treatment space, and corrections in the patient positioning are obtained by means of optical tracking. Variations in daily patient setup are promptly detected; bladder and rectum are visualized and target is shown in actual position. Comparison with CT and MRI data set obtained at the time of treatment planning provides x, y, and z alignment offsets, corresponding with a new couch position. Sensitive organs can be spared, resulting in reduced discomfort and complications, and radiation treatment increases efﬁcacy through dose escalation to the target with an improved local control. Drawbacks of this system are poor imaging quality of ultrasound in presence of organs containing air or covered by thick bony structures. Also, the procedure for image acquisition involves instruments that are not integrated in the normal treatment scenario, with consequent signiﬁcant lengthening of the treatment time.

X-ray Imaging On-Board Imager, a feature of Varian linacs, consists of an Xray 150-kV tube and an amorphous silicon ﬂat-panel X-ray detector facing it on the other side of the patient, both mounted on articulated motorized arms. It generates diagnostic quality projection images of the patient at low energy, as opposed to the megavolt images generated by the former technology of electronic portal imaging (EPI). Advantages of the new device consist of the possibility of high-quality image acquisition, optimal acquisition geometry with respect to the target, speed, and lower dose to the patient. Images are compared with the treatment planning images, and appropriate patient positioning corrections are made. Two-dimensional (2D) projection imaging is adequate when the target lies close to bony structures and its position is unaffected by respiratory or other physiologic movements or by gravity. This condition typically applies to the brain, where the skull can serve as a stereotactic frame, and, to a lesser extent, to the contents of the spinal canal and the paravertebral tissues. To image soft tissue, metal markers can be implanted and then imaged prior to treatment with biplanar X-rays. For this solution, kV imaging provides better resolution than MV and allows for the use of markers as small as 1.2 mm diameter. Still, 2D imaging does not provide sufﬁcient information regarding movement of the target. The solution has been offered by Varian and Elekta, who have equipped their latest accelerators with CT quality imaging, obtained with the “cone beam” technology. The conical beam of the accelerator is used to acquire several hundreds of images of the patient in a single 360° rotation. The result is a 3D image of soft tissue, but when the tumor to be treated lies in the lungs or in organs adjacent

to the diaphragm, the fourth dimension (time) has to be visualized. Asking the patient to hold the breath during treatment at a speciﬁc cycle (deep inspiration), teaching him to breathe only when treatment is stopped, or using physical restraints for the chest can be very demanding, especially for patients that have limited respiratory capacity. A different solution has been formulated, based on the description of the movement of the chest or abdomen during breathing, obtained with optical tracking of markers applied to the skin and observed by infrared cameras. The waveform of the patient’s breathing pattern is synchronized with the CT image acquisition, so that gating of the radiation beam can be automatically performed by the cameras during treatment when they detect the position of the markers that corresponds with the position of the tumor in the ﬁeld chosen for treatment.

Other Imaging Techniques Yet another approach to image-guided radiotherapy is that of helical tomotherapy (HT) combining a rotating intensity-modulated fan beam with integrated CT imaging. The elegant concept of imaging the body during treatment was ﬁrst described by Carol [23] but successfully exploited only later by Mackie et al. [24] after the introduction of spiral CT. The result is deﬁned as adaptive radiotherapy highly integrated adaptive radiotherapy (Hi Art). Signiﬁcant clinical follow-up documenting beneﬁts related to the introduction of these novel technologies is not yet available, but the combination of IMRT and IGRT promises improved tumor control and superior normal tissue sparing. All this is achievable at the expense of instrument cost, complication of the procedure, and treatment time. A favorable element of this equation is the enhancement of automation in patient positioning, which is the basis for the introduction of robotics in radiotherapy.

Applications of Robotics to Radiotherapy The evolution of morphologic and functional diagnostic imaging and the availability of computer power have made possible the design of complex treatment plans, highly conformal to the target and capable of sparing adjacent organs, even in cases of complex geometry like concave targets surrounding and overlapping organs at risk. Translating the treatment plan into reality with continuously improving accuracy is where a large portion of research in radiotherapy is aimed.

Radiosurgery: The Gamma Knife The principles of radiosurgery—single session, highly conformal dose delivery—are the most demanding in terms of accurate targeting. Radiosurgery was introduced in the 1950s for treatment of central nervous system lesions. It has been very efﬁciently accomplished with the Gamma Knife, a hemispheric array of cobalt-60 sources collimated to the center. Imaging of the target and positioning of the patient is obtained with reference to a frame, which is ﬁrmly applied to the patient’s skull. The ﬁxed spatial relationship between the frame and the target structure, identiﬁed by means of “localizers” applied to the frame during

12.

robotics and radiosurgery

image acquisition, translates into precise targeting of the lesion once the frame is positioned for treatment, and the target is placed at the center of the collimated sources. When the shape of the target is not spherical, the treatment is obtained with a cluster of spherical lesions that “cover” the irregularly shaped volume to be treated. The Gamma Knife is a concept that has limitations in its use: only cranial-cervical lesions with a single dose can be treated, the dose distribution to the target is less homogeneous than that obtained with other methods, but still it represents the gold standard, to which all other technologies are confronted. In recent years Cartesian robot has been incorporated in the latest version of this equipment, with the aim of positioning the frame during treatment so that each “shot” or spherical radiation dose can be automatically and precisely directed in sequence to the target. Lately, the Gamma Knife “Perfexion” has been introduced. It represents the ultimate evolution of this elegant technique. Beside signiﬁcant room added for cranial and cervical targets, a clever robotic delivery of the does by means of mechanical alignment of Co60 sources with different size collimators, combined with robotic movements of the couch head holder system, allows for improvement of does conformity with signiﬁcant shortening of treatment time. Meanwhile, other techniques have been introduced that take advantage of the enormous development of computers and imaging technology that has occurred in the past couple of decades. Physicists, radiation oncologists, and neurosurgeons having witnessed the tremendous achievements of radiosurgery, obtained with the original technique, have expanded the neurosurgical concept of stereotaxy to ﬁnd new applications for treatment of other districts of the body. This concept has also deeply modiﬁed the way fractionated radiotherapy is delivered.

Linac and CyberKnife, Single and Multiple Stereotactic Fractions Linac radiosurgery was introduced in the early 1980s. The elegant concept of multiple non-coplanar arc therapy has been complemented in the past decade with the introduction of micromultileaf collimators. Whether they are built into the linac head or added to it as an accessory, these multiaxial robots allow for highly conformed and uniform dose to target and can modulate the intensity of the dose according to treatment requirements. Dosimetric performance of some of these devices, particularly of the ones that claim better design characteristics (double focusing and interdigitation of blade travel, thinner leaves, dynamic movements, minimal transmission and leakage between the leaves), is very high, inferior only to protons [25]. Charged particle beams dosimetry is in fact superior to any other photon therapy method and can also be very effectively shaped by electromagnetic steering. Access to this treatment modality is out of reach for the large majority of radiation oncologists, being only available at present at very few centers worldwide. About a decade ago, CyberKnife was introduced in clinical practice: it was the ﬁrst device for precision radiotherapy to use an industrial robot to direct a photon beam generated by a small on-board linac to the target. Detailed description of this innovative device is dealt with in another chapter of this book; here we will only describe and comment on some of its features, which include unobstructed access to the entire body; high

167

mechanical precision; and innovative image-guided control loop, with target tracking capability. This is a device that truly represents the evolution of the Gamma Knife concept, and similarly it has been developed by a neurosurgeon [26]. It resembles the method: covering the target with high conformity, using a non-isocentric array of beams (rather then multiple treatment “shots”). The similarity ends here: the system is open and allows for the unobstructed treatment of the whole body. The source is a lightweight 6-MeV linac, with circular secondary collimators of various diameters, mounted on an industrial robot (KUKA Roboter GmbH). An X-ray imaging system is part of the system control. The treatment planning is similar to other inverse planning systems: identiﬁcation of the target and critical structures surrounding it from pretreatment morphologic and functional data set, prescription of the dose to the target, and description of constraints. A CT acquisition, among other appropriate studies, is mandatory, to generate pre- and intratreatment digitally reconstructed radiographs (DRRs), necessary for patient positioning and subsequent tracking. Planning of dose delivery is performed in an unique fashion: the process chooses a number of “nodes” laying on a sphere some 80 cm around the target volume. Guided by the prescription dose and the constraints for critical structures, the system chooses beam directions and weights for each node, to reach optimal conformity to the prescribed dose distribution. This “non-isocentric” technique is very effective in designing highly conformal plans. “Wraparound” doses for critical structures, like the spinal cord in the treatment of spinal metastases, can be obtained. A unique feature of the system is represented by the image guide loop during treatment. Two ﬂat-panel, amorphous silicon X-ray cameras are used for patient positioning and treatment tracking. The systems generates a sequence of DRRs from the pretreatment CT study that are matched with the couple of orthogonal X ray images acquired during treatment. Changes in the position of the target during treatment are disclosed with the acquisition of orthogonal projection images. The new target position is compared with the position at the planning phase, and the beam directions are corrected, accordingly. This step is repeated at each treatment node.

Frameless Head and Spine Radiosurgery The introduction of image-guided radiotherapy is certainly a major improvement compared with the simple immobilization techniques that have been used so far in the treatment of the body. The frameless solution introduced by the CyberKnife for the treatment of cerebral lesions appeals to patients and eliminates the need for a surgical act. This solution is not applicable to the Gamma Knife, an exquisitely frame-based device, but can be promptly adopted in linac sites, equipped with ﬂat screen panel and optical tracking. A similar evolution has been observed in neurosurgery. The need for stereotactic assistance in neurosurgical procedures, where location and dimensions of the lesion required it, prompted the introduction of the stereotactic frame guidance. Soon it became evident that obstruction and access limitation was a major limit of the method.

168

c. giorgi and a. cossu

Frameless solutions for surgery were anticipated in the early 1990s [27]. Large diffusion of “neuronavigation” assisted neurosurgical procedures has followed: optical tracking or magnetic ﬁeld patient’s registration allow for unobstructed access to the operative ﬁeld. This is evidently a legitimate choice, justiﬁed by incompatibility between frame-based localization and open neurosurgery and represents an evolution compared with the previously adopted frame-based technique. It needs to be pointed out, though, that when a procedure requires high precision in target localization (like deep brain stimulating electrodes in the subthalamus), no surgeon risks the potential error introduced by frameless methods. One could argue that ﬁducials glued to the skin can shift slightly and that biplanar X-ray registration is more accurate, but even in oldfashioned “functional” operating room sites, where biplanar teleradiography hardware is in use, no stereotactic procedure would take place without the reassuring presence of a frame. Naturally, if a patient is asked whether he prefers to be treated with the inconvenience of a frame or by means of a “soft” ﬁxation, he will prefer the latter, but he would certainly have doubts, if he would be plainly informed that frameless radiosurgery involves computational power, extra labor and irradiation, to match what is readily obtained with a frame. Frame positioning, in experienced hands, has never been an issue for the patient and will not take more than minutes to be done. Frameless radiosurgery of the spine, on the contrary, is without doubt perfectly adequate, and is the only sound method available, and in the past years an increasing number of body localizers devoted to body radiosurgery positioning have appeared on the market.

reconﬁguration of the beam can be performed. An organ position is best described by imaging of the organ itself. Determining the position of adequate bony landmarks or of metal ﬁducials implanted in the organ is not always feasible in the ﬁrst case, or is invasive with all related consequences in the second. Fiducial markers, if implanted in sufﬁcient number, can be automatically detected and the rototranslation of the target can be solved so rapidly that real-time ﬂuoroscopy can be used. Fluoroscopy, even performed with the best possible efﬁciency, delivers a skin dose of 2 cGy per minute, which for prolonged treatment as in hypofractionation or radiosurgery is unacceptable. A proposed solution is correlation between different data: intermittent imaging of the organ and continuous measures of movement of chest or abdomen or spirometry that is used to interpolate target position between radiographic images. This works satisfactorily for some tumor locations but not for the majority of lung tumors, where motion is too complex. Proposed solutions that develop mathematical model to predict breathing motion or empirical approaches with adaptive ﬁlters are being pursued, but still the timing of intermittent respiratory image acquisition should be well below one-half second. The issue of the dose necessary for imaging of the target remains a critical one: the accumulation of radiographic dose during treatment should not be greater than the therapy beam leakage and scatter dose. For CyberKnife, it has been calculated that the radiographic dose is smaller than the leakage dose only for treatment of the head and spine [28]. Imaging of target affected by respiratory movements is still an unsolved issue for any technology.

Other Methods for Directing the Beam Tracking Targets Being able to correctly position a patient for treatment and to compensate for movements during treatment has a number of advantages: it becomes possible to explore different treatment schemes that are difﬁcult to perform with traditional loose ﬁxation. Fractionated treatment can reduce or eliminate the need for tumor motion margin. This is particularly evident for the treatment of moving organs but can reduce the difference between target deﬁnition in radiosurgery and in fractionated radiotherapy, thanks to the ability to determine the target position, with accuracy very close to that of a stereotactic frame. Once a patient is correctly positioned, random and cyclic movements have to be taken into account. Treatment of the CNS and spine only deals with random motion both in time and direction of the target, but these can become signiﬁcant as the time for fraction increases, as in radiosurgery. The solution adopted by CyberKnife, assuming random movements of the patient occurring once every 2 minutes, sets the time for imaging interval to 1 minute. Researches state that this solution will yield misdirection of less than 1% of the dose to more than 1 mm off-target, and conclude that this ﬁgure is sufﬁcient to maintain radiosurgical standard [28]. Unfortunately, quite often patients that undergo radiosurgery are not very cooperative: fast random motion can occur, and this tracking interval could be insufﬁcient to correct the beam direction. Tracking an organ with cyclic movements (respiration and heartbeat) involves determining the actual position of the organ and predicting its future position, so that new alignment and

Determining variations in target position requires the possibility to redirect the beam. Apart from the CyberKnife, other methods have been described, namely moving the patient with a robotic couch or shifting the aperture of a multileaf collimator. Patient positioning, motion control, and tracking of the target are required to deliver the prescribed dose to the lesion with minimal irradiation of the surrounding tissue. Generation of images during treatment and robotic beam or patient positioning during single or multiple treatment fractions are the instruments currently under development. As a result, single-session treatments of head and spine lesions without the constraint of a rigid ﬁxation are already feasible, and similar stereotactic delivery methods can be applied to small lesions throughout the body, in single or multiple sessions. Robotization of the delivery of the dose by means of a robot-mounted linac is one possible solution, another one considers the automatic positioning of the couch (Table 12-3). Again, compensating for changes in patient position or random intrafraction movements is a task that can be effectively solved by different methods, but tracking moving targets increases complexity for the reasons illustrated above. Studies are being conducted regarding shifting the aperture of a multileaf collimator [29]. In the case of a conventional linac mounted unit, the alignment of the beam can only be maintained in the plane of the treatment ﬁeld. This satisﬁes the stochastic category of motion, but in case of targets rotating out of plane, it could be very difﬁcult to maintain treatment conformality, mainly due to limited leaf speed.

12.

robotics and radiosurgery

TABLE 12-3. Feature comparison between whole body radiotherapy robotic devices. Feature

CyberKnife

Robotic couch

Patient’s motion discomfort Mechanical precision Dose conformity Dose homogeneity Body access IGRT

None

Possible

High

High

Very high Poor Easy 2D low detail, X-ray quality Labor intensive Yes

Very high Very high (Tomotherapy) Easy 3D high detail, CT quality

Custom

Regular

Ready

Feasible

Limited by design

Unlimited

Safety Linac energy and weight limitations Site dimension and shielding Frameless neuro-SRS Future system development

Standard No

Robotic Couch Target positioning by means of a robotic couch presents several advantages, mainly because a robotic couch can be ﬁtted to any linac. Combined with optical tracking and IGRT, superior image acquisition solutions are complemented by a very effective tool for positioning the patient and compensating for target movement. Fast-moving and rotating targets, limited patient compliance to continuous motion, and the presence of inertia could pose serious limits to this solution alone and require the combination of the multi-leaf collimator (MLC) beam shifting. Out of the several solutions available on the market, it is noteworthy to mention the latest evolution: it has been designed by 3Dline Medical Systems, taking into account previous experience with similar parallel architectures. It can substitute existing couches, but it can also be mounted on top of one, without limiting the use of the linac for more “traditional” procedures. It has 6 degrees of freedom; linear excursions cover 80 mm; angular orientation span 4° on the three axes, in any position of the working volume. It is a “lineapode” architecture that uses six rods to obtain all translational and angular movements. All actuators are mounted on the base of the structure, thus reducing to 200 mm the height of the device at rest and reducing the inertia to a minimum.

Glossary CAS Computer-aided surgery. Initially, CAS meant a technology of surgical simulation using three-dimensional organ models and reconstructed medical imaging by computer graphics technique. In Japan, Prof. Takeyoshi Dohi and Prof. Masakazu Tsuzuki at the University of Tokyo were the ﬁrst who used this word for their research. In other countries, the term “computer-assisted surgery” is commonly used.

169

CT Computed tomography. Originally known as computed axial tomography (CAT or CAT scan) and body section roentgenography, it is a medical imaging method employing tomography where digital geometry processing is used to generate a three-dimensional image of the internals of an object from a large series of two-dimensional X-ray images taken around a single axis of rotation. DOF Degrees of freedom. The set of independent displacements that specify completely the displaced or deformed position of the body or system. In robotics, DOF is often used to describe the number of directions in which a robot can pivot or move a joint. EPR Electronic portal image IGRT Image-guided radiation therapy IMRT Intensity-modulated radiation therapy Laparoscopy Laparoscopic surgery, also called keyhole surgery (when natural body openings are not used), Band-Aid surgery, or minimally invasive surgery. MIS Minimally invasive surgery MRI Magnetic resonance imaging, formerly referred to as magnetic resonance tomography (MRT) or nuclear magnetic resonance (NMR), is a method used to visualize the inside of living organisms. It is primarily used to demonstrate pathologic or other physiologic alterations of living tissues and is a commonly used form of medical imaging. PET Positron emission tomography is a nuclear medicine medical imaging technique that produces a three-dimensional image or map of functional processes in the body. PTV Planning target volume Radiosurgery A medical procedure that allows noninvasive brain surgery (i.e., without actually opening the skull) by means of directed beams of ionizing radiation. It is a relatively recent technique (1951), which is used to destroy, by means of a precise dosage of radiation, intracranial tumors and other lesions that could be otherwise inaccessible or inadequate for open surgery. Radiotherapy The medical use of ionizing radiation as part of cancer treatment to control malignant cells (also called radiation therapy). US Ultrasonography (medical sonography) is a useful ultrasound-based diagnostic medical imaging technique used to visualize the fetus, muscles, tendons, and many internal organs; their size, structure, and any pathologic lesions.

References 1. Robotics Research Group, University of Texas at Austin. Learn More History. Available at http://www.robotics.utexas.edu/rrg/ learn_more/history. 2. Malone R. The Robot Book. New York: Push Pin Press, 1978. 3. Al-Hassani STS. Foundation for Science Technology and Civilisation. Al-Jazari—The Mechanical Genius. Available at http://www. muslimheritage.com/topics/default.cfm?ArticleID=188. 4. Spartacus Educational. Joseph Jacquard. Available at http:/www. spartacus.schoolnet.co.uk/SCjacquard.htm. 5. II Robotics. Timeline—Back to the Beginnings. Available at http:/ www.iirobotics.com/webpages/robothistory.php. 6. Tesla Society. Available at ieee.org/web/aboutus/history_center/ biography/tesla.html. 7. Vallée R. Norbert Weiner. International Society for Systems Sciences. Available at http://www.isss.org/lumwiener.htm.

170

c. giorgi and a. cossu

8. Robot Hall of Fame. The 2003 Inductees: Unimate. Carnegie Mellon University. Available at http://www.robothalloffame.org/ unimate.html. 9. SRI International. Shakey the Robot. SRI International. Available at http://www.sri.com/about/timeline/shakey.html. 10. Kwoh YS, Hou J, Jonckheere EA, et al. A robot with improved absolute positioning accuracy for CT guided stereotactic brain surgery. IEEE Trans Biomed Eng 1988; 35:153–160. 11. Benabid AL, Cinquin P, Lavallée S, et al. Computer driven robot for stereotactic surgery connected to CT scan and magnetic resonance imaging: technological design and preliminary results. Appl Neurophysiol 1987; 50:153–154. 12. Fankhauser H, Glauser D, Flury P, et al. Robot for CT-guided stereotactic neurosurgery. Stereotact Funct Neurosurg 1994; 63:93–98. 13. Masamune K, Kobayashi E, Masutani Y, et al. Development of an MRI-compatible needle insertion manipulator for stereotactic neurosurgery. J Image Guid Surg 1995; 1:242–248. 14. Steinmeier R, Rachinger J, Kaus M. Factors inﬂuencing the application accuracy of neuronavigation systems. Stereotact Funct Neurosurg. 2000; 75:4. 15. Bonev I. Delta Parallel Robot—The Story of Success. ParalleMIC—The Parallel Mechanisms Information Center. Available at http://www.parallemic.org/Reviews/Review002.html. 16. Giorgi C, Sala R, Riva D, et al. Robotics in child neurosurgery. Child’s Nerv Syst 2000; 16:832–834. 17. Ng WS, Davies BL, Hibberd RD, et al. A ﬁrst hand experience in transurethral resection of the prostate. IEEE Med Biol Soc Mag 1993; 120–125. 18. Bargar WL, Bauer A, Borner M. Primary and revision total hip replacement using the Robodoc system. Clin Orthop 1998; 354: 82–91.

19. AESOP, Computer Motion Robot. Available at http://www. timeforce.com/Medical_Robotics/Medical_Robotics_Companies/ computermotionproﬁle.html. 20. Damiano RJ Jr, Reichenspurner H, Ducko CT. Robotically assisted endoscopic coronary artery bypass grafting: current state of the art. Adv Card Surg 2000; 12:37–57. 21. Mohr FW, Falk V, Diegeler A, et al. Computer-enhanced coronary artery bypass surgery. J Thorac Cardiovasc Surg 1999; 117: 1212–1214. 22. Hootman R. Homebrewed Robots. Available at www.virtuar. com/click/2005/robonexus/. 23. Carol MP. Peacock: a system for planning and rotational delivery of intensity modulated ﬁelds. Int J Imag Sys Technol 1995; 6:56–61. 24. Mackie TR, Balog J, Ruchala K, et al. Tomotherapy. Semin Radiat Oncol 1999; 9:108–117. 25. Clivio A, Bolsi, Cozzi L, et al. Advanced radiotherapy techniques applied to brain tumours. A comparative study. Presented at 8th Biennial ESTRO Meeting on Physics and Radiation Technology for Clinical Radiotherapy, Lisbon, Spain, September, 2005. 26. Adler JR, Chang SD, Murphy MJ. The CyberKnife: A frameless robotic system for radiosurgery. Stereotact Funct Neurosurg 1997; 69:124–128. 27. Giorgi C. Imaging techniques and computers. Stereotact Funct Neurosurg 1994; 63:8–15. 28. Murphy MJ. Tracking moving organs in real time. Sem Radiat Oncol 2004; 14:91–100. 29. Jiang S, Zygmansky P, Kung J. Gated motion adaptive therapy (GMAT): modiﬁcation of IMRT MLC leaf sequence to compensate for tumor motion. Med Phys 2002; 29:1347–1348.

1 3

CyberKnife Radiosurgery John R. Adler Jr., Alexander Muacevic, and Pantaleo Romanelli

Image-Guided Robotic Radiosurgery

CyberKnife Technology

Conceptually, the lineage of CyberKnife (Accuray Inc., Sunnyvale, CA) technology derives from the clinical principles that underlie stereotactic radiosurgery. This minimally invasive procedure involves the precise delivery of large doses of ionizing radiation to destroy well-deﬁned targets without injuring the surrounding and intervening healthy tissue. This objective is achieved using large numbers of narrow beams that emanate from a wide array of directions and intersect (and therefore accumulate) within the volume selected for ablation. The cumulative dose that can be administered this way overwhelms any capacity for cellular repair, thereby typically ensuring tissue destruction. Until recently, radiosurgery was conﬁned to the brain and skull base, and for almost 30 years now, the Gamma Knife (Elekta AB, Stockholm, Sweden) system has been the standard instrument for the neurosurgical application of radiosurgery. However, modiﬁed linear accelerators (linacs) also emerged over the years to become well-accepted radiosurgical technologies. Gamma Knife and conventional linacs both require the application of an invasive frame on the patient’s head to achieve the desired stereotactic accuracy of ±1 mm. The frameless targeting system of the CyberKnife represents a radical departure from this approach. Furthermore, the combination of imagedguided targeting with robotic technology is enabling the original scope of radiosurgery to be dramatically expanded. During CyberKnife radiosurgery, real-time intraoperative imaging is used instead of a stereotactic frame to establish the tumor position with reference to skeletal anatomy. In several clinical circumstances, the combination of image guidance and robotics offers a material advantage over more conventional approaches to radiosurgery. Primary among these beneﬁts is the fact that the beam can track lesion motion in any direction throughout the body. Recently, a new tracking method based on the correlation between external (chest) and internal (lesion) motion has made it possible to also follow targets that move with respiration while the treatment beam is on. These developments have extended what was once a solely intracranial application to targets throughout the chest and abdomen. In this chapter, we describe the CyberKnife and some of its more unique clinical indications.

The CyberKnife is composed of a lightweight and compact high-energy X-ray source (6-MeV linac, dose rate 6 Gy/min) coupled to a robotic arm capable of moving with 6 degrees (6D) of freedom (Kuka GmbH, Augsburg, Germany) (Fig. 13-1). During treatment, the manipulator is capable of aiming 1200 beam directions (1600 with the newest G4 model) toward the lesion being treated. The robot is spatially calibrated to a computerized localization system consisting of two X-ray generators that are ﬁxed on the ceiling to enable orthogonal images of the target region. Images are recorded on ﬂoor-mounted silicon detectors that generate high-resolution digital images. X-rays are registered to digitally reconstructed radiographs (DRRs) derived from the planning computed tomography (CT) scan (Fig. 13-2), and deviations of the target region are corrected automatically during initial patient setup on the ﬁve-axis patient couch. During treatment, X-rays are frequently acquired and used to automatically compensate for patient movements within a range of 10 mm by adjusting the direction of the treatment beam. The targeting accuracy of this system design has been repeatedly demonstrated to be submillimetric [1, 2]. At each position of the robot, beams can be directed toward different areas of the target region. This design enables the treating surgeon to select from a large array of non-isocentric, non-coplanar beams during the process of constructing a treatment plan and thereby create dose distributions that conform to even irregularly shaped lesion volumes (Fig. 13-3). In contrast, more conventional radiosurgical devices construct spherical dose distributions around a discrete isocenter.

CyberKnife Spinal Radiosurgery Image-guided robotic radiosurgery can also be a useful tool for ablating a broad spectrum of spinal lesions. It is worth emphasizing that in terms of dose conformality and targeting accuracy, CyberKnife spinal radiosurgery compares favorably with standard frame-based intracranial radiosurgery [2]. The initial method for targeting the spine during CyberKnife radiosurgery required the percutaneous implantation of ﬁducials (typically, three to four 4 × 2 mm stainless steel screws placed in vertebrae adjacent to the target lesion). By comparing the positions of

171

172

j.r. adler jr. et al.

FIGURE 13-1. The main components of the CyberKnife system. (Image used with permission from Accuray Incorporated.)

FIGURE 13-2. A library of digitally reconstructed radiographs (DRRs) is generated from the treatment planning CT. Each of these images, which approximate an oblique projection, emulate a unique pose of the patient’s anatomy.

FIGURE 13-3. (a) With standard radiosurgery dispersed isocentrical beams, all intersect a common region (light gray). Because multiple spherical volumes are needed to cover nonround lesions, the resulting

dose distribution tends to be inhomogeneous. (b) Non-isocentric beams from various directions. Beams do not all cross in a single point and dose need not be inhomogeneous.

13.

cyberknife radiosurgery

ﬁducials in pretreatment DRRs with their location in intratreatment radiographs, it is possible to calculate a rigid-body 6D (3 × translation, 3 × rotation) offset and adjust the treatment beam accordingly. To compensate for patient motion during treatment, the imaging-and-alignment process is repeated between beams. The clinically relevant accuracy of the above approach was investigated by Yu and co-workers [2]. In this study, measurements were performed at three different CyberKnife facilities using head and torso phantoms loaded with packs of radiochromic ﬁlm. The absolute displacement of the dose contours from that deﬁned during treatment planning was reported to be submillimetric; the average offset was 0.3 ± 0.1 mm. Fiducial tracking error was below 0.3 mm for radial translations smaller than 14 mm and less than 0.7 mm for rotations up to 4.5 degrees.

Fiducial-Free Tracking More recently, ﬁducial-free tracking has been developed for CyberKnife spinal radiosurgery (Xsight; Accuray, Sunnyvale, CA). This technology enables the tracking of spinal lesions based on anatomic landmarks instead of surgically implanted ﬁducials. Similar in many ways to the head-tracking algorithm, Xsight automatically references radiographically visible skeletal structures. The fact that much of the spine is a nonrigid body introduces an added dimension of complexity to the underlying algorithm. Nevertheless, the approach has proved remarkably robust and accurate. Recent investigation demonstrates that the targeting accuracy of Xsight compares favorably with the published precision of ﬁducial-based localization [3]. The main beneﬁts of a ﬁducial-free system are (1) nonrigid deformation is accounted for, potentially improving treatment accuracy in the many cases where patient pose changes have occurred, and (2) marker insertion is not required. The latter precludes any associated risks and dramatically increases convenience for both patient and clinician. A further, yet more minor, advantage of skeletal-based localization is that the spatial ﬁdelity of targeting is maintained in the event of ﬁducial migration, the risk of which is admittedly small with spinal screws.

173

CyberKnife Body Radiosurgery The CyberKnife can also be used to ablate soft tissue targets within the major body cavities. As a prelude to such radiosurgery, a minimally invasive procedure is required to percutaneously implant several (at least 3) small (0.8 × 4 mm) gold seeds in close proximity to the treated lesion. These markers provide a radiopaque frame of reference for radiosurgical targeting. However, intrathoracic and intraabdominal organs move throughout treatment, especially because of diaphragm motion [4–9]. In an attempt to compensate, conventional radiotherapy systems may utilize respiratory gating (turning on the treatment beam when the moving tumor is in range) or breathing restriction (using frames that apply abdominal pressure) to limit tumor movement. The accuracy that can be achieved with either approach is still less than ideal and a substantial margin (1 cm or more) is commonly included around the lesion being irradiated. In contrast with standard gantry-based systems, the robotic arm of the CyberKnife can move the linac in any direction. This capability makes it feasible to compensate for complex breathing motion in real time. The Synchrony motion-tracking system was developed to accomplish this objective. This technology senses in real time three-dimensional respiratory motion by means of infrared (IR) LEDs attached to the patient’s chest or abdomen and a pair of IR cameras. The absolute location of the tumor at a given point in time is determined from the position of percutaneously implanted gold ﬁducials within the CyberKnife’s orthogonal X-ray imaging system (Fig. 13-4). The Synchrony system develops a continuously updated model that relates the position of the implanted gold markers to the moving skin surface of the patient. The correlation model built by Synchrony begins with a series of radiographs taken after patient setup and prior to the start of treatment. The positions of the ﬁducials in the digital radiographs are correlated with the time-stamped positions of the IR emitters at the moment of X-ray acquisition (Fig. 13-5). As the patient breathes, software computes a correlation model that is used to dynamically adjust the aim of the linac. In the process, the position of the treatment beam is continuously adapted to the location of the tumor. Periodically during

FIGURE 13-4. A CT image is displayed that was obtained during the process of implanting a ﬁducial into a right-sided lung metastasis from a tonsillar carcinoma. The needle with an ejected 5-mm gold seed at the outer tumor margin is shown.

174

j.r. adler jr. et al.

such inverse planning proves to be of greatest advantage for the most irregular target volumes. After the radiation oncologist, often working with a surgeon, speciﬁes the dose to a tumor and surrounding critical structures, the treatment planning system determines directions and activation durations for each non-isocentric, non-coplanar beam (selected from a universe of approximately 1200 beam directions) such that the primary speciﬁcations are met. Next, an optimization method is used to compute beam weights such that initial dose constraints are fulﬁlled. Typically, the ﬁnal plan utilizes more than 100 beam directions in a single treatment. Recently, planning has been extended to the case of moving targets [12, 13]. However, while the target may move during treatment, critical structures in the vicinity may not move with the same speed or direction. The relative motion of organs with respect to one another requires highly accurate planning. The temporal changes over the entire respiration cycle are represented in a motion model (obtained from either fourdimensional CT or by deforming an inhalation image stack into an exhalation image stack). This motion model is used in fourdimensional planning. It has been shown recently [14] that the linear programming approach used by the CyberKnife can be extended to four-dimensional inverse planning. FIGURE 13-5. This schematic illustrates several aspects of dynamic respiratory compensation with the Synchrony system. In this situation, the movement of a lung tumor is determined in real time by correlating the location of internal gold seeds (not visible) with external markers (IR emitters). Tumor position is then fed back to the robot, which adjusts accordingly for tumor motion.

treatment, radiographs are obtained and the correlation model is updated. Movements of the IR emitters that violate the correlation model (as when a patient’s breathing pattern changes) turn the treatment beam off and initiate the model-building procedure again.

CyberKnife Treatment Planning Radiation planning with the CyberKnife is generally quite different from the standard isocentrically arrayed radiosurgical system (e.g., the Gamma Knife). Although treatment plans are always developed on top of CT image data, other imaging modalities (e.g., MRI, fMRI, PET) are readily supported by means of image fusion. Like other radiosurgical devices, the CyberKnife uses secondary collimators to shape cylindrical beams with diameters ranging from 5 to 60 mm. However, during the planning phase of standard radiosurgical systems, beam directions and beam weights are iteratively deﬁned by the radiation oncologist, physicist, and/or dosimetrist, often with a surgeon, until a suitable distribution of radiation dose is achieved. Although an isocentric treatment mode for spherically symmetric lesions is available with the CyberKnife, most planning is done using an inverse approach. For more complex and irregularly shaped tumors, robotic radiosurgery provides an enormous range of possibilities for placing beams and beam arrays. In designing plans for these targets, the goal of the CyberKnife planning algorithm is to ﬁnd a scheme of beams that returns a maximally conformal distribution for a given shape of target [10, 11]. Practically speaking,

Intracranial Lesions Most conditions treated with standard frame-based radiosurgery are readily treatable with the frameless CyberKnife using the same basic radiosurgical principles. However, the combination of real-time image guidance and robotic treatment delivery can offer distinct advantages over conventional frame-based radiosurgery under some clinical circumstances. Because of its robotic nature, the CyberKnife enables an increase in both the range of beam trajectories and in some cases the volume of space over which they are distributed. Clinically speaking, this design provides greater homogeneity, when desirable, and enhanced dose conformality for irregularly shaped targets. Even more importantly, frameless stereotaxy makes it practical to both incorporate hypofractionation into a radiosurgical context and, for the ﬁrst time, perform extracranial radiosurgery with true stereotactic precision [2]. The CyberKnife can be used to treat all the common intracranial conditions currently treated radiosurgically, such as intracranial tumors, arteriovenous malformations, and functional indications such as trigeminal neuralgia. However, the use of hypofractionation makes it feasible to treat many larger tumors or lesions adherent to especially radiation-sensitive brain structures such as the anterior visual pathways. Moreover, CyberKnife radiosurgical ablation is readily extended to lesions that originate beneath or extend through the skull base and involve the hypopharynx, such as tumors of the foramen magnum or nasopharyngeal carcinoma. Although it is beyond the scope of this chapter to describe all potential intracranial indications, here we highlight a few of them.

Perioptic Tumors Hypofractionation is especially useful for treating lesions immediately adjacent to the optic pathways or lesions involving other

13.

cyberknife radiosurgery

cranial nerves. High rates of tumor control and preservation of visual function after multisession robotic radiosurgery have been reported [15–17]. In these retrospective studies, patients with meningioma, pituitary adenoma, and craniopharyngioma, all within 2 mm of the optic apparatus, were treated with CyberKnife radiosurgery delivered in two to ﬁve sessions to a mean cumulative marginal dose of 20.3 Gy. After 4 years of follow-up, this multisession approach resulted in very high levels of local tumor control (>95%) and was nearly universally safe; one patient with a history of conventional radiation therapy and three courses of perioptic radiosurgery suffered a radiationinduced optic nerve injury. This experience is remarkable for the high incidence of visual function preservation and tumor control in what was otherwise a particularly challenging group of patients; nearly all had a history of either surgical resection (sometimes multiple) and/or conventional radiotherapy.

175

Spinal Indications In the ﬁrst reported use of image-guided robotics to perform spinal radiosurgery, Ryu and co-workers demonstrated the safety and short-term efﬁcacy for a variety of neoplastic and vascular lesions [21]. Surgical implantation of ﬁducials into adjacent vertebral segments was necessary for tracking the ablated spinal lesion. There were no adverse events related to the implantation of ﬁducials. Given the appreciable uncertainty about the spinal cord’s tolerance to radiosurgery, hypofractionation, using two to ﬁve fractions, was utilized in nearly all cases. Subsequent to the above investigation, Gerstzen and coworkers found in their clinical series of 125 patients that singlefraction spinal radiosurgery was safe and effective under most circumstances [22]. A mean tumor dose of 14 Gy as prescribed to the 80% isodose was used (Fig. 13-6). Most patients’ lesions

Vestibular Schwannomas Small- to moderate-sized vestibular schwannomas can be treated with single or fractionated radiosurgical regimes. Since 1999, the CyberKnife at Stanford University has been used to treat more than 350 patients with vestibular schwannoma, delivering 18 to 21 Gy in three sessions separated by 24 hours. To date, only three patients within this cohort has shown evidence of tumor progression, and in no case was treatment-related trigeminal or facial nerve dysfunction observed. Among those patients with Gardner-Robertson grade I or II hearing preoperatively, 74% retained these hearing levels after an average 4 years of follow-up. Furthermore, there were no cases of total hearing loss [18].

Trigeminal Neuralgia Radiosurgical rhizotomy, most commonly performed with the Gamma Knife, is well-established in the management of trigeminal neuralgia. However, after more than a decade of experience, treatment latencies and the overall response rate continue to be less than ideal. With the goal of overcoming these limitations, the standard Gamma Knife trigeminal rhizotomy has been modiﬁed by using the capacity of the CyberKnife to deliver non-isocentric plans. Rapid onset of pain relief after CyberKnife treatment was ﬁrst reported in a small series of patients by Romanelli and co-workers [19]. A more recent and larger multi-institutional study conﬁrmed these results [20]. In this later study, a prescribed dose of 60 to 70 Gy was delivered to a 6- to 8-mm length of the retrogasserian region of the trigeminal nerve. The median latency to pain relief was only 7 days. Initial pain control was ranked as excellent in 88% of patients, whereas three patients reported no pain relief and two experienced only a moderate reduction of pain. Although pain relief appeared durable in 78% of this cohort, half of the patients in this series eventually developed facial numbness. Because a clear relationship was observed between the length of the trigeminal nerve treated and the onset of numbness, a gradual dose and volume de-escalation was subsequently conducted. The current parameters used for trigeminal rhizotomy at Stanford include a 6-mm length of nerve and dose prescriptions of 60 Gy marginal and 75 Gy Dmax.

FIGURE 13-6. Examples of CyberKnife treatment planning for a T12 spinal metastasis from a renal cell carcinoma. Isodose distribution is overlaid on both axial (a) and coronal (b) MRI scans. In this case, the selected treatment dose was 18 Gy prescribed to the 70% isodose. Note the steep dose gradient adjacent to the spinal cord.

176

j.r. adler jr. et al.

were metastatic (108 cases), and 59% of them had been treated previously with conventional irradiation. The authors reported an improvement in pain scores in 74 of 79 patients with axial and radicular pain prior to radiosurgery. No acute radiation toxicity or new neurologic deﬁcits occurred during the median 18-month follow-up. Meanwhile, Degen et al. [23] treated 51 patients with 72 spinal tumors (58 metastatic and 14 primary) with multiple-session CyberKnife radiosurgery. After a mean follow-up of 1 year, there were signiﬁcant and durable reductions in pain scores as well as a maintenance of physical and mental quality of life measures. Side effects were mild and selflimiting. Because by nature the spinal cord is fragile, poorly vascularized, and highly sensitive to radiation damage, the ﬁndings of both of the above studies are remarkable. The beneﬁts are even more impressive when one considers that a signiﬁcant percentage of patients in these series had undergone previous conventional irradiation. Furthermore, the relative safety of this procedure at these institutions reinforces claims about the overall system accuracy of robotic radiosurgery. In management of brain tumors, radiosurgery is often used as one component of a multimodality approach. The same philosophy can also be applied to spinal lesions. As an example, CyberKnife radiosurgery has been combined with kyphoplasty to address pathologic compression fractures [24]. This is a new treatment paradigm for metastatic spinal tumors. By coupling two minimally invasive procedures, the risk of major surgical complications is lessened, treatment costs are lowered, and quality of life is improved. Even more integrated radiosurgical approaches are likely to emerge in the future that will further change the classic surgical management of many spinal lesions. One of the more important nonneoplastic applications of CyberKnife spinal radiosurgery is the treatment of intramedullary spinal cord arteriovenous malformations (SCAVMs). Alternative therapies for this speciﬁc group of AVMs, which include embolization and surgery, have only a limited role and are almost never curative. Because there are so few viable treatment options for most cases of SCAVM, spinal radiosurgery could prove a welcome new therapeutic tool. With this in mind, 21 patients with intramedullary spinal cord AVMs (11 cervical, 7 thoracic, 3 lumbar) were treated with the CyberKnife between 1997 and 2006 as part of a gradual dose escalation study at Stanford University. Preliminary ﬁndings from this experience have been reported [25]. However, for the entire series of patients, radiosurgery was delivered in one to ﬁve sessions to an average AVM volume of 1.8 cm3 using an average marginal dose of 19.5 Gy. Patients were followed with annual MRI and angiography every 3 years. Average follow-up now exceeds 2.5 years. Among the six patients studied with posttreatment angiography, AVM obliteration was partial in four and complete in two. Signiﬁcant AVM obliteration has been observed on MRI in nearly every case that was more than 1 year from radiosurgery; AVM involution appears complete in three cases and conﬁrmatory angiography is pending. Not surprisingly, the radiologic outcome to date suggests that more aggressive radiosurgical regimens (generally using fewer fractions) correlate with a higher rate of AVM obliteration. No patients suffered post-SRS hemorrhage or any signiﬁcant neurologic deterioration attributable to SRS. Despite the still-evolving radiosurgical experience with SCAVM, it seems safe to say that the CyberKnife now offers an important new treatment option for these challenging lesions.

Intrathoracic and Intraabdominal Lesions The past half decade has witnessed an explosion of interest in using ever more precise irradiation to treat lesions of the chest and abdomen. All these procedures combine noninvasive external immobilization and targeting with conventional medical linacs [26–31]. Meanwhile, a review of the preliminary stereotactic radiotherapy experience treating liver malignancies has been recently provided by Fuss and Thomas [32]. Image-guided robotic radiosurgery adds a powerful and versatile tool to this ﬁeld. In fact, because of Synchrony’s unique capabilities for tracking and correcting for respiratory motion, the CyberKnife may eventually have its biggest clinical impact in treating lesions of the chest, abdomen, and pelvis. Consistent with this idea, multiple clinical trials are now under way worldwide to assess the utility of CyberKnife ablation of lung tumors. It is important to reiterate that tumor targeting and tracking with Synchrony requires the implantation of radiopaque ﬁducials near the lesion [33, 34]. Gold seeds are preferred because they are inert, readily inserted, and easy to image with the CyberKnife’s imaging module. The feasibility of CyberKnife radiosurgical ablation for pulmonary lesions was ﬁrst investigated in 23 patients with biopsyproven lung tumors (15 primary and 8 metastatic lesions) by Whyte and co-workers [35] in a pilot study. After CT-guided percutaneous ﬁducial placement, each patient was treated with 15 Gy in a single fraction. Although radiosurgery itself was well tolerated, several patients experienced complications as a result of the implantation of ﬁducials. After limited postsurgical follow-up, radiographic progression was found only in two patients; the follow-up period ranged from 1 to 26 months (mean, 7 months). Because several local failures were subsequently observed, the radiosurgical dose for lung cancer has been gradually escalated. Le et al. [36] conducted a phase I dose-escalation study to assess the effects of single-fraction CyberKnife radiosurgery in patients with inoperable T1–2N0 non–small cell lung cancers (NSCLCs) or solitary lung metastases. Doses ranging from 15 to 30 Gy were tested in 32 patients. At 1 year, doses greater than 20 Gy prevented local progression of NSCLC in 91% of patients, whereas doses less than or equal to 20 Gy resulted in only a 54% freedom from local progression. However, higher doses (25 to 30 Gy) were also associated with serious complications in patients with prior radiotherapy and larger midline tumors. NSCLCs responded better than metastatic tumors. The authors concluded that single-fraction CyberKnife was feasible and effective, but in selected cases, single fractions of 25 Gy or more may be unacceptably toxic. These results are consistent with those emerging during the past decade showing that high-dose stereotactic radiotherapy can be effective for NSCLC, but they also suggest that a hypofractionated approach may be required to minimize complications. The treatment of pancreatic adenocarcinomas with CyberKnife radiosurgery was ﬁrst reported by Romanelli et al. [19]. Twelve patients were treated with 15, 20, or 25 Gy delivered as a single fraction. The treatment was well tolerated, values of the pancreatic cancer marker CA 19–9 were decreased in most patients, and all patients with pain prior to treatment experienced improvement within days. These preliminary results were expanded upon by Koong and associates [37], who observed local control of tumor growth with a single dose of

13.

cyberknife radiosurgery

25 Gy, without serious gastrointestinal toxicity. In a later paper [38], CyberKnife radiosurgery was delivered as a boost after conventional intensity-modulated radiotherapy (IMRT; 45 Gy delivered in 1.8-Gy fractions) and concurrent chemotherapy (5-ﬂuorouracil or capecitabine). Although high rates of tumor control were achieved (15 of 16 patients were free from local progression until death), the incidence of complications increased signiﬁcantly above that observed with radiosurgery alone. In addition, there was no improvement in survival. This ﬁnding encouraged the authors to subsequently forego IMRT in such patients in favor of radiosurgery alone, with or without adjuvant chemotherapy. Although CyberKnife radiosurgery has proved highly effective in achieving local control and palliating symptoms, survival continues to be largely dictated by the frequent occurrence of regional metastases. Nevertheless, control of pancreatic tumors itself is an impressive accomplishment; if such an effect could be combined with more effective systemic therapies, noninvasive radiosurgical ablation holds out the promise of prolonging survival. Although preliminary, the above results with abdominal and thoracic tumors have been encouraging. Nevertheless, much more evidence is needed to establish a deﬁnite role for robotic radiosurgery in managing nonneurologic disorders. In this regard, multiple clinical trials are under way worldwide that seek to discover if CyberKnife radiosurgery can substantially impact the overall clinical outcome for a broad range of extracranial indications.

Future Directions Stated simply, the CyberKnife was developed to enable the safe, accurate, and effective application of radiosurgery throughout the body. Future enhancements in the CyberKnife will be largely dictated by the experiences of ongoing clinical studies as well as the creativity of medical professionals in radiosurgery. Nevertheless, certain improvements seem likely. For example, the radiation output of the CyberKnife linac is likely to increase. In addition to shortening the length of treatment, this development should enable progressively smaller collimators to create ever more conformal treatment volumes. Given the relative youth of the CyberKnife concept, one can readily envision how a number of similar subsystems will be optimized with time. Fiducial targeting of lesions within the major body cavities is both robust and extremely accurate. However, the implantation process can be technically demanding, adds to the complexity of the overall radiosurgical procedure, and is modestly invasive. In the case of lung treatment, pneumothorax is not uncommon. Although the challenges are considerable, a technology for targeting intraabdominal and thoracic lesions without ﬁducials will be a clear improvement in robotic radiosurgery. Recent research suggests that motion compensation without implanted gold markers may be clinically practical [39]. Such technology would considerably simplify treatment protocols and enhance patient comfort and safety. The proposed method for ﬁducial-less targeting extends the current CyberKnife X-ray image correlation targeting system by also incorporating breathing motion into the pretreatment DRR library. Using four-dimensional imaging and image registration, several pretreatment CT scans are acquired at different stages of the respiratory cycle. A large array of DRRs are

177

then computed from each CT volume. In effect, each of these synthetic images incorporates both patient and respiratory movement. Similar to the current process, throughout radiosurgery live X-ray images are compared with the precalculated DRRs. However, the new comparison group would include synthetic X-rays that also reﬂect a range of respiratory states. In performing two-dimensional–two-dimensional (2D-2D) registrations, the best matching DRR is identiﬁed, thereby enabling the current phase of the breathing cycle and the location of the tumor to be determined. With dynamic motion compensation (i.e., Synchrony), the treatment beams move synchronously with the target, and the beam strikes the target region approximately as planned. However, some tumors deform or rotate with respiration, and tissue surrounding the tumor may exhibit a different motion pattern than the tumor itself. Information on the relative motion of organs with respect to treatment beams (during the actual radiosurgery) can be incorporated into the planning process, thus further improving the overall precision and safety of the treatment. This could be of particular beneﬁt when treating lesions close to the spinal cord where the precise dose tolerance can be critical. Current robotic systems use only cylindrical collimators. However, multileaf collimators in combination with robotic systems could potentially further reduce treatment time and improve treatment conformality. Given the ﬂexibility with which the CyberKnife can conﬁgure treatment beams, a relatively simple multileaf collimator might be utilized instead of the more standard microleafed device used in conventional radiation therapy. On the other hand, the same improvement in time efﬁciency might also be achieved by further increasing linac output, which unlike multileaf collimation would not sacriﬁce beam penumbra. Whether taken alone or in aggregate, these technical improvements may in turn usher in new clinical radiosurgical applications that have yet to be envisioned. Although future technological developments have the potential to improve the process and clinical outcome of robotic radiosurgery, we are equally excited by the prospects for improving our basic radiobiological understanding of large-fraction irradiation. A number of important questions remain unanswered. The biggest of these continues to be what are the optimal doses and fractionation schemes for speciﬁc tumor entities, particularly within the thoracic and abdominal cavities. Furthermore, the combination of radiosurgery with new systemic immunotherapies and chemotherapies, which attack microscopic malignancies, and may in fact be radiation sensitizers, are likely to play a vital role for improving future clinical outcomes.

Conclusion Radiosurgery is in the midst of a technological revolution. The recent introduction of image guidance and robotic delivery has dramatically expanded the scope of this ﬁeld. As inevitable improvements in neuroimaging and computer technology emerge over the coming years, they will serve as an impetus for further improvements in robotic radiosurgical technology. This perspective stems in large part from the inherent, and somewhat unique, ﬂexibility of the CyberKnife’s basic design. As the full extent of this vision is realized, the concept of radiosurgical ablation will continue to expand into new anatomic

178

j.r. adler jr. et al.

regions and disorders. This evolution will also require that surgeons of nearly all stripes and radiation oncologists reexamine some of the basic tenets encompassed within their respective disciplines.

References 1. Chang SD, Main W, Martin DP, et al. An analysis of the accuracy of the CyberKnife: a robotic frameless stereotactic radiosurgical system. Neurosurgery 2003; 52:140–146; discussion 146–147. 2. Yu C, Main W, Taylor D, et al. An anthropomorphic phantom study of the accuracy of CyberKnife spinal radiosurgery. Neurosurgery 2004; 55:1138–1149. 3. Ho AK, Fu D, Cotrutz C, et al. A study of the accuracy of CyberKnife spinal radiosurgery using skeletal structure tracking. Neurosurgery 2007; 60(2 Suppl 1):ONS147–156. 4. Gierga DP, Chen GT, Kung JH, et al. Quantiﬁcation of respiration-induced abdominal tumor motion and its impact on IMRT dose distributions. Int J Radiat Oncol Biol Phys 2004; 58: 1584–1595. 5. Kaus MR, Netsch T, Kabus S, et al. Estimation of organ motion from 4D CT for 4D radiation therapy planning of lung cancer. Presented at Medical Image Computing and Computer-Assisted Intervention—MICCAI 2004, 7th International Conference, Saint-Malo, France, September 26–29, 2004. 6. Langen KM, Jones DT. Organ motion and its management. Int J Radiat Oncol Biol Phys 2001; 50:265–278. 7. Mageras GS, Pevsner A, Yorke ED, et al. Measurement of lung tumor motion using respiration-correlated CT. Int J Radiat Oncol Biol Phys 2004; 60:933–941. 8. Plathow C, Ley S, Fink C, et al. Analysis of intrathoracic tumor mobility during whole breathing cycle by dynamic MRI. Int J Radiat Oncol Biol Phys 2004; 59:952–959. 9. Shirato H, Seppenwoolde Y, Kitamura K, et al. Intrafractional tumor motion: lung and liver. Semin Radiat Oncol 2004; 14:10–18. 10. Webb S. Conformal intensity-modulated radiotherapy (IMRT) delivered by robotic linac—testing IMRT to the limit? Phys Med Biol 1999; 44:1639–1654. 11. Webb S. Conformal intensity-modulated radiotherapy (IMRT) delivered by robotic linac—conformality versus efﬁciency of dose delivery. Phys Med Biol 2000; 45:1715–1730. 12. Li JG, Xing L. Inverse planning incorporating organ motion. Med Phys 2000; 27:1573–1578. 13. Unkelbach J, Oelfke U. Incorporating organ movements in inverse planning: assessing dose uncertainties by Bayesian inference. Phys Med Biol 2005; 50:121–139. 14. Schlaefer A, Fisseler J, Dieterich S, et al. Feasibility of fourdimensional conformal planning for robotic radiosurgery. Med Phys 2005; 32:3786–3792. 15. Adler JR Jr, Gibbs IC, Puataweepong P, Chang SD. Visual ﬁeld preservation after multisession CyberKnife radiosurgery for perioptic lesions. Neurosurgery 2006; 59(2):244–254. 16. Mehta VK, Lee QT, Chang SD, et al. Image guided stereotactic radiosurgery for lesions in proximity to the anterior visual pathways: a preliminary report. Technol Cancer Res Treat 2002; 1:173–180. 17. Pham CJ, Chang SD, Gibbs IC, et al. Preliminary visual ﬁeld preservation after staged CyberKnife radiosurgery for perioptic lesions. Neurosurgery 2004; 54:799–810; discussion 810–812. 18. Chang SD, Gibbs IC, Sakamoto GT, et al. Staged stereotactic irradiation for acoustic neuroma. Neurosurgery 2005; 56:1254– 1261; discussion 1261–1253. 19. Romanelli P, Heit G, Chang SD, et al. CyberKnife radiosurgery for trigeminal neuralgia. Stereotact Funct Neurosurg 2003; 81: 105–109.

20. Lim M, Villavicencio AT, Burneikiene S, et al. CyberKnife radiosurgery for idiopathic trigeminal neuralgia. Neurosurg Focus 2005; 18:E9. 21. Ryu S, Fang Yin F, Rock J, et al. Image-guided and intensitymodulated radiosurgery for patients with spinal metastasis. Cancer 2003; 97:2013–2018. 22. Gerszten PC, Ozhasoglu C, Burton SA, et al. CyberKnife frameless stereotactic radiosurgery for spinal lesions: clinical experience in 125 cases. Neurosurgery 2004; 55:89–98; discussion 98–99. 23. 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; 2:540–549. 24. Gerszten PC, Germanwala A, Burton SA, et al. Combination kyphoplasty and spinal radiosurgery: a new treatment paradigm for pathological fractures. J Neurosurg Spine 2005; 3:296–301. 25. Sinclair J, Chang SD, Gibbs IC, Adler JR Jr. Multisession CyberKnife radiosurgery for intramedullary spinal cord arteriovenous malformations. Neurosurgery 2006; 58:1081–1089; discussion 1081–1089. 26. Bilsky MH, Yamada Y, Yenice KM, et al. Intensity-modulated stereotactic radiotherapy of paraspinal tumors: a preliminary report. Neurosurgery 2004; 54:823–830; discussion 830–821. 27. Herfarth KK, Debus J, Lohr F, et al. Stereotactic single-dose radiation therapy of liver tumors: results of a phase I/II trial. J Clin Oncol 2001; 19:164–170. 28. Shiu AS, Chang EL, Ye JS, et al. Near simultaneous computed tomography image-guided stereotactic spinal radiotherapy: an emerging paradigm for achieving true stereotaxy. Int J Radiat Oncol Biol Phys 2003; 57:605–613. 29. Timmerman R, Papiez L, McGarry R, et al. Extracranial stereotactic radioablation: results of a phase I study in medically inoperable stage I non-small cell lung cancer. Chest 2003; 124: 1946–1955. 30. Uematsu M, Shioda A, Suda A, et al. Computed tomographyguided frameless stereotactic radiotherapy for stage I non-small cell lung cancer: a 5-year experience. Int J Radiat Oncol Biol Phys 2001; 51:666–670. 31. Yenice KM, Lovelock DM, Hunt MA, et al. CT image-guided intensity-modulated therapy for paraspinal tumors using stereotactic immobilization. Int J Radiat Oncol Biol Phys 2003; 55: 583–593. 32. Fuss M, Thomas CR Jr. Stereotactic body radiation therapy: an ablative treatment option for primary and secondary liver tumors. Ann Surg Oncol 2004; 11:130–138. 33. Schweikard A, Glosser G, Bodduluri M, et al. Robotic motion compensation for respiratory movement during radiosurgery. Comput Aided Surg 2000; 5:263–277. 34. Schweikard A, Shiomi H, Adler J. Respiration tracking in radiosurgery. Med Phys 2004; 31:2738–2741. 35. Whyte RI, Crownover R, Murphy MJ, et al. Stereotactic radiosurgery for lung tumors: preliminary report of a phase I trial. Ann Thorac Surg 2003; 75:1097–1101. 36. Le QT, Loo BW, Ho A, et al. Results of a phase I dose-escalation study using single-fraction stereotactic radiotherapy for lung tumors. J Thorac Oncol. 2006 Oct; 1(8):802–809. 37. Koong AC, Le QT, Ho A, et al. Phase I study of stereotactic radiosurgery in patients with locally advanced pancreatic cancer. Int J Radiat Oncol Biol Phys 2004; 58:1017–1021. 38. Koong AC, Christofferson E, Le QT, et al. Phase II study to assess the efﬁcacy of conventionally fractionated radiotherapy followed by a stereotactic radiosurgery boost in patients with locally advanced pancreatic cancer. Int J Radiat Oncol Biol Phys 2005; 63:320–323. 39. Schweikard A, Shiomi H, Adler JR. Respiration tracking in radiosurgery without ﬁducials. Int J Med Robotics Comput Assist Surg 2005; 1:19–27.

PA R T IV

Treatment of Disease Types

179

1 4

Brain Metastases John H. Suh, Gene H. Barnett, and William F. Regine

Introduction Brain metastases are a signiﬁcant cause of morbidity and mortality that affects an estimated 20% to 30% of cancer patients [1]. The annual incidence in the United States has been estimated to be as high as 170,000 cases [2]. Brain metastases most commonly originate from tumors of the lung, breast, renal cells, and colon and from melanoma. In some cases, the primary is unknown. Multiple lesions are seen more frequently with melanoma and lung cancer, whereas single lesions are more common with breast, colon, and renal cell cancer. An estimated 30% to 40% of patients with a brain metastasis present with a single lesion [3]. Despite improvements in imaging, surgical technique, and radiation delivery, the prognosis for these patients remains dismal. Patients who receive no treatment have a median survival of only 1 month. Those who are treated with steroids survive a median of 1 to 2 months, and those who undergo whole-brain radiation therapy survive for a median of 4 to 7 months after therapy [4–7]. As a result, the development of brain metastasis is one of the most feared complications of cancer patients, and it represents a therapeutic challenge to neurosurgeons, radiation oncologists, medical oncologists, and neuro-oncologists. Historically, the treatment of brain metastasis has included whole-brain radiation therapy (WBRT), which was ﬁrst reported in the 1950s and early 1960s [8, 9]. Response to WBRT is best among patients with small cell lung cancer, lymphoma, and germ cell tumors. The Radiation Therapy Oncology Group (RTOG) conducted numerous trials from 1971 through 1993 to investigate various fractionation schemes and doses of WBRT, which are listed in Table 14-1 [10–18]. In addition, the RTOG has investigated the use of hyperfractionation and radiation sensitizers such as bromodeoxyuridine and found no improvement in overall survival. Based on these studies, the use of 3000 cGy in 10 fractions became a popular fractionation scheme to consider for patients with brain metastases. Although neurologic symptoms improved in the majority of patients, local control was low resulting in neurologic death in 25% to 54% of patients [10]. Given the poor prognosis for patients with brain metastases, alternative strategies to improve outcomes have been explored. Stereotactic radiosurgery (SRS) is a technique that

delivers a single, high dose of ionizing radiation using stereotactically directed narrow beams to small intracranial targets while sparing the surrounding brain tissue [19]. Since Sturm’s initial report of 12 lesions treated on a modiﬁed linear accelerator, a number of papers have corroborated the beneﬁt of SRS in newly diagnosed and recurrent brain metastases [20]. Currently, brain metastases represent the most common indication for SRS. Although SRS has become an important treatment option, its use has also sparked controversy about the appropriate treatment for patients with brain metastases. This chapter will review the prognostic factors, surgical results, rationale for SRS, results of surgery compared with SRS, the institutional results of SRS with WBRT and SRS alone, the results of completed phase II/III clinical trials, complications of SRS, and future direction of SRS for brain metastases.

Prognostic Factors The most commonly used prognostic scale for patients with brain metastasis is the RTOG recursive partitioning analysis (RPA) [21]. This scale divides patients into three classes based on 1200 consecutive patients enrolled in three RTOG trials from 1979 to 1993. The vast majority of these patients had unresectable and/or multiple metastases but received standard doses of WBRT. Table 14-2 lists the various classes and their components. The most important factors include extracranial metastases, patient age, Karnofsky performance status, and control of primary tumor. A study from Lagerwaald and colleagues reviewed 1292 patients with brain metastases [22]. In this Dutch study, lung cancer was the most common primary disease (56%). Median survival was 3.4 months, and the 1-year and 2-year survivals were 12% and 4%, respectively. The most important factors were treatment modality, performance status, extracranial disease burden, and response to steroid treatment.

Surgery Given the low local control rates associated with WBRT alone, surgical removal of tumors—particularly single or symptomatic lesions—was explored in hopes of improving local control and

181

182

j.h. suh et al.

TABLE 14-1. Prospective Radiation Therapy Oncology Group (RTOG) brain metastases studies (1971–1993). Protocol

Year

No. patients

Total dose/no. fractions

Median survival

RTOG 6901 [10]

1971–1973

RTOG 7361 [11]

1973–1976

RTOG 6901 [12] RTOG 7361 [12] Ultrarapid RTOG 7606 [13] Favorable patients RTOG 8528 [14]

1971–1973 1973–1976

233 217 233 227 447 228 227 26 33

30 Gy/10 Fx/2 weeks 30 Gy/15 Fx/3 weeks 40 Gy/15 Fx/3 weeks 40 Gy/20 Fx/4 weeks 20 Gy/5 Fx/1 week 30 Gy/10 Fx/2 weeks 40 Gy/15 Fx/3 weeks 10 Gy/1 Fx/1 day 12 Gy/2 Fx/2 days

21 weeks 18 weeks 18 weeks 16 weeks 15 weeks 15 weeks 18 weeks 15 weeks 13 weeks

RTOG 9104 [15]

1991–1995

RTOG 7916 [16] Misonidazole

1979–1983

130 125 30 53 44 36 213 216 193 200 196 190

18 weeks 17 weeks 4.8 months 5.4 months 7.2 months 8.2 months 4.5 months 4.5 months 4.5 months 4.1 months 3.1 months 3.9 months

RTOG 8905 [17] BrdU

1989–1993

30 Gy/10 Fx/2 weeks 50 Gy/20 Fx/4 weeks 48 Gy/1.6 Gy bid 54.5 Gy/1.6 Gy bid 64 Gy/1.6 Gy bid 70.4 Gy/1.6 Gy bid 30 Gy/10 Fx 54.4 Gy/1.6 Gy bid 30 Gy/10 Fx/2 weeks 5 Gy/6 Fx/3 weeks 30 Gy/10 Fx + Miso 5 Gy/6 Fx + Miso 37.5 Gy/15 Fx/3 weeks 37.5 Gy/15 + BrdU

1976–1979 1986–1989

36 34

6.1 months 4.3 months

Fx, fraction; bid, twice daily; Miso, misonidazole; BrdU, bromodeoxyuridine. Source: Adapted from Sneed PK, Larson DA, Wara WM. Neurosurg Clin N Am 1996; 7:505–515.

TABLE 14-2. RTOG RPA classes for brain metastases. Factors

Median survival (months)

Class I

Age <65 and KPS ≥70 and Controlled primary tumor and Extracranial metastases

7.1

Class II

KPS ≥70 and Age ≥65 and/or Uncontrolled primary and/or Extracranial metastases

4.2

Class III

KPS ≤60

2.3

Source: 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; 37:745–751.

survival. Surgery can also establish or conﬁrm the diagnosis, especially for patients without primary tumor conﬁrmation. In addition, surgery can provide immediate and effective palliation of symptomatic mass effect. Three randomized trials comparing WBRT with surgery followed by WBRT for patients with single metastasis have been completed and are summarized in Table 14-3 [23–25]. These trials were based on the premise that improved local control of a single brain metastasis would result in improved survival. Patchell’s study demonstrated that surgery and WBRT improved survival (40 vs. 15 weeks, p < 0.01), local control (80% vs. 48%, p < 0.02), and functional independence (38 vs. 8 weeks, p < 0.005) compared with biopsy and WBRT [23]. Noordijk’s study included 63 patients with CT-conﬁrmed single metastasis who were randomized to surgery and WBRT or WBRT alone [24]. Surgery and WBRT improved survival and functional independence more than WBRT alone (10 vs. 6 months, p = 0.04 and 7.5 vs. 3.5 months, p = 0.06, respectively). Patients with

TABLE 14-3. Phase III trials of WBRT versus surgery and WBRT. Author

Year

Surgery

WBRT Dose (cGy)/no. fractions/ no. weeks

N

MS (weeks)

KPS ≥70 (weeks)

CNS death

Local control

Patchell [23]

1990

Noordijk [24]

1994

Mintz [25]

1996

Yes No Yes No Yes No

3600 / 12 / 2.5 weeks 3600 / 12 / 2.5 weeks 4000 / 20 / 2 weeks 4000 / 20 / 2 weeks 3000 / 10 / 2 weeks 3000 / 10 / 2 weeks

25 23 32 31 41 43

40 15 43 26 22 25

38 8 33 15 N/A* N/A*

29% 52% 35% 33% 14% 28%

80% 48% N/A N/A N/A N/A

N, number of patients; MS, median survival; N/A, not applicable; N/A*, the mean proportion of days that the patients had KPS ≥70 was not different.

14.

brain metastases

active systemic disease did not beneﬁt from surgery. The third trial, which was from Mintz, did not demonstrate a survival beneﬁt for patients undergoing surgery (6.3 vs. 5.6 months, p = 0.24) [25], which was most likely a result of two factors: (1) the study consisted of a large percentage of patients with active systemic disease and (2) patients with lower baseline Karnofsky performance status (KPS) were included.

Rationale for Using SRS for Brain Metastases Brain metastases have features that make them ideal targets for SRS. The majority of these tumors are pseudospherical in shape, are located in the gray-white junction, have a maximum tumor diameter of less than 4 cm, and, unlike primary gliomas, are noninﬁltrative. These features also allow for accurate target delineation, planning, and treatment delivery. In addition, a single large fraction of radiation appears to have an equal effect in all tumor types, even among the “radioresistant” tumors such as renal cell carcinoma and melanoma [26–28] (Case Study 14-1). The use of SRS alone or as an adjunct to WBRT is based on the premise that improved local control will reduce morbidity, improve quality life, and prolong survival. Nieder reviewed the computed tomography (CT) scans of 332 patients with brain metastases to evaluate local control and time to local progression based on dose [29]. A biologically effective dose (BED) using the linear quadratic model with an alpha/beta of 10 was derived for each patient. The partial response rates signiﬁcantly improved for patients with higher BED suggesting that higher doses were needed to improve local control. As mentioned, two

183

Case Study 14-1 This is a 73-year-old male with a history of renal cell carcinoma s/p right nephrectomy 10 years ago. He presented with left-sided weakness. MRI was obtained, which revealed a 1.7-cm right frontal lesion (Fig. 14-1a, b). Staging workup was negative. He underwent Gamma Knife radiosurgery to the right frontal lesion (Fig. 14-1c): 2400 cGy was prescribed to the 50% local isodose line, which covered 100% of the target. The plan utilized 8 shots using an 8-mm helmet without plugs. Tumor volume = 2.0 cm3. Maximum dose = 4800 cGy. Maximum diameter = 1.9 cm. MD/PD = 2.00. PIV/TV = 1.70. MRI 3 months after Gamma Knife radiosurgery (Fig. 14-1d, e) shows the patient’s left-sided weakness improved.

phase III trials comparing WBRT versus surgical resection and WBRT demonstrated that improved local tumor control of a single metastasis led to improved survival. Given the potential advantages of SRS, interest developed in using SRS rather than surgery to improve local control of brain metastasis. Normal brain tolerance to radiation therapy is related to total dose, fractional dose, elapsed time of radiation delivery, volume of normal brain irradiated, and the amount of and interval from exposure of prior radiation therapy [30]. From a radiobiologic standpoint, brain metastases are considered category IV targets, which are targets with early-responding tissue

FIGURE 14-1. MRI of male with a history of renal cell carcinoma. See Case Study 14-1.

184

j.h. suh et al.

surrounded by late-responding normal tissue [31]. Because the tumor cells are within the target, one would expect a high therapeutic index for these targets given the steep dose gradient associated with SRS. Unlike surgery, SRS has the potential to sterilize cells that may not be contained within the high-dose region.

Surgery Versus SRS Based on results of surgical resection for single metastasis, investigators became interested in evaluating the use of SRS for brain metastases. This interest was based on the potential advantages of SRS, which include outpatient delivery, elimination of general anesthesia, minimal risk for bleeding or infection, lower costs, short convalescence time, and lower risk for complications. In addition, SRS is not limited by location, number of lesions, comorbidity, or coagulopathy. For these various reasons, SRS has become a viable alternative to surgery given the potential cost savings and quality-of-life beneﬁts [32, 33]. A multi-institutional retrospective review from the University of Wisconsin at Madison, University of Florida at Gainesville, and Joint Center for Radiation Therapy included 122 patients with newly diagnosed, potentially resectable single brain metastases treated by SRS and WBRT [34]. The local control rate was 86%. The median survival was 56 weeks with 25% of deaths related to brain tumor progression. Based on results of this retrospective review, the investigators concluded that SRS could be used in place of surgery for patients with brain metastasis. Another study from the M.D. Anderson Cancer Center compared 13 patients treated by SRS with 62 patients treated by surgery [35]. These patients were retrospectively matched for important prognostic factors. The median survival was 7.5 months for the radiosurgery group and 16.4 months for the surgery group. The local recurrence for SRS was 21% versus 8% for surgery. Based on these results, they concluded surgery should be the preferred treatment for surgically accessible lesions. A study from the Mayo Clinic compared the outcomes of radiosurgery versus surgery [36]. This retrospective review included patients with tumors that would have allowed for either type of treatment. No difference in 1-year survival was noted between the groups. A retrospective study from Muacevic compared surgery and radiotherapy with Gamma Knife radiosurgery for patients with a single tumor 3.5 cm or less in diameter [37]. The 1-year survival rates (53% vs. 43%, p = 0.19), 1-year local control rates (75% vs. 83%, p = 0.49), and 1-year neurologic death rates (37% vs. 39%, p = 0.8) for the surgical group and SRS group, respectively, were not statistically different. Because the previously described studies are all retrospective, patient selection bias was probably the key contributor to the mixed results and controversy regarding the beneﬁts of surgery versus SRS. The Joint Center for Radiation Therapy attempted to enroll patients onto a phase III trial comparing surgery to SRS in the 1990s. After 3 years, only six patients were enrolled because of patient or physician preference [38]. M.D. Anderson Cancer Center is currently performing a randomized

trial of surgery versus radiosurgery for patients with a single brain metastasis who are considered eligible for either treatment. Primary end points are survival and local control. This study was opened in 1997. For patients with resectable tumors larger than 3.5 cm or those without histologic conﬁrmation of a primary tumor, most would agree that surgical resection should be considered rather than radiosurgery. Surgical removal of the tumor offers faster improvement in neurologic function and minimizes the use of chronic steroid use. Prolonged use of steroids can result in a number of problems including diabetes, proximal muscle weakness, peripheral edema, psychosis, susceptibility to infections, and gastrointestinal perforation [39]. Surgery is also more effective in alleviating mass effect.

Treatment of Recurrent Metastatic Disease One of the initial reports of SRS for brain metastases consisted of patients with recurrent brain metastases after WBRT or surgery. Loefﬂer reported the Joint Center for Radiation Therapy (JCRT) retrospective results of 18 patients with recurrent metastases after previous surgery, WBRT, or both [40]. The majority of patients improved neurologically with local control being achieved in all patients (Case Study 14-2). A follow-up report from Alexander consisting of mostly patients with recurrent brain metastases reported tumor control rates of 85% at 1 year and 65% at 2 years, although the rate was lower for recurrent lesions [41]. Median survival was 9 months after SRS. Another series from the University of Cincinnati reported the results of 84 patients with 1 to 6 lesions recurrent after WBRT [42]. Median survival was 43 weeks from SRS. Median time to local failure was 35 weeks for all lesions and 52 weeks

Case Study 14-2 This is a 57-year-old female with history of a T3N1M0, stage III adenocarcinoma of the esophagus s/p radiation therapy (3000 cGy/15 fractions, 150 cGy/fraction twice daily) to the distal esophagus with concurrent chemotherapy s/p esophageal resection followed by an additional 3000 cGy/15 fractions, 150 cGy/fraction twice daily to the tumor bed with concurrent chemotherapy. She did well for 10 months until she presented with a right-sided seizure. MRI revealed a 2-cm left frontoparietal lesion (Fig. 14-2a, b). She underwent WBRT (3750 cGy/15 fractions using 250 cGy) followed by Gamma Knife radiosurgery within 1 week of completing WBRT (Fig. 14-2c). Treatment planning information: 1800 cGy was prescribed to the 50% isodose line, which covered 100% of the target. The plan utilized 6 shots using an 18-mm helmet without plugs. Tumor volume = 5.3 cm3. Maximum dose = 36 Gy. Maximum diameter = 2.5 cm. MD/PD = 2.00. PIV/TV = 1.64. Film 3 months after Gamma Knife radiosurgery (Fig. 14-2d, e); she had excellent response to WBRT and Gamma Knife radiosurgery.

14.

brain metastases

185

FIGURE 14-2. MRI of female with a history of adenocarcinoma of the esophagus. See Case Study 14-2.

for lesions treated with greater than or equal to 18 Gy. A study from Noël of 54 consecutive patients reported 1- and 2-year local control rates of 91.3% and 84% and 1- and 2-year brain control rates of 65% and 57%, respectively [43]. Unfortunately, no prospective study has evaluated the use of SRS with other treatment options such as surgery, conformal radiation therapy, or chemotherapy for patients with recurrent brain metastases after WBRT.

Radiosurgery with WBRT Based on the results of SRS for recurrent brain metastasis, the use of SRS with WBRT was investigated for newly diagnosed patients with the hypothesis that the combination of WBRT with SRS would improve local and regional control of brain metastases. A number of the institutional studies have demonstrated high local control rates with the addition of SRS to WBRT [41, 44–52]. These local control rates range from 63% to 97% versus 42% to 87% with use of SRS alone and are summarized in Table 14-4 [26, 50, 52–54].

TABLE 14-4. Retrospective studies of SRS versus SRS and WBRT (local tumor control).

Flickinger [26] Chidel [50] Shehata [52]* Pirzkall [53] Sneed [54]

SRS alone (%)

SRS+WBRT (%)

p value

47 52 87 72 42

82 80 97 86 63

0.008 0.034 0.0001 0.13 0.008

*For metastases less than or equal to 2 cm in diameter.

Shehata evaluated the optimal SRS dose and inﬂuence of WBRT on tumor control among 160 patients with 468 recurrent and newly diagnosed metastases ≤2 cm in diameter treated between October 1992 and May 2001 [52]. On multivariate analysis, the most important factor for local tumor control was the addition of WBRT to SRS (97% vs. 87% for those who did not undergo WBRT; p = 0.001). For patients undergoing WBRT, the use of doses >20 Gy resulted in higher-grade 3 and 4 neurotoxicity (5.9% vs. 1.9% for those receiving <20 Gy) and did not result in better local control. Thus, the authors recommended using 20 Gy for metastases ≤2 cm in diameter with planned combined WBRT. A multi-institutional retrospective study from 10 institutions analyzed outcomes for 502 patients with newly diagnosed brain metastases treated by SRS and WBRT. Patients were stratiﬁed by the RTOG recursive partitioning analysis [55]. For all RPA classes, survival improved for the SRS and WBRT patients compared with WBRT-alone patients. Median survival was 16, 10, and 8 months for the SRS and WBRT patients versus 7.1, 4.2, and 2.3 months for the WBRT patients, respectively. Three prospective trials have been completed comparing the use of SRS with WBRT versus WBRT alone. The ﬁrst trial from the University of Pittsburgh trial enrolled 27 patients with two to four brain metastases measuring less than 2.5 cm in maximum diameter [56]. The study was stopped at an interim analysis at 60% because a signiﬁcant difference in the 1-year local failure rate was noted (8% in the SRS arm vs. 100% in the WBRT arm). The median time to local failure was 6 months for WBRT versus 36 months for the WBRT and SRS arm. Median survival was not signiﬁcantly different for the SRS and WBRT versus WBRT arm (11 months vs. 7.5 months,

186

j.h. suh et al.

respectively). The small sample size as well as possible selection and statistical bias may have distorted the results. The Brown University trial enrolled 96 patients onto a three-armed, prospective randomized trial of Gamma Knife radiosurgery, WBRT, or both [57]. Fifty-one patients underwent surgical resection of large symptomatic lesions prior to randomization, which was not evenly distributed among the treatment arms. No difference in overall survival was noted among the three study arms (7, 5, and 9 months for the SRS alone, WBRT plus SRS, and WBRT arms, respectively). The local control rates were 87%, 91%, and 62%, respectively, for the three arms. The risk for developing new brain lesions was higher for the patients who did not receive WBRT (43% vs. 19% and 23% for the WBRT groups). Unfortunately, the study had major treatment bias given the unequal distribution of patients undergoing surgical resection in the three arms. The RTOG recently reported on a large prospective trial (RTOG 9508) of 333 patients with one to three newly diagnosed brain metastases on magnetic resonance imaging (MRI), each ≤4 cm in diameter, randomized to WBRT (3750 cGy/15 fractions) and SRS versus WBRT alone [58]. Patients had a minimum KPS of 70 and were excluded if the metastases were located in the brain stem or within 1 cm from the optic apparatus. Survival signiﬁcantly improved for the patients with a single brain metastasis treated with SRS plus WBRT versus WBRT (6.5 months vs. 4.9 months, p = 0.04) despite the fact that 19% of patients failed to receive the intended treatment. On multivariate analysis, RPA class (1 vs. 2) and tumor type (squamous or non–small cell vs. other) had a statistically signiﬁcant effect on survival. Overall survival did not improve for the patients with two to three lesions although KPS at 6 months had stabilized or improved with the addition of SRS plus WBRT versus WBRT alone (43% vs. 27%, respectively; p = 0.03). Steroid use was also lower in the SRS and WBRT arm. Local control was improved at 1 year (82% vs. 71%; p = 0.01). Using a Cox model, the risk of developing local recurrence was 43% greater with WBRT alone (p = 0.0021). The type of SRS unit did not inﬂuence results. Based on the results of RTOG 9508, a phase III trial (RTOG 0320) is under way for NSCLC patients with one to three brain metastases. This trial will randomize patients to SRS plus WBRT, SRS plus WBRT and geﬁtinib, or SRS plus WBRT and temozolomide.

and Women’s Hospital of surgically staged IIIA patients with non–small cell lung cancer (NSCLC) found that the most common site of recurrence was the brain, in particular for patients with residual nodal disease after neoadjuvant therapy and nonsquamous histology who had a risk of 53% at 3 years [64]. The RTOG is currently performing a phase III trial (RTOG 0212) for patients with locally advanced NSCLC; patients are randomized to PCI (3000 cGy/15 fractions) versus observation. Several prospective trials have been performed to evaluate whether patients with brain metastases have neurocognitive deﬁcits prior to the initiation of treatment. A recent prospective phase III trial evaluated patients with brain metastasis with monthly neurocognitive testing consisting of memory, ﬁne motor speed, executive function, and global neurocognitive testing [65]. Table 14-5 lists the various neurocognitive testing that was performed for patients enrolled on this study. This trial demonstrated that 90.5% of patients had impairment of one or more neurocognitive tests at baseline and 42.4% of patients had impairment in four or more tests. Another study from RTOG demonstrated that tumor progression resulted in a signiﬁcant decline in the mini-mental status exam at 3 months that was much more pronounced for the patients enrolled on the phase III trial of 30 Gy/10 fractions versus 54.4 Gy/30 fractions using 1.6 Gy twice daily [66]. Chang reported on a prospective trial of neurocognitive testing for patients with 1 to 3 newly diagnosed brain metastases to determine if neurocognitive function was spared with SRS alone [67]. Eighteen of the 27 enrolled patients had evaluable neurocognitive function data. At baseline, 66% of patients had some degree of impairment that was related to total brain metastases volume. After SRS, acute and subacute improvements in executive function and processing speed occurred, but learning and memory skills declined. Seven to 9 months after SRS, there was a suggestion of global decline of neurocognitive function. Recently, the RTOG reported on the feasibility of performing ﬁve neurocognitive measures and administering a qualityof-life instrument in patients with brain metastases [68]. The overall compliance rate for administration and completion of these measures and instrument prior to treatment, at the completion of WBRT, and 1 month after WBRT was ≥95%, ≥84%, and ≥70%, respectively. Based on the encouraging results of

Radiosurgery Alone Despite evidence supporting the use of WBRT for patients with single brain metastasis, the use of SRS alone for treatment of brain metastases has gained popularity and has become a contentious issue. This has been mostly driven by concerns regarding quality-of-life issues and the potential side effects of WBRT, especially cognitive effects for long-term survivors [59–60]. The commonly referenced paper by DeAngelis and colleagues used fractionation schemes (400 to 600 cGy/fraction in 10/12 patients) that are used for patients with a poor prognosis; with clinical dementia limited to only those patients treated with >3 Gy per fraction [59]. In addition, some believe that patients can be effectively managed by repeat SRS rather than WBRT. Several trials of prophylactic irradiation (PCI) have demonstrated a decrease in the development of brain metastases but no survival advantage [61–63]. A recent study from Brigham

TABLE 14-5. Neurocognitive testing used for phase III brain metastases trial. Test

Neurocognitive domain

Hopkins Verbal Learning Test (recall) Hopkins Verbal Learning Test (recognition) Hopkins Verbal Learning Test (delay) Trailmaking A Trailmaking B Controlled oral word association Grooved pegboard dominant hand Grooved pegboard nondominant hand

Memory Memory Memory Executive Executive Executive Fine motor Fine motor

Source: Umsawasdi T, Valdivieso M, Chen TT, et al. Role of elective brain irradiation during combined chemoradiotherapy for limited disease non-small cell lung cancer. J Neurooncol 1984; 2:253–259.

14.

brain metastases

this study, and others of similar design, current and future developing brain tumor/metastases studies have and will more routinely incorporate baseline and follow-up neuropsychometric testing as an inherent part of study design and patient outcome evaluation. This is particularly important as it is becoming increasingly evident that neurologic/neurocognitive decline seen in cancer patients is multifactorial and not due to the effects of radiation therapy alone. This has been demonstrated in recent publications implicating the potential signiﬁcant impact of other treatment modalities such as chemotherapy and/or surgery on neurologic/neurocognitive function [69–73]. In addition to the potential toxicity concerns of WBRT, a randomized trial from Patchell that showed no survival beneﬁt to WBRT is often quoted as a reason why WBRT should not be given. This phase III randomized trial of 95 patients compared surgery alone with surgery followed by WBRT for patients with a single brain metastasis [74]. The addition of WBRT to surgery decreased the chance for neurologic death (14% vs. 44%, p = 0.003), decreased local recurrence (10% vs. 46%, p < 0.01), and decreased tumor recurrence anywhere in the brain (XX% vs. 70%, p < 0.001). The majority of patients (61%) received WBRT at time of recurrence resulting in a large crossover of the observation arm to WBRT. Survival was not different, although the study’s end point was local control and thus it was not powered to demonstrate a survival advantage [75]. It is also important to note that patients who undergo SRS alone may be at higher risk for morbidity associated with brain tumor recurrence. A retrospective study from the University of Kentucky reviewed 36 patients with newly diagnosed brain metastases—22 had a single lesion [76]. These patients were treated with SRS alone. Seventeen of the 36 (44%) patients developed recurrent brain metastases. Of the 17 patients, 12 (71%) were symptomatic and 10 (59%) had neurologic deﬁcits.

Large Institutional/Multi-Institutional Results of SRS Alone Sneed reported on a UCSF retrospective study of outcomes of patients with one to four metastases treated by SRS alone compared with WBRT plus SRS [54]. Physician preference and referral patterns inﬂuenced the use of upfront WBRT. The patients had similar median survivals (11.1 months for SRS alone vs. 11.3 months for WBRT plus SRS) and local tumor control rates (71% vs. 79%). The incidence of distant brain metastasis was, however, signiﬁcantly higher in the SRS-only group versus WBRT and SRS (72% vs. 31%). Despite the higher rate of distant brain metastases, these metastases were controlled with salvage therapies including WBRT, partial brain radiotherapy, SRS, and surgery. The authors concluded that survival was not compromised by the omission of WBRT. Another retrospective paper from Hasegawa reviewed 172 patients with brain metastases (3.5 cm or less in diameter) managed by SRS alone [77]. One hundred twenty-one patients had follow-up imaging with 80% of patients having solitary lesions. The median survival was 8 months, and the local tumor control rate was 87%. At 2 years, the local control rate, distant brain control, and total intracranial control rates were 75%, 41%, and 27%, respectively. Tumor volume signiﬁcantly

187

predicted for local control (p = 0.02) with tumor volumes of at least 4 cm3 having local control rate of 49% at 1 and 2 years. Based on these results, the authors advocated avoiding WBRT for selected patients with one or two tumors with good control of their primary cancer, better KPS (90 or higher), and younger age (<60 years old). One study prospectively evaluated the role of SRS alone for patients with up to three brain metastases measuring less than 3 cm in maximum diameter [78]. The study consisted of 101 patients with a total of 155 lesions. At 1 year, the local control, distant brain freedom from progression, and overall brain freedom from progression rates were 91%, 53%, and 51%, respectively. The best predictor for improved overall brain freedom from progression was the presence of a single lesion and a greater than 2-year interval from primary diagnosis to diagnosis of brain metastases. The overall survival stratiﬁed by the RTOG RPA class was 13.4, 9.3, and 1.5 months for class I, II, and III, respectively (p < 0.0001). Median survival was 7.6 months. These results suggested that SRS alone can provide good local control but that marked distant failure can be expected. Another study from Pirzkall reviewed their experience of 236 patients with 311 brain metastases who underwent SRS alone (158 patients) or SRS with WBRT (73 patients). No difference was noted in terms of survival or local control. The SRS with WBRT group had a trend toward better survival compared with SRS alone (15.4 vs. 8.3 months; p = 0.08) for the patients with no evidence of extracranial disease [73]. The local control and distant brain control rates were worse with SRS alone. A multi-institutional retrospective review from 10 institutions analyzed the results of 569 patients with newly diagnosed single or multiple brain metastases treated initially with SRS alone versus SRS and WBRT [79]. For all RPA classes, the survival was comparable between the two treatment groups (RPA class I, 14 vs. 15 months; RPA class II, 8 vs. 7 months; RPA class III, 5 vs. 6 months) suggesting that upfront WBRT does not improve survival better than SRS alone.

Phase II/III Trials of SRS with or Without WBRT The Eastern Cooperative Oncology Group (ECOG) phase II feasibility trial of SRS alone for renal cell carcinoma, sarcoma, and melanoma patients with one to three brain metastases was presented at the 2004 ASCO meeting [80]. The objective of the study was to determine the radiographic and neurologic progression rates at 3, 6, and 9 months. The trial enrolled 36 patients; 32 were eligible by entry criteria. The SRS dose was based on tumor size (24, 18, or 15 Gy). The median survival was 8.3 months. Progression occurred in 41% of the cases. The authors concluded that omission of WBRT was associated with a high brain failure rate and should be approached judiciously. The interim report of the Japanese Radiation Oncology Study Group’s (JROSG) phase III study of SRS versus SRS plus WBRT (3000 cGy/10 fractions) was presented at the 2004 ASCO meeting [81]. This trial randomized patients with one to four brain metastases to SRS alone (61 patients) or SRS plus WBRT (59 patients). The primary end point of the study was survival with the secondary points including cause of death, freedom from new brain metastases, KPS preservation rate, local tumor control, and late radiation morbidity. No signiﬁcant

188

j.h. suh et al.

difference in survival, cause of death, or preservation of neurologic or systemic function was noted. At 12 months, the local control was 86% for the WBRT and SRS group versus 70% for the SRS alone group (p = 0.019). The freedom from new brain metastasis was also signiﬁcantly better at 6 months for the WBRT and SRS group (82%) compared with the SRS alone group (49%) (p = 0.003). The American College of Surgical Oncology Group (ACOSOG) initiated a phase III trial (ACOSOG Z0300) in February 2002, which randomized patients with one to three brain metastases to SRS versus SRS plus WBRT (3000 cGy in 12 fractions). The primary end point of the study was to determine if overall survival was equal or better for patients undergoing SRS alone than for patients undergoing SRS and WBRT. Secondary end points included evaluation of local failure, quality of life, functional independence, neurocognitive status, and toxicities associated with each treatment arm. Patients were stratiﬁed by age (≥60 vs. <60), extracranial disease (control ≤3 months vs. >3 months), and number of lesions (1 vs. >1). The goal was to accrue 480 patients. The study is temporarily suspended secondary to poor accrual.

Complications Associated with SRS Acute Complications Acute complications during the ﬁrst week after SRS are uncommon. Some of the complications include headache after frame removal, infection of the pin sites, nausea and vomiting, seizures, transient worsening of preexisting neurologic conditions, and fatigue. The incidence of severe headaches after frame removal is low [51]. Nausea/vomiting can be minimized if the dose to the area postrema is kept below 375 cGy [39]. The risk for seizures has been reported to range from 2% to 6% [41, 42, 51]. The risk for seizures is higher for patients with cortical lesions and for those with history of seizures. For these patients, anticonvulsants should be therapeutic or considered, and the steroid dose should be increased.

Subacute Complications These complications occur within the ﬁrst 6 months after SRS. Alopecia has been reported in 5.6% of patients [41]. These patients had superﬁcial tumors that resulted in a dose of 4.4 Gy to the scalp. Neurologic deterioration can occur in some patients, which is usually treated with steroids.

Chronic Complications Radiation necrosis represents the most serious chronic complication. In general, the risk of radiation necrosis increases with higher doses, prior radiation therapy, and larger tumor volumes. This entity can be difﬁcult to distinguish from tumor recurrence on MRI and may require the use of surgery, positron emission tomography (PET), and/or magnetic resonance spectroscopy (MRS). RTOG 9005 was a phase I/II trial to determine the maximum tolerated radiosurgery dose for patients with recurrent primary brain tumors or brain metastases treated previously with

TABLE 14-6. RTOG CNS toxicity criteria used for RTOG 9005. Grade 1 Grade 2 Grade 3 Grade 4

Grade 5

Mild neurologic symptoms; No medication required Moderate neurologic symptoms; Outpatient medication required (e.g., steroids) Severe neurologic symptoms; Outpatient or inpatient medication required Life-threatening neurologic symptoms (e.g., uncontrolled seizure, paralysis, coma); includes clinically and radiographically suspected radiation necrosis and histologically proven radiation necrosis at time of operation Death

Source: Hasegawa T, Kondziolka D, Flickinger JC, et al. Brain metastases treated with radiosurgery alone: an alternative to whole brain radiotherapy? Neurosurgery 2003; 52:1318–1326.

fractionated radiation therapy [82]. In this trial, the maximum tolerated doses were inversely correlated with the maximum tumor diameter. The doses were 24 Gy for a tumor ≤20 mm in diameter, 18 Gy for a tumor 21 to 30 mm in diameter, and 15 Gy for a tumor 31 to 40 mm in diameter. Of note, investigators were reluctant to escalate the dose for tumors ≤20 mm in diameter even though the maximum tolerated dose was not reached. The rate of radiation necrosis was 5%, 8%, 9%, and 11% at 6, 12, 18, and 24 months after SRS, respectively. Other factors that predicted for grades 3 to 5 neurotoxicity were tumor dose and KPS. Table 14-6 lists the RTOG CNS toxicity criteria used for RTOG 9005. The incidence of complications was also associated with tumor dose and KPS. The study also reported on physics and quality control assessments (MD/PD ratio, a measure of dose homogeneity, and PIV/TV ratio, a measure of conformity of the treated volume relative to target volume) that are useful for all tumors treated by SRS [83]. Valery reported on 377 patients with 760 lesions treated with linear accelerator–based SRS. Seven patients had severe complications including nine patients who developed radiation necrosis. The median tumor volume was 4.9 cm3. The median prescribed tumor dose was 15.6 Gy. The only factor that inﬂuenced the risk for radiation necrosis was the conformality index [84]. The general management strategy for radiation necrosis is to decrease the edema and necrosis. Usually, high doses of steroids are used to minimize neurologic deterioration. If the patient becomes symptomatic despite steroids, surgical resection can be considered. In some cases, hyperbaric oxygen has been used for patients who are poor surgical candidates, have multiple areas of radiation necrosis, or have a surgically inaccessible lesion. In a retrospective review of 40 patients, 90% reported subjective improvement and 80% had objective neurologic improvement [85]. Steroids were discontinued or decreased in two thirds of patients. A randomized trial is under way at the University of Cincinnati comparing surgery with hyperbaric oxygen.

Future Directions Based on the results of RTOG 9508, a randomized trial, RTOG 0525, is ongoing for NSCLC patients with one to three brain metastases. The primary end point of the study is survival.

14.

brain metastases

Another phase II study (PCYC-0224) is evaluating the use of motexaﬁn gadolinium, a radiation sensitizer, with WBRT (37.5 Gy in 15 fractions) with SRS for patients with one to four metastases. Others are incorporating newer imaging modalities such as MRS and PET to aid in planning or to deliver higher doses to certain regions of the target [86–87].

Conclusion Stereotactic radiosurgery has an important role in the management of brain metastases. Retrospective studies have shown the beneﬁts of SRS as sole treatment, as an adjunct to WBRT, and as salvage treatment after WBRT in some patients. Retrospective studies suggest that patients who undergo SRS or surgery have comparable outcomes. The results of RTOG 9508 provide level 1 evidence of a survival beneﬁt of SRS and WBRT compared with WBRT alone for patients with a single lesion. For patients with two to three lesions, the use of SRS and WBRT can be considered based on performance status, extent and activity of extracranial disease, and steroid use. The omission of WBRT has been driven by the concerns of the potential risks of WBRT and apparent lack of survival beneﬁt of WBRT rather than evidence from prospective clinical studies. The potential side effects of WBRT need to be balanced with the risk for neurologic and neurocognitive decline of uncontrolled brain metastases and the additional cost of more frequent scans. Recent phase III trials have shown that many patients with brain metastases have neurocognitive deﬁcits prior to WBRT. As the prognosis improves for cancer patients, the challenge of improving survival while limiting acute and long-term side effects will continue to make the management of brain metastases controversial. We encourage the participation of patients with brain metastases in clinical trials to improve outcomes and to help answer many important questions about management of this very common disease.

References 1. Wen PY, Loefﬂer JS. Management of brain metastasis. Oncology (Huntingt) 1999; 13(7):941–954. 2. Greenberg H, Chandler WF, Sandler HM. Brain metastases. In Greenberg H, Chandler WF, Sandler HM, eds. Brain Tumors. Oxford: Oxford University Press, 1999:299–317. 3. Lohr F, Pirzkall A, Hof H, et al. Adjuvant treatment of brain metastases. Semin Surg Oncol 2001; 20:50–56. 4. Zimm S, Wampler GL, Stablein D, et al. Intracerebral metastases in solid-tumor patients: natural history and results of treatment. Cancer 1981; 48(2):384–394. 5. Posner JB. Intracranial metastases. Neurologic Complications of Cancer, vol. 45. Philadelphia: FA Davis Company, 1995:77–110. 6. Nussbaum ES, Djalilian HR, Cho KH, et al. Brain metastasis: histology, multiplicity, surgery and survival. Cancer 1996; 78: 1781–1788. 7. Hazra T, Mullins GM, Lott S. Management of cerebral metastases from bronchogenic carcinoma. Johns Hopkins Med J 1972; 130: 377–383. 8. Chao JH, Phillips R, Nickson JJ. Roentgen-ray therapy of cerebral metastases. Cancer 1954; 7:682–689. 9. Chu FCH, Hilaris BB. Value of radiation therapy on the management of intracranial metastases. Cancer 1961; 4:577–581. 10. Borgelt B, Gelber R, Kramer S, et al. The palliation of brain metastasis: ﬁnal results of the ﬁrst two studies by the Radiation

11.

12.

13.

14.

15.

16.

17.

18. 19. 20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

189

Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1980; 6:1–9. Gelber RD, Larson M, Borgelt BB, et al. Equivalence of radiation schedules for the palliative treatment of brain metastases in patients with favorable prognosis. Cancer 1981; 48:1749–1753. Borgelt B, Gelber R, Larson M, et al. Ultra-rapid high-dose irradiation schedules for the palliation of brain metastasis: ﬁnal result of the ﬁrst two studies by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1981; 7:1633–1638. Kurtz JM, Gelber R, Brady LW, et al. The palliation of brain metastasis in a favorable patient population: a randomized clinical trial by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1981; 7:891–895. Sause WT, Scott C, Krisch R, et al. Phase I/I1 trial of accelerated fractionation in brain metastasis RTOG 85-28. Int J Radiat Oncol Biol Phys 1993; 26:653–657. Murray KJ, Scott C, Greenberg HM, et al. A randomized phase III study of accelerated hyperfractionated versus standard in patients with unresected brain metastasis: a report of the Radiation Therapy Oncology Group (RTOG) 9104. Int J Radiat Oncol Biol Phys 1997; 39:571–574. Komarnicky LT, Philips TL, Martz K, et al. A randomized phase III protocol for the evaluation of misonidazole combined with radiation in the treatment of patients with brain metastases (RTOG 7916). Int J Radiat Oncol Biol Phys 1991; 20:53–58. Phillips TL, Scott CB, Leibel S et al. Results of a randomized comparison of radiotherapy and bromodeoxyuridine with radiotherapy alone for brain metastasis: a report of RTOG trial 89-05. Int J Radiat Oncol Biol Phys 1995; 33:339–348. Sneed PK, Larson DA, Wara WM. Radiotherapy for cerebral metastases. Neurosurg Clin N Am 1996; 7:505–515. Leksell L. The stereotactic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316–319. 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; 13:279–282. 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; 37:745–751. Lagerwaald FJ, Levendag PC, Novak PJ, et al. Identiﬁcation of prognostic factors in patients with brain metastases: a review of 1292 patients. Int J Radiat Oncol Biol Phys 1999, 43:795–801. 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. Noordijk EM, Vecht CJ, Haaxma-Reiche H, et al. The choice of treatment of single brain metastasis should be based on extracranial tumor activity and age. Int J Radiat Oncol Biol Phys 1994; 29:711–717. 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; 78:1470–1476. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multi-institutional experience with stereotactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol Biol Phys 1994; 28:797–802. Yu C, Chen JCT, Apuzzo MLJ, et al. Metastatic melanoma to the brain: prognostic factors after gamma knife radiosurgery. Int J Radiat Oncol Biol Phys 2002; 52:1277–1287. Goyal LK, Suh JH, Reddy CA, et al. The role of whole brain radiotherapy and stereotactic radiosurgery on brain metastasis from renal cell carcinoma. Int J Radiat Oncol Biol Phys 2000; 47:1007–1012. Nieder C, Nestle U, Walter K, et al. Dose/effect relationships for brain metastases. J Cancer Res Clin Oncol 1998; 1214:346– 350.

190

j.h. suh et al.

30. Leibel SA, Sheline GE. Radiation therapy for neoplasms of the brain. J Neurosurg 1987; 66:1–22. 31. Larson DA, Flickinger JC, Loefﬂer JS, et al. The radiobiology of radiosurgery. Int J Radiat Oncol Biol Phys 1993; 25:557–561. 32. Mehta M, Noyes W, Craig B, et al. A cost-effectiveness and cost-utility analysis of radiosurgery vs. resection for single brain metastases. Int J Radiat Oncol Biol Phys 1997; 39:445–454. 33. Sperduto PW, Hall WA. Radiosurgery, cost-effectiveness, gold standards, the scientiﬁc method, cavalier cowboys, and the cost of hope. Int J Radiat Oncol Biol Phys 1996; 36:511–513. 34. 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. 35. Bindal AK, Bindal RK, Hess KR, et al. Surgery versus radiosurgery in the treatment of brain metastases. J Neurosurg 1996; 84:748–754. 36. O’Neill BP, Iturria NJ, Link MJ, et al. A comparison of surgical resection and stereotactic radiosurgery in the treatment of solitary brain metastases. Int J Radiat Oncol Biol Phys 2003; 55:1169– 1176. 37. Muacevic A, Kreth FW, Horstmann GA, et al. Surgery and radiotherapy compared with gamma knife radiosurgery in the treatment of solitary cerebral metastases of small diameter. J Neurosurg 1999; 91:35–43. 38. Loefﬂer JS, Shrieve DC. What is appropriate therapy for a patient with a single brain metastasis? Int J Radiat Oncol Biol Phys 1994; 29:915–917. 39. Weissman DE. Glucocorticoid treatment of brain metastases and epidural spinal cord compression: a review. J Clin Oncol 1988; 6:543–551. 40. Loefﬂer JS, Kooy HM, Wen PW, et al. The treatment of recurrent brain metastasis with stereotactic radiosurgery. J Clin Oncol 1990; 8:576–582. 41. Alexander E 3rd, Moriarty TM, Davis RB, et al. Stereotactic radiosurgery for the deﬁnitive, noninvasive treatment of brain metastasis. J Natl Cancer Inst 1995; 87:34–40. 42. Breneman JC, Warnick RE, Albright RE Jr, et al. Stereotactic radiosurgery for the treatment of brain metastases: results of a single institution series. Cancer 1997; 79:551–557. 43. Noël G, Proudhom MA, Valery CA, et al. Radiosurgery for reirradiation of brain metastases: results in 54 patients. Radiother Oncol 2001; 60:61–67. 44. Kihlstrom L, Karlsson B, Lindquist C. Gamma knife surgery for cerebral metastases. Implications for survival based on 16 years experience. Stereotact Funct Neurosurg 1993; 61:45–50. 45. Valentino V. The results of radiosurgical management of 139 single cerebral metastases. Acta Neurochir (Suppl) 1995; 63:95– 100. 46. Shiau CY, Sneed PK, Shu HK, et al. Radiosurgery for brain metastases: relationship of dose and pattern of enhancement to local control. Int J Radiat Oncol Biol Phys 1997; 37:375–383. 47. Matsuo T, Shibata S, Yasunaga A, et al. Dose optimization and indication of Linac radiosurgery for brain metastases. Int J Radiat Oncol Biol Phys 1999; 45:931–939. 48. Schoeggl A, Kitz K, Ertl A, et al. Prognostic factor analysis for multiple brain metastases after gamma knife radiosurgery: results in 97 patients. J Neurooncol 1999; 42:168–175. 49. Simonova G, Liscak R, Novotny J. Solitary brain metastases treated with the Leksell gamma knife: prognostic factors for patients. Radiother Oncol 2000; 57:207–213. 50. Chidel MA, Suh JH, Reddy CA, et al. 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; 47:993–999.

51. Joseph J, Adler JR, Cox RS, et al. Linear accelerator-based stereotaxic radiosurgery for brain metastases: the inﬂuence of number of lesions on survival. J Clin Oncol 1996; 14:1085–1092. 52. Shehata MK, Young B, Reid B, et al. Stereotactic radiosurgery of 468 brain metastases < or = 2 cm: implications for SRS dose and whole brain radiation therapy. Int J Radiat Oncol Biol Phys 2004; 59:87–93. 53. Pirzkall A, Debus J, Lohr F, et al. Radiosurgery alone or in combination with whole-brain radiotherapy for brain metastases. J Clin Oncol 1998; 16:3563–3569. 54. Sneed PK, Lamborn KR, Forstner JM, et al. Radiosurgery for brain metastases: is whole brain radiotherapy necessary? Int J Radiat Oncol Biol Phys 1999; 43:549–558. 55. Sanghavi SN, Miranpuri SS, Chappell R, et al. 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; 51:426–434. 56. Kondziolka D, Patel A, Lunsford LD, et al. Stereotactic radiosurgery plus whole brain radiotherapy versus radiotherapy alone for patients with multiple brain metastases. Int J Radiat Oncol Biol Phys 1999; 45:427–434. 57. Chougule P, Burton-Williams M, Saris S, et al. Randomized treatment of brain metastasis with Gamma Knife radiosurgery, whole brain radiotherapy or both. Int J Radiat Oncol Biol Phys 2000; 48(Suppl):114. 58. 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: 1622–1672. 59. DeAngelis LM, Posner LB. Radiation-induced dementia in patients cured brain metastases. Neurology 1989; 39:789–796. 60. So NK, O’Neill BP, Frytak S, et al. Delayed leukoencephalopathy in survivors with small cell lung cancer. Neurology 1987; 37:1198– 1201. 61. Cox JD, Stanley K, Petrovich Z, et al. Cranial irradiation in cancer of the lung of all cell types. JAMA 1981; 245:469–472. 62. Russell AH, Pajak TE, Selim HM, et al. Prophylactic cranial irradiation for lung cancer patients at high risk for development of cerebral metastasis: results of a prospective randomized trial conducted by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1991; 21:637–643. 63. Umsawasdi T, Valdivieso M, Chen TT, et al. Role of elective brain irradiation during combined chemoradiotherapy for limited disease non-small cell lung cancer. J Neurooncol 1984; 2:253– 259. 64. Mamon HJ, Yeap BY, Reblando J, et al. High risk of brain metastases in surgically staged IIIA non-small cell lung cancer patients treated with surgery, chemotherapy, and radiation. J Clin Oncol 2005; 23:1530–1537. 65. 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; 22:157–165. 66. Regine WF, Scott C, Murray K, et al. Neurocognitive outcome in brain metastases patients treated with accelerated-fractionation vs. accelerated hyperfractionation radiotherapy: an analysis from the Radiation Therapy Oncology Group 91-04. Int J Radiat Oncol Biol Phys 2001; 51:711–717. 67. Chang EL, Wefel JS, Meyers CA, et al. Prospective neurocognitive evaluation of patients with 1 to 3 newly diagnosed brain metastases treated with stereotactic radiosurgery alone. Int J Radiat Oncol Biol Phys 2004; 60:S164–S165. 68. Regine WF, Schmitt FA, Scott CB, et al. Feasibility of neurocognitive outcome evaluations in patients with brain metastases in a multi-institutional cooperative group setting: results of Radiation

14.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

brain metastases

Therapy Oncology Group trial BR-0018. Int J Radiat Oncol Biol Phys 2004; 58:1346–1352. Sonderkaer S, Schmiegelow M, Carstensen H, et al. The long-term neurologic outcome of childhood brain tumors treated by surgery alone. J Clin Oncol 2003; 21:1347–1351. Tchen N, Juffs HG, Downie FP, et al. Cognitive function, fatigue, and menopausal symptoms in women receiving adjuvant chemotherapy for breast cancer. J Clin Oncol 2003; 21:4175– 4183. Schagen SB, van Dam FSAM, Muller MJ, et al. Cognitive deﬁcits after postoperative adjuvant chemotherapy for breast cancer. Cancer 1999; 85:640–650. van Dam FSAM, Schagen SB, Muller MJ, et al. Impairment of cognitive function in women receiving adjuvant treatment for high risk breast cancer: high-dose versus standard dose chemotherapy. J Natl Cancer Inst 1998; 90:210–218. Brezden CB, Phillips KA, Abdolell M, et al. Qaeda to function and breast cancer patients receiving adjuvant chemotherapy. J Clin Oncol 2000; 18:2695–2701. Patchell RA, Tibbs PA, Regine WF, et al. Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial. JAMA 1998; 280:1485–1489. Patchell RA, Regine WF. The rationale for adjuvant whole brain radiation therapy with radiosurgery in the treatment of single brain metastases. Technol Ca Res Treat 2003; 2:111–115. Regine WF, Huhn JL, Patchell RA, et al. 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– 338. Hasegawa T, Kondziolka D, Flickinger JC, et al. Brain metastases treated with radiosurgery alone: an alternative to whole brain radiotherapy? Neurosurgery 2003; 52:1318–1326. Lutterbach J, Cyron D, Henne K, et al. Radiosurgery followed by planned observation in patients with one to three brain metastases. Neurosurgery 2003; 52:1066–1074.

191

79. Sneed PK, Suh JH, Goesch SJ, et al. A multi-institutional review of radiosurgery alone versus radiosurgery with whole brain radiotherapy as the initial management of brain metastases. Int J Radiat Oncol Biol Phys 2002; 53:519–526. 80. Manon RR, Oneill A, Mehta M, et al. Phase II trial of radiosurgery (RS) for 1-3 newly diagnosed brain metastases from renal cell, melanoma, and sarcoma: An Eastern Cooperative Oncology Group Study (E6397). Proceedings of the 40th ASCO meeting [abstract no. 1507]. J Clin Oncol 2004. 81. Aoyama H, Shirato H, Nakagawa K, et al. Interim report of the JROSG99-1 multi-institutional randomized trial, comparing radiosurgery alone versus radiosurgery plus whole brain irradiation for 1-4 brain metastases. Proceedings of the 40th ASCO meeting [abstract no. 1506]. J Clin Oncol 2004. 82. 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; 47:291–298. 83. Shaw E, Scott C, Souhami L, et al. Radiosurgery for the treatment of previously irradiated recurrent primary brain tumors and brain metastases: initial report of the Radiation Therapy Oncology Group protocol 90-05. Int J Radiat Oncol Biol Phys 1996; 34: 647–654. 84. Valery CA, Cornu P, Noël G, et al. Predictive factors of radiation necrosis after radiosurgery for cerebral metastases. Stereotact Funct Neurosurg 2003; 81:115–119. 85. Warnick RE, Darakchiev BJ, Breneman JC. Stereotactic radiosurgery for patient with solid brain metastases: current status. J Neuro-oncol 2004; 69:125–137. 86. Ma L, Chin LS, DiBiase SJ, et al. Concomitant boost of stratiﬁed target area with gamma knife radiosurgery: a treatment planning study. Am J Clin Oncol 2003; 26:e100–e105. 87. Levivier M, Massager N, Wikler D, et al. Use of stereotactic PET images in dosimetry planning of radiosurgery for brain tumors: clinical experience and proposed classiﬁcation. J Nucl Med 2004; 45:1146–1154.

1 5

Metastatic Brain Tumors: Surgery Perspective Raymond Sawaya and David M. Wildrick

Introduction Metastatic brain tumors occur more frequently than other intracranial neoplasms and are a serious complication of systemic cancer [1]. Their annual incidence exceeds 100,000 [2, 3] and is on the rise as patients live longer from improved treatments for extracranial cancer [4, 5]. Thus, patients with brain metastases constitute a signiﬁcant fraction of the case load in the neurosurgery services at many oncologic care centers. Early in the 1900s, the results of surgical treatment of single brain metastases were disappointing [6], with patients seldom showing a median survival time of longer than 6 months [7]. Although this was longer than the 4- to 6-week survival time typical of patients with untreated brain metastases, the surgical morbidity was high (15% to 50%) [8], and whole-brain radiation therapy (WBRT) alone was frequently favored for management of patients with brain metastases. Since then, the Radiation Therapy Oncology Group (RTOG) has demonstrated that treatment of brain metastasis patients with WBRT alone can extend their median survival time to 16 weeks [9–11]. In addition, two independent, prospective randomized studies showed that surgery plus WBRT offered a survival advantage superior to WBRT alone in such patients [12, 13]. Since Leksell demonstrated the utility of focused beams of high-energy X-rays in the ablation of intracranial tumors [14, 15], the technique of stereotactic radiosurgery (SRS) has continued to evolve. During the 1990s, SRS was increasingly used in the management of brain metastases, with some investigators recently claiming that SRS alone produces results similar to those obtained by combining SRS and WBRT, or potentially even equivalent to the outcome with surgical resection [16–20]. With this in mind and because of the ease of use of SRS and the perception that it costs less than conventional surgery, some have even suggested that most metastatic brain tumors should be managed exclusively with SRS. This controversial notion has already led to treatment of some series of brain metastasis patients with SRS alone followed by observation [18, 19]. At The University of Texas M. D. Anderson Cancer Center (“M. D. Anderson”), SRS is regarded as a specialized tool to be used judiciously, as a given patient’s situation dictates, rather than as a generalized treatment modality for brain metastases. We continue to see surgical resection as playing the central role in the management of patients with a limited number of brain

metastases [4, 21–24]. In this chapter, we present our view of the relative roles of surgery and SRS in terms of issues concerning patient selection, treatment outcome, cost-effectiveness of the treatment, and the patients’ quality of life. These differences are summarized in Table 15-1 [25–27].

Patient Selection When a patient presents with a brain metastasis and symptoms of mass effect, there is seldom a dispute that the lesion should be removed surgically. Similarly, when a patient with a brain metastasis presents in too poor a medical condition to be a surgical candidate (or declines surgery), SRS represents a logical treatment modality. Other patients with single brain metastases can be sorted into three groups. The ﬁrst group has relatively large lesions (exceeding 3 cm in maximum diameter) that can only be effectively removed by surgical resection. Treatment of these tumors with SRS is not effective because the radiation dose must be reduced as the tumor size increases to prevent damage to surrounding brain tissue. This relationship was clearly demonstrated by Mehta et al. [25], who showed that with SRS, the rate of complete response (CR; total disappearance of the magnetic resonance [MR] image of the lesion after SRS) falls off dramatically with tumor volume such that if a tumor 2 cm3 in volume has a 50% CR rate, an 8- to 9-cm3 lesion shows about a 20% CR rate (Fig. 15-1). Patients in the second group have very small lesions (less than 10 mm in maximum diameter) that are not surgically accessible and are located deep within the brain. In this situation, SRS provides an effective alternative to resection, superior to WBRT, which was the only treatment previously available. The third group of patients are those who have metastases that are less than 3 cm in maximum diameter and are surgically accessible. Whether surgery or SRS is the best treatment for these lesions is the subject of much current debate. The 3-cm upper size limit referred to above is probably too high for adequate SRS treatment. At M. D. Anderson, a recent study of 153 brain metastases from melanoma treated with SRS [28] showed that the 1-year local control rate of smaller tumors with a maximum diameter of no more than 1.5 cm (volume = 2 cm3) was superior to that of larger lesions (75.2% and 42.3%, respectively; p < 0.05). Moreover, another

193

194

r. sawaya and d.m. wildrick

TABLE 15-1. Differences between surgery and stereotactic radiosurgery in the treatment of brain metastasis.

Patient selection Tissue diagnosis Lesion size Surgical candidate Treatment outcome Tumor status Local control Local recurrence rate Median survival time Complications Presentation Major neurological Mortality (30-day)†

Surgical resection

Stereotactic radiosurgery

Conﬁrms the lesion is a tumor Large, ≥1.5 cm*; especially if there is mass effect Yes

Cannot conﬁrm the lesion is a tumor Small, ≤2.0 cm*; signiﬁcant mass effect absent No (or patient declines surgery)

Removed (≥94% on average) [4] 85% for >40 months (up to 5 years) [23] 8% [23] to 12% [4] 10 [13] to 16.4 [23] months

Not removed 85% at 12 months; 65% at 24 months [25] 30% [26] to 47% [27] 7.5 [23] to 14 [16] months

Usually immediate

Frequently delayed; necrosis may necessitate surgical resection 25% in eloquent brain (RTOG grade 3) [24] 1.8%

Cost-effectiveness

7% in eloquent brain; 6% overall [4] <2% Based on 3-day hospital stay

Quality of life Relief from mass effect Steroid use Follow-up visits Patient debilitation

Immediate Tapered off over 2–4 weeks Few Not debilitated at home

Based on 1-day outpatient stay; includes no costs of maintenance steroids, follow-up ofﬁce visits, or follow-up MRI Delayed Can last for months; may cause dependence Many Debilitated at home

*Maximum diameter. †

Occurring within 30 days after the procedure.

recent study at M. D. Anderson used SRS to treat brain metastases from different primary tumor types [26] and found a 1year actuarial local control rate of 86% for lesions that were 1 cm or less in maximum diameter (0.5 cm3) but only a 56% local control rate for lesions larger than this (p = 0.0016). Furthermore, a slightly earlier study by a radiosurgery group in Pittsburgh [29] included only brain metastases that were no

100

GTR

Treatment Outcome

80

Survival

Percent

60

40 CR 20

0 0

larger than 2.5 cm in maximum diameter and found that treatment with SRS plus WBRT was superior to WBRT-alone. It is important to note that control rates similar to surgery for lesions as large as 2 cm (volume ∼5 cm3) can be achieved with the addition of fractionated WBRT [30]. Collectively, these ﬁndings clearly indicate that for effective use of SRS alone (i.e., without the addition of WBRT), a cutoff point in brain metastasis diameter that is lower than 3 cm is warranted.

2

4

6

8

10

12

14

16

18

20

Tumor Size (cc) FIGURE 15-1. Tumor size versus response to radiosurgery and surgery for brain metastases. Comparison of a representative radiosurgery complete response (CR) curve [25] with a surgery gross-total resection (GTR) curve (average percentage, 94% from Table 15-2), showing that CR rate with radiosurgery is tumor size dependent, whereas GTR rate with surgery is not. (From Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg 1999; 45:41–47. Used with permission.)

The controversy remains as to whether surgery or SRS is more appropriate and effective to treat patients in the group with single brain metastases that are surgically accessible and range between 1 and 3 cm in maximum diameter. Because no one has yet been able to complete a prospective randomized study comparing treatment of such patients with surgical resection and SRS, the question of which of the two is more effective lacks a deﬁnitive answer. In lieu of this, many retrospective studies have been performed comparing surgical resection and SRS for brain metastasis treatment. Among the more carefully constructed studies of this type are those of Auchter et al. [16], Bindal et al. [23], and Cho et al. [31]. Auchter and coworkers [16] performed a multiinstitutional retrospective outcome and prognostic factor analysis of patients with single cerebral metastases who were treated with SRS plus WBRT. From their database of 533 patients with brain metastases treated with SRS and WBRT, they selected 122 patients who fulﬁlled the criteria for surgical resection established in the prospective randomized trial of Patchell et al.

15.

195

metastatic brain tumors: surgery perspective

[13], including having a single brain metastasis that was surgically resectable (but no urgent need for surgery), no prior radiotherapy or surgical treatment, independent functional status (a Karnofsky Performance Scale [KPS] score >70), and a non– radiosensitive tumor. They then compared the outcome of these 122 patients with that of the patients in the prospective randomized surgical series of Patchell et al. [13] and Noordijk et al. [12]. The actuarial median survival time after treatment with SRS plus WBRT was 56 weeks compared with 40 weeks [13] and 43 weeks [12], respectively, after surgery plus WBRT. The median duration of functional independence was 44 weeks after SRS and WBRT [16] compared with 38 weeks [13] and 33 weeks [12], respectively, after surgery and WBRT. These results indicated to Auchter et al. [16] that SRS plus WBRT produced outcomes for patients with single brain metastases that were comparable with, if not better than, surgery plus WBRT. Cho and coworkers [31] retrospectively reviewed a series of 225 single brain metastases in patients treated with either WBRT alone, SRS plus WBRT, or surgery and WBRT. There was a similar distribution of prognostic factors such as age, sex, KPS score, and metastasis location in these three groups except for extracranial cancer, which was 14% higher in the group treated with SRS and WBRT. The actuarial median survival times for patients treated with SRS and WBRT (9.8 months) and surgery plus WBRT (10.5 months) were nearly identical, but both were signiﬁcantly better than for those treated with WBRT alone (3.8 months). Other more recent retrospective studies have also labeled resection and SRS as equally effective in the treatment of small to moderately sized brain metastases [32]. At M. D. Anderson, Bindal and coworkers [23] also retrospectively compared surgical resection and SRS treatment of brain metastases. Thirty-one patients treated with SRS were followed prospectively, and 62 patients treated with conventional surgery were retrospectively matched to those in the SRS group. Patients were matched according to primary tumor histology, extent of systemic disease, preoperative KPS score, time to brain metastasis, number of brain metastases, and patient age and sex. WBRT treatment was similar in both groups, and patient eligibility criteria for SRS were the same as for surgery. Lesions treated by SRS were limited to those <3 cm in maximum diameter, and 81% of them were deemed surgically resectable. There was a statistically signiﬁcant survival advantage in the surgically treated group (16.4 months median survival) relative to the group treated with SRS (7.5 months median survival) (p = 0.0009 by multivariate analysis). In contrast to the conclusions of Auchter et al. [16] and Cho et al. [31], Bindal’s group [23] concluded that surgery was superior to SRS in clinically similar patients in terms of survival, local recurrence, and morbidity. Thus, determination of whether treatment of brain metastases with SRS is better than, equivalent to, or worse than surgical resection must await a phase III prospective trial comparing SRS with conventional surgery for the treatment of single brain metastases.

Tumor Local Control and Recurrence When brain metastases are removed surgically, local control is produced by total removal of the tumor and elimination of surrounding edema (Fig. 15-2). At M. D. Anderson, gross-total

FIGURE 15-2. (A) Preoperative and (B) postoperative gadolinium contrast-enhanced MR images of a brain metastasis in a patient with non–small cell lung cancer. In each panel, the left and right images are T1- and T2-weighted, respectively. (From Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg 1999; 45:41–47. Used with permission.)

resection (equivalent to complete response in radiosurgery) of metastatic tumors is achieved in 90% to 96% of instances, in both eloquent and noneloquent brain regions (Table 15-2). With surgery, local control (absence of recurrent tumor growth after gross-total resection) typically persists at an 85% level at 1 year and remains there for an interval ranging from 40 months to 5 years (Fig. 15-3). TABLE 15-2. Gross-total resections performed for metastatic tumors in different brain regions (gross-total resection is equivalent to a complete response in radiosurgery). Metastasis location (Tumor functional grade*)

No. of patients

Noneloquent (I) Near-eloquent (II) Eloquent (III) Total

79 61 54 194

GTR (%)

75 (95) 55 (90) 52 (96) Average, 94

GTR, gross-total resection. *Grade according to functional location. Source: Adapted from Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg 1999; 45:41–47. Used with permission.

Percent Free From Local Recurrence

196

r. sawaya and d.m. wildrick

100 MDACC Surgery 80 BRIGHAM Radiosurgery 60 40 MDACC Radiosurgery 20 0 0

10

20

30

40

Follow-up Time (months) FIGURE 15-3. Freedom from local recurrence (local control) of brain metastases with radiosurgery and with surgery. Curve showing time from radiosurgical treatment to local failure (BRIGHAM Radiosurgery) in 42 patients treated at Brigham and Women’s hospital [25] superimposed on curves showing time from surgical resection (MDACC Surgery) to local recurrence in 62 patients and time to local failure in 31 patients radiosurgically treated (MDACC Radiosurgery) at The University of Texas M. D. Anderson Cancer Center [23]. (From Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg 1999; 45: 41–47. Used with permission.)

In contrast, when brain metastases are treated with SRS, it is common for the tumor to remain visible on computed tomography (CT) or MR images. Local tumor control with SRS is less stringently deﬁned and, in addition to tumor regression includes lesions remaining unchanged in size as well as those undergoing no more than a 25% increase in size/volume above the baseline measurement. This makes it difﬁcult and misleading to compare values published for local control of brain metastases by surgical resection and SRS. In some patients who have a brain metastasis that is stable on MR images or increases in size by less than 25%, surgery may be required to relieve neurologic deﬁcits resulting from mass effect and persistent edema (Fig. 15-4) [33].

After surgical resection of brain metastases at M. D. Anderson, we have observed local tumor recurrence at a rate ranging from about 8% to 12% [4, 23], which is consistent with the observations of others. In patients treated for brain metastases with SRS, similar 1-year local control failure rates (6% to 15%) are often reported [16, 18, 19, 34–38]. These numbers reported for SRS are ﬂawed not only because of the less stringent deﬁnition of local control with SRS but also because tumor recurrence in brain metastasis patients is a function of survival time, and many radiosurgical series of these patients report such short survival intervals (8 months on average) that failure of local tumor control cannot be reliably determined. Treatment of brain metastasis patients with SRS at Brigham and Women’s Hospital [25] gave an 85% local control rate at 1 year, but this dropped steadily to 65% after 2 years (Fig. 15-3). More recent series of brain metastasis patients treated with SRS have presented overall recurrence values (30% to 47%) [26, 27] that are much higher than those typical of surgically treated patients. These numbers may be more realistic, reﬂecting a trend toward increased follow-up monitoring of patients, with more routine neuroimaging [28].

Complications Complications arising from surgical resection of brain metastases are usually evident immediately. Deleterious effects of SRS treatment are frequently delayed and may be missed in follow-up visits, leading to an underreporting of complications. A recent study using SRS alone to treat brain metastases showed a complication rate of only 8% [19]; nevertheless, 70% of the complications were acute and included increases in seizures and worsening of preexisting neurologic symptoms. Similarly, in a study of recurrence of brain metastasis after SRS, Regine et al. [27] observed that recurrence was symptomatic in 71% of patients and was associated with a neurologic deﬁcit in 59%. With the advent of modern neurosurgical techniques, especially intraoperative stereotaxy and ultrasonography, few metastatic brain tumors are surgically inaccessible. In a study of 194 patients at M. D. Anderson [4], the 30-day morbidity rate for any major neurologic deﬁcit after brain metastasis resection was 6% (Table 15-3). Even for resections in eloquent brain regions, this value only rose to 7%. The common perception that SRS is safer to use than conventional surgery to treat brain metastases within or near eloquent brain areas may not be warranted. We recently performed a retrospective study of neurologic complications in patients

TABLE 15-3. Major neurologic complications of surgery for metastases in different brain regions. Metastasis location

FIGURE 15-4. CT scan of a patient with non–small cell lung cancer who had a 3-cm brain metastasis in the right anterior parietal lobe (left) that was treated with radiosurgery. Three months later, the tumor shrank by 23%, but the edema and the patient’s symptoms were unchanged (right). The patient subsequently underwent a craniotomy and total resection of the mass. The results of the surgery are shown in Fig. 15-2. (From Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg 1999; 45:41–47. Used with permission.)

Noneloquent Near-eloquent Eloquent Average major neurologic deﬁcit rate

No. patients*

No. complications (%)

79 61 54

1 (1) 6 (10) 4 (7) (6)

*Total = 194 patients. Source: Adapted from Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg 1999; 45:41–47. Used with permission.

15.

197

metastatic brain tumors: surgery perspective 100

A.Pre

B.Post 80

% of Patients

60

28≤ 26 24 22 20 18 16 14 12 10 8

6

4

2

40

20 0 0 2

4

6

8 10 12 14 16 18 20 22 24 26 ≥28

Time (Weeks) FIGURE 15-5. Graphs depicting the duration (in weeks) of steroid facilitated tapering off of steroid use within 2 to 4 weeks. (From Vecil treatment from the time of conventional resection of the index lesion GG, Suki D, Maldaun MV, et al. Resection of brain metastases pre(Time 0). The duration of continuous treatment is shown both before viously treated with stereotactic radiosurgery. J Neurosurg 2005; 102(2): (A) and after (B) conventional resection. Note that surgical resection 209–215. Used with permission.)

treated with SRS for brain metastases at M. D. Anderson and found that such complications (including radiation necrosis, seizures, and neurologic symptom development or worsening) were highest for tumors in eloquent brain, reaching 25% for complications of RTOG grade 3 and above [24]. At M. D. Anderson, the current mortality rate for surgical resection of brain metastases is less than 2%. Although it has been claimed that SRS poses little risk of permanent neurologic morbidity in brain metastasis treatment, mortalities have occurred. In the study by Bindal et al. [23], spontaneous intratumoral hemorrhage occurred in 3 of 31 patients treated with SRS, causing the death of one. Kondziolka and coworkers [29] compared brain metastasis treatment with SRS plus WBRT versus WBRT alone and reported a 1.8% mortality rate within 4 weeks after SRS. Thus, it appears that mortality rates for SRS and surgical resection of brain metastases are similar. Another potential source of complications when treating brain metastases with SRS stems from its inability to provide a tissue biopsy for pathologic diagnosis. Because 5% to 11% of patients with a prior history of cancer and a brain lesion that appears consistent with a metastasis may actually have nonmetastatic disease (primary brain tumor, abscess, multiple sclerosis plaque, or a hemorrhage) [13, 39], SRS treatment of such lesions can lead to an adverse outcome. With surgical resection of brain metastases, it is easy to histologically conﬁrm that the lesion is metastatic cancer, and this situation does not arise.

Cost Effectiveness SRS is frequently promoted as a lower-cost alternative to conventional surgery in the management of brain metastases. Two publications have emerged using retrospective analyses to compare relative costs of SRS and surgery in attempt to bolster this position [40, 41]. Unfortunately, these cost analyses have relied on surgical expenses and data on length of hospital stay from publications on patient series that were already more than a decade old, such as that of Patchell et al. [13], and have not taken into account the drastic cost-cutting measures adopted by

most hospitals and surgeons around the country during the past decade. Moreover, these cost-comparison studies [40, 41] make the assumption that SRS and surgery are “equiefﬁcacious” in the management of patients with cancer metastatic to the brain. They also assign inferior median survival values to studies employing surgical resection plus WBRT for brain metastasis treatment relative to SRS, thereby inﬂating the average cost of surgery. Because there have been no prospective randomized studies comparing SRS and surgery with respect to patient outcomes, there is no accurate statistical basis for the presumed “equiefﬁcaciousness” or the survival numbers put forth. To accurately reﬂect the total costs of SRS and surgical resection for brain metastases, the expenses involved in patient follow-up must be included. Typically, with SRS, only the 1-day cost of treatment is considered in cost-effectiveness studies, with follow-up being ignored. At M. D. Anderson, follow-up costs after surgery for brain metastases have been severely reduced because most patients are now discharged from the hospital after 3 days. Follow-up expenses after SRS are more signiﬁcant than for surgery because many routine magnetic resonance imaging (MRI) studies and ofﬁce visits are required to monitor the delayed effects of the radiation. These delayed effects require extended administration of medications such as dexamethasone, which becomes costly. In our recent study of patients undergoing resection of brain metastases after failure of SRS to control their lesions [33], after their SRS treatment and prior to surgery, 40% of 53 patients required continuous steroid administration for 12 weeks, which is consistent with steroid dependency and is not a trivial expense. After surgery, more than 95% of these patients were able to cease steroid usage within 2 to 4 weeks (Fig. 15-5).

Quality of Life Surgical resection of a brain metastasis usually provides immediate relief of neurologic symptoms produced by mass effect of the tumor and edema surrounding it. Thus, the patient can be

198

r. sawaya and d.m. wildrick

on his feet that much sooner, and few follow-up visits are required. With SRS treatment of a brain metastasis, radiation damage to the tumor frequently takes time to develop, and the therapeutic effect with consequent resolution of neurologic symptoms may be delayed for days or weeks. Multiple followup visits are required to monitor the tumor’s response to SRS with MRI or CT. For symptom relief, the patient must receive continuous steroid administration that can last for months, while remaining debilitated at home.

Conclusion The debate continues whether surgical resection or SRS is better in managing tumors metastatic to the brain. Settling the issue of which modality affords patients the longer median survival time awaits the outcome of a prospective randomized trial comparing the two. Both methods have clear beneﬁts. Surgery has the advantages of immediate resolution of mass effect, procurement of tissue for pathologic diagnosis, and lack of the risk of radiation necrosis [42]. SRS carries a decreased risk of hemorrhage and infection, has no risk of tumor seeding, and is less invasive, potentially less costly, and requires a shorter hospital stay than standard craniotomy. Disadvantages of SRS include potential radiation necrosis or exacerbation of peritumoral edema and a requirement for long-term steroid administration [43, 44]. In our experience at M. D. Anderson, we believe that the size and location of the metastases along with the patient’s clinical presentation can be used to recommend one modality or the other. We almost always favor surgery in the case of brain metastases larger than 3 cm in maximum diameter, whereas deeply located lesions smaller than 1 to 1.5 cm in maximum diameter are usually treated with SRS. If the lesions lend themselves to treatment by either method, we let the patient’s symptoms be the guide; lesions that produce symptoms are more frequently treated surgically, and lesions that do not can be treated with SRS. Of course, this approach may be modiﬁed depending upon the patient’s systemic cancer status or medical condition. Patients who cannot tolerate surgery, who have progressive systemic disease, or who are expected to live for less than 3 months are treated with SRS. The best management strategy for patients with brain metastases should involve the complementary use of surgery, SRS, and WBRT. Acknowledgment. We thank the Barbara Falik Research Fund for helping make this work possible.

References 1. Posner JB, Chernik NL. Intracranial metastases from systemic cancer. Adv Neurol 1978; 19:579–592. 2. Tsukada Y, Fouad A, Pickren JW, Lane WW. Central nervous system metastasis from breast carcinoma. Autopsy study. Cancer 1983; 52(12):2349–2354. 3. Sawaya R, Bindal RK, Lang FF, Abi-Said D. Metastatic brain tumors. In: Kaye AH, Laws ER, eds. Brain Tumors: An Encyclopedic Approach, 2nd ed. Edinburgh: Churchill Livingstone, 2001:999–1026. 4. Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg 1999; 45:41–47.

5. Young RF. Radiosurgery for the treatment of brain metastases. Semin Surg Oncol 1998; 14(1):70–78. 6. Grant FC. Concerning intracranial malignant metastases: their frequency and the value of surgery in their treatment. Ann Surg 1926; 84:635–646. 7. Cairncross JG, Posner JB. The management of brain metastases. In: Walker MD, ed. Oncology of the Nervous System. Boston: Martinus Nijhof, 1983:341–377. 8. Horwitz NH, Rizzoli HV. Postoperative complications of intracranial neurological surgery. Baltimore: Williams & Wilkins, 1982. 9. Borgelt B, Gelber R, Kramer S, et al. The palliation of brain metastases: ﬁnal results of the ﬁrst two studies by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1980; 6(1):1–9. 10. Diener-West M, Dobbins TW, Phillips TL, Nelson DF. Identiﬁcation of an optimal subgroup for treatment evaluation of patients with brain metastases using RTOG study 7916. Int J Radiat Oncol Biol Phys 1989; 16(3):669–673. 11. Sneed PK, Larson DA, Wara WM. Radiotherapy for cerebral metastases. Neurosurg Clin N Am 1996; 7(3):505–515. 12. Noordijk EM, Vecht CJ, Haaxma-Reiche H, et al. The choice of treatment of single brain metastasis should be based on extracranial tumor activity and age. Int J Radiat Oncol Biol Phys 1994; 29(4):711–717. 13. 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(8):494–500. 14. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316. 15. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry 1983; 46(9):797–803. 16. 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(1):27–35. 17. Flickinger JC, Kondziolka D. Radiosurgery instead of resection for solitary brain metastasis: the gold standard redeﬁned [editorial] [see comments]. Int J Radiat Oncol Biol Phys 1996; 35(1):185–186. 18. Hasegawa T, Kondziolka D, Flickinger JC, et al. Brain metastases treated with radiosurgery alone: an alternative to whole brain radiotherapy? Neurosurgery 2003; 52(6):1318–1326; discussion 1326. 19. 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):1066–1073; discussion 1073–1074. 20. Sneed PK, Suh JH, Goetsch SJ, et al. 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(3):519–526. 21. Bindal RK, Sawaya R, Leavens ME, Lee JJ. Surgical treatment of multiple brain metastases. J Neurosurg 1993; 79(2):210–216. 22. Bindal RK, Sawaya R, Leavens ME, et al. Reoperation for recurrent metastatic brain tumors. J Neurosurg 1995; 83(4):600–604. 23. Bindal AK, Bindal RK, Hess KR, et al. Surgery versus radiosurgery in the treatment of brain metastasis. J Neurosurg 1996; 84(5):748–754. 24. Dare AO, Sawaya R. Part II: Surgery versus radiosurgery for brain metastasis: surgical advantages and radiosurgical myths. Clin Neurosurg 2004; 51:255–263. 25. 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(4):619–625. 26. Chang EL, Hassenbusch SJ 3rd, Shiu AS, et al. The role of tumor size in the radiosurgical management of patients with ambiguous brain metastases. Neurosurgery 2003; 53(2):272–280.

15.

metastatic brain tumors: surgery perspective

27. Regine WF, Huhn JL, Patchell RA, et al. 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– 338. 28. Selek U, Chang EL, Hassenbusch SJ 3rd, et al. Stereotactic radiosurgical treatment in 103 patients for 153 cerebral melanoma metastases. Int J Radiat Oncol Biol Phys 2004; 59(4):1097– 1106. 29. 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):427–434. 30. Shehata MK, Young B, Reid B, et al. Stereotatic radiosurgery of 468 brain metastases ≤ 2 cm: implications for SRS dose and whole brain radiation therapy. Int J Radiat Oncol Biol Phys 2004; 59(1): 87–93. 31. Cho KH, Hall WA, Lee AK, et al. Stereotactic radiosurgery for patients with single brain metastasis. J Radiosurg 1998; 1(2):79– 85. 32. O’Neill BP, Iturria NJ, Link MJ, et al. 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– 1176. 33. Vecil GG, Suki D, Maldaun MV, et al. Resection of brain metastases previously treated with stereotactic radiosurgery. J Neurosurg 2005; 102(2):209–215. 34. Alexander E 3rd, Moriarty TM, Davis RB, et al. Stereotactic radiosurgery for the deﬁnitive, noninvasive treatment of brain metastases. J Natl Cancer Inst 1995; 87(1):34–40. 35. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multiinstitutional experience with stereotactic radiosurgery for solitary

36.

37.

38.

39.

40.

41.

42.

43.

44.

199

brain metastasis. Int J Radiat Oncol Biol Phys 1994; 28(4):797– 802. Kihlstrom L, Karlsson B, Lindquist C, et al. Gamma knife surgery for cerebral metastasis. Acta Neurochir Suppl (Wien) 1991; 52:87–89. Muacevic A, Kreth FW, Horstmann GA, et al. 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. Petrovich Z, Yu C, Giannotta SL, et al. Survival and pattern of failure in brain metastasis treated with stereotactic gamma knife radiosurgery. J Neurosurg 2002; 97(5 Suppl):499–506. Voorhies RM, Sundaresan N, Thaler HT. The single supratentorial lesion. An evaluation of preoperative diagnostic tests. J Neurosurg 1980; 53(3):364–368. Mehta M, Noyes W, Craig B, et al. A cost-effectiveness and costutility analysis of radiosurgery vs. resection for single-brain metastases. Int J Radiat Oncol Biol Phys 1997; 39(2):445–454. Rutigliano MJ, Lunsford LD, Kondziolka D, et al. The cost effectiveness of stereotactic radiosurgery versus surgical resection in the treatment of solitary metastatic brain tumors. Neurosurgery 1995; 37(3):445–453; discussion 453–455. Loefﬂer JS, Alexander E 3rd. Radiosurgery for the treatment of intracranial metastases. In: Alexander E 3rd, Loefﬂer JS, Lunsford LD, eds. Stereotactic Radiosurgery. New York: McGraw-Hill, 1993:197–206. Sturm V, Kimmig B, Engenhardt R, et al. Radiosurgical treatment of cerebral metastases. Method, indications and results. Stereotact Funct Neurosurg 1991; 57(1–2):7–10. Kondziolka D, Lunsford LD. Brain metastases. In: Apuzzo MLJ, ed. Brain Surgery: Complication Avoidance and Management. New York: Churchill Livingstone, 1993:615–641.

1 6

Brain Metastases: Whole-Brain Radiation Therapy Perspective Roy A. Patchell and William F. Regine

Introduction Whole-brain radiation therapy (WBRT) is the mainstay of treatment for most patients with intracerebral metastases. It is useful for most newly diagnosed tumors and sometimes also as salvage therapy for recurrent tumors.

Efﬁcacy of WBRT Given the general consensus that WBRT is an effective treatment for brain metastases, it is surprising to note that there have been no randomized trials comparing WBRT with best supportive care or no treatment [1]. The evidence for the effectiveness of WBRT is derived from retrospective studies, nonrandomized prospective trials, and commonsense reasoning. Perhaps the best support for the general efﬁcacy of WBRT comes from comparing the outcomes of patients after WBRT treatment with the natural history of untreated brain metastases. Data from large retrospective studies [1–3] have shown that WBRT increases the median survival time to 3 to 6 months and that more than half of patients treated with WBRT die ultimately of progressive systemic cancer and not as a direct result of brain metastases. This contrasts with the natural history of untreated patients, which is that patients live for a median of 1 to 2 months and virtually all die as result of their brain lesions [4–7]. There is little doubt that WBRT does control or improve symptoms, at least temporarily, in the majority of patients with brain metastases. That WBRT has been accepted as effective should not be a serious problem when we remember that the efﬁcacy of most of the treatments used in medicine has not been established by the use of randomized trials.

Unsettled Issues Although there is general agreement on efﬁcacy, no consensus exists on the optimum WBRT dose and schedule for the treatment of brain metastases. The best available data on the effect of dose and schedule for the treatment of brain metastases comes from several large-scale multi-institutional trials conducted by the Radiation Therapy Oncology Group (RTOG) [1,

8–11]. These studies have shown no signiﬁcant difference in the frequency and duration of response for total radiation doses ranging from 2000 cGy over 1 week to 5000 cGy over 4 weeks. Regimens of 1000 cGy in a single dose or 1200 cGy in two doses were less effective and are no longer in use. Typical radiation treatment schedules for brain metastases consist of short courses (7 to 15 days) of WBRT with relatively high doses per fraction (150 to 400 cGy per day) and total doses in the range of 3000 to 5000 cGy. These schedules minimize the duration of treatment while still delivering adequate amounts of radiation to the tumor.

Attempts to Improve WBRT Different fractionation and dosing schemes have been tried in order to improve the effectiveness of WBRT with varying results. Epstein et al. [12] reported a phase I/II dose-escalation study of hyperfractionated radiotherapy using total doses of 48, 54.4, 64, and 70.4 Cy. No increased toxicity was identiﬁed and the three highest dose arms had a statistically signiﬁcant improvement in median survival over the lowest dose arm. This study suggested a dose-dependent effect with the use of hyperfractionated radiation in unresected, single brain metastasis. Unfortunately, a large randomized RTOG trial, studying both single and multiple brain metastases, failed to show improvement in overall survival using a hyperfractionated schedule of 54.5 Gy in 34 fractions compared with a standard regimen of 30 Gy in 10 fractions [13]. Another attempt to improve the efﬁcacy of WBRT was the addition of a boost dose to the tumor. However, a retrospective study [14] has found that increased conventional focal irradiation to the tumor, or “boost” dosing, did not increase survival or time to neurologic recurrence when compared with WBRT-alone. Radiation cell sensitizing agents have also been used in an attempt to increase tumor cell death. The rationale was based on the observation that hypoxic cells (often found centrally in a tumor) are more resistant to the effects of ionizing radiation. Agents such as misonidazole have the potential to increase cell sensitivity to irradiation. In the past, none of the radiation cell sensitizers has been shown to provide any additional beneﬁt

201

202

r.a. patchell and w.f. regine

over conventional radiotherapy [1, 15, 16]. However, two recent trials involving motexaﬁn gadolinium in lung cancer [17] and RSR13 in breast cancer [18] have shown guardedly promising results, and further investigation of these agents is needed to determine if they are effective.

Complications of WBRT WBRT has complications. Almost all patients experience a temporary loss of hair; hair usually returns 6 to 12 months after completing therapy. Also, in the short-term, patients may have a transient worsening of neurologic symptoms while receiving therapy. Many physicians believe that maintaining patients on steroids during radiotherapy will minimize radiation complications, although conclusive proof of this has not been forthcoming. The long-term side effects of radiotherapy are usually not a signiﬁcant issue in the treatment of patients with brain metastases because of the relatively short survival time of these patients. The frequency of serious long-term complications is unknown. One often quoted retrospective study by DeAngelis et al. [19] suggests that as many as 11% of long-term survivors (>12 months) of brain metastases treated with WBRT develop dementia. However, virtually all of the patients in that sample who developed dementia had been treated with atypically large radiation fractionation schedules. The patients treated with fraction sizes less than 3.0 Gy per day did not develop clinically apparent dementia. Thus, the actual frequency of radiationrelated dementia when using convention fractionation schedules is not known but is certainly less than 11%. In any event, the frequency of long-term neuropsychological side effects of WBRT in adult brain metastases patients appears to have been overestimated and seems to be within the acceptable range when modern fractionation schemes are employed.

WBRT in the Treatment of Multiple Brain Metastases Because the majority of patients have multiple metastases, the efﬁcacy data for WBRT comes in a large part from treatment of patients with multiple brain metastases. There is little doubt that WBRT is effective for multiple metastases, and it is used routinely. The main controversy regarding the treatment of multiple metastases involves the use of radiosurgery (either with or without WBRT). To date, there have been three randomized trials [20–22] assessing the efﬁcacy of radiosurgery in the treatment of multiple metastases. The ﬁrst randomized trial was reported by Kondziolka et al. [20]. In that study, 27 patients with multiple brain metastases were randomized to treatment with WBRTalone or WBRT plus a radiosurgery boost. The study was stopped early because the authors claimed to have found a large difference in the recurrence rates in favor of radiosurgery. Unfortunately, the study used nonstandard end points to measure recurrence. The investigators used any change in measurement of the lesion rather than the more usual 25% increase in diameter. No attempt was made to control for steroid use, radiation changes, or other factors that might produce small ﬂuctuations in lesion size on magnetic resonance imaging

(MRI). Also, a study with only 27 patients in it lacked the statistical power to support any meaningful conclusion, regardless of p values. As a result, this study was uninterpretable. A second study reported in abstract form by Chougule et al. [21] randomized patients with one to three brain metastases to treatment with radiosurgery-alone, radiosurgery plus WBRT, or WBRT-alone. The study had 109 patients. There was no statistically signiﬁcant difference in survival among the three treatment arms. Median survival times for the radiosurgery, radiosurgery plus WBRT, and the WBRT-alone treated groups were 7, 5, and 9 months, respectively. Local control rates in the brain were also not signiﬁcantly different. This trial suffered from several methodological problems. The most serious error was that 51 of the patients had had surgery for at least one symptomatic brain metastasis prior to entry into the study. No attempt was made to stratify for previous surgery or to otherwise ensure that surgical patients were equally distributed among the treatment groups. The inclusion of the surgical patients effectively made this a six-arm trial (the original three subdivided again into surgically treated patients and nonsurgically treated patients), and therefore, the size of this trial was not large enough to support a meaningful analysis. Also, because surgery is in all probability an effective therapy for brain metastases, the nonrandom distribution of surgically treated patients among the treatment arms substantially weakened the trial. Therefore, this study, although ostensibly negative, is really uninterpretable. A third study was reported by Andrews et al. [22] This study (RTOG 9508) contained 333 evaluable patients with one to three brain metastases who were randomized to treatment with either WBRT (37.5 Gy) plus radiosurgery or WBRT (37.5 Gy) alone. The primary end point was survival. Overall, there was no signiﬁcant difference in survival between the two treatment groups (median, 6.5 months for radiosurgery plus WBRT and 5.7 months for WBRT-alone, p = 0.1356). There was no survival beneﬁt from radiosurgery in patients with multiple metastases (median, 5.8 months for radiosurgery plus WBRT and 6.7 months for WBRT-alone, p = 0.9776). (However, for patients with single metastases, there was a signiﬁcant survival advantage favoring radiosurgery, median 6.5 months vs. 4.9 months, p = 0.0393). Lower posttreatment Karnofsky scores and steroid dependence were more common in the WBRT-alone group. Multiple subgroup analyses were made and a beneﬁt for radiosurgery plus WBRT was found in several subgroups that included patients with single and multiple metastases. These subgroups were RPA class 1 patients, patients with metastases size equal to or larger than 2 cm, and lung cancer patients with squamous cell histology. However, these subset analyses were not prespeciﬁed, and the p values needed for signiﬁcance should have been adjusted for inﬂation of type I error. When this was done, none of these subgroup analyses showed a positive beneﬁt for radiosurgery [23]. So, for multiple brain metastases, this was a completely negative trial with regard to the major end points prevention of death due to neurologic causes and overall survival. Radiosurgery has been put to the test in the treatment of multiple metastases and has not been established as effective. Therefore, based on the best available evidence, WBRT-alone is the treatment of choice for most patients with multiple brain metastases (and the word “multiple” in this context means more than one).

16.

brain metastases: whole-brain radiation therapy perspective

WBRT in the Treatment of Single Brain Metastases For the treatment of single brain metastases, randomized trials have established the superiority of focal treatment (either conventional surgery or radiosurgery) plus WBRT over treatment with WBRT-alone. Therefore, good-prognosis patients with single brain metastases should be treated with upfront surgery or radiosurgery. However, with the establishment of the efﬁcacy of focal treatments for single brain metastases, a new controversy has arisen as to whether adjuvant WBRT is really necessary after a “complete resection” or “successful” treatment with radiosurgery. Adjuvant WBRT is thought to be of beneﬁt because there may be residual disease in the tumor bed or at distant microscopic sites in the brain. However, brain metastases tend to be discrete masses that are theoretically capable of being removed totally or destroyed, and so WBRT may not be necessary after “successful” focal therapy. There are several reasons for eliminating WBRT. First, WBRT has adverse, long-term neuropsychological side effects. Second, there are also the costs and time commitment of the patient that must be considered. And ﬁnally, there is the possibility that WBRT may simply not be needed at all. It is theoretically possible to remove single brain metastases by surgery totally or to control them with radiosurgery. Furthermore, neuroimaging has improved, and it may now be possible to detect reliably additional metastases that may be present and treat these with additional focal therapy. If these last two statements are true, then there would be little justiﬁcation for adjuvant WBRT. On the other hand, compelling reasons exist for giving adjuvant WBRT. As a practical matter, it is probably impossible to remove completely all metastases with conventional surgery, and radiosurgery does not completely control the tumors. In addition, neuroimaging may not have reached the point yet where we can be absolutely certain that all metastases are being detected, and therefore some type of additional treatment may be needed. Also, although WBRT does have side effects, these side effects may not be as severe or as common as was previously thought. Furthermore, most patients with brain metastases have relatively limited overall survival times, and so the really serious long-term side effects are usually not an issue in their care. Two randomized trials [24, 25] have addressed the question of adjuvant WBRT in conjunction with focal treatment. A study published in 1999 by Patchell et al. [24] examined the effect of WBRT in conjunction with conventional surgery. In that study, 95 patients who had single brain metastases that were completely surgically resected were randomized to treatment with postoperative WBRT (50.4 Gy) or to observation with no further treatment of the brain metastasis (until recurrence). Recurrence of tumor anywhere in the brain was less frequent in the radiotherapy group than in the observation group (18% vs. 70%, p < 0.001). Postoperative radiotherapy prevented brain recurrence at the site of the original metastasis (10% vs. 46%, p < 0.001) and at other sites in the brain (14% vs. 37%, p < 0.01). As a result, patients in the radiotherapy group were less likely to die of neurologic causes than patients in the observation group (6 of 43 who died [14%] vs. 17 of 39 [44%]; p = 0.003). There was no signiﬁcant difference between the two

203

groups in overall length of survival or the length of time that patients remained functionally independent. The effect of WBRT in association with radiosurgery was examined in a randomized trial conducted by the Japanese Radiation Oncology Study Group and reported in abstract form by Aoyama et al. [25] In that study, 132 patients with one to four brain metastases were randomized to treatment WBRT plus radiosurgery or with radiosurgery-alone. The WBRT total dose was 30 Gy. The radiosurgery dose was 18 to 25 Gy at the periphery of the lesion in the radiosurgery-alone group and was reduced by 30% in the WBRT plus radiosurgery group. The median survival time was 7.5 months in the WBRT plus radiosurgery group and 8.0 months in the radiosurgery-alone group, p = 0.42. The 12-month brain metastases recurrence rates were signiﬁcantly (p < 0.001) different (47% in the WBRT plus radiosurgery group and 76% in the radiosurgery-alone group). Death due to neurologic causes and neurologic functioning were not signiﬁcantly different between the two groups. Despite the fact that both of the randomized trials [24, 25] showed clearly that WBRT prevented recurrences, these studies have actually provoked controversy rather than settling the issue. Results of these trials have been used as reasons both to give and not to give adjuvant WBRT. The justiﬁcation for not giving WBRT holds that because no survival difference was found in either of the trials, WBRT really adds nothing to the treatment. This argument fails on several counts. The Patchell study [24] used tumor recurrence as the primary end point and was not designed either to show a difference in survival or to rule one out. There was actually an increase of 11% in survival time in the WBRT group when compared with the observation group. The relative risk of improved survival with WBRT was 1.1. However, this was not a statistically signiﬁcant difference. Because there was a statistically signiﬁcant reduction in death due to neurologic causes, ultimately adjuvant WBRT might have had some positive impact on overall survival time. The estimated sample size required to detect a signiﬁcant difference of 11% in overall survival with adequate power would have been 1005 patients per group or 2010 patients total. For practical reasons, the study could not be designed to have this large of a sample size and, therefore, was not designed to detect moderate differences in survival, even one as large as 11%. There is an even stronger reason for discounting the apparent lack of efﬁcacy of postoperative WBRT with regard to length of survival in the Patchell trial [24]. Recurrence of tumor in the brain was the primary end point of that randomized trial, and this end point was the only truly direct measure of the effects of adjuvant WBRT. Up until recurrence of tumor, the two treatment groups were distinct, and the patients in each had received the treatment assigned by randomization. However, at recurrence, no speciﬁc treatment was mandated by the study design, and as a result, patients received a variety of additional treatments. There was an extremely large crossover of the observation group to WBRT. Of the 32 patients in observation group who developed recurrent brain metastases, 28 patients got WBRT. Overall, that means that 61% (28 patients of 46 total) in the “no WBRT” observation arm were, in fact, treated ultimately with WBRT. For the purposes of length of survival and functional independence, the study was virtually a comparison of surgery plus immediate WBRT versus surgery plus delayed

204

r.a. patchell and w.f. regine

WBRT. This substantially diluted the effect of WBRT given immediately postoperatively because WBRT probably improved the length of survival and functional independence in the observation group. Therefore, for all the reasons given above, the Patchell study did not “prove” that WBRT had no positive effect on survival. The Aoyama trial [25] also failed to ﬁnd any difference in survival with the addition of WBRT. This study differed in design from the Patchell study and was devised with survival as the primary end point; it was originally powered to be able to detect a 30% or greater difference with 89 patients per treatment arm. (Given that the Patchell study [24] and numerous retrospective studies had failed to ﬁnd a difference in survival even close to 30%, the Aoyama study [25] should have been powered as an equivalence trial [if survival was going to be used as the primary end point].) The study was stopped after an interim analysis at 132 patients indicated that the study was not going to show a statistically signiﬁcant difference in survival, even if all of the 178 patients originally projected to be in the trial had been randomized. Therefore, any negative conclusion based on the original power calculation, that is, that there was no more than a 30% difference in survival (or any other reasonable difference), cannot be supported. This study also had an 11% crossover that further eroded the strength of any conclusions about survival. In addition, about half of the patients in both treatment arms had multiple brain metastases. The study by Andrews et al. [22] failed to establish the efﬁcacy of radiosurgery in the treatment of multiple metastases, and so the inclusion of these patients is problematic and further reduces the number of valid patients on which to base a conclusion about survival beneﬁt. Thus, like the Patchell trial [24] before it, this Aoyama study [25] was unable to show that there was no difference in survival. Therefore, arguments based on the supposed lack of efﬁcacy of (immediate) postoperative WBRT on survival are based on a misunderstanding of the design and limits of the randomized trials. In addition, the Aoyama study demonstrated that omitting WBRT does not produce any difference in either gross neurologic or neurocognitive functioning. From this information, Aoyama et al. [25] and the Journal of the American Medical Association (JAMA) editorial writer [26] concluded that the addition of WBRT is not necessary and can be safely omitted in the treatment of most patients with brain metastases. However, even if one takes the data presented in the paper at face value, it is possible to draw exactly the opposite conclusion. As stated by the authors and the editorial writer, the main reason for not giving WBRT is to avoid the long-term neurotoxic effects of WBRT. Yet, this study found no difference in neurologic functioning, neurocognitive functioning, gross radiation-induced side effects, or survival times between the two groups. In fact, deterioration in neurologic function attributable to progression of brain metastases was observed in 59% of patients in the WBRT group and 86% in the SRS-alone group (p = 0.05) indicating a signiﬁcantly higher rate of neurologic deterioration as a consequence of tumor progression in patients when WBRT is withheld. Thus, at very least, WBRT appears to signiﬁcantly reduce the recurrence of brain metastases without demonstrable neurotoxicity. Therefore, the trial by Aoyama et al. [25] seems to support strongly the use of WBRT

as upfront treatment in the management of most patients with brain metastases. The most forceful argument in favor of adjuvant WBRT involves an examination of the effects of not giving WBRT. Patients who do not receive adjuvant WBRT suffer substantially more recurrent brain metastases than patients who are treated with WBRT. As previously noted, the harmful side effects of WBRT appear to have been overestimated in the past and are probably in the acceptable range. Unfortunately, the same cannot be said of the side effects of recurrence of brain metastases. Several studies [27, 28] have demonstrated that the recurrence of brain metastases has a negative effect on the neurocognitive functioning of patients. A study by Regine et al. [27] found that in 36 patients with brain metastases treated with SRS alone, 47% had recurrence of brain metastases and 71% of the recurrences were symptomatic. Signiﬁcantly, 59% of the patients with recurrent tumors had associated neurologic deﬁcits and 17% were unable to undergo salvage brain therapy because of their overall poor general status associated with brain tumor recurrence. These ﬁndings are now substantiated by the level 1 evidence provided by the Aoyama phase III trial where deterioration in neurologic function attributable to progression of brain metastases was observed in 59% of patients in the WBRT group versus 86% in the SRS-alone group (p = 0.05); indicating a signiﬁcantly higher rate of neurologic deterioration as a consequence of tumor progression in patients when WBRT is withheld [25]. Another study by Regine et al. [28] showed that, at 3 months after treatment, patients treated for brain metastases with WBRT had greater negative changes in their mini-mental status examinations with uncontrolled brain tumors than they did with controlled brain tumors (−6.3 points versus −0.5 points, p = 0.02). Also relevant (but perhaps somewhat farther aﬁeld) was a study by Taylor et al. [29] showing that, in patients with primary brain tumors at 12 months after treatment, changes in mini-mental status examinations were worse in patients with uncontrolled tumors (−2.42 points) than in patients with controlled tumors (+0.076 points) (p = 0.0046). All of the patients in this study had received large total doses of conventional radiation therapy. These studies all strongly suggest that uncontrolled brain tumors result in a substantial decrease in mental performance and that this reduction far outweighs any decrement seen with cranial radiation therapy. Therefore, the side effects of recurrent tumors are worse than the side effects of preventive treatment. This is an extremely strong argument for the use of adjuvant WBRT in association with focal therapy.

WBRT for Recurrent Brain Metastases Brain metastases often recur, and CNS progression may be accompanied by systemic tumor progression and a decline in functional status. In general, the same types of treatment used for newly diagnosed brain metastases are also available for recurrent tumors. However, the type of previous therapy may limit the therapeutic options available at recurrence, and the development of radioresistance is not uncommon.

16.

brain metastases: whole-brain radiation therapy perspective

If patients have not already had WBRT, they should be treated with it on recurrence. However, often patients with recurrences have already been treated with WBRT, and this limits the amount of subsequent radiation that can be given safely. The amount of additional radiation that can be offered is usually in the range of 1500 to 2500 cGy, a total dose usually inadequate to control tumor growth. Several retrospective studies [30–34] have attempted to asses the efﬁcacy of salvage WBRT. It is difﬁcult to assess efﬁcacy from these reports. Rates of improvement ranged from 27% to 70%; however, the range of duration of response was fairly uniform and was 2.5 to 3 months. The median survival ranged from 1.8 to 4.0 months. Relatively few long-term complications were reported; however, because the median survival is quite short, most patients did not live long enough to develop the long-term complications of radiation. The problem with the interpretation of these studies is that they often used different end-point measurements for improvement and had heterogeneous patient populations. Some included patients with poor performance status and extensive disease and others selected out favorable subgroups for radiation. One recommendation based on a retrospective study [30] is to restrict reirradiation to patients who showed an initial favorable response to radiotherapy, had a longer disease-free interval, and who remain in good general condition when the cerebral recurrence develops. However, even in this favorable subgroup, only 42% of patients showed symptomatic improvement, and the median survival after reirradiation was 5 months. Despite such relatively poor results, additional radiotherapy is frequently one of the few treatment options for patients with recurrent disease.

References 1. Berk L. An overview of radiotherapy trials for the treatment of brain metastases. Oncology 1995; 9:1205–1219. 2. Posner JB. Management of brain metastases. Rev Neurol 1992; 148:477–487. 3. Cairncross JG, Posner JB. The management of brain metastases. In: Walker MD, ed. Oncology of the Nervous System. Boston: Martinus Nijhoff, 1983:341–377. 4. Horton J, Baxter DH, Olson KB. The management of metastases to the brain by irradiation and corticosteroids. Am J Roentgenol Radium Ther Nucl Med 1971; 111:334–335. 5. Markesbery WR, Brooks WH, Gupta GD, et al. Treatment for patients with cerebral metastases. Arch Neurol 1978; 35:754–756. 6. Ruderman NB, Hall TC. Use of glucocorticoids in the palliative treatment of metastatic brain tumors. Cancer 1965; 18:298–306. 7. Chang DB, Yang PC, Luh KT, et al. Late survival of non-small cell lung cancer patients with brain metastases. Chest 1992; 101: 1293–1297. 8. Borgelt B, Gelber R, Kramer S, et al. The palliation of brain metastases: ﬁnal results of the ﬁrst two studies by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1980; 6: 1–9. 9. Gelber RD, Larson M, Borgelt BB, et al. Equivalence of radiation schedules for the palliative treatment of brain metastases in patients with favorable prognosis. Cancer 1981; 48:1749–1753. 10. Kurtz JM, Gelber RD, Brady LW, et al. The palliation of brain metastases in a favorable patient population: a randomized clinical trial by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1981; 7:891–895.

205

11. Deiner-West M, Dobbins TW, Phillips TL, et al. Identiﬁcation of an optimal subgroup for treatment evaluation of patients with brain metastases using RTOG study 7916. Int J Radiat Oncol Biol Phys 1989; 16:669–673. 12. Epstein BE, Scott CB, Sause WT, et al. Improved survival duration in patients with unresected solitary brain metastasis using accelerated hyperfractionated radiation therapy at total doses of 54.4 gray and greater. Results of the Radiation Therapy Oncology Group 85–28. Cancer 1993; 7:1362–1367. 13. Murray KJ, Scott C, Greenberg H, et al. A randomized phase III study of accelerated hyperfractionated versus standard radiation in patients with unresected brain metastases: a report of the Radiation Therapy Oncology Group (RTOG) 9104. Int J Radiat Oncol Biol Phys 1997; 39:571–574. 14. Hoskins PJ, Crow J, Ford HT. The inﬂuence of extent and local management on the outcome of radiotherapy for brain metastases. Int J Radiat Oncol Biol Phys 1990; 19:111–115. 15. Eyre HJ, Ohlsen JD, Frank J, et al. Randomized trial of radiotherapy versus radiotherapy plus metronidazole for the treatment metastatic cancer to the brain. J Neuro-Oncology 1984; 2:325–330. 16. Komarnicky LT, Phillips TL, Martz K, et al. A randomized phase III protocol for the evaluation of misonidazole combined with radiation in the treatment of patients with brain metastases (RTOG-7916). Int J Radiat Oncol Biol Phys 1991; 20:53–58. 17. 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; 21:2529–2536. 18. Suh JH, Stea B, Nabid A, et al. Phase III study of efaproxiral as an adjunct to whole brain radiation therapy for brain metastases. J Clin Oncol 2006; 24:106–114. 19. DeAngelis LM, Mandell LR, Thaler HT, et al. The role of postoperative radiotherapy after resection of single brain metastases. Neurosurgery 1989; 24:798–805. 20. 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; 45:427–434. 21. Chougule PB, Burton-Williams M, Saris S, et al. Randomized treatment of brain metastases with gamma knife radiosurgery, whole brain radiotherapy or both. Int J Radiat Oncol Biol Phys 2000; 48:114. 22. Andrews DW, Scott CB, Sperduto PW, et al. Whole brain radiation therapy with and without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomized trial. Lancet 2004; 363:1665– 1672. 23. Elveen T, Andrews DW. Summary of RTOG 95-01 phase III randomized trial of whole brain radiation with and without stereotactic radiosurgery boost, including presentation of a clinical case study. Am J Oncol Rev 2004; 3:592–600. 24. Patchell RA, Tibbs PA, Regine WF, et al. Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial. JAMA 1998; 280:1485–1489. 25. Aoyama H, Shirato H, Tago M, et al. Stereotactic radiosurgery plus whole-brain radiation therapy vs Stereotactic radiosurgery alone for the treatment of brain metastases. JAMA 2006; 295: 2483–2491. 26. Raizer J. Radiosurgery and whole-brain radiation therapy for brain metastases. JAMA 2006; 295:2535–2536. 27. Regine WF, Scott C, Murray K, Curran W. Neurocognitive outcome in brain metastases patients treated with acceleratedfraction vs. accelerated-hyperfractionated radiotherapy: an analysis from Radiation Therapy Oncology Group Study 91-04. Int J Radiat Oncol Biol Phys 2001; 51:711–717.

206

r.a. patchell and w.f. regine

28. Regine WF, Huhn JL, Patchell RA, et al. Risk of symptomatic brain tumor recurrence and neurologic deﬁcit after radiosurgery al in patients with newly diagnosed brain metastases: results and implications. Int J Radiat Oncol Biol Phys 2002; 52:333–338. 29. Taylor BV, Buckner JC, Cascino TL, et al. Effects of radiation and chemotherapy on cognitive function in patients with highgrade gliomas. J Clin Oncol 1998; 16:2195–2201. 30. Hocht S, Wiegel T, Hinkelbein W. Reirradiation for recurrent brain metastases. Controversies in Neuro-Oncology. Front Radiat Oncol Basal 1999; 33:327–331.

31. Hazuka MB, Kinzie JJ. Brain metastases: results and effects of reirradiation. Int J Radiat Oncol Biol Phys 1988; 15:433–437. 32. Kurup P, Reddy S, Hendrickson FR. Results of re-irradiation for cerebral metastases. Cancer 1980; 46:2587–2589. 33. Wong WW, Schild SE, Sawyer TE, et al. Analysis of outcome in patients reirradiated for brain metastases. Int J Rad Oncol Biol Phys 1996; 34:585–590. 34. Cooper JS, Steinfeld AD, Lerch IA. Cerebral metastases: value of reirradiation in selected patients. Radiology 1990; 174:883– 885.

1 7

High-Grade Gliomas David Roberge and Luis Souhami

Introduction Primary brain tumors are classiﬁed according to their predominant cell type. Glial neoplasms are the most common primary intracranial malignancies and are classiﬁed as astrocytic tumors, oligodendroglial tumors, and ependymal tumors. Astrocytomas are the most common variant of glioma, and most adult astrocytomas are of high grade. Of these, glioblastoma multiforme (GBM), or astrocytoma grade IV, represents the most common subtype [1], making this the most common malignant tumor in adults. Despite the best contemporary use of surgery, radiation, and chemotherapy, long-term survival of patients with GBM has been distinctly uncommon. In unselected series, the 5-year survival has been as low as 2% [2, 3]. Anaplastic astrocytoma (AA), or astrocytoma grade III, represents less than 20% of malignant gliomas and is associated with a more favorable outcome [1, 4]. For the scope of this chapter, only high-grade astrocytic tumors will be discussed. High-grade astrocytic tumors are inﬁltrating tumors that can rapidly enlarge, resulting in various signs and symptoms, such as focal or generalized seizures, headache, visual disturbances, speech disturbances, changes in mental status, and motor or sensory deﬁcits. Malignant glial tumors are thought to evolve from an accumulation of multiple genetic aberrations in normal precursor cells. In a stepwise fashion, this accumulation of deleterious genetic alterations may lead to transformation to a low-grade glioma and to the subsequent aggressive phenotype associated with high-grade tumors. Several molecular studies have generated multiple markers linked with malignant gliomas, including chromosomal deletion, addition, mutation, and gene ampliﬁcation, which may have important clinical implications [5]. Both GBM and AA can be of two types based on their clinical presentation. Primary high-grade astrocytomas occur de novo and are associated with short duration of symptoms and worse prognosis, and secondary high-grade gliomas often occur in patients with a previous low-grade astrocytoma, suggesting a different pathogenesis [6]. High-grade glioma (HGG) is a local disease. Distant spread is a rare event, and 90% of recurrences are located within 2 cm of the original enhancing lesion [7–10]. Overall, the prognosis of HGG is related to tumor grade, performance status, age, and treatment. Five other prognostic factors were

incorporated in a 1993 prognostic scheme based on recursive partitioning analysis (RPA) of the individual patient data from three Radiation Therapy Oncology Group (RTOG) trials [4]. Using several prognostic variables (Table 17-1), patients were divided into six classes with median survivals ranging from 4.7 to 58.6 months. The RPA classiﬁcation is now frequently used in the comparison of treatment results from different series. Despite intermittent waves of enthusiasm regarding various treatment modalities, the survival of patients with HGG has not increased substantially from 1950 to 2000. In this chapter, we describe and contextualize the use of stereotactic radiosurgery (SRS) and fractionated stereotactic radiotherapy (F-SRT) in the management of primary and recurrent malignant gliomas.

Historical Perspective The poor prognosis of HGG is known for many years. In 1926, Bailey and Cushing, in reference to what they at the time called “spongioblastoma multiforme,” made the following observation: “Operative procedures, howsoever radical [block extirpations repeated on signs of recurrence; saturation with X-rays or radium emanations after wide decompression with or without surgical interference with the tumor], have apparently done little more than to prolong life, save vision, and alleviate headache for an average of a few months. . . . Whether deep Roentgenization ever does more than hold the growth temporarily in check is problematical” [11]. By the time Elvidge, Penﬁeld, and Cone published their “McGill series” in 1937, the name “glioblastoma multiforme” had become widely accepted, and neurosurgery was still faced with the same “difﬁcult human problem” posed by these patients [12]. Thirteen years after the publication of Elvidge and colleagues [12], a case history was published as part of a paper reviewing 70 glioblastoma multiforme patients treated at the Monteﬁore Hospital [13] (Case Study 17-1). The described case represented the longest surviving patient in the literature of the day [13]. Prognostic factors for

207

208

d. roberge and l. souhami

TABLE 17-1. RTOG recursive partitioning analysis classes. Class

Deﬁnition

Median survival (months)

Two-year survival (%)

I

Age <50, AA, normal mental status

58.6

76

II

Age ≥50, KPS 70–100, AA, ≥3 months from ﬁrst symptoms to treatment

37.4

68

III

Age <50, AA, abnormal mental status or Age <50, GBM, KPS 90–100

17.9

35

IV

Age <50, GBM, KPS <90 or Age ≥50, KPS 70–100, AA, ≥3 months from ﬁrst symptoms to treatment or Age >50, GBM, surgical resection, good neurologic function

11.1

15

V

Age ≥50, KPS 70–100, GBM, either surgical resection and poor neurologic function or biopsy only followed by ≥54.4 Gy EBRT or Age ≥50, KPS <70, normal mental status

8.9

6

VI

Age ≤50, KPS <70, abnormal mental status Age ≥50, KPS 70–100, GBM, biopsy only, <54.4 Gy EBRT

4.6

4

Source: 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; 85:704–710.

the Monteﬁore series were reviewed: “An analysis of factors in survival revealed that operation had no signiﬁcant effect on longevity and irradiation a slight effect, if any. The longest survivals occurred in patients with onset at 24 to 42 years of age. However, age at onset was not a constant prognostic factor. The sex of the patient, the location of the neoplasm, or the histologic appearance gave no indication as to longevity. A signiﬁcant lengthening of survival was found in those patients whose initial symptom was a motor seizure.” If patients dying within 1 week of surgery were excluded, the median overall survival was 13.1 months with postoperative irradiation and 12

Case Study 17-1 Case 1. S.S. Autopsy no. 9588. “In 1929, at the age of 37, this right-handed white male had a convulsive seizure lasting a few minutes. . . . In 1931, following an attack, he noticed loss of sensation on the right side of the body and a right hemiparesis. In 1933, a partial motor aphasia developed. Dr. Paul C. Bucy operated on him in October 1933. A ﬁrm area measuring 2 cm in diameter was found on the left side in the gyrus just anterior to the precentral gyrus. . . . An attempt at complete removal was not made because of the location. A small piece was taken for study, . . . The tissue was diagnosed by Dr. Percival Bailey as a malignant glioma, probably a glioblastoma multiforme. . . . Following the operation, the patient was given a course of roentgen therapy for a total of 6068 r. . . . In September 1937, he was struck by an automobile, and was unconscious for 24 hours. Sometimes thereafter his seizures returned. He gradually lost power on the right side, and the aphasia became worse. In October 1939, he was seen by Dr. Leo Davidoff. . . . An encaphalogram was interpreted as showing left-sided cerebral atrophy without evidence of regrowth. Nothing further was done.

months without. It must however be noted that for all cases of prolonged survival reported in this paper, the patient had received irradiation. As demonstrated by the previous series, irradiation had been in common use long before randomized trials demonstrating its efﬁcacy. In 1947, Bush and Christensen wrote: “While it is very difﬁcult to judge of the effect of radiation in these cases, it is our impression that this therapy is of deﬁnite value. A halfhearted attempt of giving every other patient Roentgen therapy was never carried through as we felt ourselves unable to deprive the particular patient of what we felt was the best chance” [14].

He entered Monteﬁore Hospital in August 1939 at the age of 49. . . . there were now astereognosis in the right hand, a partial motor aphasia, and some blurring of the right disc margin. . . . During his stay in the hospital, both focal motor and generalized seizures were noted. Irradiation was not deemed advisable, and he was discharged to be followed in the outpatient clinic. He had difﬁculties in expressing himself. Because of this, he insisted on being re-admitted in July 1941. At this time it was felt that the paretic side was more spastic, and that the motor aphasia now had a sensory component. . . . He was given a course of roentgen-therapy amounting to 3000 r. with slight improvement. Thereafter he continued to have right-sided seizures until July 1943, when he rapidly became worse and died. . . . Autopsy. A broncho-pneumonia was found to be the immediate cause of death. The brain weighed 1420 gm. . . . There was a hemorrhagic tumor nodule, measuring 3 × 2 cm. in the left 3rd frontal convolution. . . . The tumor was a glioblastoma multiforme. There were some spongioblastic portions, but there were, in addition, extensive necrosis, pseudopallisading around focal necrotic zones, thrombosis of vessels and endothelial proliferation, perivascular lymphocytes, numerous mitotic ﬁgures, and pleomorphism.”

17.

high-grade gliomas

It was only in the early 1980s that both the Brain Tumor Cooperative Group (BTCG) and the Scandinavian Glioblastoma Study Group (SGSG) published results of randomized trials of adjuvant irradiation [15, 16]. These trials conﬁrmed a large and signiﬁcant increase in median survival after the administration of postoperative irradiation compared with surgery alone. In the Scandinavian trial, 45 Gy of whole-brain irradiation increased the median survival from 5.2 months to 10.8 months. In BTCG study 6901, median survival increased from 3.2 months to 8.1 months after 50 to 60 Gy of whole-brain radiotherapy. A randomized trial by the Medical Research Council [17] conﬁrmed an improvement in median survival from 9 to 12 months when 60 Gy was compared with 45 Gy. This trial suggested a possible dose-control relationship. For HGG, 60 Gy has been considered the standard postoperative radiotherapy dose by most investigators. Initially as part of the same trials looking at adjuvant irradiation, a large number of randomized trials have been conducted testing the use of concurrent and/or adjuvant chemotherapy agents and radiation sensitizers in the management of HGG [15, 18–22]. In 1993, a meta-analysis by Fine et al. summarized data from 16 randomized trials involving more than 3000 patients [23]. Optimistic physicians saw a large (52%) difference in 2-year survival; pessimistic physicians saw a relatively small difference in median survival. A subsequent meta-analysis from the Glioma Meta-analysis Trialists group conﬁrmed a 1.5-month increase in median survival and failed to demonstrate any beneﬁt of multidrug regimens over singleagent nitrosourea compounds [24]. Adjuvant BCNU chemotherapy was considered standard of care by most although it was used variably in clinical practice. Recently, in a joint European Organization for Research and Treatment of Cancer (EORTC) and National Cancer Institute of Canada (NCIC) randomized trial [25], concomitant and adjuvant temozolomide signiﬁcantly increased the median survival of patients diagnosed with glioblastoma multiforme from 12.1 months to 14.6 months and more than doubled the 2-year survival (26.5% vs. 10.4%). It is of interest to note that the control arm (radiotherapy-alone) had the same median survival as reported in the 1949 series from the Monteﬁore Hospital, highlighting how little progress had been made in the past 50 years. Based on this EORTC/NCIC trial, the current standard of care for GBM patients with a performance status of 0 to 1 (on subgroup analysis, there was no beneﬁt for patients with an ECOG performance status of 2) is external beam radiotherapy to 60 Gy and temozolomide. Although patients with AA histology were not included in the trial, this regimen will likely also become the de facto standard for these patients as well.

Rationale for Stereotactic Radiosurgery Inﬁltrating high-grade glial neoplasms would appear to be poor candidates for the stereotactic application of single-fraction irradiation. These tumors are hypoxic, acute-responding [26], and admixed with normal tissue [27, 28]. Despite these biological roadblocks, stereotactic radiation has been pursued in the management of HGG. The use of stereotactic radiosurgery

209

(SRS; deﬁned here as a radiotherapy technique characterized by accurate delivery of high doses of radiation in a single session to small intracranial targets in such a way that the dose fall-off outside the target volume is very sharp) has been based on the pattern of failure of this disease, the dose-response data from external beam radiation, and early data from interstitial brachytherapy trials [29, 30]. As mentioned previously, 90% of patients recur within 2 cm of the contrast-enhancing lesions–-this despite the fact that tumor cells can be found pathologically at larger distances, often following the peritumoral edema. Furthermore, multicentric or metastatic disease is rare [10]. Moreover, in an analysis of the Brain Tumor Study Group data, a dose-response relationship has been shown for doses of 50 to 60 Gy [31]. In this data set, the median survival increases from 28 weeks at 50 Gy to 42 weeks at the 60 Gy level. Signiﬁcant improvement in median survival was also observed in the randomized British trial comparing 45 Gy to 60 Gy [17]. Thus, all of these characteristics made SRS an attractive option to be used as a focal boost in selected patients with HGG. In the early 1990s, phase I/II data suggested that interstitial brachytherapy improved local control and survival in selected primary and recurrent HGG patients. At University of California San Francisco (UCSF), the median survival of the ﬁrst 18 patients treated with interstitial brachytherapy for recurrent GBM was 52 weeks with 2 patients surviving more than 5 years [32]. In a subsequent North Carolina Oncology Group (NCOG) study [30], 107 patients with HGG were enrolled in a program of brachytherapy added to EBRT and adjuvant procarbazine, lomustine and vincristine (PCV) chemotherapy. In the 63 evaluable patients who were actually implanted, the median survivals were 157 weeks and 88 weeks, respectively, for patients with grade III astrocytoma and GBM. Brachytherapy was associated with a high risk of radionecrosis, and it was believed that SRS might offer the focal dose escalation beneﬁts of implants with lessened toxicity. Unfortunately, concerns about selection bias [33], as a possible explanation for these improved median survivals after brachytherapy, appear to be have been conﬁrmed by two subsequent negative phase III trials of brachytherapy by the BTCG [34] and the Princess Margaret Hospital [35] groups. In these randomized studies, the added use of a focal brachytherapy boost did not lead to an improvement in survival in patients harboring a HGG. In the Toronto trial, the median survival was 13.2 months on the standard arm and 13.8 months on the brachytherapy arm, and in the BTCG study it was 58.8 weeks on the standard arm and 64.1 weeks on the brachytherapy arm. The cumulative proportion surviving between the two treatment groups was not statistically signiﬁcantly different in either study. In 1985, Columbo et al. reported on the ﬁrst patients treated in Vinceza, Italy, with a new technique for linear accelerator (linac)-based radiosurgery using non-coplanar arcs [36]. Only 6 of 22 patients had adequate follow-up. Of these, one patient was treated for a 3.5-cm grade III astrocytoma. Two doses of 20 Gy were delivered during separate procedures. This patient worsened within 2 months and was reoperated. After this early experience, several institutional reports have been published on the use of SRS boost in the management, at presentation or at time of recurrence, of patients with HGG. The noninvasive

210

d. roberge and l. souhami

FIGURE 17-1. Publications on the radiosurgical treatment of HGG.

nature of SRS coupled with the promising early results of interstitial brachytherapy led to the inﬂated enthusiasm for the technique in patients with HGG and a growing number of publications became available over the years (Fig. 17-1). Most of these reports were published before the results of the randomized trials of brachytherapy and SRS were available. Single-fraction stereotactic radiation has predominately been used in one of two scenarios: in primary lesions as a supplement to wider volume external-beam radiation therapy (EBRT) and in recurrent lesions as a single modality. With the advent of noninvasive immobilization devices, the use of fractionated stereotactic radiation therapy (F-SRT) has become more common. Initially, F-SRT was explored to increase the therapeutic ratio in previously irradiated patients; it has since also been explored as a boost in the treatment of newly diagnosed HGG patients.

Treatment Planning Although each case must be approached on an individual basis, the following general guidelines represent a reasonable approach to the delivery of stereotactic therapy for HGG.

possible rapid tumor progression. Completely resected tumors are generally not considered for SRS boosts. The maximum clinically tolerated doses for the treatment of HGG has been derived from the RTOG 90-05 phase I study and is tumor size dependent. In the study by Shaw et al. [37], single doses of 24, 18, and 15 Gy were found to be the maximum tolerated doses for tumor diameters of ≤20 mm, 21 to 30 mm, and 31 to 40 mm, respectively. These doses are usually prescribed to an isodose surface between 50% and 90%. Caution should be exercised when treating brain-stem lesions and lesions within 10 mm of the optic chiasm, as these lesions were not included in the RTOG study and the maximum tolerated radiosurgery dose has not been clearly established for these structures. For recurrent disease, similar principles apply. It is a general rule that only lesions smaller than 4 cm are considered for SRS. In these cases, the GTV is the enhancing lesion on a gadolinium-enhanced T1 MRI (or contrast-enhanced CT scan). On a case-by-case basis, a clinical target volume (CTV) of a few millimeters (∼2 to 5 mm) can be added. For rigid immobilization systems, no additional margin is added for the PTV. Table 17-2 contrasts commonly used treatment parameters for stereotactic irradiation in HGG.

F-SRT For both primary and recurrent lesions, larger volumes can be considered for F-SRT than SRS. Depending on the planned dose of F-SRT and the use (concurrent or prior) of EBRT, volumes of up to approximately 100 cm3 can be considered for treatment. Margins will depend on the volume to be treated, proximity of critical structures, prior treatment, and immobilization device used. A reasonable schema would have the GTV equal the enhancing lesion on a gadolinium-enhanced T1 MRI (or contrast-enhanced CT scan), no CTV margin, and a PTV of 2 to 3 mm (corresponding with the reproducibility of a common three-ply thermoplastic mask-based immobilization system [38, 39]). Table 17-3 [40–47] contains various published fractionation schemes. Depending on the number of isocenters, dose would be prescribed to the 50% to 90% isodose surface after assessing the plan with regard to volume irradiated, subjective isodose distribution, conformity, homogeneity, and dose to critical structures. Currently, there is no evidence to suggest that higher MDPD (maximum dose/prescription dose) ratios should be favored.

SRS In the primary treatment of HGG, SRS is used as a boost to EBRT. The goal of radiosurgery is to inﬂict precise damage to tissue within the target volume, in this case proliferating glial cells. Thus, only tumors with a limited diameter (at presentation or postoperatively) should be considered for SRS. The RTOG considered a diameter of 40 mm the maximum diameter allowed for patient entry into the SRS trials. The target volume (PTV) is the tumor (as viewed on computed tomography [CT] or magnetic resonance imaging [MRI]) without margins. If MRI registration is used, a contrast-enhanced CT scan is obtained on the day of the radiosurgery to prevent mistargeting because of

TABLE 17-2. Comparison of common parameters for stereotactic irradiation modalities. Parameter

SRS

F-SRT

Tumor size

≤40 mm 15 to 24 Gy Single 50% to 80% None Rigid

≤60 mm Variable Multiple 80% to 90% 2 to 5 mm Relocatable

Total dose Fractions Prescription isodose Margins Stereotactic frame

17.

211

high-grade gliomas

TABLE 17-3. Results of F-SRT for primary and recurrent high-grade gliomas. Lederman [43]‡

Regine [44, 45]

Baumert [46]

Cho [47]

Staten Island

Kentucky

Zurich

Minnesota

2000 88 recurrent

2000 8 primary 1 recurrent GBM 6 AA 2 LGG 1

2003 17 primary

2004 10 primary

GBM 15 AA 2

GBM 10

Author

Glass [40]*

Shepherd [41]

Hudes [42]

Institution

Temple

Date of publication Number of patients

1997 20 recurrent

Royal Marsden 1997 33 recurrent

Thomas Jefferson 1999 20 recurrent

Histology

GBM 13 AA 7

AA 29 AO 3

GBM 19 AA 1

GBM

Median age (range)

44.5 (6–73)

37 (19–55)†

52 (26–77)

56 (21–82)

≥18

51 (25–64.8)

57 (17–80)

KPS

—

80 (60–100)†

80 (60–100)

70 (50–90)

≥60

70 (60–90)

Time from Diagnosis (for recurrent lesions) Median follow-up (months) Median tumor volume (range) Median peripheral dose (range) Median survival (months) 1-year survival (%) 2-year survival (%) Median prescription isodose (range) EBRT

—

29†

3.1 (0.8–45.5) from EBRT

N/A

(5/12 WHO 0/1, 7 WHO 2) N/A

N/A

—

—

>12

15†

25 (9–50)

—

14.3 (1.76–122)

24 (3–93)†

12.66

32.7 (1.5–150.3)

7.4 (1.5–27.2)†

42/7 (—)

35/7 (20/4–50/10) AA 11

30/10 (21/7–35/10) 10.5

24/4 (18/4–36/4) 7

14–28/2–4

— — 90 (80–90)

20 — 89 (80–95)

17 2 90 (80–90)

56 44 50

All prior EBRT (45–60) 6†

All prior EBRT 60 (44–72) 25 (for progression) No GR III

99%

12.7 — — ≥70 All prior EBRT?

Reoperation (%)

25

Toxicity

15% necrosis

45% at 24 months†

8.5

6.5

34 (4–70) 15 20/5 2 10/2 20

27.5/11 (20–35/8–14) 15.9 67 — 88 (75–90)

59.4/33

77 42 Prescribed at isocenter 60

60 (50.4–60)

12

57†

11/17

20

—

4/15 GR IV†

6% necrosis

10% necrosis

11

*With cis-platinum. †Of a larger group including other histologies. ‡With Taxol.

Linac Versus Gamma Knife HGGs are large, inﬁltrating tumors where submillimeter accuracy in delivery should not be an issue. Because of the necrotic and presumed radioresistant core of high-grade tumors, there might be a theoretical advantage to more inhomogeneous plans. As Gamma Knife (GK) treatments commonly use multiple isocenters and are prescribed to lower isodose surfaces, they might offer a theoretical advantage over single-isocenter plans (if this were ever clinically veriﬁed, linac centers could then choose to prescribe treatments to lower isodose surfaces). This hypothesis was tested by Larson and colleagues at UCSF in a phase II trial [48]. In this trial, prescriptions were made to isodoses as low as 25%. The results, however, appear no different than would be expected with conventional treatment (15 weeks median survival for patients with recurrent grade IV tumors).

A review of the single-institution series will reveal a mix of GK and linac-based treatments. The two largest series are from Pittsburgh [49] and Boston [50]–-prototypical GK and linacbased programs. Not unexpectedly, there is no difference in result between these two series–-both reporting a median survival of 20 months for the treatment of primary tumors. In the initial RTOG phase I trial of radiosurgery for recurrent tumors, there appeared to be a large difference in results favoring GK in the treatment of a mixed bag of primary brain tumors [51]. This is not borne out in the much larger and more homogeneous experience of the RTOG 93-05 randomized trial [52]. In this trial, in a subgroup analysis, patients treated with linac systems had a 14-month survival, not statistically different from the 12.1-month survival of GK patients (Fig. 17-2). If patients are to be treated with radiosurgery, the delivery system should not be an issue.

212

d. roberge and l. souhami 100 Linac

Survival Rate

80

Gamma Knife

60

40

20

0 0

6

12

18 24 30 36 Months FIGURE 17-2. Survival by SRS treatment delivery system in RTOG 93-05. (Reprinted from 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; 60:853–860. With permission from Elsevier.)

SRS in the Primary Management of HGGs The 1990s saw the publication of several small, retrospective and prospective, series of patients treated with SRS boost in the primary treatment of malignant glioma (Table 17-4) [53–58]. Ages ranged from 3 to 84 years and Karnofsky performance score (KPS) from 50 to 100. Overall, the median survival of patients treated with SRS was quite encouraging, although some series reported no beneﬁt in survival when comparing results with historical controls. However, in a disease where more than 95% of patients will ultimately die of their disease, the median survival is dictated more by patient-related than therapy-related factors [4]. By the very nature of SRS, patients are selected for small tumors, more complete resection (size determination being based on postoperative imaging), and a good response to initial therapy (when SRS is administered after EBRT ± chemotherapy, patients with progressive disease or decreasing performance status are likely to be excluded). In addition, trials commonly require a good performance status (KPS of 50 [49], 70 [50, 55], or 90 [56]). Other potential sources of bias include high patient motivation, more favorable tumor biology, and aggressive treatment of recurrences. Taking these selection factors into account, only one-tenth to one-quarter of GBM patients will be eligible for radiosurgical boost [59]. The results of radiosurgical series were viewed with skepticism. In an attempt to reduce bias, authors sought out retrospective control populations. One of the largest reviews [60] attempted to retrospectively stratify 115 patients from 3 institutions (Harvard, the University of Florida, and the University of Wisconsin) according to the prognostic classes of the RTOG recursive partitioning analysis of patients enrolled on RTOG 74-01, 79-18, and 82-02 [61–63]. This analysis concluded that there was a signiﬁcant improvement in both 2-year and median survival favoring SRS-treated patients (Table 17-5). Unfortunately, along with other ﬂaws, the RTOG classes are broad, do

not include all known prognostic factors (most notably tumor size), and convert important linear variables into binary ones (age, mental status, and KPS). Thus, this approach, along with all retrospective comparisons, is inherently ﬂawed. Irish et al. [59] published an elegant analysis of 101 consecutive patients seen at the London regional cancer center. Of these, 27% were deemed eligible for radiosurgery. The median survival of these patients was 23.4 months compared with 8.6 for the radiosurgery-ineligible patients. This study highlights the signiﬁcant effects of patient selection. It also demonstrates that a patient group can be selected that performs better than the most favorable group of grade IV tumors in the RTOG recursive partitioning analyses–-the RTOG class III patients (age >50, KPS ≥90) for whom the expected median survival is 18 months. The bottom line is that only randomization can control efﬁciently for all confounding factors. Having previously demonstrated the feasibility of a multiinstitutional radiosurgery trial [51] and encouraged by the favorable survival of patients treated on phase I/II protocols, the RTOG opened protocol 93-05 in 1994. This was a prospective randomized trial evaluating upfront SRS followed by EBRT with carmustine (BCNU) (arm 1) versus EBRT and BCNU (arm 2). Eligibility criteria required patients to be at least 18 years of age, have a KPS of at least 60%, and to have a histologically proved supratentorial, unifocal GBM. All lesions were to be 4 cm or less in maximal diameter. Patients presenting with tumors larger than 40 mm preoperatively were eligible only if the postoperative imaging studies showed residual tumors of ≤40 mm in maximal diameters. In the investigational arm, radiosurgery was to be given up front, within 5 weeks of surgery. External beam radiation was the same in both arms: a volume encompassing the tumor, surrounding edema (for the ﬁrst 46 Gy), and a margin was treated to a total of 60 Gy with daily fractions of 2 Gy. BCNU was administered at a dose of 80 mg/m2 on days 1, 2, and 3 of EBRT and then repeated every 8 weeks for a total of 6 cycles. The ﬁrst patient was treated in February 1994 and the study was closed in June 2000 after the 203rd patient was enrolled (thus meeting the accrual target of 200 patients). Results were published in 2004 [52]. Figure 17-3 illustrates the case of a patient treated on the experimental arm. At a median follow-up of 61 months (Fig. 17-4), the median survival for arm 1 was 13.5 months (95% CI: 11.0 to 14.9) and it was 13.6 months (95% CI: 11.3 to 15.2) for arm 2 (p = 0.53). The 2-year actuarial survival rates were 21% for arm 1 and 19% for arm 2. Disappointingly, patterns of failure were not inﬂuenced by the therapy with 90% of patients presenting with a component of local failure. There was also no difference in the mini-mental or Spitzer [64] indices (a validated quality of life index). Despite the exclusion from analysis of seven patients having progressed prior to SRS, no subgroup was identiﬁed as beneﬁting from SRS. An intention-to-treat analysis including all randomized patients produced nearly identical results. Both acute and late RTOG grade 3 toxicities were more frequent on arm 1, although not signiﬁcantly so. There were no incidences of radiation-related grade 4 toxicity in either arm. Five patients did die of chemotherapy-related toxicity on arm 1 compared with two patients on arm 2.

17.

213

high-grade gliomas

TABLE 17-4. SRS in the primary management of HGGs. Authors

Selch [53]

Masciopinto [54]

Gannett [45]

Buatti [56]

Kondziolka [49]

Shenouda [57]

Shrieve [50]

Nwokedi [58]

Institution

UCLA

University of Wisconsin

University of Arizona

University of Florida

University of Pittsburgh

McGill University

Harvard University

University of Maryland

Date of publication

1993

1995

1995

1995

1997

1997

1999

2002

Number of patients

18

31

30

11

107 (65 primary lesions)

14

78

31

Histology

12 GBM 6 AA

GBM

17 GBM 10 AA

6 GBM 5 AA

45 GBM 20 AA

GBM

GBM

GBM

KPS

100% >70

57% >70

97% >70

All >90%

Mean KPS 90 (50–100)†

79% >70

Median 90 (50–100)

61% >70

Median tumorvolume (cm3)

20 (8–46)

16 (2–60)

24 (2–115)

14 (6–23)

6.5 (1–31)†

<34

10

25

Median MPD (range)

30 (15–35)

12 (10–20)

10 (0.5–18)

13 (10–15)

16 (12–25)†

20

12 (6–24)

17

Sequence

Post-EBRT

Pre-EBRT 12 Post-EBRT17 No EBRT 2

Within 8 weeks (median 4)

12–109 days post-EBRT (typically 2–3 weeks)

GBM median 6.2 months postdiagnosis; AA median 3.9 months postdiagnosis

Pre-EBRT

Post-EBRT, median 14.2 weeks from diagnosis (range 1–42)

Within 4 week spostEBRT

GBM 13 AA 28

17

GBM 20 AA 56

10

19.9

25

Median survival (months)

9

9.5

1-year survival (%)

GBM 33 AA 100

37

GBM 43 AA 64.5

—

—

43

88.5

—

2-year survival (%)

GBM 33 AA 100

—

GBM 8 AA 53

—

GBM 41 AA 88

—

35.9

—

Median prescription isodose (%)

62 mean

67 (40–80) mean

70

80 (70–90)

50 (40–90)†

—

85 (60–100)

50 (45–65)

Median age (range)

56 (35–79)

57.7 (20.4–78.6)

54 (5–74)

42.1 (15–77)

GBM 51 (3–72) AA 45 (3–73)†

67.5 (45–78)

51 (12–84)

21/31 >50

EBRT

45–60 Gy in 14/18 patients

0–66

Median 59.4 (44–62)

Median 60 (54–60)

GBM all, mean 60 Gy

60 Gy accelerated

73/78

All 24/31 >59 Gy

Median followup (months)

10 (3–22)

30 (living)

6 (minimum)

—

25 (minimum)

—

Number of reoperation

0/2

—

0/10

0/4

3/22†

1/14

20/39

—

Prognostic factors on multivariate analysis

—

Age, KPS, extent of surgery*

KPS*

Reoperation*

Age, KPS†

None signiﬁcant

Age

Age, SRS

9.5

8.8

*Univariate analysis. †

For entire group, including recurrent tumors.

Questions may arise about the timing of the radiosurgery boost in the RTOG trial. Although both pre- and post-EBRT boosts have been used, a look at Table 17-4 will reveal that post-EBRT boosts are more common in reported phase II data. At the time of the study design in 1993, the choice of an upfront boost for the RTOG 93-05 trial was motivated primarily by three factors: (1) to beneﬁt from controlling potential accelera-

TABLE 17-5. Comparison of survival results between radiosurgery and RTOG treated patients based on recursive partitioning analysis. Radiosurgery RPA

3 4 5

RTOG

Median survival (months)

Two-year OS

Median survival (months)

Two-year OS

38.1 19.6 13.1

75% 34% 21%

17.9 11.1 8.9

35% 15% 6%

ted tumor repopulation, (2) to avoid exclusion of any patient entering the radiosurgical arm because of tumor progression during the EBRT, and (3) to avoid the bias inherent in selecting patients after fractionated radiotherapy. Two reports on the use of a focal boost post-EBRT published at that time [59, 65] showed that a large proportion of patients (10% to 47%) developed tumor progression while undergoing the EBRT. Even with the strategy of an upfront SRS boost, 7% of the patients in the SRS arm had to be excluded because of tumor progression. Biologically early intensiﬁcation of the radiation seems favorable, and there is no convincing evidence that a postEBRT SRS boost is radiobiologically advantageous over a preEBRT SRS boost. Case series of post-EBRT boosts cannot report patient outcomes on an intent to treat basis and have included patients receiving radiosurgery up to a mean of 6.2 months after their diagnosis [49]. The additional bias introduced by including patients who still have small nonprogressing tumors and a good

214

d. roberge and l. souhami

FIGURE 17-3. Patient H.L., a 73-year-old gentleman treated for glioblastoma multiforme on the radiosurgery arm of RTOG 93-05. Prior to 60 Gy of EBRT, the residual tumor in the left temporal lobe was boosted using a three-isocenter plan using the McGill dynamic stereo-

tactic technique. (a) Fifteen grays was delivered to the 50% isodose surface, as per protocol guidelines. (b) The tumor failed locally 13 months later.

performance status several months after their initial diagnosis is obvious. Of interest, despite using a different temporal sequence of their boost implants, in the two randomized trials of stereotactic brachytherapy boost [35, 66], both had similar results, failing to demonstrate a survival beneﬁt for this technique when given either before or after the EBRT. A summary of the relative merits of pre- and post-EBRT SRS is presented in Table 17-6. Although the phase III data do not unequivocally disprove any beneﬁt to post-EBRT radiosurgical boosts, there is no reason to believe that this approach is superior. Thus, the best evidence currently available does not support an improvement in median survival after the addition a SRS to

a standard treatment regimen for GBM. There is also no evidence of an improvement in quality of life or a change in pattern of failure. There is, however, an increase in both acute and late radiation-related toxicity. These results are not surprising in a disease where surgical excision of the radiosurgery target is not curative.

100

Single Fraction Stereotactic Radiosurgery for Recurrent HGGs Treatment of recurrent HGG is essentially palliative; cases of patients surviving more than 5 years are anecdotal. Strategies available [67] include reoperation, chemotherapy, EBRT reirradiation, SRS, F-SRT, interstitial brachytherapy, supportive care, and so forth. Each of these modalities has its own toxicities and indications. Currently published phase III studies support

RT Survival Rate

80

SRS+RT TABLE 17-6. Timing of SRS in the primary management of HGG.

60

Pros

40

20

SRS before EBRT

SRS after EBRT

Increases number of eligible patients All patients will receive protocol program Decreases selection bias

Better deﬁnition of tumor margins Target may be smaller

0 0

6

12

18 24 30 36 Months FIGURE 17-4. Overall survival by treatment arm on RTOG 9305. (Reprinted from 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; 60:853–860. With permission from Elsevier.)

Cons

Patients not fully recovered from surgery Radiosurgery potentially delivered to a larger target Because of postoperative changes, imaging less satisfactory

SRS may be planned far ahead Patients fully recovered from surgery Target may enlarge Tumor may grow beyond 4 cm Performance status may deteriorate Selection bias

17.

high-grade gliomas

215

FIGURE 17-5. Patient C.B., a 55-year-old, was treated for a left frontal GBM on the radiotherapy-only arm of the EORTC/NCIC protocol. His disease progressed soon after the end of radiotherapy, and he was given temozolomide. The tumor progressed after the second cycle of temozolomide. The patient was switched to BCNU chemotherapy, and his disease remained stable. Ten months after EBRT, there was evidence of further disease progression and he was treated with SRS. (a) Two

isocenters were treated using the McGill dynamic stereotactic technique. A dose of 15 Gy was prescribed to the 50% isodose surface. (b) He subsequently developed what was believed to be an area of asymptomatic radionecrosis, managed conservatively. (c) He was well until the tumor progressed a second time, 3 years after SRS. This second failure was at an edge of the tumor bed outside the radiosurgery target volume (arrow), and repeat SRS was recently performed.

an improvement in median survival from the use of BCNU polymers [68] and an improvement in softer measures of outcome from the use of oral temozolomide [67, 69, 70]. There are no phase III trials of radiosurgery for recurrent tumors. Only very selected cases are candidates for SRS at recurrence. Figure 17-5 illustrates the case of a patient treated at our institution with apparent clinical beneﬁt. A 1994 series from the University of California San Diego included 15 patients with recurrent HGG [71]. Doses prescribed ranged from 12 to 15 Gy based on the target volume. Of the total group of 20 patients, 7 have suffered intracranial pressure–related acute toxicity (fatal in 1 patient), and 1 has suffered from a late somnolence syndrome. Outcome was not reported by tumor type. In 1995, the experience of the University of Minnesota was published [72]. A total of 35 recurrent tumors (26 GBM, 9 AA) had been treated with single doses of radiation (7.5 to 40 Gy). The median survival for all patients from the time of SRS was 8 months. The actuarial rate of necrosis requiring surgery was 14% (the overall reoperation rate was 31%). Despite SRS, 85% of recurrences were local or marginal. In 1999, this experience was updated to include a total of 46 patients and then had a median survival of 11 months [73]. Kondziolka et al. reported on the University of Pittsburgh experience treating 42 patients with recurrent HGG [49]. Median survival from radiosurgery was favorable, 30 months for GBM, 31 for AA. There were no cases of acute toxicity, and three patients experiences late radiosurgery-related morbidity. Overall 22 of 107 (21%) of patients in this series required craniotomy after SRS. Additional institutional series can be found in Table 17-7 [74]. Common themes from these series are as follows:

Fractionated Stereotactic Radiotherapy

• Despite median survivals of 7.5 to 30 months [49, 71], only 1 patient is reported as surviving 5 years after radiosurgery (0.5%) [49]. • Toxicity is incompletely reported with up to 46% of patients experiencing treatment complications [71]. • Reported prognostic factors include histology, age, KPS, and tumor volume [49, 73, 74].

When introduced by Leksell more than 50 years ago [75], stereotactic radiosurgery was intended as a means of creating a “radiation injury.” The intent at that time was to destroy brain tissue, not selectively inactivate tumor cells intertwined with normal brain tissue. In an attempt to decrease complications through normal tissue repair and to potentially increase efﬁcacy by allowing reoxygenation and cell-cycle reassortment, several centers have treated small series of patients with F-SRT. Often, these series are of patients with recurrent, previously irradiated lesions where there was a concern for toxicity. Beginning in 1987, patients were treated at our institution with a regimen of 42 Gy given in fractions of 7 Gy on alternating days over 2 weeks [76]. For these treatments, rigid immobilization was performed with a halo-type frame [77]. Since this early experience, the general trend for F-SRT has been toward smaller fractions using noninvasive immobilization. Authors from the Royal Marsden Hospital have reported a series of 33 patients treated with F-SRT for recurrent primary brain tumors [41]. This was a dose-escalation study conducted from January 1989 to July 1994. All of these patients had been previously irradiated. Doses ranged from 4 to 10 daily fractions of 5 Gy. The median survival of the 21 patients with recurrent HGG (11 AA, 10 GBM) was 9.6 months. A matched-pair analysis was performed using patients treated for their recurrence with nitrosourea-based chemotherapy. There was a small, statistically signiﬁcant difference in median survival favoring the RT cohort. Late toxicity (not graded) was seen in 36% of the patients. At Thomas Jefferson University, from November 1994 to September 1996, 20 patients with recurrent (or persistent) HGG after external beam radiation were treated with F-SRT [42]. Three different regimens were used: 24 Gy in daily fractions of 3 Gy, 30 Gy in daily fractions of 3 Gy, and 35 Gy in daily fractions of 3.5 Gy. Median survival from the completion of F-SRT was 10.5 months. There were no grade 3 to 4 acute or late toxicities. Between April 1991 and January 1998, 71 patients with recurrent tumors (13% anaplastic oligodendrogliomas, 87%

216

d. roberge and l. souhami

TABLE 17-7. Results of SRS for recurrent HGGs. Author

Selch [82]

Chamberlain [59]

Shrieve [62]

Kondziolka [41]

Cho [86]

Larson [40]

Larson [40]

Institution

UCLA

UCSD

Harvard

University of Pittsburgh

University of Minnesota

UCSF

UCSF

Date of publication

1993

1994

1995

1997

1999

2002

2002

Number of patients

17

13

86

42

46

26*

54

Histology

12 GBM 5 AA

5 GBM 8 AA

86 GBM (14 transformed from LGG)

19 GBM 23 AA

27 GBM 15 AA 4 AO

GBM 14, GRIII 12

GBM 39, GRIII 15

Time from diagnosis (months)

Median 8

Median 15 (4–96)

Median 10.3 (2.3–115)

GBM mean 18.9, AA mean 19.8

Median 10 (1–166)

GBM median 12 (3–50), GRIII median 43 (7–175)

GBM median 7 (2–70), GRIII median 35 (2–145)

KPS

100% > 70

Median 70 (50–100)

90 (70–100)

Mean KPS 90 (50–100)†

Median 70 (40–90)

90 (70–100)

Median 70 (40–90)

Median tumor volume, cm3 (range)

28.6 (6–121)

36 (3–50.6)

10.1 (2.2–83)

6.5 (0.9–31)†

30 (3–125)

GBM 8 (1.6–29.7), GRIII 2.7 (0.4–13.4)

GBM 9.1 (0.3–29.1), GRIII 6 (0.3–20.3)

Median peripheral dose Gy (range)

24.4 (mean)

12.5 (12–15.7)

13 (6–20)

15.5 (12–25)†

17 (9–40)

GBM 15 (12–17.5), GRIII 16.5 (15–18)

GBM 16 (10–20), GRIII 17 (15–18.5)

Median survival (months)

10

GBM 15 AA 7.5

10.2

GBM 30 AA 31

11 (1–36)

GBM 38 weeks, GRIII 68 weeks

GBM 44 weeks, GRIII 59 weeks

1-year survival (%)

—

GBM 40 AA 12.5

—

45

42

—

42

2-year survival (%)

—

—

19

—

—

—

—

Median prescription isodose (range)

64

80

80 (50–100)

50 (40–90)†

50 (40–90)

25–30

50 (45–55)

Median age (range)

49 (27–79)

36 (17–62)

46 (9–77)

GBM 51 (3–72), AA 45 (3–73)†

48 (16–75)

GBM 53 (22–74), GRIII 44 (24–62)

GBM 50 (21–77), GRIII 35 (24–58)

Prior Chemo

—

100%

33%

—

57%

—

—

Prior EBRT

—

100%

100%

—

100% (median 60 Gy)

100%

100%

Median follow-up (months)

7 (3–14)

17.5 (6.8–45.1)

—

—

—

—

Reoperations

5/17

(46% complication rate)

19/86

3/22†

22%

—

—

Prognostic factors on multivariate analysis

—

—

Age, tumor volume‡

Age, KPS†

KPS, tumor grade

—

—

Survivors >5 years

None

None

None

1 (74 months)

None

None

None

8

GRIII, grade III glial tumors. *Patients treated on a protocol with Marimastat. †Including patients treated for a primary glioma. ‡Univariate analysis.

high-grade astrocytomas, 15% of these were dedifferentiated low-grade tumors) were treated at the University of Minnesota Hospital. A total of 46 of the patients in this retrospective review were treated with SRS and 25 with F-SRT [73]. For the SRS cases, the median minimum peripheral dose was 17 Gy. F-SRT was delivered to a median of 37.5 Gy in 15 fractions. The median survival from stereotactic irradiation was 11 months for SRS and 12 for F-SRT (p = 0.3). Acute complications were seen in 40% of both groups. Late complications (clinical/pathologic necrosis or cranial nerve palsy) were seen more commonly in the SRS group, 30% versus 8% (p < 0.05). The authors concluded that as the F-SRT group had worse prognostic factors, a similar median survival, and less late toxicity, it might represent a better treatment option for recurrent HGGs.

Table 17-3 summarizes some of the larger published singleinstitution series of F-SRT for HGG. Fraction sizes range from 2.5 Gy to 7 Gy for what are mostly recurrent tumors. Median survival for variable patient populations ranges from 7 to 20 months. Complication rates are incompletely reported and range from none to 27% grade IV [41, 42]. The same issues of bias that plagued the experience with single-fraction SRS apply to the younger experience utilizing F-SRT. Treatment options for recurrent lesions are numerous. If stereotactic radiation is used, the choice of fractionation may represent a compromise between convenience and risk of radionecrosis. A more deﬁnitive answer may not soon come for F-SRT. The EORTC and the Medical Research Council (MRC) had

17.

high-grade gliomas

217

FIGURE 17-6. Patient T.V., a 71-year-old, was treated on RTOG BR0023. As per protocol, ﬁve weekly fractions of 5 Gy were delivered concurrently with 50 Gy of EBRT. (a) The F-SRT was prescribed to the

80% isodose and delivered using ﬁve static, non-coplanar micro-multileaf beams. (b) The patient declined adjuvant chemotherapy and recurred locally within weeks of radiotherapy.

jointly initiated a randomized trial of F-SRT in the primary management of HGGs [78]. The patients eligible for this study were those with ≤4 cm WHO grade III to IV gliomas, age <65 years with a good performance status. These patients were to be randomized to partial brain radiation (54 to 60 Gy) followed or not by F-SRT (four × 5 Gy daily). This study closed in December 2001 because of poor accrual (M. Brada, personal communication). The use of hypofractionated F-SRT in combination with standard radiation therapy has been tested in phase I/II trials in selected patients with primary GBM. Regine et al. [79] treated 18 previously unirradiated patients with brain tumors using a combination of conventional external beam radiation therapy and a split-course F-SRT. Of these 18 patients, 11 were malignant gliomas. F-SRT dose schedule was based in the residual tumor volume and varied from 7 Gy given two to four times. F-SRT was delivered before and after external beam radiation therapy. Of 15 patients evaluated so far, at a median follow-up of 15 months excessive toxicity was seen only in the group with larger tumor volumes. Of interest, 38% of the patients requiring reoperation (3/8) were found to have only necrosis without evidence of tumor cells. Cardinale et al. [80] reported on 20 selected patients entered on a phase II trial of F-SRT for previously untreated GBM. Eligibility criteria mandated that the size of the postoperative tumor cavity, including any enhancing areas, be 6 cm or less. F-SRT was combined with standard radiation therapy (50 Gy) and was given once per week during weeks 1 through 4. F-SRT dose was tumor size–dependent and ranged from 6 to 10 Gy per fraction. At a median follow–up time of 21 months, the median survival was 19 months. Radiation necrosis was found in 62% of patients submitted to reoperation (eight patients). Although these results are encouraging and the biological concept of F-SRT is interesting, attempting to deliver a higher dose of irradiation while shortening the overall treatment time to decrease the opportunity of accelerated repopulation of clonogenic cells, it is unlikely that a major improvement in tumor control or survival will be achieved. Recently, the RTOG conducted a phase II trial (RTOG BR-0023) testing Dr.

Cardinale’s scheme in patients with primary supratentorial GBM [81]. This study was open from March 2001 to June 2003 and was performed to assess the feasibility, toxicity, and efﬁcacy of this approach. On this study, a relocatable immobilizer was used for the four weekly fractions (5 Gy each for targets >40 mm, 7 Gy for smaller targets) of F-SRT that were delivered using 3D conformal radiotherapy (3DCRT) (GK and cone-based SRS systems were not permitted). EBRT (50 Gy) was followed by 6 cycles of BCNU (80 mg/m2 for 3 days every 8 weeks) chemotherapy. A total of 76 patients were analyzed. Treatments were well tolerated with only one acute grade 4 (lethargy) and one late grade 3 (necrosis) toxicity observed. The median survival was 12.5 months, and no survival difference was seen when results were compared with the RTOG historical database. Thus, although feasible and well tolerated, this dose-intense, accelerated regimen does not appear to improve survival. This larger RTOG trial represents more deﬁnitive evidence that a hypofractionated boost strategy may be of limited value. The case of a patient treated on this trial is presented in Figure 17-6.

Treatment Toxicities Signiﬁcant acute complications (i.e., those occurring within days to weeks of treatment) are unusual and generally self-limited. Occasionally, an exacerbation of existing symptoms may occur, particularly in patients with moderate brain edema and with larger lesions (3 to 4 cm). Most patients will respond to increasing doses of corticosteroids. Late complications attributable to SRS are usually deﬁned as necrosis within the treatment volume. In patients with GBM, distinguishing radiation-induced necrosis from tumor recurrence is not infrequently a very difﬁcult problem and may lead to an overestimation of the rate of treatment-induced complications. The rate of reoperation for ranges from 0 [55] to 54% [50]. However, viable cells are identiﬁed in the majority of such reoperation specimens. Pure necrosis without residual tumor cells is a rare event.

218

d. roberge and l. souhami

In the review involving three institutions and 115 patients [60], late complications occurred in 16% of the patients. Radiation necrosis was diagnosed in the majority of the patients (17/19) either by reoperation or by imaging evaluation. Prolonged corticosteroid therapy was required in 47% of the patients. In the recent RTOG randomized trial (RTOG 93-05), there were four cases (5%) of grade III radiation toxicity in the SRS arm and no cases in the standard arm. Of the 28 patients in the SRS arm undergoing subsequent salvage surgery, 7 (25%) had only necrosis in the resection specimen compared with 3 of 31 patients (9.6%) in the standard arm. Clark et al. [82], from McGill University, have described a single number parameter for use during F-SRT that may be used to represent effective dosage to intracranial structures that are irradiated with inhomogeneous dose distribution having steep dose gradients. Evaluating biologically effective dosevolume histograms, the authors obtained an integral biologically effective dose (IBED) for each case and demonstrated a threshold value for late damage to the brain stem consistent with similar thresholds that have been determined for external beam radiation therapy. Using α/β ratio of 2.5 as representative of the dose-response of brain-stem tissue, they were able to establish a threshold value in the region of 60 Gy2.5. The determination of the IBED can be a valuable parameter to be used as a guide to determine the appropriate treatment volume and fractionation regimen that will minimize toxicity to surrounding vital structures in patients undergoing F-SRT.

Future Directions in Stereotactic Radiation for Glial Neoplasms New strategies in the application of stereotactic radiation for malignant gliomas include changes in planning and fractionation; concurrent use of chemotherapy; and use of radiation modiﬁers and biological agents. Results of Temple University and of Staten Island looking at chemotherapy (cis-platinum and Taxol [Taxol; BristolMeyers–Squibb, Princeton, NJ]) and F-SRS for recurrent glioma have been reported [40, 83]. Retrospectively, these results are similar to patients treated without chemotherapy, and further use of these agents with SRS is unlikely. Other treatment strategies that were brewing in the laboratories of the University of Pennsylvania include radioprotection of normal tissue [84, 85] or tumor sensitization through gene transfer [86]. In view of the recent data regarding temozolomide, new trials in the treatment of HGG will have to incorporate this agent. Because the European intergroup randomized trial was closed prematurely and therefore will be unable to demonstrate a level 1 beneﬁt to F-SRT, the approach will likely fade in favor of integrating new agents with conventional EBRT. At McGill University, we are currently treating selected patients with a hypofractionated regimen of 60 Gy in 20 fractions. The ﬁrst 20 patients treated using this dose scheme without chemotherapy had a median survival of 8.1 months [87]. With a median followup of 7 months, no late toxicity was observed. With the addition of concurrent temozolomide for the latest 35 patients, the median survival is now 14.4 months. A phase I trial (NCI-T99-0041) investigating the sensitization of SRS with gadolinium texaphyrin was recently closed. In

this trial, GBM patients with a KPS >60 and a tumor <40 mm were ﬁrst treated with conventional EBRT. After EBRT, an intravenous infusion of gadolinium texaphyrin was given 3 hours prior to SRS. The UCSF group has published its experience using an matrix metalloproteinase inhibitor (MMPI) inhibitor (Marimastat [British Biotech, Inc. Oxford, United Kingdom] starting on the day of SRS and continued until progression, death, or toxicity) in combination with SRS for recurrent HGG [48]. In this phase I trial, patients’ outcomes were not clearly different than expected for patients with recurrent HGG. There is growing interest in new functional imaging modalities for HGG. This might be of interest in target deﬁnition, especially for recurrent tumors treated with SRS where nonenhancing tumor may be missed by gadolinium-enhanced T1 MRI. The group from UCSF has compared SRS treated volumes with metabolically active tumor volumes deﬁned by proton magnetic resonance imaging (1H-MRS) [88]. A total of 26 patients were retrospectively divided into high-risk or low-risk groups according to the amount of overlap between the metabolic target and the SRS target. The median survival was 15.7 months for patients in the low-risk group and 10.4 months for those in the high-risk group. This data will likely lead at UCSF to prospective evaluation of the use of 1H-MRS targets for SRS. Combining new strategies in HGG stereotactic radiotherapy, Neider et al. have presented the results of a phase II study looking at adding chemotherapy and molecular targeting to FSRT for recurrent HGG [89]. Forty-six patients were treated with F-SRT (30 Gy in 6 fractions) to a target deﬁned using C11methionine PET, CT and MRI information. In 30 patients, FSRT was sandwiched in a 6-month course of temozolomide. Median survival was 11 months with temozolomide, 6 months without. Toxicity was acceptable with three reoperations for apparent radionecrosis, which, on pathology, was combined with recurrent tumor. The bottom line is that, despite past failures, as long as 90% of HGGs continue to fail locally, there will be interest in dose intensiﬁcation. As new techniques to achieve this are tested [90], it will be important to better report toxicity and remain wary of selection bias to which small cohorts are prone.

Conclusion The best level 1 evidence currently available does not support an improvement in outcome from the use of SRS for primary GBM. Despite the modest beneﬁts achieved by escalating EBRT dose to match normal tissue tolerance, it is unlikely that variations in fractionation, total dose, or treatment delivery will lead to a signiﬁcant change in the outcome for patients with HGG. For a very select group of patients with recurrent disease, SRS or F-SRT may represent a safe, reasonable, but palliative treatment. For recurrent HGG, other treatment modalities have been tested in phase III trials, and only prospective investigation can deﬁne the exact role of stereotactic radiation.

References 1. Surawicz TS, McCarthy BJ, Kupelian V, et al. Descriptive epidemiology of primary brain and CNS tumors: results from the Central Brain Tumor Registry of the United States, 1990–1994. Neuro-oncology 1999; 1:14–25.

17.

high-grade gliomas

2. Melendon R, Halperin EC. Long term survival after treatment of glioblastoma multiforme: is it overstated. 44th Annual Meeting of the American Society for Therapeutic Radiology and Oncology (ASTRO). New Orleans; Louisiana, 2002. 3. Scott JN, Rewcastle NB, Brasher PM, et al. Which glioblastoma multiforme patient will become a long-term survivor? A population-based study. Ann Neurol 1999; 46:183–188. 4. 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: 704–710. 5. Rao RD, James CD. Altered molecular pathways in gliomas: an overview of clinically relevant issues. Semin Oncol 2004; 31:595– 604. 6. Kitange GJ, Templeton KL, Jenkins RB. Recent advances in the molecular genetics of primary gliomas. Curr Opin Oncol 2003; 15:197–203. 7. Wallner KE, Galicich JH, Krol G, et al. Patterns of failure following treatment for glioblastoma multiforme and anaplastic astrocytoma. Int J Radiat Oncol Biol Phys 1989; 16:1405–1409. 8. Hochberg FH, Pruitt A. Assumptions in the radiotherapy of glioblastoma. Neurology 1980; 30:907–911. 9. Cerame MA, Guthikonda M, Kohli CM. Extraneural metastases in gliosarcoma: a case report and review of the literature. Neurosurgery 1985; 17:413–418. 10. Erlich SS, Davis RL. Spinal subarachnoid metastasis from primary intracranial glioblastoma multiforme. Cancer 1978; 42:2854 – 2864. 11. Bailey P, Cushing H. Classiﬁcation of the Tumors of the Glioma Group on a Histogenetic Basis with a Correlation Study of Prognosis. Philadelphia: Lippincott, 1926. 12. Elvidge A, Wilder P, Cone W. The gliomas of the central nervous system. Res Publ Ass Nerv Ment Dis 1937; 16:107–181. 13. Netsky MG, August B, Fowler W. The longevity of patients with glioblastoma multiforme. J Neurosurg 1950; 7:261–269. 14. Bush E, Christensen E. Three types of glioblastoma. J Neurosurg 1947; 4:200–220. 15. Walker MD, Alexander E Jr, Hunt WE, et al. Evaluation of BCNU and/or radiotherapy in the treatment of anaplastic gliomas. A cooperative clinical trial. J Neurosurg 1978; 49:333–343. 16. Kristiansen K, Hagen S, Kollevold T, et al. Combined modality therapy of operated astrocytomas grade III and IV. Conﬁrmation of the value of postoperative irradiation and lack of potentiation of bleomycin on survival time: a prospective multicenter trial of the Scandinavian Glioblastoma Study Group. Cancer 1981; 47: 649–652. 17. 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; 64:769–774. 18. 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; 303:1323–1329. 19. Shapiro WR, Green SB, Burger PC, et al. Randomized trial of three chemotherapy regimens and two radiotherapy regimens and two radiotherapy regimens in postoperative treatment of malignant glioma. Brain Tumor Cooperative Group Trial 8001. J Neurosurg 1989; 71:1–9. 20. Green SB, Byar DP, Walker MD, et al. Comparisons of carmustine, procarbazine, and high-dose methylprednisolone as additions to surgery and radiotherapy for the treatment of malignant glioma. Cancer Treat Rep 1983; 67:121–132. 21. Leibel SA, Pajak TF, Massullo V, et al. A comparison of misonidazole sensitized radiation therapy to radiation therapy alone for the palliation of hepatic metastases: results of a Radiation Therapy Oncology Group randomized prospective trial. Int J Radiat Oncol Biol Phys 1987; 13:1057–1064.

219

22. Chang CH, Horton J, Schoenfeld D, et al. Comparison of postoperative radiotherapy and combined postoperative radiotherapy and chemotherapy in the multidisciplinary management of malignant gliomas. A joint Radiation Therapy Oncology Group and Eastern Cooperative Oncology Group study. Cancer 1983; 52: 997–1007. 23. Fine HA, Dear KB, Loefﬂer JS, et al. Meta-analysis of radiation therapy with and without adjuvant chemotherapy for malignant gliomas in adults. Cancer 1993; 71:2585–2597. 24. Chemotherapy for high-grade glioma. Cochrane Database Syst Rev 2002:CD003913. 25. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005; 352:987–996. 26. Zietman AL, Suit HD, Tomkinson KN, et al. The response of two human tumor xenografts to fractionated irradiation. The derivation of alpha/beta ratios from growth delay, tumor control, and in vitro cell survival assays. Int J Radiat Oncol Biol Phys 1990; 18: 569–575. 27. Marin LA, Smith CE, Langston MY, et al. Response of glioblastoma cell lines to low dose rate irradiation. Int J Radiat Oncol Biol Phys 1991; 21:397–402. 28. Raza SM, Lang FF, Aggarwal BB, et al. Necrosis and glioblastoma: a friend or a foe? A review and a hypothesis. Neurosurgery 2002; 51:2–12; discussion 12–13. 29. Loefﬂer JS, Alexander E 3rd, Hochberg FH, et al. Clinical patterns of failure following stereotactic interstitial irradiation for malignant gliomas. Int J Radiat Oncol Biol Phys 1990; 19:1455– 1462. 30. Gutin PH, Prados MD, Phillips TL, et al. External irradiation followed by an interstitial high activity iodine-125 implant “boost” in the initial treatment of malignant gliomas: NCOG study 6G-82-2. Int J Radiat Oncol Biol Phys 1991; 21:601–606. 31. Walker MD, Strike TA, Sheline GE. An analysis of dose-effect relationship in the radiotherapy of malignant gliomas. Int J Radiat Oncol Biol Phys 1979; 5:1725–1731. 32. Gutin PH, Leibel SA, Wara WM, et al. Recurrent malignant gliomas: survival following interstitial brachytherapy with highactivity iodine-125 sources. J Neurosurg 1987; 67:864–873. 33. Florell RC, Macdonald DR, Irish WD, et al. Selection bias, survival, and brachytherapy for glioma. J Neurosurg 1992; 76: 179–183. 34. Shapiro WR, Green SB, Burger PC. A randomized trial of interstitial radiotherapy (IRT) boost for the treatment of newly diagnosed malignant glioma (glioblastoma multiforme, anaplastic astrocytoma, anaplastic oligodendroglioma, malignant mixed glioma): BTCG Study 8701. Neurology 1994; 44(Suppl 2):A263. 35. 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– 1011. 36. Colombo F, Benedetti A, Pozza F, et al. External stereotactic irradiation by linear accelerator. Neurosurgery 1985; 16:154– 160. 37. Shaw E, Scott C, Souhami L, et al. Radiosurgery for the treatment of previously irradiated recurrent primary brain tumors and brain metastases: initial report of radiation therapy oncology group protocol (90-05). Int J Radiat Oncol Biol Phys 1996; 34:647–654. 38. Willner J, Flentje M, Bratengeier K. CT simulation in stereotactic brain radiotherapy–analysis of isocenter reproducibility with mask ﬁxation. Radiother Oncol 1997; 45:83–88. 39. Alheit H, Dornfeld S, Dawel M, et al. Patient position reproducibility in fractionated stereotactically guided conformal radiotherapy using the BrainLab mask system. Strahlenther Onkol 2001; 177:264–268. 40. Glass J, Silverman CL, Axelrod R, et al. Fractionated stereotactic radiotherapy with cis-platinum radiosensitization in the treatment

220

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

d. roberge and l. souhami of recurrent, progressive, or persistent malignant astrocytoma. Am J Clin Oncol 1997; 20:226–229. Shepherd SF, Laing RW, Cosgrove VP, et al. Hypofractionated stereotactic radiotherapy in the management of recurrent glioma. Int J Radiat Oncol Biol Phys 1997; 37:393–398. Hudes RS, Corn BW, Werner-Wasik M, et al. A phase I dose escalation study of hypofractionated stereotactic radiotherapy as salvage therapy for persistent or recurrent malignant glioma. Int J Radiat Oncol Biol Phys 1999; 43:293–298. 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; 23:155– 159. Regine WF, Patchell RA, Strottmann JM, et al. Combined stereotactic split-course fractionated gamma knife radiosurgery and conventional radiation therapy for unfavorable gliomas: a phase I study. J Neurosurg 2000; 93(Suppl 3):37–41. Regine WF, Patchell RA, Strottmann JM, et al. 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: 421–426. Baumert BG, Lutterbach J, Bernays R, et al. Fractionated stereotactic radiotherapy boost after post-operative radiotherapy in patients with high-grade gliomas. Radiother Oncol 2003; 67:183– 190. Cho KH, Hall WA, Lo SS, et al. Stereotactic radiosurgery versus fractionated stereotactic radiotherapy boost for patients with glioblastoma multiforme. Technol Cancer Res Treat 2004; 3: 41–49. Larson DA, Prados M, Lamborn KR, et al. Phase II study of high central dose Gamma Knife radiosurgery and marimastat in patients with recurrent malignant glioma. Int J Radiat Oncol Biol Phys 2002; 54:1397–1404. Kondziolka D, Flickinger JC, Bissonette DJ, et al. Survival beneﬁt of stereotactic radiosurgery for patients with malignant glial neoplasms. Neurosurgery 1997; 41:776–783; discussion 783– 775. Shrieve DC, Alexander E 3rd, Black PM, et al. Treatment of patients with primary glioblastoma multiforme with standard postoperative radiotherapy and radiosurgical boost: prognostic factors and long-term outcome. J Neurosurg 1999; 90:72–77. 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; 47:291–298. 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; 60:853–860. Selch MT, Ciacci JD, De Salles AA, et al. Radiosurgery for primary malignant brain tumors. In: De Salles AAF, Goetsch SJ, eds. Stereotactic Surgery and Radiosurgery. Madison, WI: Medical Physics, 1993:335–352. Masciopinto JE, Levin AB, Mehta MP, et al. Stereotactic radiosurgery for glioblastoma: a ﬁnal report of 31 patients. J Neurosurg 1995; 82:530–535. Gannett D, Stea B, Lulu B, et al. 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; 33:461–468. Buatti JM, Friedman WA, Bova FJ, et al. Linac radiosurgery for high-grade gliomas: the University of Florida experience. Int J Radiat Oncol Biol Phys 1995; 32:205–210.

57. Shenouda G, Souhami L, Podgorsak EB, et al. Radiosurgery and accelerated radiotherapy for patients with glioblastoma. Can J Neurol Sci 1997; 24:110–115. 58. Nwokedi EC, DiBiase SJ, Jabbour S, et al. Gamma knife stereotactic radiosurgery for patients with glioblastoma multiforme. Neurosurgery 2002; 50:41–46; discussion 46–47. 59. Irish WD, Macdonald DR, Cairncross JG. Measuring bias in uncontrolled brain tumor trials–to randomize or not to randomize? Can J Neurol Sci 1997; 24:307–312. 60. Sarkaria JN, Mehta MP, Loefﬂer JS, et al. Radiosurgery in the initial management of malignant gliomas: survival comparison with the RTOG recursive partitioning analysis. Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1995; 32:931–941. 61. Werner-Wasik M, Scott CB, Nelson DF, et al. Final report of a phase I/II trial of hyperfractionated and accelerated hyperfractionated radiation therapy with carmustine for adults with supratentorial malignant gliomas. Radiation Therapy Oncology Group Study 83-02. Cancer 1996; 77:1535–1543. 62. Nelson DF, Diener-West M, Weinstein AS, et al. A randomized comparison of misonidazole sensitized radiotherapy plus BCNU and radiotherapy plus BCNU for treatment of malignant glioma after surgery: ﬁnal report of an RTOG study. Int J Radiat Oncol Biol Phys 1986; 12:1793–1800. 63. 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; 6:279–284. 64. Spitzer WO, Dobson AJ, Hall J, et al. Measuring the quality of life of cancer patients: a concise QL-index for use by physicians. J Chronic Dis 1981; 34:585–597. 65. Loefﬂer JS, Alexander E 3rd, Shea WM, et al. Radiosurgery as part of the initial management of patients with malignant gliomas. J Clin Oncol 1992; 10:1379–1385. 66. Selker RG, Shapiro WR, Burger P, et al. The Brain Tumor Cooperative Group NIH Trial 87-01: a randomized comparison of surgery, external radiotherapy, and carmustine versus surgery, interstitial radiotherapy boost, external radiation therapy, and carmustine. Neurosurgery 2002; 51:343–355; discussion 355–347. 67. Nieder C, Grosu AL, Molls M. A comparison of treatment results for recurrent malignant gliomas. Cancer Treat Rev 2000; 26:397– 409. 68. Brem H, Piantadosi S, Burger PC, et al. Placebo-controlled 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– 1012. 69. Osoba D, Brada M, Yung WK, et al. Health-related quality of life in patients treated with temozolomide versus procarbazine for recurrent glioblastoma multiforme. J Clin Oncol 2000; 18:1481– 1491. 70. Yung WK, Albright RE, Olson J, et al. A phase II study of temozolomide vs. procarbazine in patients with glioblastoma multiforme at ﬁrst relapse. Br J Cancer 2000; 83:588–593. 71. Chamberlain MC, Barba D, Kormanik P, et al. Stereotactic radiosurgery for recurrent gliomas. Cancer 1994; 74:1342–1347. 72. Hall WA, Djalilian HR, Sperduto PW, et al. Stereotactic radiosurgery for recurrent malignant gliomas. J Clin Oncol 1995; 13:1642– 1648. 73. Cho KH, Hall WA, Gerbi BJ, et al. Single dose versus fractionated stereotactic radiotherapy for recurrent high-grade gliomas. Int J Radiat Oncol Biol Phys 1999; 45:1133–1141. 74. Shrieve DC, Alexander E 3rd, Wen PY, et al. Comparison of stereotactic radiosurgery and brachytherapy in the treatment of recurrent glioblastoma multiforme. Neurosurgery 1995; 36:275– 282; discussion 282–274.

17.

high-grade gliomas

75. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chirurgica Scand 1951; 102:316–319. 76. Souhami L, Olivier A, Podgorsak EB, et al. Fractionated stereotactic radiation therapy for intracranial tumors. Cancer 1991; 68: 2101–2108. 77. Clark BG, Podgorsak EB, Souhami L, et al. A halo-ring technique for fractionated stereotactic radiotherapy. Br J Radiol 1993; 66:522–527. 78. Brada M, Baumert B. Focal fractionated conformal stereotactic boost following conventional radiotherapy of high-grade gliomas: a randomized phase III study. A joint study of the EORTC (22972) and the MRC (BR10). Front Radiat Ther Oncol 1999; 33:241– 243. 79. Regine WF, Patchell RA, Strottmann JM, et al. Combined stereotactic split-course fractionated gamma knife radiosurgery and conventional radiation therapy for unfavorable gliomas: a phase I study. J Neurosurg 2000; 93(Suppl 3):37–41. 80. Cardinale RM, Broaddus WC, Benedict SH, et al. Accelerated radiotherapy using weekly stereotactic comformal boosts for glioblastoma multiforme. Int J Radiat Oncol Biol Phys 2000; 48(Suppl 1):257. 81. Cardinale RM, Won W, Choucair A, et al. A phase II trial of accelerated radiotherapy using weekly stereotactic conformal boost for supratentorial glioblastoma multiforme. RTOG 0023. ASCO Proc 2005: Abstract 1511. 82. Clark BG, Souhami L, Pla C, et al. The integral biologically effective dose to predict brain stem toxicity of hypofractionated stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 1998; 40:667–675. 83. Lederman G, Arbit E, Odaimi M, et al. Fractionated stereotactic radiosurgery and concurrent taxol in recurrent glioblastoma mul-

84.

85.

86.

87.

88.

89.

90.

221

tiforme: a preliminary report. Int J Radiat Oncol Biol Phys 1998; 40:661–666. Kondziolka D, Mori Y, Martinez AJ, et al. Beneﬁcial effects of the radioprotectant 21-aminosteroid U-74389G in a radiosurgery rat malignant glioma model. Int J Radiat Oncol Biol Phys 1999; 44:179–184. Kondziolka D, Somaza S, Flickinger JC, et al. Cerebral radioprotective effects of high-dose pentobarbital evaluated in an animal radiosurgery model. Neurol Res 1994; 16:456–459. Niranjan A, Moriuchi S, Lunsford LD, et al. Effective treatment of experimental glioblastoma by HSV vector-mediated TNF alpha and HSV-tk gene transfer in combination with radiosurgery and ganciclovir administration. Mol Ther 2000; 2:114–120. Sultanem K, Lambert C, Gopaul D, et al. The use of hypofractionated accelerated intensity modulated irradiation in the treatment of glioblastoma multiforme: Preliminary results of a phase I trial. Int J Radiat Oncol Biol Phys 2001; 51:204. Chan AA, Lau A, Pirzkall A, et al. Proton magnetic resonance spectroscopy imaging in the evaluation of patients undergoing gamma knife surgery for Grade IV glioma. J Neurosurg 2004; 101:467–475. Nieder C, Weidenmann S, Stark S, et al. Phase II trial of functional imaging-optimized stereotactic fractionated radiotherapy plus temozolomide for recurrent high-grade glioma. J Clin Oncol [ASCO Annual Meeting Proceedings (Post-Meeting Edition)] 2004; 22:1552. Tatter SB, Shaw EG, Rosenblum ML, et al. An inﬂatable balloon catheter and liquid 125I radiation source (GliaSite Radiation Therapy System) for treatment of recurrent malignant glioma: multicenter safety and feasibility trial. J Neurosurg 2003; 99: 297–303.

1 Malignant Glioma: 8 Chemotherapy Perspective Roger Stupp and J. Gregory Cairncross

Introduction Despite optimal surgery and adjuvant radiotherapy, malignant gliomas almost always recur. One cause of failure is explained by the diffusely inﬁltrating pattern of growth displayed by most gliomas, with malignant cells disseminated far beyond the initial bulky tumor. Although most relapses occur at or near the initial site of tumor, either along the resection margin or at the edge of the radiation ﬁeld, distant recurrences or seeding of the cerebrospinal ﬂuid (CSF) occurs in long-surviving patients. Malignant gliomas are not a localized disease but instead affect the whole brain and may display wide dissemination even at early stages. Thus, therapeutic strategies aiming only at bulky local disease are doomed to fail. The successful treatments of the future will need to eradicate microscopic disease in many sites within the brain and do so without neurotoxic effects. Here, we summarize progress in the medical treatment of malignant glioma.

Obstacles to Drug Therapy of Brain Tumors The Blood-Brain Barrier The brain is considered a pharmacological sanctuary by virtue of the blood-brain barrier. Many chemotherapeutics, those that are large or hydrophilic, are unable to cross this protective barrier. In most malignant gliomas, the blood-brain barrier is at least partially disrupted, as signaled by contrast enhancement on computed tomography (CT) and magnetic resonance imaging (MRI) studies. The blood-brain barrier is not an impediment to the identiﬁcation of effective drug therapies, because radiographic responses can be seen, but it is a signiﬁcant obstacle to the development of curative drug therapies. Inﬁltrating microscopic glioma hidden behind intact portions of the blood-brain barrier escaping exposure to drugs that do not cross the barrier set the stage for tumor regrowth, despite effective multimodality local therapies. Furthermore, gliomas may induce aberrant capillary growth, which in turn may limit the penetration of chemotherapeutics into the tumor and adjacent brain. Other factors such as tumor size, cell density, and increased intracerebral pressure can inhibit blood ﬂow and reduce drug delivery to malignant gliomas.

Hence, a prerequisite for a highly efﬁcacious antiglioma drug, administered intravenously or by mouth, is the ability to cross the blood-brain barrier. Although barrier permeability can be inferred from the chemical structure of the drug, measurements of drug concentrations in brain and glioma tissue are difﬁcult. It is neither easy, nor ethical, to perform repeated brain biopsies to establish drug tissue levels, and animal models of glioma are often too artiﬁcial to be helpful. Measurements of drug levels in the CSF may be a valuable surrogate for brain and glioma exposure, but false readings may occur due to the effects of tumor edema or compartmentalization of CSF spaces that result from pressure effects or circulating tumor cells that disrupt the ﬂow of CSF [1].

Concomitant Medications Many chemotherapy agents are metabolized by P450dependent hepatic enzymes. Antiepileptics and corticosteroids, commonly prescribed for patients with gliomas, induce these hepatic enzymes, leading to rapid degradation of certain classes of chemotherapeutics, undermining effective cancer treatment. The optimal concentrations of chemotherapeutics may be difﬁcult to achieve when P450-inducing drugs are coadministered. The pharmacokinetic proﬁles of new drug therapies for malignant glioma must be assessed in the presence of concomitant antiepileptic and steroid medications especially. Although the increased metabolism can be overcome by increasing the doses of the chemotherapy agents by several-fold, this approach will make drug therapy prohibitively expensive and—more importantly—potentially dangerous. Should the antiepileptic treatment be discontinued inadvertently, a massive overdose of chemotherapy could ensue. To avoid such pitfalls, the use of third-generation, non–enzyme-inducing antiepileptics [e.g., lamotrigine (Lamictal, GlaxoSmithKline, Research Triangle Park, NC), gabapentin (Neurontin, Pﬁzer, New York, NY), levetiracetam, (Keppra , UCB, Brussels, Belgium), topiramate (Topamax, Ortho-McNeil Neurologics, Titusville, NJ)] should be considered [2–4]. In clinical practice, the enzymeinducing effect of dexamethasone and the enzyme-inhibiting action of the anticonvulsant valproic acid are of lesser concern.

223

224

r. stupp and j.g. cairncross

How to Measure Antitumor Activity? In oncology, new agents are initially tested for antitumor activity in trials that assess whether the tumor type in question shrinks noticeably after drug exposure. Shrinkage is usually determined radiographically using CT or MRI and described as a rate of response. By convention, a 25% response rate means that 25% of the tumors tested were smaller after drug exposure. The World Health Organization (WHO) criteria specify measurement of perpendicular tumor diameters, generating a baseline maximum surface area; a reduction of the surface area by ≥50% is considered a partial response; disappearance of all tumor is deﬁned as complete response, and an increase by 25% is considered tumor progression. Regressions of <50% or increases <25% are called stable disease. For neurooncology, these stringent imaging criteria have been supplemented by an assessment of the use of corticosteroids, which must be stable or decreasing in dose in order to qualify for a response or progression to be declared (the so-called Macdonald criteria) [5]. In recent years in general oncology, unidimensional tumor measurements have been proposed to simplify and speed up the process of assessing response, however in neurooncology these so-called RECIST criteria have not yet been adopted or validated [6]. Because responses are uncommon in neurooncology (i.e., most tumors are resistant to most drugs) and may be difﬁcult to measure, even when they occur, research reports on the medical treatment of brain tumor frequently interpret stable disease or minor response (less than 50%) as evidence of antitumor activity. This practice inﬂates response rates and is of uncertain merit. With MRI having replaced CT, the matter of response assessment has been further complicated. Enhanced CT is a relatively simple imaging tool. MRI, on the other hand, is a better and more complicated machine with a wide assortment of T1, T2, ﬂuid-attenuated inversion recovery (FLAIR), diffusion, and other sequences: Which ones should be used to assess the response of gliomas to new drugs? For now, the WHO and Macdonald style response criteria, currently usually applied to T1 sequences on gadolinium-enhanced MRI, continue to be useful [7, 8]. Because many gliomas are difﬁcult to measure using perpendicular diameters, and to acknowledge the possibility that prolonged disease stabilization may indeed be a genuine indicator of antiglioma activity, the percentage of patients who are progression-free at 6 months has been proposed as an alternative measurement of drug efﬁcacy in neurooncology [9]. This end point, although never validated, has been widely adopted in recent years in chemotherapy trials for recurrent malignant glioma. Of course, overall survival remains the most important and reliable measurement of treatment outcome and randomized trials the best way to test a new drug for glioma [10].

Active Agents (Table 18-1) Nitrosoureas Nitrosoureas were the principal drugs for malignant glioma for 20 years. These lipophilic alkylating agents readily cross the blood-brain barrier and have shown some activity in patients

TABLE 18-1. Selected active cytotoxic agents. Mechanism of action

Toxicity

Carmustine (BCNU)

Chlorethylating agent

Lomustine (CCNU)

Chlorethylating agent

Nimustine (ACNU)

Chlorethylating agent

Fotemustine

Chlorethylating agent

Temozolomide

Methylating agent

Irinotecan (investigational)

Topoisomerase 1 inhibitor

Myelosuppression Nausea/vomiting Pulmonary ﬁbrosis Myelosuppression Nausea/vomiting (Pulmonary ﬁbrosis) Myelosuppression Nausea/vomiting Thrombocytopenia Nausea/vomiting Thrombocytopenia Nausea/vomiting Diarrhea Nausea/vomiting Neutropenia

with recurrent malignant glioma. However, for most of these agents, no recent trial data is available, and the reports in the literature are based on older pathology criteria and classiﬁcations. Further, most trials were conducted before the availability of modern MRI. Carmustine (BCNU) has been the most widely used chemotherapy for malignant glioma. In recent years, lomustine (CCNU) or nimustine (ACNU; 1-[(4-amino-2-methyl-5pyrimidinyl)methyl]-1-(2-chloroethyl)-3-nitrosourea) have in part replaced BCNU, to lessen pulmonary toxicity. Fotemustine is a second-generation nitrosourea that has single-agent activity against recurrent glioma [11]. In Japan and Germany, nimustine (ACNU) is preferred, and fotemustine is most popular in France [11–15].

Procarbazine Procarbazine is an alkylating agent that is given orally but requires hepatic activation. In a randomized trial comparing procarbazine versus BCNU versus BCNU and methylprednisolone versus methylprednisolone-alone, the chemotherapy arms were equivalent and slightly superior to methylprednisolone alone [16]. Median survival was only 9 to 12 months from diagnosis. In the setting of recurrent malignant glioma, response and tumor control rates of 30% have been reported [17, 18].

The PCV Regimen In the 1970s, investigators from University of California San Francisco (UCSF) developed the PCV regimen, a combination of procarbazine, lomustine (CCNU), and vincristine. In 1990, Levin et al. reported that adjuvant chemotherapy with procarbazine, CCNU, and vincristine (PCV) yielded better survival results than adjuvant BCNU for newly diagnosed malignant glioma. This assertion was based on a re-analysis of a randomized trial conducted between 1977 and 1983 [19]. The survival difference favoring PCV was only statistically signiﬁcant for the subgroup of patients (n = 73) with anaplastic astrocytomas (total patients randomized, 148) and good performance status, who had received at least 1 cycle of chemotherapy [19]. The

18.

malignant glioma: chemotherapy perspective

superiority of PCV could not be conﬁrmed in a subsequent pooled analysis of 432 patients treated on Radiation Therapy Oncology Group (RTOG) protocols [20]. The Medical Research Council (UK) conducted a randomized trial of standard focal radiotherapy (RT) versus RT plus adjuvant PCV chemotherapy [21], ﬁnding no difference in overall survival. Patients with grade 3 tumors also did not beneﬁt from PCV.

Temozolomide Over the past decade, temozolomide (TMZ) has emerged as an active agent against malignant glioma [22]. TMZ is an oral alkylating agent, which is rapidly and completely absorbed and spontaneously converts to the active metabolite, MTIC. TMZ has linear pharmacokinetics with maximum plasma concentrations 30 to 90 minutes after oral intake. Although recommended to be taken in a fasting state (at least 1 hour before and after intake), food will only lead to a 10% reduced area under the concentration curve (AUC) and a delayed peak concentration [23]. TMZ has an excellent penetration into all body tissues. Pharmacokinetics of TMZ in the CSF have been recently reported. The AUC in the CSF as a surrogate for brain tissue penetration corresponded with approximately 20% of the AUC in plasma [24]. Similar to nitrosoureas, TMZ is a DNA alkylating agent. Methylation of the O-6 position of guanine by TMZ is an especially important biological action, although not the most frequent adduct. If left unrepaired, the guanine O-6 lesion triggers cytotoxic responses and apoptosis. Three pivotal phase II trials led to the approval of temozolomide (Temodar; Temodal; Schering-Plough, Kenilworth, NJ). The U.S. FDA granted provisional (accelerated) FDA approval in 1999 for the treatment of recurrent anaplastic astrocytoma. In Europe, it was also approved for the treatment of recurrent glioblastoma [25–27]. Although the objective response rate in the glioblastoma trials was low (5% and 7%, respectively) and was only 35% in the anaplastic astrocytoma studies, each trial suggested an increase in the fraction of patients being progression-free at 6 months compared with a historical database [9]. This observation encouraged further evaluation of TMZ as a therapy for both recurrent and newly diagnosed malignant glioma. For recurrent glioma, a randomized phase III trial of TMZ versus PCV is being conducted by the British National Cancer Research Institute (BR12 trial chair: Michael Brada). This trial will also evaluate a potential improvement with a dose-intense TMZ schedule (3 weeks out of 4). A deﬁnitive role of TMZ in the initial treatment of newly diagnosed glioblastoma was demonstrated in a randomized phase III trial by the European Organisation for Research and Treatment of Cancer (EORTC) and the NCI Canada Clinical Trials Group [28].

Primary Treatment of Gliobla stoma The value of radiotherapy in the treatment of malignant glioma (both glioblastoma and anaplastic astrocytoma) was established in randomized trials almost 30 years ago [29–31]. Compared with supportive care or nitrosourea-based chemotherapy alone, radiotherapy increased median survival from 4 to 8 months. Subgroup analyses suggested that the addition of chemotherapy

225

to RT might further improve survival for patients with grade 3 and 4 tumors, but none of the randomized trials demonstrated an unequivocal beneﬁcial role for nitrosourea-based chemotherapy. A meta-analysis based on studies published to 1993 [32] and a more recent analysis of pooled individual data from more than 3000 patients in 12 randomized trials [33] found a small but signiﬁcant survival advantage, favoring chemotherapy in addition to surgery and RT at initial diagnosis. For patients with both grade 3 and 4 tumors, a hazard ratio of 0.85 for chemotherapy equated to a 5% improvement in overall survival from 15% to 20% at 2 years [33]. The ﬁnal report of a German multicenter trial with patient accrual from 1994 to 1999 was published in September 2003 [15]. In this trial, BCNU was replaced by ACNU, another nitrosourea with less pulmonary toxicity. Three hundred seventyﬁve patients were randomized to radiotherapy and either ACNU/cytarabine or ACNU/teniposide (there was no radiotherapy-alone control arm). Although there was no difference in survival between the two groups, the survival outcomes were impressive. There was a median survival of 16 months and a 2year survival rate of 27% among 302 patients with glioblastoma. These results are the best yet reported in a multicenter trial for glioblastoma. More than 90% of patients had undergone debulking surgery, and accrual occurred over 6 years in 15 centers (average of 3 patients per year per center). Thus, a strong selection bias cannot be excluded. Finally, in 2005, the EORTC and NCIC Intergroup trial demonstrated beyond doubt that chemotherapy (i.e., TMZ during RT and for 6 months afterward) improved overall survival in patients with glioblastoma [28]. In this phase III trial, 573 patients were randomized in less than 18 months. Patients who received initial treatment with RT alone (but may have received TMZ or other chemotherapy at progression) had a median survival of 12 months compared with 15 months for patients treated with TMZ/RT at diagnosis. More importantly, the chances of survival at 2 years were only 10% in the RT arm compared with 26% for those treated with TMZ/RT. This trial establishes a new standard of care for patients with glioblastoma and serves as the benchmark to which newer strategies must be compared. These results compare favorably with reported results among patients with small tumors eligible for stereotactic radiosurgery. The good outcome in the control arm also reﬂects the, albeit modest, activity of the TMZ when administered as second-line therapy after tumor progression. A companion translational research study to this trial was especially illuminating. Patients whose tumors had a methylated MGMT gene promoter, resulting in gene silencing and less effective repair of DNA damage caused by TMZ, were the subset who beneﬁted most from the addition of chemotherapy to RT [34]. Patients with tumors in which the MGMT gene promoter was unmethylated and presumably fully expressed derived no beneﬁt from TMZ. Patients with a silenced gene treated with TMZ/RT had a 2-year survival rate of 46% compared with 14% in unmethylated tumor group. If these important ﬁndings can be conﬁrmed in the ongoing RTOG/EORTC Intergroup trial (RTOG0525/EORTC26053), it may allow for the selection of patients most likely to beneﬁt from chemotherapy, whereas for the other patients different strategies or novel drugs should be developed.

226

r. stupp and j.g. cairncross

ized trial comparing adjuvant BCNU with adjuvant TMZ in newly diagnosed patients who are also being irradiated is currently being conducted by the RTOG.

Treatment Options (Table 18-2) Anaplastic Astrocytoma Unlike glioblastoma, there are no randomized data providing level I evidence on the role of chemotherapy in the initial treatment of anaplastic astrocytoma. Some trials, however, have suggested a beneﬁt from adjuvant nitrosourea therapy for patients with grade III tumors [31, 35]. One of the few positive trials was reported by the EORTC. Two hundred sixty-nine patients with malignant glioma were randomized to focal RT versus RT + dibromodulcitol and BCNU [35]. The longer survival in the experimental arm was mainly due to the subgroup of 29 (!) patients with non-glioblastoma high-grade tumors. A conﬁrmatory trial for only grade 3 tumors was initiated by the EORTC and completed accrual (after 10 years) in 2003. No outcome data are available yet. A randomized trial comparing adjuvant PCV versus adjuvant PCV plus diﬂuoromethyl ornithine (DFMO), an ornithine decarboxylase inhibitor, has been reported. Patients were randomized at the end RT. Those who had already progressed or had declining performance status were not included in the study. For those with grade 3 gliomas, survival and time to progression were longer in the PCV + DFMO arm [36], a difference that was not statistically signiﬁcant. An identical trial in patients with glioblastoma was negative [37]. In recurrent anaplastic astrocytoma, TMZ has shown a high objective response rate, 35% in one phase II trial. A random-

Anaplastic Oligoastrocytoma and Oligodendroglioma Based on the 2000 WHO criteria and greater awareness of this subgroup, oligoastrocytoma and oligodendroglioma are increasingly identiﬁed as separate and distinct tumor entities [38]. While mixed oligoastrocytomas behave like astrocytomas, pure oligodendrogliomas have a more favorable natural history. Speciﬁc molecular changes, in particular the loss of genetic information on chromosomes 1p and 19q, have been associated with high sensitivity to chemotherapy and especially long survival times [39]. Two recently reported randomized trials investigated neoadjuvant (before RT) or adjuvant (after completion of RT) PCV chemotherapy in patients with newly diagnosed anaplastic oligoastrocytomas and oligodendrogliomas [40, 41]. Both trials conﬁrmed a more favorable prognosis for patients with 1p/19q loss of heterozygosity (LOH), however overall survival was not signiﬁcantly different whether patients received the chemotherapy upfront or only at progression. Even when the subgroup of patients with 1p/19q LOH was analyzed independently, no improvement in overall survival could be demonstrated with the early administration of PCV chemotherapy, although a signiﬁcant improvement in progression-free survival was seen in this genetic subset associated with PCV treatment at diagnosis. These trials emphasize that the subgroup of

TABLE 18-2. Selected treatment options and potential alternatives. Treatment

Evidence level

References

Alternatives

Evidence level

References

Remarks

I

[28]

ACNU/RT → ACNU + VM26 Hypofractionated RT

III II

[15] [43, 44]

For elderly patients

BCNU/RT → BCNU TMZ/RT → TMZ

II V

[16] [68]

Analogy to GBM

PCV × 4 before RT

I

[40]

RT → PCV × 6

I

[41]

TMZ/RT

V

[68]

Procarbazine PCV

IV IV

[18] [39, 70]

Irinotecan* Erlotinib, geﬁtinib* Imatinib*

IV IV IV

[71–73] [74, 75] [76, 77]

Glioblastoma

TMZ/RT → TMZ × 6 Anaplastic astrocytoma

RT

I

[29, 30, 67]

Anaplastic oligoastrocytoma

RT

Prolongation of progression-free survival Prolongation of progression-free survival Analogy to GBM

Recurrent glioma

TMZ TMZ

III IV

[25] [69]

In oligos both as ﬁrst- or second-line chemotherapy Disease stabilizations in GBM, but not in AA

RT, radiotherapy; TMZ, temozolomide; ACNU, nimustine; VM26, teniposide; BCNU, carmustine; PCV, procarbazine; CCNU (lomustine), vincristine. *Investigational.

18.

227

malignant glioma: chemotherapy perspective

TABLE 18-3. Recursive partitioning analysis of the EORTC/NCIC trial. Standard arm: Radiotherapy

Experimental arm: Combined TMZ/RT

RPA class

Median survival

Two-year survival

Median survival

Two-year survival

3 4 5

14.8 (10.9–17.0) 13.3 (12.0–15.0) 9.1 (8.0–11.7)

19.6 (6.9–32.3) 10.9 (5.8–16.0) 6.2 (1.1–11.2)

21.4 (15.3–N*) 16.3 (14.1–18.3) 10.3 (8.6–12.4)

43.4 (28.0–58.9) 27.9 (20.5–35.3) 16.5 (8.8–24.2)

RPA, recursive partitioning analysis; TMZ/RT, temozolomide and radiotherapy; N*, upper conﬁdence boundary not reached. Source: Mirimanoff RO, Gorlia T, Mason W, et al. Radiotherapy and temozolomide for newly diagnosed glioblastoma: recursive partitioning analysis of the EORTC 26981/22981-NCIC CE3 phase III randomized trial. J Clin Oncol 2006; 24(16):2563–2569.

patients with 1p/19q LOH represent a distinct pathologic entity.

How to Select Patients for Chemotherapy? When considering patients with malignant glioma for a speciﬁc therapy, treatment-related, tumor-related, and patient-related factors need to be considered. One of the most important deterrents to chemotherapy is poor performance status. The patient with poor neurologic function is unlikely to improve substantially with chemotherapy. Similarly, the patient with poor general function is unlikely to tolerate sufﬁcient chemotherapy to beneﬁt, especially when the response of malignant gliomas to such treatment is often slow or protracted. Tumor size and resectability are also prognostic factors [42]. For large, deep tumors and tumors crossing the midline, only a biopsy under stereotactic guidance is performed. Tumors in these locations are often accompanied by rapidly progressive neurologic symptoms, steroid dependence, and deteriorating performance status. In all clinical trials, these types of patients have very poor outcomes irrespective of treatment with TMZ/ RT or other measures. In the recent EORTC/NCIC trial, a trend for longer survival after TMZ/RT could be demonstrated among patients with a tumor biopsy only (median survival 9.4 vs. 7.4 months, p = 0.09) (see supplementary appendix to [28] at http://content.nejm.org/cgi/content/full/352/10/987/DC1). Age has been repeatedly seen to be an important prognostic factor for survival. The overall poor outcome for patients over the age of 60 or 70 years suggested that these patients should be only managed with supportive care. Two recent trials have evaluated the role of radiotherapy in elderly patients with glioblastoma. A French randomized trial compared radiotherapy (28 × 1.8 Gy, 50.4 Gy) with best supportive care in 84 patients over the age of 70 years (median, 73 years). Overall survival in patients receiving radiotherapy was 6.8 months compared with 4.0 months with best supportive care only (p = 0.002) [43]. Roa et al. reported equivalent survival when treating glioblastoma patients over the age of 60 years with either standard fractionated radiotherapy (30 × 2 Gy, 60 Gy in 6 weeks) or hypofractionated radiotherapy (15 × 2.66 Gy, 40 Gy in 3 weeks) [44]. Median survivals were 5.1 and 5.6 months, respectively (p = 0.57). In the TMZ/RT trial, patients over the age of 50 years had inferior overall survival, but the beneﬁt afforded by TMZ/RT persisted across all age groupings. (Note: This trial excluded patients who were older than 70 years.) Accordingly, treatment should not

be withheld based on age alone. A randomized trial by the Nordic Clinical Brain Tumor Study Group comparing radiotherapy alone with TMZ chemotherapy alone in elderly patients is ongoing, and a planned NCIC/EORTC Intergroup study will compare TMZ/RT to radiation alone in patients >65 years of age.

RPA Analysis (Table 18-3) Grouping of the patients according to the RTOG recursive partitioning analysis based on age, performance status, neurologic function, treatment, and tumor grade allows for some comparison between trials and variable distribution of prognostic factors [45, 46]. In the future, individual tumor characteristics like MGMT promoter methylation status or gene expression proﬁling may allow the selection of patients likely to beneﬁt from a speciﬁc therapeutic approach [34, 47, 48]. In our clinical practice, we require that the patients are able to come to the clinic to receive both chemotherapy and radiation as an outpatient. For patients with a severely impaired performance status, we frequently recommend single-modality therapy or supportive care only.

Chemotherapy Toxicity and Quality of Life Systemic administration of chemotherapy is always associated with a risk of other organ toxicity. All cytotoxic agents currently used for the treatment of malignant glioma may induce moderate or severe myelosuppression. Carmustine may also cause severe interstitial pneumonitis in 20% to 30% of patients [49– 51]. A major limitation of the PCV chemotherapy regimen is the cumulative hematotoxicity, which frequently prohibits delivering the planned number of cycles and dose intensity. Although temozolomide is considered a particularly welltolerated chemotherapy, severe myelosuppression and, particularly thrombocytopenia, may occur in up to 5% of patients. Continuous low-dose TMZ (with or without concomitant use of corticosteroids) will frequently induce profound lymphocytopenia [52–54]. CD4 counts may fall well below 200/mm3, the threshold recommended for Pneumocystis jiroveci (ex carinii) prophylaxis in HIV-positive patients [55]. By extrapolation, close monitoring of CD4 counts or prophylaxis with pentamidine inhalations or trimethoprim-sulfamethoxazole is advised for those on continuous low-dose TMZ.

228

r. stupp and j.g. cairncross

Quality of life using the EORTC quality of life questionnaire QLQ-30 and a speciﬁc brain cancer module (BM-20) has been evaluated in recent TMZ trials [56, 57]. In a randomized phase II trial, TMZ administration after ﬁrst-line radiotherapy (± chemotherapy) was compared with procarbazine [26]. Quality of life was superior in several domains in patients treated with TMZ compared with procarbazine [56]. In the EORTC/NCIC trial, no detrimental effect on quality of life was observed in the longer surviving TMZ/RT patient group [58]. Except for a slight increase in fatigue (6% with RT vs. 9% with TMZ/RT), there was no treatment difference in non-hematologic toxicities [28].

Chemotherapy or Stereotactic Radiosurgery Chemotherapy Versus Radiosurgery (Table 18-4) Stereotactic radiosurgery, despite promising reports on selected patients in phase II trials and in retrospective series, failed to demonstrate any prolongation of survival, quality of life, or local control and was the object of a recent systematic review by the American Society for Therapeutic Radiology and Oncology (ASTRO) [59]. Apparent improvements in survival in some noncontrolled trials may simply reﬂect patient selection [10, 60]. Only one randomized trial evaluated a stereotactic boost of 15 to 24 Gy administered to tumors of ≤4 cm in size before standard external beam radiotherapy [61]. Another randomized trial evaluating fractionated stereotactic boost (5 × 4 Gy) administered after the end of external beam radiotherapy was discontinued early due to a lack of accrual [62]. Why do increasing radiation doses not confer a signiﬁcant improvement in outcome [63]? Glioblastoma is an inﬁltrating tumor affecting the whole brain, and the tumor enhancement seen on

MRI is “only the tip of the iceberg.” One limitation of stereotactic radiosurgery is tumor size. The technique can only be applied to tumors with a maximum diameter of 3 to 4 cm. In reality, malignant gliomas are large and diffuse. Tumor cells can be found far beyond the initial tumor site, explaining the inevitable failure of all exclusively local treatment approaches to this disease. Even in older times, when surgery was the only treatment modality in use, heroic lobectomies and hemispherectomies were performed in vain; tumor recurrences at distant brain sites invariably occurred. The dogma of regional tumor recurrence popularized by Hochberg and Pruitt [64] may need to be revisited in the near future as patients begin to live longer after combined chemoradiotherapy and other truly effective therapies. Stereotactic radiosurgery not only requires equipment and expertise, but also it is only applicable to a minority of 10% to 25% of patients with malignant glioma. Patients considered for radiosurgery need to have an adequate performance status, and tumors should not be in proximity of vital structures. In comparison, chemotherapy with TMZ can be prescribed without signiﬁcant toxicity to 90% of patients with malignant glioma. For TMZ chemotherapy, level I evidence is available for the treatment of newly diagnosed glioblastoma. In this trial with 85 centers, many smaller hospitals participated, conﬁrming the applicability of this treatment to everyday practice. (Note: One must not extrapolate these results to the treatment of anaplastic gliomas and low-grade gliomas.) In reality, these very different treatments need not be mutually exclusive. The combination of stereotactic radiosurgery with standard chemoradiotherapy with TMZ is likely feasible and could be properly studied. In vitro, temozolomide has been shown to inhibit tumor cell migration. Similarly, the combination of temozolomide and radiotherapy together with carmustine-impregnated wafers (Gliadel; MCI Pharma, Baltimore, MD) might improve local control while eradicating inﬁltrating tumor cells [65].

TABLE 18-4. Advantages of chemotherapy and SRS. Chemotherapy

SRS

Level of evidence Tumor size

I (glioblastoma) Any

Tumors <3–4 cm

Karnofsky performance status

>60% No

>80% Yes

Manageable None Possible None High Long Large scale, anytime Minor, mainly performance status

Absent Moderate None Possible Moderate Short Limited, restrictive Highly selected, applicable to only 10% to 25% of high-grade glioma patients

Substitute for surgery Toxicity Systemic Local Drug interaction Edema induction Costs Treatment duration Availability Patient selection

Remarks

IV Chemotherapy will treat microscopic inﬁltrative disease, whereas SRS will only eliminate visible tumor SRS requires anesthesia If tissue previously available for histologic diagnosis SRS risk of late radionecrosis

18.

malignant glioma: chemotherapy perspective

References 1. Balis FM, Blaney SM, McCully CL, et al. Methotrexate distribution within the subarachnoid space after intraventricular and intravenous administration. Cancer Chemother Pharmacol 2000; 45(3):259–264. 2. Dichter MA, Brodie MJ. New antiepileptic drugs. N Engl J Med 1996; 334(24):1583–1590. 3. Kanner AM, Balabanov AJ. The use of monotherapy in patients with epilepsy: an appraisal of the new antiepileptic drugs. Curr Neurol Neurosci Rep 2005; 5(4):322–328. 4. Vecht CJ, Wagner GL, Wilms EB. Interactions between antiepileptic and chemotherapeutic drugs. Lancet Neurol 2003; 2(7): 404–409. 5. Macdonald DR, Cascino TL, Schold SC Jr, Cairncross JG. Response criteria for phase II studies of supratentorial malignant glioma. J Clin Oncol 1990; 8(7):1277–1280. 6. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000; 92(3):205–216. 7. Leimgruber A, Ostermann S, Yeon E, et al. Perfusion and diffusion MRI of glioblastoma progression in a 4 years prospective temozolomide clinical trial. Int J Radiat Oncol Biol Phys 2006; 64(3):869–875. 8. Moffat BA, Chenevert TL, Lawrence TS, et al. Functional diffusion map: a noninvasive MRI biomarker for early stratiﬁcation of clinical brain tumor response. Proc Natl Acad Sci U S A 2005; 102(15):5524–5529. 9. Wong ET, Hess KR, Gleason MJ, et al. Outcomes and prognostic factors in recurrent glioma patients enrolled onto phase II clinical trials. J Clin Oncol 1999; 17(8):2572. 10. 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):307–312. 11. Frenay M, Giroux B, Khoury S, et al. Phase II study of fotemustine in recurrent supratentorial malignant gliomas. Eur J Cancer 1991; 27(7):852–856. 12. Takakura K, Abe H, Tanaka R, et al. Effects of ACNU and radiotherapy on malignant glioma. J Neurosurg 1986; 64(1):53–57. 13. Yoshida J, Kajita Y, Wakabayashi T, Sugita K. Long-term followup results of 175 patients with malignant glioma: importance of radical tumour resection and postoperative adjuvant therapy with interferon, ACNU and radiation. Acta Neurochir 1994; 127(1–2): 55–59. 14. Kochii M, Kitamura I, Goto T, et al. Randomized comparison of intra-arterial versus intravenous infusion of ACNU for newly diagnosed patients with glioblastoma. J Neurooncol 2000; 49(1): 63–70. 15. Weller M, Muller B, Koch R, et al. Neuro-Oncology Working Group 01 trial of nimustine plus teniposide versus nimustine plus cytarabine chemotherapy in addition to involved-ﬁeld radiotherapy in the ﬁrst-line treatment of malignant glioma. J Clin Oncol 2003; 21(17):3276–3284. 16. Green SB, Byar DP, Walker MD, et al. Comparisons of carmustine, procarbazine, and high-dose methylprednisolone as additions to surgery and radiotherapy for the treatment of malignant glioma. Cancer Treat Rep 1983; 67(2):121–132. 17. Rodriguez LA, Prados M, Silver P, Levin VA. Reevaluation of procarbazine for the treatment of recurrent malignant central nervous system tumors. Cancer 1989; 64(12):2420–2423. 18. Newton HB, Junck L, Bromberg J, et al. Procarbazine chemotherapy in the treatment of recurrent malignant astrocytomas after radiation and nitrosourea failure. Neurology 1990; 40(11):1743– 1746.

229

19. Levin VA, Silver P, Hannigan J, et al. Superiority of postradiotherapy adjuvant chemotherapy with CCNU, procarbazine, and vincristine (PCV) over BCNU for anaplastic gliomas: NCOG 6G61 ﬁnal report. Int J Radiat Oncol Biol Phys 1990; 18(2): 321–324. 20. Prados MD, Scott C, Curran WJ Jr, et al. Procarbazine, lomustine, and vincristine (PCV) chemotherapy for anaplastic astrocytoma: A retrospective review of radiation therapy oncology group protocols comparing survival with carmustine or PCV adjuvant chemotherapy. J Clin Oncol 1999; 17(11):3389–3395. 21. Medical Research Council Brain Tumor Working Party. Randomized trial of procarbazine, lomustine and vincristine in the adjuvant treatment of high-grade astrocytoma: A Medical Research Council trial. J Clin Oncol 2001; 19(2):509–518. 22. Stupp R, Gander M, Leyvraz S, Newlands E. Current and future developments in the use of temozolomide in the treatment of brain tumors. Lancet Oncol 2001; 2(9):552–560. 23. 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– 1030. 24. 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):3728–3736. 25. Yung WK, Prados MD, Yaya-Tur R, et al. Multicenter phase II trial of temozolomide in patients with anaplastic astrocytoma or anaplastic oligoastrocytoma at ﬁrst relapse. Temodal Brain Tumor Group. J Clin Oncol 1999; 17(9):2762–27671. 26. Yung WK, Albright RE, Olson J, et al. A phase II study of temozolomide vs. procarbazine in patients with glioblastoma multiforme at ﬁrst relapse. Br J Cancer 2000; 83(5):588–593. 27. Brada M, Hoang-Xuang K, Rampling R, et al. Multicenter phase II trial of temozolomide in patients with glioblastoma multiforme at ﬁrst relapse. Ann Oncol 2001; 12:259–266. 28. 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):987–996. 29. Walker MD, Alexander E Jr, Hunt WE, et al. Evaluation of BCNU and/or radiotherapy in the treatment of anaplastic gliomas. A cooperative clinical trial. J Neurosurg 1978; 49(3):333–343. 30. 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(23):1323–1329. 31. Kristiansen K, Hagen S, Kollevold T, et al. Combined modality therapy of operated astrocytomas grade III and IV. Conﬁrmation of the value of postoperative irradiation and lack of potentiation of bleomycin on survival time: a prospective multicenter trial of the Scandinavian Glioblastoma Study Group. Cancer 1981; 47(4): 649–652. 32. Fine HA, Dear KB, Loefﬂer JS, et al. Meta-analysis of radiation therapy with and without adjuvant chemotherapy for malignant gliomas in adults. Cancer 1993; 71(8):2585–2597. 33. Stewart LA. Chemotherapy in adult high-grade glioma: a systematic review and meta-analysis of individual patient data from 12 randomised trials. Lancet 2002; 359(9311):1011–1018. 34. 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. 35. Hildebrand J, Sahmoud T, Mignolet F, et al. Adjuvant therapy with dibromodulcitol and BCNU increases survival of adults with malignant gliomas. EORTC Brain Tumor Group. Neurology 1994; 44(8):1479–1483. 36. Levin VA, Hess KR, Choucair A, et al. Phase III randomized study of postradiotherapy chemotherapy with combination alphadiﬂuoromethylornithine-PCV versus PCV for Anaplastic Gliomas. Clin Cancer Res 2003; 9(3):981–990.

230

r. stupp and j.g. cairncross

37. Levin VA, Uhm JH, Jaeckle KA, et al. Phase III randomized study of postradiotherapy chemotherapy with alpha-diﬂuoromethylornithine-procarbazine, N-(2-chloroethyl)-N′-cyclohexyl-Nnitrosurea, vincristine (DFMO-PCV) versus PCV for glioblastoma multiforme. Clin Cancer Res 2000; 6(10):3878–3884. 38. Kleihues P, Cavenee WK. Pathology and Genetics of Tumours of the Nervous System. Lyon: IARC Press, 2000. 39. 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(19): 1473–1479. 40. Cairncross G, Berkey B, Shaw E, et al. Phase III Trial of Chemotherapy Plus Radiotherapy Compared With Radiotherapy Alone for Pure and Mixed Anaplastic Oligodendroglioma: Intergroup Radiation Therapy Oncology Group Trial 9402. J Clin Oncol 2006; 24(18):2707–2714. 41. van den Bent MJ, Carpentier AF, Brandes AA, et al. Adjuvant Procarbazine, Lomustine, and Vincristine Improves ProgressionFree Survival but Not Overall Survival in Newly Diagnosed Anaplastic Oligodendrogliomas and Oligoastrocytomas: A Randomized European Organisation for Research and Treatment of Cancer Phase III Trial. J Clin Oncol 2006; 24(18):2715–2722. 42. 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(2):190–198. 43. Keime-Guibert F, Chinot O, Taillandier F, et al. Radiotherapy for glioblastoma in the elderly. N Engl J Med 2007; 356(15): 1527–1535. 44. Roa W, Brasher PM, Bauman G, et al. Abbreviated course of radiation therapy in older patients with glioblastoma multiforme: a prospective randomized clinical trial. J Clin Oncol 2004; 22(9):1583–1588. 45. 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):704–710. 46. Scott CB, Scarantino C, Urtasun R, et al. Validation and predictive power of Radiation Therapy Oncology Group (RTOG) recursive partitioning analysis classes for malignant glioma patients: a report using RTOG 90-06. Int J Radiat Oncol Biol Phys 1998; 40(1):51–55. 47. Hegi ME, Diserens AC, Godard S, et al. Clinical trial substantiates the predictive value of O-6-methylguanine-DNA methyltransferase promoter methylation in glioblastoma patients treated with temozolomide. Clin Cancer Res 2004; 10(6):1871–1874. 48. Godard S, Getz G, Delorenzi M, et al. Classiﬁcation of human astrocytic gliomas on the basis of gene expression: a correlated group of genes with angiogenic activity emerges as a strong predictor of subtypes. Cancer Res 2003; 63(20):6613–6625. 49. O’Driscoll BR, Hasleton PS, Taylor PM, et al. Active lung ﬁbrosis up to 17 years after chemotherapy with carmustine (BCNU) in childhood. N Engl J Med 1990; 323(6):378–382. 50. Kreisman H, Wolkove N. Pulmonary toxicity of antineoplastic therapy. Semin Oncol 1992; 19(5):508–520. 51. Schmitz N, Diehl V. Carmustine and the lungs. Lancet 1997; 349(9067):1712–1713. 52. Brock CS, Newlands ES, Wedge SR, et al. Phase I trial of temozolomide using an extended continuous oral schedule. Cancer Res 1998; 58(19):4363–4367. 53. Stupp R, Dietrich P, 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):1375–1382. 54. Su YB, Sohn S, Krown SE, et al. Selective CD4+ lymphopenia in melanoma patients treated with temozolomide: a toxicity with therapeutic implications. J Clin Oncol 2004; 22(4):610–616.

55. Kovacs JA, Masur H. Prophylaxis against opportunistic infections in patients with human immunodeﬁciency virus infection. N Engl J Med 2000; 342(19):1416–1429. 56. Osoba D, Brada M, Yung W, Prados M. Health-related quality of life in patients treated with temozolomide versus procarbazine for recurrent glioblastoma multiforme. J Clin Oncol 2000; 18(7):1481– 1491. 57. Osoba D, Brada M, Yung WK, Prados MD. Health-related quality of life in patients with anaplastic astrocytoma during treatment with temozolomide. Eur J Cancer 2000; 36(14):1788–1795. 58. Taphoorn MJ, Stupp R, Coens C, et al. Health-related quality of life in patients with glioblastoma: a randomised controlled trial. Lancet Oncol 2005; 6(12):937–944. 59. Tsao MN, Mehta MP, Whelan TJ, 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(1):47–55. 60. Curran WJ Jr, Scott CB, Weinstein AS, et al. Survival comparison of radiosurgery-eligible and -ineligible 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):857–862. 61. 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):853–860. 62. Brada M, Baumert B. Focal fractionated conformal stereotactic boost following conventional radiotherapy of high-grade gliomas: a randomized phase III study. A joint study of the EORTC (22972) and the MRC (BR10). Front Radiat Ther Oncol 1999; 33:241– 243. 63. Stupp R, Weber DC. The role of radio- and chemotherapy in glioblastoma. Onkologie 2005; 28(6–7):315–317. 64. Hochberg FH, Pruitt A. Assumptions in the radiotherapy of glioblastoma. Neurology 1980; 30(9):907–911. 65. Stupp R, Hegi ME. Treatment of brain tumors (author reply). N Engl J Med 2005; 352(22):2350–2353. 66. Mirimanoff RO, Gorlia T, Mason W, et al. Radiotherapy and temozolomide for newly diagnosed glioblastoma: recursive partitioning analysis of the EORTC 26981/22981-NCIC CE3 phase III randomized trial. J Clin Oncol 2006; 24(16):2563–2569. 67. Walker MD, Strike TA, Sheline GE. An analysis of dose-effect relationship in the radiotherapy of malignant gliomas. Int J Radiat Oncol Biol Phys 1979; 5(10):1725–1731. 68. Stupp R, van den Bent MJ, Hegi ME. Optimal role of temozolomide in the treatment of malignant gliomas. Curr Neurol Neurosci Rep 2005; 5(3):198–206. 69. van den Bent MJ, Taphoorn MJ, Brandes AA, et al. Phase II study of ﬁrst-line chemotherapy with temozolomide in recurrent oligodendroglial tumors: the European Organization for Research and Treatment of Cancer Brain Tumor Group Study 26971. J Clin Oncol 2003; 21(13):2525–2528. 70. Triebels VH, Taphoorn MJ, Brandes AA, et al. Salvage PCV chemotherapy for temozolomide-resistant oligodendrogliomas. Neurology 2004; 63(5):904–906. 71. Friedman HS, Petros WP, Friedman AH, et al. Irinotecan therapy in adults with recurrent or progressive malignant glioma. J Clin Oncol 1999; 17(5):1516–1525. 72. Raymond E, Fabbro M, Boige V, et al. Multicentre phase II study and pharmacokinetic analysis of irinotecan in chemotherapy-naive patients with glioblastoma. Ann Oncol 2003; 14(4):603–614. 73. Buckner JC, Reid JM, Wright K, et al. Irinotecan in the treatment of glioma patients: current and future studies of the North Central Cancer Treatment Group. Cancer 2003; 97(9 Suppl):2352–2358.

18.

malignant glioma: chemotherapy perspective

74. Raizer JJ, Abrey LE, Wen P, et al. A phase II trial of erlotinib (OSI-774) in patients with recurrent malignant glioma not on EIAEDs. J Clin Oncol [Proc Am Soc Clin Oncol, Post-Meeting Edition] 2004; 22(14S):107s. 75. Rich JN, Reardon DA, Peery T, et al. Phase II trial of geﬁtinib in recurrent glioblastoma. J Clin Oncol 2004; 22(1):133–142. 76. Raymond E, Brandes A, Van Osterom A, et al. Muticentre phase II study of imatinib mesylate in patients with recurrent glioblastoma: An EORTC NCCG/BTG Intergroup study. J Clin Oncol

231

[Proc Am Soc Clin Oncol, Post-Meeting Edition] 2004; 14S: 107s. 77. van den Bent M, Brandes A, Frenay M, et al. Multicentre phase II study of imatinib mesylate in patients with recurrent anaplastic oligodendroglioma (AOD)/mixed oligoastrocytoma (MOA) and anaplastic astrocytoma (AA)/low grade astrocytoma (LGA): An EORTC New Drug Development Group (NDDG) and Brain Tumor Group (BTG) study. J Clin Oncol [Meeting Abstracts] 2005; 23(16 Suppl):118s.

1 9

Meningioma Carlos A. Mattozo and Antonio A.F. de Salles

Introduction Meningiomas are generally benign lesions that account for 15% to 20% of primary brain tumors, affect predominately middle-aged patients, and occur predominately in females [1, 2]. The atypical and malignant meningiomas are characterized by successive recurrences and an aggressive behavior. Among all meningiomas, their incidence varies in the literature ranging from 4.7% to 7.1% and 1.0% to 3.7% for atypical and malignant, respectively [3–6]. There is absence of female predominance in malignant meningiomas suggesting that endocrinologic inﬂuences apparently correlated with the genesis of benign meningiomas are not active in malignant ones [4]. In general, meningiomas are usually well circumscribed, globular, and slow-growing tumors that are though to arise from mesodermal arachnoid cells. These arachnoid cap cells, localized on the apex of arachnoid granulations, are exposed to the venous blood of the sinus and have a low rate of cell division. The induction of cell division is thought to be an early important step in the neoplastic transformation of these cells. They are structurally similar to those of meningiomas, and both tend to form whorls. However, this presumed origin does not explain some unusual sites such as the ventricles or within the cerebral parenchyma, where they probably arise from perivascular arachnoidal cells [7]. Meningiomas are considered indolent tumors, but the natural history of any particular case is unpredictable. Although three well-deﬁned groups of meningiomas are currently used (named grade I, benign; grade II, atypical; and grade III, malignant), there were a great number of classiﬁcations adopted in the literature, and, as a consequence, detailed correlative clinicopathologic studies on this subject are difﬁcult to perform, and it is not possible to compare data from different centers. The last review of the World Health Organization (WHO) 2000 classiﬁcation provides histologic criteria of meningiomas with reasonably good prognostic correlation [8]. However, the majority of publications are not based on the last classiﬁcation of WHO 2000, which deﬁned the increased mitotic count as the main differentiation between atypical and anaplastic lesions.

Additionally, another characteristic of meningiomas is the histologic dedifferentiation, increasing the grade as the successive recurrences occur. This aspect is not often discussed in the literature, and the existing series report an incidence of 4.4% to 18% of cases [3, 5, 9, 10]. Ikeda et al. [11] demonstrated that the major phenotypic changes in the progression of meningiomas from the classic to the anaplastic type are the loss of the meningioma’s architecture, a decreased expression of epithelial membrane antigen (EMA), an increased expression of vimentin, and an expression of abnormal intermediate ﬁlament proteins. Recently, Al-Mefty et al. [12] published a series of 11 (6%) meningiomas with malignant progression out of 175 recurrent lesions. Interestingly, they found a complex karyotype present in the histologically lower-grade tumor, contradicting the stepwise clonal evolution model.

Surgical Considerations Microsurgery remains the best option for symptomatic intracranial meningiomas if complete resection can be achieved with low morbidity. Simpson [13] has proposed a system that grades the extent of tumor removal and correlates with tumor relapse. Grade 1 resection required gross total removal of all visible tumor plus dural attachments and was associated with 9% local recurrence. Gross total removal with dural cautery constituted grade 2 resection and was associated with a 19% relapse. A partial resection was a grade 3 procedure with a relapse rate of 40%. Poorer control rates occurred after grade 4 (decompression) or 5 (biopsy) procedures. Since this system was proposed, many other reports have conﬁrmed that local relapse of meningiomas depends upon the extent of ﬁrst resection [14–16]. After gross total removal, the local recurrence rate at 5 and 10 years approximate 10% and 20%, respectively [16]. After less than gross complete resection, relapse rates at 5, 10, and 15 years were recently reported at 25%, 70%, and 90%, respectively [16]. Jääskeläinen et al. [3] reported a recurrence rate 5 years after complete resection of 3% for benign, 38% for atypical, and 78% for anaplastic meningiomas. Nevertheless, total excision including the dural site of origin is seldom possible, especially when important vascular and/or

233

234

c.a. mattozo and a.a.f. de salles

neural structures are encased within the tumor or if the tumor inﬁltrates adjacent anatomic structures [17–19]. The success rates associated with microsurgical excision of benign skull base tumors has varied from 26% to 89% [20–25]. Although the mortality rate has decreased in recent reports, a signiﬁcant number of patients continue to experience adverse effects related to surgical procedures with rates ranging from 15% to 26% [24, 26–29].

Radiation Therapy The beneﬁt of conventional radiation therapy (XRT) to complement subtotal resection of meningiomas is well established [10, 20, 30]. External beam radiation therapy in doses of 45 to 50 Gy reduces the progression rate after partial resection of meningiomas [30, 31]. A control rate of 72% to 98% has been reported, depending on the timing of XRT after surgery, the extent of resection, the presence of cell atypia, the XRT dose, the tumor location, and the XRT technique used [30–34]. Glaholm et al. [32] reported 5-, 10-, and 15-year relapse-free survival rates of 78%, 67%, and 56%, respectively, after postoperative radiotherapy. Goldsmith et al. [33], in a retrospective review of 117 patients receiving conventional RT after subtotal resection, reported a signiﬁcant impact on progression-free survival (PFS) after XRT related to total dose. The 10-year relapse-free rate in those receiving ≤52 Gy was 65% compared with 93% for those receiving >52 Gy. For skull base benign meningiomas, a 5- and 10-year PFS rate of 92% and 83%, respectively, has been reported after XRT [35]. Atypical and malignant meningiomas differ from their benign counterparts in that they frequently recur after complete excision and are more often fatal. Milosevic et al. [36] studied the role of XRT for those lesions and found that young

FIGURE 19-1. (A) MRI scan from case 1 showing recurrence of an atypical meningioma in left tentorium 45 months after initial resection. (B) Treatment plan with shaped beans, 14 Gy, 90% isodose line, volume

Case Study 19-1 A 47-year-old woman had a left frontotemporal atypical meningioma that was resected on January 1998. On October 2001, she presented with headaches, and the MRI scan revealed a small recurrence (Fig. 19-1A). She was treated by SRS, 14 Gy delivered to 90% isodose line, volume of 1.07 cm3 (Fig. 19-1B). The follow-up scan 33 months after treatment showed a stable tentorial lesion (not shown) but a new lesion (arrow) in the ﬂoor of the left medial cranial fossa (Fig. 19-2A). The patient was retreated with 18 Gy delivered to 90% isodose line, volume of 0.65 cm3 (Fig. 19-2B). The last follow-up MRI scan in January 2005 showed both lesions decreasing in size (Fig. 19-3).

age (<58 years), modern imaging and treatment planning techniques, and a postoperative radiation dose of at least 50 Gy contribute to improved outcome in patients with atypical and malignant meningiomas. They also recommended that all patients should receive radiation therapy immediately after initial surgery. Dziuk et al. [37] stated that adjuvant XRT, recurrence status, and extent of resection were independent predictors of recurrence for malignant meningiomas. They found that the 5-year disease-free survival after an initial total resection increased from 15% to 80% with radiation therapy. Recently, Ware et al. [38] published their experience treating atypical and malignant meningiomas with resection followed by brachytherapy with permanent low-dose 125I. The median freedom from progression of the combined group of patients was 10.4 months. Furthermore, this approach resulted in a high complication rate. Their radiation necrosis rate was 27%, with 13% of patients requiring additional surgery (Case Study 19-1).

1.07 cm3. Note that there is a side inversion in the BrainLAB treatment plan software.

19.

meningioma

235

FIGURE 19-2. (A) MRI scan at 33 month follow-up showing a new lesion in the ﬂoor of the left medial cranial fossa. (B) Treatment plan with shaped beans, 18 Gy, 90% isodose line, volume 0.65 cm3.

The main criticism for XRT as a complement to surgical resection has been a high long-term complication rate, which has been reported to be from 17% to 38% when patients were followed longer than 5 years [39]. Nutting et al. [35] in their skull base benign meningioma series reported that complications such as retinopathy, cataract, hypopituitarism, and shortterm memory deﬁcit were likely to have been related to irradiation, occurring in 13.4% of the patients. The true incidence of neuropsychologic side effects for patients who survive more than 6 months without tumor recurrence, as in the case of meningiomas, has been signiﬁcantly underestimated. A study of 748 patients showed some degree of neuropsychologic deﬁcit in 29%, and 12% were classiﬁed as having post-radiation dementia [40].

FIGURE 19-3. MRI scans 6 months after the second treatment by SRS showing both lesions with decreased size.

Stereotactic Radiation Radiation therapy to large extents of the brain is necessary only in selected meningiomas. Malignant meningiomas with partial resection, as well as atypical meningiomas with risk of dural spread, require more extensive radiation ﬁelds. Benign lesions, well demarcated, must be managed with focal radiation, either stereotactic radiosurgery (SRS) or stereotactic radiotherapy (SRT) depending on involvement of eloquent structures or volume of the lesion. The application of SRS and SRT for treatment of meningiomas has been extensively studied at our institution. For the radiosurgery procedures at UCLA, a Clinac 18 linear accelerator (Varian, Inc., Palo Alto, CA) adapted to a Philips SRS 200 was used from 1990 to 1996. A dedicated Varian 600 SR and X-Knife software (Radionics, Inc., Burlington, MA) were introduced for the development of SRT, being used from 1996 to 1998. The Novalis dedicated system with miniature multileaf collimator capability has been used since 1997 (BrainLAB, Munich, Germany). The whole UCLA meningioma series was published by Torres et al. [41] with a review of 161 patients who underwent SRS or SRT for treatment of 194 meningiomas in all locations. Clinical and radiologic follow-up were obtained in 128 patients harboring 156 lesions. Stereotactic radiosurgery was used to treat 79 lesions, and SRT was used to treat 77 lesions. The overall mean follow-up period was 32.5 months. The mean prescribed dose for the SRS group was 15.6 Gy, whereas for the SRT group, the dose was 48.4 Gy. For benign tumors, the overall local control rate was 94.8%. The control rates for the SRS and SRT were 92% and 97.2%, respectively (Table 19-1). The overall rate of clinical improvement was 33.5%. Worsening of the clinical symptoms was observed in 7.9% of patients who underwent SRS and in 3.1% who underwent SRT (Table 19-2). The incidence of clinical complications in the SRS

236

c.a. mattozo and a.a.f. de salles

TABLE 19-1. Tumor growth control after stereotactic radiation of benign meningiomas.

TABLE 19-3. Tumor growth control after stereotactic radiation of atypical meningiomas.

Tumor size

SRS (%)

SRT (%)

Total (%)

Tumor size

SRS (%)

SRT (%)

Total (%)

Decreased No change Increased Total Tumor control

22 36 5 63 58

24 (33.3) 46 (63.8) 2 (2.7) 72 (100) 70 (97.2)

46 (34.1) 82 (60.7) 7 (5.2) 135 (100) 128 (94.8)

Decreased No change Increased Total Tumor control

1 (6.2) 4 (25) 11 (68.7) 16 (100) 5 (31.2)

1 (20) 2 (40) 2 (40) 5 (100) 3 (60)

2 (9.5) 6 (28.5) 13 (62) 21 (100) 8 (38.1)

(34.9) (57.1) (7.9) (100) (92)

and SRT groups was 5% and 5.2%, respectively. For the group of 21 atypical meningiomas, the tumor growth control was 31.2% and 60% for the SRS and SRT groups, respectively (Table 19-3). The analysis of the long-term follow-up (mean, 72.5 months) group of 32 patients revealed tumor growth control rate of 92.3% and 100% for the SRS and SRT treated patients, respectively. Worsening of previous neurologic deﬁcit was identiﬁed in two (7.9%) patients treated with SRS. No complications were found in the SRT patients [42]. De Salles et al. [43] studied the role of SRS and SRT in the management of skull base meningiomas and proposed a grade system (see the section “Patient Algorithm” later). They studied 40 patients (34 SRS; 6 SRT) treated during an 8-year period. Tumor control rates were 90% for grade I, 86% for grade II, 86% for grade III, 42% for grade IV, and no control for grade V (Table 19-4). The control rate correlated well with the tumor grade (p < 0.036). Additionally, the complication rate was 7.5% (three cases) occurring only in SRS treatments. Since this report, grades III, IV, and V are treated with SRT at our institution. The experience of treatment of cavernous sinus meningiomas with SRT was recently reviewed by Selch et al. [44]. They studied 45 patients treated with a median prescribed dose of 50.4 Gy delivered to a median tumor volume of 14.5 cm3. The 3-year progression-free survival rate was 97.4% and the preexisting neurologic complaints improved in 20% of patients. One patient had an ipsilateral cerebrovascular accident 6 months after SRT, although this complication is not likely related to the procedure.

Stereotactic Radiosurgery SRS has become a well-accepted modality for the treatment of a number of neurologic indications, ranging from primary and metastatic malignancies to benign tumors, arteriovenous malformations (AVMs), pain, and other neurologic conditions. Unlike conventional fractionated radiotherapy, SRS does not exploit differences in the repair capacities of normal and neoplastic cells after repeated exposure to sublethal doses of irra-

TABLE 19-2. Clinical follow-up results. Neurologic ﬁndings

SRS (%)

SRT (%)

Total (%)

Improved Unchanged Worsened Number of cases

22 (35) 36 (57.1) 5 (7.9) 63 (100)

21 (32.3) 42 (64.6) 2 (3.1) 65 (100)

43 (33.5) 78 (60.9) 7 (5.4) 128 (100)

diation. SRS achieves a therapeutic ratio by virtue of the steep dose gradient between the isocenter and surrounding normal parenchyma, inevitably produced by the convergence of multiple small radiotherapy beams. The biologic effect is irreparable cellular damage and vascular occlusion within the high dose volume. Meningiomas have several characteristics that make them excellent tumors for radiosurgery [45]. First, they usually grow slowly, which allows time for the effects of irradiation to materialize. Second, they are easily imaged with both computed tomography (CT) and magnetic resonance imaging (MRI), thereby facilitating accurate stereotactic targeting and appropriate dose planning. Finally, a high dose with rapid fall-off can be directed to the imaging-deﬁned tumor margin with conﬁdence because meningiomas are usually well-demarcated and rarely invade surrounding brain tissues (Fig. 19-4). The application of stereotactic radiation using all available techniques, including linear accelerator (linac) [41, 43, 44, 46–48], Gamma Knife [49–54], and proton beam [55], has been reported in the literature. The effectiveness of SRS and the risks of complication depend on the size of the tumor. Lesions larger than 3 cm in diameter must be treated with signiﬁcantly lower doses of radiation to avoid excessive risk of complications. Such small radiation doses may make radiosurgery ineffective. Tumor location also affects the treatment decision, depending on the proximity to eloquent structures (especially the optic apparatus). Singledose SRS is indicated for meningiomas smaller than 3 cm or 20 cm3 in volume and with a minimal distance from the optic apparatus of between 2 and 4 mm [54, 56, 57]. Furthermore, tumors causing signiﬁcant brain-stem compression may not be suitable for single-dose radiosurgery because a slight swelling during the course of tumor necrosis induced by radiosurgery may not be tolerable. Many alternatives were attempted in order to overcome such limitations, usually resulting in treat-

TABLE 19-4. Tumor control rates for cavernous sinus meningiomas according to the UCLA grading system. Grade

Meningioma radiologic aspect

I II

Conﬁned to the cavernous sinus Involvement of the petroclival region without brain-stem compression Extension to and compression of the optic nerve, chiasm, or tract Involvement of the petroclival region with compression of the brain stem Extensive involvement of both cavernous sinus

III IV V

Tumor control (%)

90 86 86 42 0

19.

meningioma

FIGURE 19-4. (A) MRI scan showing a incidental ﬁnding of a small falx meningioma in a 77-year-old woman with high operative risk. The lesion had a slight increase in size during 1 year of follow-up and was treated with 14 Gy, 90% isodose line. (B) Stable size of the tumor 40 months after treatment.

ment failure. This included two-staged procedures [49, 58] or a decrease in the ideal peripheral dose [59]. A lack of homogeneity in the radiosurgery dose has been a controversial issue in SRS treatment of brain tumors. Dose homogeneity has been sacriﬁced to obtain conformal tumor planning, thereby avoiding nearby normal structures. However, those structures encased by the tumor, such as cranial nerves, the carotid artery, or even the optic apparatus, may be jeopardized with inhomogeneous radiation treatment. Multiple isocenters restrict the prescription to low isodose volumes with a very high gradient of dose between the periphery and interior of the tumor. According to Nedzi et al. [60], dose inhomogeneity ≥5 Gy was the most signiﬁcant factor associated with morbidity after SRS. Within the past few years, however, the development of miniature multileaf collimators has altered the way in which radiosurgery is performed with linac systems. This technique allows the beam to conform to the target closely and exclude normal structures. In addition to high target conformity, the “single isocenter shaped beam approach” results in a homogeneous dose within the target volume, as well as a highly efﬁcient planning and treatment process. SRS has already proved to be effective in the treatment of meningiomas. Kondziolka et al. [52] evaluated the long-term outcome of 85 patients (90% of them skull base located) who underwent Gamma Knife radiosurgery. Tumor control was obtained in 93% of patients. Tumor resection was necessary in 5 (5.8%) patients because of delayed tumor growth or persistent symptoms. New or worsened neurologic deﬁcits occurred in 5 (5.8%) patients during the follow-up. Thirty-four percent to 71% of meningiomas treated by SRS are expected to decrease in size, and 37% to 57% are expected to stabilize [41, 51, 61, 62]. DiBiase et al. [63] studied the factors predicting local tumor control after Gamma Knife radiosurgery in 162 patients harboring meningiomas. The median prescribed dose was 14 Gy to the 50% isodose line and the median tumor volume was 4.5 cm3. The 5-year disease-free survival was 86.2%. The factors correlated with a worse prognosis in their study were male patients, conformity index less than 1.4, and tumor volume greater than 10 cm3. However, patients who were treated with less conformal plans that covered the dural tail had better outcomes (Fig. 19-5).

237

Hakim et al. [46] published the results of 127 patients with 155 meningiomas treated by SRS with linac, including atypical and malignant meningiomas. The median marginal dose was 15 Gy and the median volume was 4.1 cm3. Freedom from progression was observed in 84.3% of patients at a median time of 23 months. The actuarial tumor control for the patients with benign meningiomas was 89.3% at 5 years. Permanent complications were observed in 4.7% of patients. SRS for patients with small- to medium-sized meningiomas was recently reported to provide tumor control equivalent to Simpson grade 1 resection. Pollock et al. [64] compared two groups of patients with lesions smaller than 35 mm. The 62 patients who underwent SRS with a mean margin dose of 17.7 Gy had no statistically signiﬁcant difference in the 3- and 7-year actuarial PFS when compared with 57 patients who underwent complete surgical resection, including their dural attachments and bone abnormalities (Simpson 1) with no radiation. Furthermore, SRS provided a higher PFS rate compared with that in patients in whom resections were Simpson grade 2 (p < 0.05) and grade 3 to 4 (p < 0.001). The complication rate was 10% for SRS and 22% for the surgical group (p = 0.06). Patients treated by radiosurgery for meningioma must be followed carefully. Flickinger et al. [65] studied a group of 219 meningiomas without pathologic veriﬁcation. The actuarial rate of identifying a diagnosis other than meningioma was 2.3% at 5 and 10 years. Nakatomi et al. [66] reported one patient treated by SRS for a supposed cavernous sinus meningioma that required surgical resection 17 months after because of recurrence. The postoperative diagnosis was primary cavernous sinus malignant lymphoma. Kim et al. [67] recently analyzed the role of Gamma Knife radiosurgery in 23 patients with 26 superﬁcially located meningiomas. The median tumor volume was 4.7 cm3, and the mean margin dose was 16 Gy at the 50% isodose line. The local control rate was 95% with stable MRI imaging in 65% of the cases. Symptomatic and transient peritumoral high signal on T2-weighted magnetic resonance (MR) images was observed in 43% of patients. This complication was associated with a high integral dose and a large tumor volume. Moreover, tumor shrinkage was more prominent in the patients with symptomatic high signal (p = 0.03). MRI showed peritumorous imaging changes in 24% of patients treated by SRS in the series published by Chang et al.

FIGURE 19-5. (A) MRI scan from a 74-year-old German woman with an incidental found meningioma in the tip of the left temporal fossa. She was followed with serial MRIs and growth was documented. The patient refused surgery and accepted radiosurgery. (B) Treatment plan showing inclusion of the dural tail within the prescription isodose line (90%, 14 Gy, shaped beam).

238

c.a. mattozo and a.a.f. de salles

[61]. Their overall rate of symptomatic edema was 9.3%. Furthermore, these imaging ﬁndings were related to tumor dose in univariate analysis, and with tumor location in the multivariate analysis. Radiation-induced imaging changes were more often seen in convexity, parasagittal, and falx meningiomas than in those of the skull base.

Skull Base Meningiomas–SRS

of patients. Additionally, all patients who developed optic system neuropathy (6.4%) received radiation doses higher than 8 Gy to optic apparatus. Leber et al. [75] correlated the doses delivered in the treatment of cavernous sinus lesions by SRS with incidence of optic neuropathy. The actuarial rate was zero for patients who received a dose of less than 10 Gy, 26.7% for patients receiving 10 to less than 15 Gy, and 77.8% for those who received doses of 15 Gy or more (p < 0.0001). Stafford et al. [76] reported 218 parasellar SRS procedures with 73% of patients receiving more than 8 Gy to a short segment of the optic apparatus. The overall incidence of radiation neuropathy was 1.9%; the risk of neuropathy was 1.1% for patients receiving 12 Gy or less. Morbidity after SRS is strongly associated with tumor volume and location [15, 77]. For these reasons, application of SRS to cavernous sinus meningiomas has generally been restricted to tumors <3 cm in greatest dimension and located several millimeters from the optic apparatus [56, 77–79]. These are the grade I and grade II of the UCLA classiﬁcation [43]. Tumors grade III, IV, and V are candidates for SRT.

Meningiomas in the skull base account for the largest group of incompletely resected lesions usually due to cavernous sinus involvement, brain-stem adhesion, and inﬁltration of adjacent structures. Complete excision in these situations is unlikely without causing morbidity [68–70]. Meningiomas are the most frequent benign tumors treated by radiosurgery, and the majority of them are located on the skull base (Fig. 19-6). Liscak et al. [71] studied 176 patients with meningiomas located on the skull base. They obtained local control in 98% of patients and a persistent morbidity rate of 4.5%. The rate of radiation-induced edema was 11%. Signiﬁcantly lower edema occurrence was observed when the dosage to the tumor margin was lower than or equal to 14 Gy. Chuang et al. [72] studied the linac-based SRS for 43 unresectable or partially resected meningiomas in the base of the skull. The 7-year local control in the SRS-alone group and surgical excision with SRS group was 100% and 84.4%, respectively (p = 0.21). Using a mean margin dose of 13 Gy with a Gamma Knife unit for treatment of 32 petroclival meningiomas, Roche et al. [73] achieved local control in all patients. Although no new cranial deﬁcit was observed, two patients developed hemiparesis due to pontine infarction. Spiegelmann et al. [48] studied 42 patients with meningiomas involving the cavernous sinus treated with linac. A mean dose of 14 Gy was delivered to the tumor margin and the median tumor volume was 8.2 cm3. Tumor control was observed in 97.5% of the patients. Cranial nerve complications included new trigeminal neuropathy in 4.7% and a new visual defect in 2.8%. Despite the dosimetric advantages of SRS, permanent cranial nerve deﬁcit is more common in the treatment of cavernous sinus meningiomas. Tishler et al. [74] studied 62 patients (42 with meningiomas) inside or near the cavernous sinus treated by SRS. New cranial neuropathies were found in 19%

SRT is a treatment option that may be undertaken when the risk of SRS is unacceptably high; for instance, in cases in which tumors involve the optic pathways, the brain stem, or lesions with larger volume near eloquent structures. This technique combines the physical dose localization advantage of SRS with the radiobiologic beneﬁts of dose fractionation [80]. Unlike SRS, the treatment planning ﬁducial system and patient immobilization device can be applied noninvasively [81]. The accuracy of patient repositioning during SRT has been documented [82–84]. A total fractionated dose of ionizing radiation, known from conventional radiation therapy experience to be effective for meningioma yet tolerated by the central nervous system, can be delivered by the SRT technique. Lo et al. [85] studied a group of 18 patients treated by SRT for meningiomas with a large volume or adjacent to eloquent structures. The median dose was 54 Gy and the median volume was 8.8 cm3. The 3-year local control rate was 92.3%; one complication was reported (5.5%). These ﬁndings were not statistically signiﬁcant different when compared with another group

FIGURE 19-6. (A) MRI scan showing a cavernous sinus and petroclinoid ligament meningioma in a 71-year-old woman with pain in V2 and V3 region. (B) Treatment plan with multiple isocenters, 16 Gy,

50% isodose line. (C) MRI scan showing stable image after 9 years of follow-up. The patient is requiring only sporadic medication for facial pain.

Stereotactic Radiotherapy

19.

meningioma

239

FIGURE 19-7. (A) MRI scans of a 52-year-old woman experiencing double vision and lack of balance showing a large meningioma involving the left cavernous sinus, left petroclival region extending into the prepontine and premedullary cisterns. The patient refused

surgery and was treated by SRT. (B) Treatment plan using shaped beams with 50.4 Gy in 28 fractions, 90% isodose line, volume 39.22 cm3. (C) MRI scans showing decrease in tumor size 43 months after treatment.

of 35 patients who underwent SRS with a median dose of 14 Gy delivered to a median volume of 6.8 cm3. The application of SRT was recently reviewed by Zabel et al. [86] in 317 patients with benign and atypical meningiomas in all intracranial sites. The median total dose was 57.6 Gy delivered to a median volume of 33.6 cm3. The overall local control rate was 93.1%. As expected, the local tumor failure was signiﬁcantly greater in patients with WHO grade 2 meningiomas than in patients with WHO grade 1 or unconﬁrmed histology (p < 0.002). Moreover, patients with a tumor volume larger than 60 cm3 had a higher recurrence rate (15.5%) when compared with patients with tumor volumes smaller than 60 cm3 (p < 0.001). They observed worsening of preexisting symptoms and new clinical symptoms in 8.2% and 2.5% of patients, respectively.

The tumor control rates for WHO grade 1 and WHO grade 2 tumors were 98.3% and 77.7%, respectively. Preexisting cranial nerve symptoms resolved completely in 28% of the patients. Only 7% of patients showed an impairment of preexisting neurologic symptoms, and 1.6% experienced new clinically signiﬁcant deﬁcits. These excellent results corroborate the UCLA experience with skull base meningiomas treated with SRT [44].

Skull Base Meningiomas–SRT Skull base meningioma patients for whom microsurgery and/or SRS are contraindicated seem to be the ideal application of SRT (Fig. 19-7). Alheit et al. [82] studied a group of 24 patients (21 of them harboring skull base meningiomas) treated by SRT with doses ranging from 50 to 55 Gy delivered to a median volume of 21 cm3. The progression-free survival rate was 100%, and two (8%) patients experienced early side effects, including one VII nerve palsy and one Addisonian state. Debus et al. [87] published a large series of 189 patients treated by SRT for skull base meningiomas with a mean radiation dose of 56.8 Gy. The median target volume was 52.5 cm3.

Intensity-Modulated Radiation Therapy The intensity-modulated radiation therapy (IMRT) technique modulates the intensity of the photon beam across the treatment portal. It has shown the ability to conform the dose to complex-shaped target volumes. Baumert et al. [88] made comparisons between the treatment plans of SRT and IMRT for skull base lesions. The conformity index and the sparing of the organs at risk were better for the IMRT plan. The largest improvement was observed for multifocal and irregular tumors. Pirzkall et al. published the results of IMRT treatment on 20 patients with benign skull base meningiomas. They used a median minimum dose of 40 Gy delivered to a median volume of 108 cm3. In a median follow-up of 36 months, the local tumor control rate was 100%. The preexisting neurologic symptoms improved in 60% and worsened in 5% of patients. They also observed no radiation-induced peritumoral edema or new onset of neurologic deﬁcits.

240

c.a. mattozo and a.a.f. de salles

Case Study 19-2 A 51-year-old woman had surgical removal of a frontal fossa meningioma in January 2001 at an outside facility. The patient was stable until early 2003 when she started experiencing fatigue and headaches. The MRI scan showed a residual tumor arising from the olfactory groove and extending into the right cribriform plate and right ethmoid. She underwent a craniofacial approach with incomplete tumor resection in August 2003. The pathology report revealed meningioma with atypical features, and the MRI scan showed residual lesion in frontal fossa (Fig. 19-8). Given the aggressive behavior of this residual tumor and the vision preservation, the treatment option was intensity-modulated stereotactic radiotherapy (IMSRT). The patient was treated with 33 fractions of 1.8 Gy with a total radiation dose of 59.4 Gy delivered to 90% isodose line, volume of 24.83 cm3 (Fig. 19-9). The MRI scans 1 year after the treatment showed the lesion stable in size (Fig. 19-10).

Interestingly, the application of intensity-modulated stereotactic radiosurgery (IMSRS) for treatment of small skull base lesions was compared with dynamic conformal arc technique (DCA). Superior homogeneity and coverage for DCA was observed. The IMSRS also increased the time for planning, dose delivery, and integral dose to the brain [89]. The intensity-modulated stereotactic radiation therapies (IMSRT or IMSRS) have limited application when treating intracranial meningiomas; speciﬁc cases where the eyes or the brain stem are surrounded by the lesion may beneﬁt from this technique (Case Study 19-2).

Special Applications of Stereotactic Radiation Optic Sheath Optic nerve sheath meningiomas are uncommon, slow-growing tumors that can be categorized as primary and secondary according to their origin. Although conservative management can be considered, the most common treatment is resection, which may lead to blindness due to damage to the pial vascular

FIGURE 19-8. Coronal and sagittal MRI scans from case 2 showing residual meningioma in frontal fossa and olfactory groove.

supply of the optic nerve [90]. Radiation therapy has been used in recent decades to treat these tumors [91, 92]. SRT has been used as one alternative to avoid radiationinduced optic neuropathy. Andrews et al. [93] treated 30 patients with 33 optic nerve sheath meningiomas with SRT. Local tumor control was observed in all patients, and they observed 92% of preserved vision among the optic nerves with vision before treatment. Improvement in visual acuity and/or visual ﬁeld occurred in 42% of nerves. Visual loss was observed in 2 (6.6%) patients and 1 (3.3%) developed optic neuritis responsive to steroids. Becker et al. studied the role of SRT in 39 patients with either primary or secondary optic nerve sheath meningiomas using a median dose of 54 Gy. The tumor local control was 100% after a median follow-up of 35 months. The rates of visual ﬁeld stabilization, improvement, and worsening were 70%, 25%, and 4%, respectively. Additionally, new endocrinologic deﬁcits after treatment were found in 14% of patients with primary tumors and in 8% of patients with secondary tumors. The visual improvement observed in patients treated by SRT for optic sheath meningiomas was recently reported to occur within 1 to 3 months after treatment in 81% of patients who experienced visual improvement [94]. Homogeneity of the dose distribution becomes important when planning SRT for these tumors, as frequently the optic nerve is surrounded by the tumor. This is possible with shaped beams.

Atypical and Malignant Meningiomas The role of conventional radiation therapy and brachytherapy in the management of atypical and malignant meningiomas is well established [33, 36–39]. Although a large number of series are focused on stereotactic radiation for treatment of meningiomas, relatively few published studies have deﬁned its application in atypical and malignant histologic features (Fig. 19-11). At our institution, SRS and SRT are used as rescue of recurrences after surgery and regional radiation. Ojemann et al. [95], using a Gamma Knife unit, treated 22 patients with primary or recurrent malignant meningiomas. The overall 5-year survival and progression-free survival rates were 40% and 26%, respectively. Age (<50 years) and tumor volume (<8 cm3) were signiﬁcant predictors of better prognosis. Complications were observed in 5 (23%) patients who developed radiation necrosis. In the meningioma series published by Stafford et al. [54], 22 patients harboring either atypical (n = 13) or malignant (n = 9) meningiomas were treated by Gamma Knife SRS. They obtained 5-year survival rates of 76% and 0% for atypical and malignant meningiomas, respectively. The 5-year local control rates were 68% and 0% for atypical and malignant tumors, respectively. Univariate analyses of tumor histologic features demonstrated no correlation between complications arising after radiosurgery. Harris et al. [96] studied the results of Gamma Knife SRS for treatment of 18 atypical and 12 malignant meningiomas. The 5- and 10-year overall actuarial survival rates for atypical meningiomas were 59% (±13), and malignant lesions had 5- and 10-year overall actuarial survival rates of 59% (±16) and 0. These rates were not signiﬁcantly different from one another. Atypical meningiomas had a 5-year PFS of 83% (±7), and

19.

meningioma

241

FIGURE 19-9. Treatment plan with intensity-modulated stereotactic radiotherapy showing the sparing of eyes. The total dose of 59.4 Gy was given with 33 fractions of 1.8 Gy delivered to 90% isodose line for a volume of 24.83 cm3.

malignant tumors had a PFS of 72% (±10) (p = 0.018). They also found a signiﬁcantly better PFS for lesions treated without evidence of progression before SRS. The results published recently by Huffmann et al. [97] using Gamma Knife SRS for treatment of atypical meningiomas is the ﬁrst report where the WHO 2000 classiﬁcation criteria was

clearly applied in a radiosurgery study. They treated 15 patients with 21 lesions using a median prescription dose of 16 Gy. The local tumor control and marginal/distant tumor control were 95% and 60%, respectively. The only patient with local relapse was treated with a prescription dose of 15 Gy. Zabel et al. [86] published a long-term follow-up to meningiomas treated by SRT. A group of 26 patients harboring atypical lesions had 3-, 5-, and 10-year recurrence-free survival rates of 96%, 89%, and 67%, respectively. Atypical histologic grade and tumor volume larger than 60 cm3 were correlated with a signiﬁcantly worse local control in comparison with benign meningiomas.

Combination of Surgery and Radiosurgery

FIGURE 19-10. MRI scans 1 year after treatment showing the lesion stable in size.

The concern of some neurosurgeons about better functional outcomes has led some groups to adopt more conservative approaches for skull base meningiomas. Radiosurgery has become more accepted and recently is being combined with surgery. Aziz et al. [98] reviewed 38 patients who underwent surgical resection for large sphenoid wing meningiomas that invaded the cavernous sinus. They proposed that tumor

242

c.a. mattozo and a.a.f. de salles

FIGURE 19-11. (A) MRI scan showing a left frontoparietal dural based recurrent atypical meningioma. (B) Treatment plan with shaped beans (with a left to right inversion), 47.6 Gy, 85% isodose line in 28

fractions, volume 20.68 cm3. (C) MRI scan showing the tumor increasing in size after 20 months of follow-up. The pathology report revealed malignant progression to WHO grade III meningioma.

portions involving the medial compartment of cavernous sinus, as well as portions that inﬁltrate the superior orbital ﬁssure, should be left for observation or stereotactic radiation. Maruyama et al. [99] proposed a treatment strategy combining nonradical surgery and radiosurgery for the management of cavernous sinus meningiomas. Twenty-three patients had partial or subtotal surgical resection with the main aim to remove the tumor components extending outside the cavernous sinus. This was performed when the tumor was attached to or displacing the optic apparatus and the brain stem, in patients with tumors larger than 3 cm in mean diameter, or in tumors suspected of malignancy with the aim to obtain histologic veriﬁcation. The residual tumors underwent radiosurgery or fractionated radiotherapy depending on whether they were larger or smaller than 3 cm. Another group of 17 patients underwent SRS-only, given the presence of tumors smaller than 3 cm that were distant from optic apparatus. The 5-year actuarial local control rate was 94%.

Recently, cytology of tumors located in the cavernous sinus became possible [100]. Frameless stereotaxis allows for needle biopsy of these tumors through the foramen ovale. Biopsy of these tumors became part of the management of tumors in the cavernous sinus requiring a differential diagnosis (Case Study 19-3).

Case Study 19-3 A 48-year-old man with diagnosis of AIDS had a 3-month history of double vision and pain in the left periorbital area as well as some decreased sensation in the left face. The MRI scan showed an extraaxial mass in the cavernous sinus region extending into the left cerebellopontine angle cistern and the prepontine cistern. The radiology report was suggestive of a meningioma (Fig. 19-12). Given the patient’s history of HIV, the option was to perform a frameless cavernous sinus biopsy through the foramen ovale (Fig. 1913). The pathology report revealed that the tumor was likely of lymphocytic and histiocytic origin. It was believed that the mass represented a lymphoma. The patient underwent stereotactic radiotherapy with a total dose of 45 Gy in 25 fractions of 1.8 Gy delivered to the 85% isodose line, volume of 17.0 cm3 (Fig. 19-13). The MRI scans 6 months after treatment showed a remarkable decrease in size of the lesion. This rapid response conﬁrms the diagnosis of a cavernous sinus lymphoma (Fig. 19-14).

Recommendations Patient Algorithm Treatment decisions for meningiomas are best made by a team of neurosurgeons and radiation oncologists representing expertise in microsurgery, radiosurgery, and radiation therapy. A combined approach to a tumor with microsurgery and radiosurgery is promising. Decompression of critical structures by microsurgery facilitates and increases the effectiveness of radiosurgery. The literature suggests that conservative resection followed by radiation therapy leads to better preservation of neural function [24, 26, 101, 102]. When choosing a course of treatment, the patient’s age and clinical status must also be considered. Radiosurgery should be considered the primary option in elderly patients in which the risks of surgery are too great or when the potential morbidity of surgery may be unacceptable (Fig. 19-15). Because radiosur-

FIGURE 19-12. MRI scans from case 3 showing a cavernous sinus region tumor extending into the left cerebellopontine angle cistern and the prepontine cistern. The images were suggestive of a meningioma.

19.

meningioma

243

FIGURE 19-13. (A) Screen shot of the navigation system showing the needle trajectory for the frameless foramen ovale biopsy. (B) Treatment plan with dynamic arcs. The tumor was treated with stereotactic

radiotherapy with a total dose of 45 Gy, 85% isodose line. The volume treated was 17.0 cm3.

gery only requires local anesthesia, several patients in their eighties and one in her nineties have successfully undergone radiosurgery for their meningiomas at UCLA. The clinical status of a patient should always be thoroughly assessed prior to treatment. When choosing the best treatment, it is important to note the onset, severity, and rate of progression of any symptoms. A patient who is asymptomatic may choose to postpone any treatment until symptoms develop or until imaging provides evidence of growth. A patient’s clinical status may weigh heavily on the choice between radiosurgery or microsurgery. Patients with parasellar lesions with acute neurologic deterioration and previous history of malignancy, immunodeﬁciency, or inﬂammatory disease could beneﬁt from histologic diagnosis before treatment. Since 2003, we have used an image-guided frameless biopsy procedure via the foramen ovale, avoiding a major microsurgical procedure. This procedure is well tolerated by the patient, with a short hospital stay [100]. The effectiveness of radiosurgery and the risks of complication depend upon the size of the tumor. Lesions larger than 3 cm in diameter must be treated with signiﬁcantly lower doses of radiation to avoid excessive risk of complications. Such small radiation doses may make radiosurgery ineffective; therefore surgical resection remains the best treatment for large tumors. Patients who are not surgical candidates who have tumors with large size and/or close to optic apparatus or brain stem are given stereotactic radiotherapy instead of SRS. Both modalities are used in different clinical scenarios in our department and should be available to complement each other, providing safer treatment options for patients (Fig. 19-16). Additionally, for cavernous sinus meningiomas, the decision between SRS or SRT is made with the help of our grading

system [43]. Tumors conﬁned to the cavernous sinus (grade I) or involving the petroclival region without brainstem compression (grade II) are suitable for SRS. Lesions with extension to the optic apparatus (grade III), involvement of petroclival region with compression of the brainstem (grade IV), or involvement of both cavernous sinus (grade V) are recommended to receive SRT (Table 19-4). Surgical efforts should be concentrated in turning a high-grade cavernous sinus meningioma into a lower-grade tumor that may be controlled by stereotactic radiation. Optic sheath meningiomas are also indicated for SRT unless the lesion has small size in a patient without useful vision, when SRS or microsurgery may be used.

Procedure and Treatment Planning At UCLA, the SRS procedure begins early in the morning with the ﬁxation of a stereotactic frame to the patient’s skull.

FIGURE 19-14. MRI scans 6 months after treatment showing a remarkable tumor response to radiation. This behavior is compatible with the diagnosis of a cavernous sinus lymphoma.

244

c.a. mattozo and a.a.f. de salles

FIGURE 19-15. (A) MRI scan from a 71-year-old man with diagnosis of Parkinson disease showing multiples meningiomas. (B) The patient was followed and the MRI scan showed increase in size of the right

sylvian ﬁssure lesion after 30 months. Only this lesion was treated with 12 Gy, 85% isodose line. (C) MRI scan showing the tumor slightly smaller in size 32 months after treatment.

Routinely, a BRW (Radionics) or a BrainLAB stereotactic frame is applied after administration of a local anesthetic. A stereotactic contrast-enhanced CT scan is obtained and fused with a previously acquired 3-mm slice MR image. Image fusion is performed using a mutual information algorithm, which allows geometric alignment between the two image sets. Thus, the CT scan is used for the precision of their anatomic localization, whereas high-resolution MR images provide the anatomic details necessary for treatment planning. The same protocol is used for the SRT-treated patients; however, the patients are immobilized with a custom-ﬁtted thermoplastic face mask (U-PLAST; BrainLAB) instead of a regular frame [82]. Before each daily treatment, reproducibility of patient positioning is evaluated using the depth helmet method of Kooy et al. [83]. For both SRS and SRT techniques, the three-dimensional treatment plans were generated using a BrainLAB planning system (versions 3.5 to 4.03; BrainLAB). The treatment plan is safely performed using the “beam’seye-view” conformal technique (Fig. 19-17). According to the

tumor size and location, the radiation is delivered using dynamic arcs or static shaped beams using a dedicated 6-MV linear accelerator (Novalis; BrainLAB). This equipment utilizes a micro-multileaf collimator, obviating the need for multiple isocenters for dose conformation.

Complication Avoidance The dose is chosen based on tumor volume using an isoeffect line generated from an integrated logistic formula. The formula relates dose and volume to the incidence of complications after SRS for AVMs [103]. In general, the authors attempted to remain below the 3% isoeffect line for prevention of permanent symptomatic brain injury. On a log-log plot of dose versus volume, this line is straight with a negative slope and has proved to be a reliable predictor of SRS complications, as previously described by Kjelberg et al. [104] Dose was also tailored to the surrounding anatomy and typically reduced for lesions near the brain stem or cranial nerves; the optic chiasm was restricted to 8 Gy.

Intracranial Meningioma Symptomatic Mass Effect

Incidental Finding

Observation No

Microsurgery

High Operative Risk

Progression

Yes

Complete Resection

No

Stereotactic Radiation

Yes

• Not Eloquent • Grades I - II • < 30 mm

Observation

SRS

• Eloquent • Grades III - V • > 30 mm

SRT

FIGURE 19-16. Flowchart depicting selection criteria for stereotactic radiation of intracranial meningiomas used in our clinic.

19.

meningioma

245

predict meningioma behavior has been used. The human telomerase catalytic subunit (hTERT) expression has shown signiﬁcant correlation with recurrent meningiomas [106, 107]. Another promising tool is the use of DNA microarray assay technology, which detects numerous genetic abnormalities, to differentiate meningioma subtypes [108]. Likewise, the surgery ﬁeld has been taking advantage of new technology in recent years. The advances in neuronavigational computer technologies have allowed surgeons to display three-dimensional reconstructions of the lesions in relation to the vessels, nerves, and eloquent tracts [109]. The diffusiontensor imaging tractography has also been used in treatment planning for radiosurgery [110]. There have been signiﬁcant advances in minimally invasive skull base surgery. Standard approaches to difﬁcult regions have been developed to allow safer and more complete resection [111–113]. As we better understand meningioma biology and genetics, and master the surgical techniques, we can improve patient outcomes and quality of life.

References

FIGURE 19-17. The beam’s-eye-view image (above) as seen in the treatment plan of a skull base meningioma (below) using the shaped beam technique with micro-multileaf collimator.

Using the beam’s-eye-view technique, the radiation arcs or beams avoid the eyes and eloquent structures whenever possible. Relative doses and sizes are designed to prevent maximum dose to critical structures within or in close relationship to the tumor.

Future Directions Decisions regarding conservative management versus aggressive surgery and adjuvant therapy should rely on precise parameters. The use of dynamic contrast-enhanced perfusion MRI with T2-weighted sequences measuring the endothelial permeability is useful in distinguishing atypical meningiomas from typical meningiomas [105]. Although the new WHO classiﬁcation recommends more objective criteria [8], there is still lack of correlation between grade and behavior in some cases. Recently, a new marker to

1. Bondy M, Ligon BL. Epidemiology and etiology of intracranial meningiomas: a review. J Neurooncol 1996; 29:197–205. 2. Rohringer M, Sutherland GR, Louw DF, et al. Incidence and clinicopathological features of meningioma. J Neurosurg 1989; 71:665–672. 3. Jaaskelainen J, Haltia M, Servo A. Atypical and anaplastic meningiomas: radiology, surgery, radiotherapy, and outcome. Surg Neurol 1986; 25:233–242. 4. Mahmood A, Caccamo DV, Tomecek FJ, et al. Atypical and malignant meningiomas: a clinicopathological review. Neurosurgery 1993; 33:955–963. 5. Palma L, Celli P, Franco C, et al. Long-term prognosis for atypical and malignant meningiomas: a study of 71 surgical cases. J Neurosurg 1997; 86:793–800. 6. Perry A, Scheithauer BW, Stafford SL, et al. “Malignancy” in meningiomas: a clinicopathologic study of 116 patients, with grading implications. Cancer 1999; 85:2046–2056. 7. Drummond KJ, Zhu JJ, Black PM. Meningiomas: updating basic science, management, and outcome. Neurologist 2004; 10:113– 130. 8. Kleihues P, Louis DN, Scheithauer BW, et al. The WHO classiﬁcation of tumors of the nervous system. J Neuropathol Exp Neurol 2002; 61:215–225; discussion 226–219. 9. Hug EB, Devries A, Thornton AF, et al. Management of atypical and malignant meningiomas: role of high-dose, 3D-conformal radiation therapy. J Neurooncol 2000; 48:151–160. 10. 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. 11. Ikeda H, Yoshimoto T. Immunohistochemical study of anaplastic meningioma with special reference to the phenotypic change of intermediate ﬁlament protein. Ann Diagn Pathol 2003; 7:214–222. 12. al-Mefty O, Kadri PA, Pravdenkova S, et al. Malignant progression in meningioma: documentation of a series and analysis of cytogenetic ﬁndings. J Neurosurg 2004; 101:210–218. 13. Simpson D. The recurrence of intracranial meningiomas after surgical treatment. J Neurochem 1957; 20:22–39. 14. Adegbite AB, Khan MI, Paine KW, et al. The recurrence of intracranial meningiomas after surgical treatment. J Neurosurg 1983; 58:51–56.

246

c.a. mattozo and a.a.f. de salles

15. Marks LB, Spencer DP. The inﬂuence of volume on the tolerance of the brain to radiosurgery. J Neurosurg 1991; 75:177–180. 16. Mirimanoff RO, Dosoretz DE, Linggood RM, et al. Meningioma: analysis of recurrence and progression following neurosurgical resection. J Neurosurg 1985; 62:18–24. 17. Sekhar LN, Patel S, Cusimano M, et al. Surgical treatment of meningiomas involving the cavernous sinus: evolving ideas based on a ten year experience. Acta Neurochir Suppl 1996; 65:58–62. 18. Sen C, Hague K. Meningiomas involving the cavernous sinus: histological factors affecting the degree of resection. J Neurosurg 1997; 87:535–543. 19. Sindou MP, Alaywan M. Most intracranial meningiomas are not cleavable tumors: anatomic-surgical evidence and angiographic predictibility. Neurosurgery 1998; 42:476–480. 20. al-Mefty O, Ayoubi S, Smith RR. The petrosal approach: indications, technique, and results. Acta Neurochir Suppl (Wien) 1991; 53:166–170. 21. Bricolo AP, Turazzi S, Talacchi A, et al. Microsurgical removal of petroclival meningiomas: a report of 33 patients. Neurosurgery 1992; 31:813–828; discussion 828. 22. Couldwell WT, Fukushima T, Giannotta SL, et al. Petroclival meningiomas: surgical experience in 109 cases. J Neurosurg 1996; 84:20–28. 23. DeMonte F, Smith HK, al-Mefty O. Outcome of aggressive removal of cavernous sinus meningiomas. J Neurosurg 1994; 81:245–251. 24. O’Sullivan MG, van Loveren HR, Tew JM Jr. The surgical resectability of meningiomas of the cavernous sinus. Neurosurgery 1997; 40:238–244; discussion 245–237. 25. Sekhar LN, Swamy NK, Jaiswal V, et al. Surgical excision of meningiomas involving the clivus: preoperative and intraoperative features as predictors of postoperative functional deterioration. 1994; J Neurosurg 81:860–868. 26. De Jesus O, Sekhar LN, Parikh HK, et al. Long-term follow-up of patients with meningiomas involving the cavernous sinus: recurrence, progression, and quality of life. Neurosurgery 1996; 39:915– 919; discussion 919–920. 27. Little KM, Friedman AH, Sampson JH, et al. Surgical management of petroclival meningiomas: deﬁning resection goals based on risk of neurological morbidity and tumor recurrence rates in 137 patients. Neurosurgery 2005; 56:546–559; discussion 546–559. 28. Nanda A, Jawahar A, Sathyanarayana S. Microsurgery for potential radiosurgical skull base lesions: a retrospective analysis and comparison of results. Skull Base 2003; 13:131–138. 29. Soyuer S, Chang EL, Selek U, et al. Radiotherapy after surgery for benign cerebral meningioma. Radiother Oncol 2004; 71:85– 90. 30. Taylor BW Jr, Marcus RB Jr, Friedman WA, et al. The meningioma controversy: postoperative radiation therapy. Int J Radiat Oncol Biol Phys 1988; 15:299–304. 31. Barbaro NM, Gutin PH, Wilson CB, et al. Radiation therapy in the treatment of partially resected meningiomas. Neurosurgery 1987; 20:525–528. 32. Glaholm J, Bloom HJ, Crow JH. The role of radiotherapy in the management of intracranial meningiomas: the Royal Marsden Hospital experience with 186 patients. Int J Radiat Oncol Biol Phys 1990; 18:755–761. 33. 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. 34. Wilson CB. Cavernous sinus meningiomas. Surg Neurol 1996; 46:191–192. 35. Nutting C, Brada M, Brazil L, et al. Radiotherapy in the treatment of benign meningioma of the skull base. J Neurosurg 1999; 90:823–827.

36. Milosevic MF, Frost PJ, Laperriere NJ, et al. Radiotherapy for atypical or malignant intracranial meningioma. Int J Radiat Oncol Biol Phys 1996; 34:817–822. 37. Dziuk TW, Woo S, Butler EB, et al. Malignant meningioma: an indication for initial aggressive surgery and adjuvant radiotherapy. J Neurooncol 1998; 37:177–188. 38. Ware ML, Larson DA, Sneed PK, et al. Surgical resection and permanent brachytherapy for recurrent atypical and malignant meningioma. Neurosurgery 2004; 54:55–63; discussion 63–54. 39. al-Mefty O, Kersh JE, Routh A, et al. The long-term side effects of radiation therapy for benign brain tumors in adults. J Neurosurg 1990; 73:502–512. 40. Crossen JR, Garwood D, Glatstein E, et al. Neurobehavioral sequelae of cranial irradiation in adults: a review of radiationinduced encephalopathy. J Clin Oncol 1994; 12:627–642. 41. Torres RC, Frighetto L, De Salles AA, et al. Radiosurgery and stereotactic radiotherapy for intracranial meningiomas. Neurosurg Focus 2003; 14:e5. 42. Torres RC, De Salles AA, Frighetto L, et al. Long-term follow-up using linac radiosurgery and stereotactic radiotherapy as a minimally invasive treatment for intracranial meningiomas. Radiosurgery 2004; 5:115–123. 43. De Salles AA, Frighetto L, Grande CV, et al. Radiosurgery and stereotactic radiation therapy of skull base meningiomas: proposal of a grading system. Stereotact Funct Neurosurg 2001; 76:218–229. 44. Selch MT, Ahn E, Laskari A, et al. Stereotactic radiotherapy for treatment of cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys 2004; 59:101–111. 45. Kondziolka D, Lunsford LD. Radiosurgery of meningiomas. Neurosurg Clin N Am 1992; 3:219–230. 46. Hakim R, Alexander E 3rd, Loefﬂer JS, et al. Results of linear accelerator-based radiosurgery for intracranial meningiomas. Neurosurgery 1998; 42:446–453; discussion 453–444. 47. Shafron DH, Friedman WA, Buatti JM, et al. Linac radiosurgery for benign meningiomas. Int J Radiat Oncol Biol Phys 1999; 43:321–327. 48. Spiegelmann R, Nissim O, Menhel J, et al. Linear accelerator radiosurgery for meningiomas in and around the cavernous sinus. Neurosurgery 2002; 51:1373–1379; discussion 1379–1380. 49. Iwai Y, Yamanaka K, Ishiguro T. Gamma knife radiosurgery for the treatment of cavernous sinus meningiomas. Neurosurgery 52003; 2:517–524; discussion 523–514. 50. Kobayashi T, Kida Y, Mori Y. Long-term results of stereotactic gamma radiosurgery of meningiomas. Surg Neurol 2001; 55:325– 331. 51. Kondziolka D, Levy EI, Niranjan A, et al. Long-term outcomes after meningioma radiosurgery: physician and patient perspectives. J Neurosurg 1999; 91:44–50. 52. Kondziolka D, Nathoo N, Flickinger JC, et al. Long-term results after radiosurgery for benign intracranial tumors. Neurosurgery 2003; 53:815–821; discussion 821–812. 53. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry 1983; 46:797–803. 54. Stafford SL, Pollock BE, Foote RL, et al. Meningioma radiosurgery: tumor control, outcomes, and complications among 190 consecutive patients. Neurosurgery 2001; 49:1029–1037; discussion 1037–1028. 55. Vernimmen FJ, Harris JK, Wilson JA, et al. Stereotactic proton beam therapy of skull base meningiomas. Int J Radiat Oncol Biol Phys 2001; 49:99–105. 56. Nicolato A, Foroni R, Alessandrini F, et al. The role of Gamma Knife radiosurgery in the management of cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys 2002; 53:992–1000. 57. Subach BR, Lunsford LD, Kondziolka D, et al. Management of petroclival meningiomas by stereotactic radiosurgery. Neurosurgery 1998; 42:437–443; discussion 443–435.

19.

meningioma

58. Pendl G, Schrottner O, Eustacchio S, et al. Stereotactic radiosurgery of skull base meningiomas. Minim Invasive Neurosurg 1997; 40:87–90. 59. Liscak R, Simonova G, Vymazal J, et al. Gamma knife radiosurgery of meningiomas in the cavernous sinus region. Acta Neurochir (Wien) 1999; 141:473–480. 60. Nedzi LA, Kooy H, Alexander E 3rd, et al. Variables associated with the development of complications from radiosurgery of intracranial tumors. Int J Radiat Oncol Biol Phys 1991; 21:591–599. 61. Chang JH, Chang JW, Choi JY, et al. Complications after gamma knife radiosurgery for benign meningiomas. J Neurol Neurosurg Psychiatry 2003; 74:226–230. 62. Valentino V, Schinaia G, Raimondi AJ. The results of radiosurgical management of 72 middle fossa meningiomas. Acta Neurochir (Wien) 1993; 122:60–70. 63. DiBiase SJ, Kwok Y, Yovino S, et al. Factors predicting local tumor control after gamma knife stereotactic radiosurgery for benign intracranial meningiomas. Int J Radiat Oncol Biol Phys 2004; 60:1515–1519. 64. Pollock BE, Stafford SL, Utter A, et al. 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; 55:1000–1005. 65. Flickinger JC, Kondziolka D, Maitz AH, et al. Gamma knife radiosurgery of imaging-diagnosed intracranial meningioma. Int J Radiat Oncol Biol Phys 2003; 56:801–806. 66. Nakatomi H, Sasaki T, Kawamoto S, et al. Primary cavernous sinus malignant lymphoma treated by gamma knife radiosurgery: case report and review of the literature. Surg Neurol 1996; 46:272– 278; discussion 278–279. 67. Kim DG, Kim Ch H, Chung HT, et al. Gamma Knife surgery of superﬁcially located meningioma. J Neurosurg 2005; 102(Suppl): 255–258. 68. Kotapka MJ, Kalia KK, Martinez AJ, et al. Inﬁltration of the carotid artery by cavernous sinus meningioma. J Neurosurg 1994; 81:252–255. 69. Larson JJ, van Loveren HR, Balko MG, et al. Evidence of meningioma inﬁltration into cranial nerves: clinical implications for cavernous sinus meningiomas. J Neurosurg 1995; 83:596–599. 70. Zentner J, Meyer B, Vieweg U, et al. Petroclival meningiomas: is radical resection always the best option? J Neurol Neurosurg Psychiatry 1997; 62:341–345. 71. Liscak R, Kollova A, Vladyka V, et al. Gamma Knife radiosurgery of skull base meningiomas. Acta Neurochir Suppl 2004; 91:65– 74. 72. Chuang CC, Chang CN, Tsang NM, et al. Linear accelerator-based radiosurgery in the management of skull base meningiomas. J Neurooncol 2004; 66:241–249. 73. Roche PH, Pellet W, Fuentes S, et al. Gamma Knife radiosurgical management of petroclival meningiomas results and indications. Acta Neurochir (Wien) 2003; 145:883–888; discussion 888. 74. 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; 27:215–221. 75. 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. 76. Stafford SL, Pollock BE, Leavitt JA, et al. A study on the radiation tolerance of the optic nerves and chiasm after stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2003; 55:1177–1181. 77. Chen JC, Giannotta SL, Yu C, et al. Radiosurgical management of benign cavernous sinus tumors: dose proﬁles and acute complications. Neurosurgery 2001; 48:1022–1030; discussion 1030–1022. 78. De Salles AA, Bajada CL, Goetsch S, et al. Radiosurgery of cavernous sinus tumors. Acta Neurochir Suppl (Wien) 1993; 58: 101–103.

247

79. Lee JY, Niranjan A, McInerney J, et al. Stereotactic radiosurgery providing long-term tumor control of cavernous sinus meningiomas. J Neurosurg 2002; 97:65–72. 80. Tome WA, Mehta MP, Meeks SL, et al. Fractionated stereotactic radiotherapy: a short review. Technol Cancer Res Treat 2002; 1:153–172. 81. Gill SS, Thomas DG, Warrington AP, et al. Relocatable frame for stereotactic external beam radiotherapy. Int J Radiat Oncol Biol Phys 1991; 20:599–603. 82. Alheit H, Dornfeld S, Dawel M, et al. Patient position reproducibility in fractionated stereotactically guided conformal radiotherapy using the BrainLab mask system. Strahlenther Onkol 2001; 177:264–268. 83. Kooy HM, Dunbar SF, Tarbell NJ, et al. Adaptation and veriﬁcation of the relocatable Gill-Thomas-Cosman frame in stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 1994; 30:685–691. 84. Solberg T, Ford J, Medin P, et al. Reproducibility of frame positioning for fractionated stereotactic radiosurgery. J Radiosurg 1999; 2:57–64. 85. Lo SS, Cho KH, Hall WA, et al. Single dose versus fractionated stereotactic radiotherapy for meningiomas. Can J Neurol Sci 2002; 29:240–248. 86. Milker-Zabel S, Zabel A, Schulz-Ertner D, et al. Fractionated stereotactic radiotherapy in patients with benign or atypical intracranial meningioma: long-term experience and prognostic factors. Int J Radiat Oncol Biol Phys 2005; 61:809–816. 87. Debus J, Wuendrich M, Pirzkall A, et al. High efﬁcacy of fractionated stereotactic radiotherapy of large base-of-skull meningiomas: long-term results. J Clin Oncol 2001; 19:3547–3553. 88. Baumert BG, Norton IA, Davis JB. Intensity-modulated stereotactic radiotherapy vs. stereotactic conformal radiotherapy for the treatment of meningioma located predominantly in the skull base. Int J Radiat Oncol Biol Phys 2003; 57:580–592. 89. Ernst-Stecken A, Lambrecht U, Ganslandt O, et al. Radiosurgery of small skull-base lesions no advantage for intensity-modulated stereotactic radiosurgery versus conformal arc technique. Strahlenther Onkol 2005; 181:336–344. 90. Gabibov GA, Blinkov SM, Tcherekayev VA. The management of optic nerve meningiomas and gliomas. J Neurosurg 1988; 68:889–893. 91. Kupersmith MJ, Warren FA, Newall J, et al. Irradiation of meningiomas of the intracranial anterior visual pathway. Ann Neurol 1987; 21:131–137. 92. Wara WM, Sheline GE, Newman H, et al. Radiation therapy of meningiomas. Am J Roentgenol Radium Ther Nucl Med 1975; 123:453–458. 93. Andrews DW, Faroozan R, Yang BP, et al. Fractionated stereotactic radiotherapy for the treatment of optic nerve sheath meningiomas: preliminary observations of 33 optic nerves in 30 patients with historical comparison to observation with or without prior surgery. Neurosurgery 2002; 51:890–902; discussion 903– 894. 94. Baumert BG, Villa S, Studer G, et al. Early improvements in vision after fractionated stereotactic radiotherapy for primary optic nerve sheath meningioma. Radiother Oncol 2004; 72:169– 174. 95. Ojemann SG, Sneed PK, Larson DA, et al. Radiosurgery for malignant meningioma: results in 22 patients. J Neurosurg 2000; 93(Suppl 3):62–67. 96. Harris AE, Lee JY, Omalu B, et al. The effect of radiosurgery during management of aggressive meningiomas. Surg Neurol 2003; 60:298–305; discussion 305. 97. Huffmann BC, Reinacher PC, Gilsbach JM. Gamma Knife surgery for atypical meningiomas. J Neurosurg 2005; 102(Suppl):283– 286.

248

c.a. mattozo and a.a.f. de salles

98. Abdel-Aziz KM, Froelich SC, Dagnew E, et al. Large sphenoid wing meningiomas involving the cavernous sinus: conservative surgical strategies for better functional outcomes. Neurosurgery 2004; 54:1375–1383; discussion 1383–1374. 99. Maruyama K, Shin M, Kurita H, et al. Proposed treatment strategy for cavernous sinus meningiomas: a prospective study. Neurosurgery 2004; 55:1068–1075. 100. Frighetto L, De Salles AA, Behnke E, et al. Image-guided frameless stereotactic biopsy sampling of parasellar lesions. Technical note. J Neurosurg 2003; 98:920–925. 101. Duma CM, Lunsford LD, Kondziolka D, et al. Stereotactic radiosurgery of cavernous sinus meningiomas as an addition or alternative to microsurgery. Neurosurgery 1993; 32:699–704; discussion 704–695. 102. Suzuki M, Mizoi K, Yoshimoto T. Should meningiomas involving the cavernous sinus be totally resected? Surg Neurol 1995; 44:3–10; discussion 10–13. 103. Flickinger JC. An integrated logistic formula for prediction of complications from radiosurgery. Int J Radiat Oncol Biol Phys 1989; 17:879–885. 104. Kjellberg RN, Davis KR, Lyons S, et al. Bragg peak proton beam therapy for arteriovenous malformation of the brain. Clin Neurosurg 1983; 31:248–290. 105. Yang S, Law M, Zagzag D, et al. Dynamic contrast-enhanced perfusion MR imaging measurements of endothelial permeability: differentiation between atypical and typical meningiomas. AJNR Am J Neuroradiol 2003; 24:1554–1559.

106. Kalala JP, Maes L, Vandenbroecke C, et al. The hTERT protein as a marker for malignancy in meningiomas. Oncol Rep 2005; 13:273–277. 107. Maes L, Lippens E, Kalala JP, et al. The hTERT-protein and Ki67 labelling index in recurrent and non-recurrent meningiomas. Cell Prolif 2005; 38:3–12. 108. Wrobel G, Roerig P, Kokocinski F, et al. Microarray-based gene expression proﬁling of benign, atypical and anaplastic meningiomas identiﬁes novel genes associated with meningioma progression. Int J Cancer 2005; 114:249–256. 109. Rohde V, Spangenberg P, Mayfrank L, et al. Advanced neuronavigation in skull base tumors and vascular lesions. Minim Invasive Neurosurg 2005; 48:13–18. 110. Kamada K, Todo T, Masutani Y, et al. Combined use of tractography-integrated functional neuronavigation and direct ﬁber stimulation. J Neurosurg 2005; 102:664–672. 111. Cook SW, Smith Z, Kelly DF. Endonasal transsphenoidal removal of tuberculum sellae meningiomas: technical note. Neurosurgery 2004; 55:239–244; discussion 244–236. 112. Couldwell WT, Weiss MH, Rabb C, et al. Variations on the standard transsphenoidal approach to the sellar region, with emphasis on the extended approaches and parasellar approaches: surgical experience in 105 cases. Neurosurgery 2004; 55:539–547; discussion 547–550. 113. Jho HD, Alﬁeri A. Endoscopic glabellar approach to the anterior skull base: a technical note. Minim Invasive Neurosurg 2002; 45:185–188.

2 0

Meningioma: Surgery Perspective Lawrence S. Chin, Pulak Ray, and John Caridi

Introduction The ﬁrst successful documented resection of a meningioma— a benign, slow-growing tumor—was performed by Zanobi Pecchiolo (1801–1866) at Siena University. In a vast surgical series published in 1847, 1524 cases were described, one of which was a large meningioma removed from the right sinciput through a triangular ﬂap. Then on December 15, 1887, William W. Keen (1837–1932) was the ﬁrst neurosurgeon to successfully resect a meningioma in the United States. The patient proceeded to survive for 30 years without any clinical signs of recurrence [1]. Because of much debate over the etiology and pathogenesis of this tumor, it was not until 1922 that Harvey Cushing, in his text entitled Meningiomas, Their Classiﬁcation, Regional Behavior, Life History, and Surgical End Results, characterized the tumor as a meningioma, thus leaving room for the histogenetic type to be clariﬁed [2, 3]. It is now known that meningiomas arise from transformed arachnoid cap cells that form slow-growing tumors that are attached to the overlying dura [4]. Whereas most are well circumscribed, approximately 10% spread diffusely following and sometimes invading the contours of the underlying bone. These are termed meningioma en plaque and are much more difﬁcult to resect. Histologically, the World Health Organization (WHO) has separated meningiomas into three grades based on microscopic appearance [5]. WHO grade I includes several variants, the most common being the transitional, ﬁbroblastic, and meningothelial. Grade I tumors generally follow a benign course and make up nearly 90% of meningiomas. Grade II tumors, which account for 5% to 7% of meningiomas, are associated with greater cellular atypia and include the clear cell and chordoid variants. These tumors have a recurrence rate 1.5 to 5 times greater than that of grade I varieties. Grade III meningiomas (1% to 3% of total meningiomas) include the anaplastic, papillary, and rhabdoid variants. These tumors have a high rate of invasion into surrounding tissues (including brain), which leads to a higher incidence of recurrence, and may metastasize. Meningiomas comprise more than 30% of primary intracranial tumors, and the incidence of meningiomas in the population is 4.5 per 100,000 person-years with females outnumbering males by 2:1 [6]. The incidence has also been shown to increase with age. Autopsy studies show that approximately 2.5% of people have an incidental meningioma [7]. Anecdotal evidence has suggested several correlations with meningiomas. Cushing

believed that head trauma was afﬁliated with tumorigenesis, but no deﬁnitive study has conﬁrmed this despite multiple attempts to ﬁnd a correlation [3, 8, 9]. Although several studies have attempted to link viral infection to the development of meningiomas, none have proved the causative effect [10]. The link between breast cancer and meningioma is also tenuous as only one paper has documented an association [11]. There are two factors that do seem to lead to meningioma formation: radiation and chromosome 22 mutation. The incidence of meningioma has been found to be four times higher in patients who had tinea capitis treated with local radiation compared with control patients [12]. Another deﬁnitive correlation with meningiomas is deletions in the long arm of chromosome 22. Numerous patients with meningiomas have been found to have monosomy 22, and the gene thought responsible for meningioma formation has been localized to the same region as that of neuroﬁbromatosis type 2 (NF2) [13]. This may explain the high correlation between NF2 and the development of meningiomas. The signs and symptoms associated with a meningioma depend primarily on its site of origin. The most common presentation is a new-onset seizure, occurring approximately 30% of the time [14]. These seizures are most often generalized but can be partial or complex partial. Headache, mental status changes, hemiplegia, isolated cranial nerve deﬁcits, and signs of increased intracranial pressure may also be seen in meningioma patients. The radiographic diagnosis of meningioma is made with computed tomography (CT) or magnetic resonance imaging (MRI). Unenhanced CT shows meningiomas to be slightly hyperdense compared with normal brain, often with surrounding edema and variable mass effect; adding intravenous contrast results in homogeneous enhancement of the mass. MRI is the preferred method of diagnosis because of its superior soft tissue resolution. The tumor appears isointense to brain on an unenhanced T1-weighted image but hyperintense with surrounding white matter edema on a T2-weighted image. The addition of gadolinium contrast reveals a uniformly enhancing mass with a broad-based connection to the underlying dura. A dural tail is variably seen and represents growth of meningioma along the dura away from the main tumor. The signiﬁcance of this ﬁnding is that if left untreated by either radiosurgery or surgery as both are focal treatment modalities, the chance for recurrence increases [15]. These tumors are nearly always extraaxial,

249

250

l.s. chin et al.

showing displacement of the cortex. Reactive sclerosis (hyperostosis) of the overlying bone can also be seen in some cases. Conventional catheter angiography, although increasingly supplanted by magnetic resonance (MR) angiography and CT angiography, is often critical in the surgical planning because it can determine the blood supply to the tumor, degree of vascularity, patency of dural venous sinuses, and suitability for endovascular embolization. Meningiomas are primarily fed by branches from the external carotid artery, commonly the middle meningeal or occipital artery, although large tumors usually parasitize a signiﬁcant blood supply from small pial arteries.

Surgical Decision-Making and Techniques Numerous factors must be addressed when evaluating a patient for surgery: age, location, size, symptoms, neurologic deﬁcit, tumor growth rate, proximity to critical structures, blood supply, venous drainage and sinus patency, previous treatments, and, most importantly, patient expectations. A multidisciplinary approach with input from other disciplines such as otolaryngology, radiation oncology, and radiology are important. The distinguishing advantage of surgery over radiosurgery is the ability to achieve a gross total resection. The incidence of tumor shrinkage after radiosurgery varies between 20% and 40%, and the actual amount of decrease is fairly minor. Thus, overriding any other consideration is the fact that some patients will insist on having a tumor, even a benign tumor that is fairly small, totally removed from their brain. These patients should be offered an opportunity for surgical resection provided that the risks are acceptable. An optimal resection requires complete removal of the meningioma with all of its dural and bony attachments. In a classic paper by Simpson from 1957, he describes the recurrence rates for meningiomas after surgical resection [16]. A ﬁve-tier grading system was devised to describe the extent of tumor removal, with grade I being the most complete and grade V being a simple decompression. Simpson noted that the recurrence rate after grade I resection was 9%, grade II was 19%, and grade III was 29%. Later, a grade 0 was added to the Simpson scale [17, 18]. This added a 2-cm margin of normal-appearing dura to a Simpson I resection. The recurrence rate of a patient undergoing a grade 0 resection of a benign meningioma is essentially zero. Clearly, the goal of surgery is to perform a grade 0 or a grade 1 resection; unfortunately, this may not be possible despite surgical expertise because the most important variable that inﬂuences extent of resection is tumor location.

Case Study 20-1 A 40-year-old man presented with seizures and mild leftsided weakness. His MRI scan showed a right convexity meningioma (Fig. 20-1A, B). He underwent a gross total resection (Fig. 20-1C), but pathology indicated an atypical meningioma (grade 2). He developed a recurrence 9 months later and underwent Gamma Knife radiosurgery to a dose of 16 Gy at 50% (Fig. 20-1D). In the ﬁrst 6 months after radiosurgery, he developed transient radiation-induced swelling associated with focal seizures and left arm weakness (Fig. 20-1E, F). Approximately 1 year after treatment, the tumor has regressed, and the patients is neurologically normal (Fig. 20-1G).

rupted by incising the dura circumferentially as the ﬁrst step in tumor removal. Pial feeders are too small to be embolized and are divided during the course of tumor dissection off the brain. Large tumors may be adherent to the pia causing unavoidable cortical injury in the process of a complete removal. This superﬁcial invasion does not constitute a sign of malignancy, but neurologic deﬁcit is possible, and patients should be counseled appropriately. Surgery should be considered anytime a patient wants a cure of their convexity meningioma regardless of size and also recommended strongly in patients with a tumor larger than 3 cm. Radiosurgery is reserved for asymptomatic, small meningiomas and the rare cases of recurrence after resection (Case Study 20-1).

Parasagittal Region Parasagittal meningiomas account for 20% to 30% of meningiomas and describe a location without intervening brain tissue or dura between the tumor and the superior sagittal sinus (Case Study 20-2). Conventional angiography or MR venography is helpful in determining if the sinus is patent. When there is invasion of a venous sinus, a complete resection requires removal of the inﬁltrated sinus as well. This is ordinarily safe only with the anterior one-third segment of the sagittal sinus. The scalp and bone ﬂap will need to cross the midline if the sagittal sinus is to be resected. Patients with middle and posterior third sagittal sinus involvement and a large tumor can beneﬁt from a combined approach where the bulk of the tumor is removed, which relieves mass effect on the brain while leaving the residual tumor in the sinus to be treated by radiosurgery.

Convexity Region Convexity meningiomas are found on the dura over the hemispheres without extension into the dural sinuses or invasion of the skull base. They represent approximately 15% of meningiomas and are the least likely to recur because they are nearly always amenable to a grade 0 or 1 resection. Intraoperative stereotactic guidance is helpful in localizing the projection of the tumor on the scalp. A generous scalp and craniotomy ﬂap can then be tailored to allow for a 2-cm margin around the tumor. Preoperative embolization is less helpful than in other areas as the blood supply from dural feeders is easily inter-

Case Study 20-2 This 37-year-old woman presented with seizures and altered mental status. A parasagittal meningioma occluding the sagittal sinus was seen in the anterior third of the sinus (Fig. 20-2A, B). A bifrontal craniotomy was performed, and the tumor was totally resected along with the sagittal sinus (Fig. 20-2C, D). One year later, she has no recurrence of tumor.

20.

meningioma: surgery perspective

FIGURE 20-1. MRI scans of a convexity atypical meningioma. Preoperative (A) axial and (B) coronal scans showing a convexity meningioma with mass effect. (C) Postoperative axial scan showing complete resection of tumor. (D) Gamma Knife treatment MRI shows recur-

251

rence of convexity tumor. (E, F) Three months after treatment, the tumor is unchanged in size but there is an increase in surrounding edema as seen on ﬂuid-attenuated inversion recovery (FLAIR) images. (G) One year later, the tumor has regressed.

FIGURE 20-2. An anterior parasagittal meningioma involving the sagittal sinus. (A) Coronal and (B) sagittal MRI scans show a meningioma arising from the sagittal sinus and anterior falx. (C, D) Postoperative MRI scans indicate complete tumor resection.

252

l.s. chin et al.

Case Study 20-3

Case Study 20-4

A 63-year-old woman was admitted with severe headaches and confusion. MRI showed a large, right sphenoid wing meningioma (Fig. 20-3A). After preoperative embolization, a right pterional craniotomy was performed with total excision of the tumor (Fig. 20-3B). One year follow-up shows no recurrence of tumor.

This 47-year-old woman presented with headaches and was found to have a right medial sphenoid wing meningioma (Fig. 20-4A). A right pterional craniotomy was performed to attempt a gross total resection. The entire tumor was removed except for the portion in the cavernous sinus. This ensured adequate separation between the tumor and the optic nerves thus allowing the residual tumor to be treated by Gamma Knife radiosurgery (Fig. 20-4B). Follow-up 5 years later shows no change in the small residual cavernous sinus mass (Fig. 20-4C).

Preoperative embolization may be useful in reducing blood loss during the debulking of large tumors. To prevent the tumor from recanalizing its dural feeders, embolization and surgery should be coordinated so that no more than 24 hours elapse between the procedures.

Sphenoid Wing and Anterior Skull Base Sphenoid wing meningiomas are divided into lateral and medial depending on the site of origin. Lateral sphenoid wing meningiomas may invade the sphenoid bone, but complete resection is still possible provided the tumor is not big and the dura of the temporal or frontal ﬂoor not extensively invaded (Case Study 20-3). Complete removal of a meningioma originating from the medial sphenoid wing is a difﬁcult task for the surgeon because the tumor likely involves the dura around the optic nerve, superior orbital ﬁssure, and cavernous sinus (Case Study 20-4). Meningiomas from the anterior cranial ﬂoor (olfactory groove, planum sphenoidale, tuberculum sella, and diaphragma sella) offer similar challenges because of the nearby optic nerves, chiasm, and pituitary stalk (Case Study 20-5). In these tumors, a grade 2 resection (coagulation of the dural attachment) is usually the best possible outcome. An overly aggressive resection may injure cranial nerves II, III, IV, V, and VI and/or the carotid artery and pituitary stalk. Identifying and maintaining the arachnoid planes around the meningioma and observing the lateral border of the cavernous sinus can help protect neurovascular structures. Early interruption of feeding arteries (e.g., anterior ethmoidal branches) at the base of the meningioma diminishes tumor bleeding during the debulking phase of the operation. Preoperative embolization is rarely

helpful in these cases because of the risk of blindness from embolization of collateral blood supply to the optic nerve from external carotid branches. In general, the technique for meningioma removal involves dissection around the periphery of the tumor with coagulation of feeding arteries while avoiding signiﬁcant brain retraction. Once this can no longer be accomplished safely because of tumor bulk, the meningioma is incised and internally debulked using coagulation and ultrasonic aspirators. The capsule of the tumor can then be infolded and dissected off the remaining brain and neurovascular structures. The attachment of the tumor is coagulated and, if possible, cut off the skull base. The top priority when removing a meningioma from the sellar/parasellar region is to decompress the optic nerves and chiasm. Not only does this address the most serious neurologic deﬁcit caused by the tumor, but also any residual tumor can be treated by radiosurgery. A separation of at least 2 mm is needed to permit an adequate dose of radiation to the tumor edge (12 Gy or higher) while keeping the optic nerve dose to safe levels (8 Gy). Because the cranial nerves in the cavernous sinus are relatively insensitive to radiation effects, there is little need to aggressively resect tumor in this location. Visible residual tumor should be treated aggressively with early radiosurgery because having a maximum amount of separation between the tumor edge and the optic nerve or brain stem is critical in determining the dose that can be safely delivered. A mistake is often made in waiting to see if the residual tumor will start to grow. Invariably it does, and when ﬁnally recognized, the tumor

FIGURE 20-3. A right lateral sphenoid wing meningioma. (A) Preoperative axial MRI scan shows a large tumor compressing the right frontal lobe. (B) Postoperative MRI scan shows a gross total resection.

20.

meningioma: surgery perspective

FIGURE 20-4. A large right medial sphenoid wing meningioma. (A) Preoperative axial MRI scan shows a large tumor arising from the medial sphenoid wing and cavernous sinus. (B) Only a small amount

Case Study 20-5 A 28-year-old woman presented with a bitemporal hemianopsia and headaches. Preoperative MRI showed a diaphragma sella meningioma (Fig. 20-5A). A right pterional craniotomy with gross total resection of tumor was performed (Fig. 20-5B). The optic nerves were well decompressed, and her vision returned to normal. Three years later, she has no recurrence of tumor, and her vision remains normal.

253

of tumor remains in the cavernous sinus, which is treated by Gamma Knife radiosurgery. (C) Five years later, there is no change in the residual tumor.

may be too close to the optic nerve to be treated by radiosurgery. At this point, the only options are fractionated radiation therapy or repeat surgery.

Petroclival Region This broad category includes meningiomas found in the posterior cavernous sinus, tentorium, cerebellopontine angle (CPA), and foramen magnum. Nearly all of these types of tumors, particularly when large, will require a skull base approach. Any combination of transpetrosal, combined subtemporal and

FIGURE 20-5. A tuberculum sella meningioma causing a bitemporal visual ﬁeld cut. (A) Sagittal MRI scan shows a meningioma arising from the diaphragma sella with elevation of the optic chiasm. (B) Postoperative MRI scan shows that the entire tumor has been removed.

254

l.s. chin et al.

Case Study 20-6 This 39-year-old woman presented with severe headaches and a large tumor based on the left petrous portion of the temporal bone and tentorium (Fig. 20-6A–C). The tumor appeared to involve the left mastoid air cells and had a dural tail along the tentorium. After preoperative embolization, a radical excision was performed using a combined subtemporal/transmastoid approach (Fig. 20-6D, E). The bulk of the tumor was removed en bloc along with a portion of the tentorium. The pathology was atypical meningioma (grade 2). Although no obvious residual tumor was seen, fractionated radiation therapy was administered because likely some tumor remained, and the atypical pathology greatly increased the risk of recurrence.

radiographic gross total resections are possible, it is rare to accomplish better than a grade II or III resection (Case Study 20-6). Thus, frequent MRI follow-up (yearly intervals) even after a gross total resection is essential to catch an early recurrence. When residual tumor is seen on postoperative imaging, it should be treated with radiosurgery. When residual tumor is known to be left behind but cannot be seen radiographically or if the histology is grade II or III, fractionated radiation therapy should be considered.

Conclusion

transmastoid, retromastoid, and extreme lateral transcondylar approaches may be needed and should be part of the skull base neurosurgeon’s arsenal. In most situations, a skull base specialist from ENT will be needed for their expertise in the bony anatomy of the temporal bone. Preoperative embolization is frequently helpful in larger tumors in this region. Although

The ideal approach to a patient with a meningioma is to consider both surgical and radiosurgical options. A multidisciplinary team often accomplishes this goal best, although good balanced opinions can be given in any clinical setting. Surgery and radiosurgery should be thought of as complementary procedures rather than competing treatment modalities. When relief of mass effect is needed because of tumor size or patient symptoms, then craniotomy is the only option. Radiosurgery can be used in smaller tumors and can treat residual tumor, thus freeing the surgeon from having to resect tumors from areas of the brain and skull that may cause signiﬁcant morbidity.

FIGURE 20-6. A very large petrous/tentorial meningioma. (A) Axial and (B) coronal MRI scans show a meningioma inﬁltrating the mastoid and compressing the left cerebellum and temporal lobe. (C) External

carotid angiogram shows feeders from the middle meningeal artery and occipital artery. (D, E) Postoperative MRI scans show a gross total resection.

20.

meningioma: surgery perspective

References 1. Keen W. Three successful cases of cerebral surgery. J Med Sci 1888; 96:329–357, 452–465. 2. Al-Mefty O. Meningiomas. New York: Raven Press, 1991. 3. Cushing H, Eisenhardt L. Meningiomas: Their Classiﬁcation, Regional Behavior, Life History, and Surgical End Results. Springﬁeld, IL: Charles C. Thomas, 1938. 4. Kallio M, Sankila R, Hakulinen T, Jaaskelainen J. Factors affecting operative and excess long-term mortality in 935 patients with intracranial meningioma. Neurosurgery 1992; 31:2–12. 5. Kleihues P, Louis DN, Scheithauer BW, et al. The WHO classiﬁcation of tumors of the nervous system. J Neuropathol Exp Neurol 2002; 61:215–225; discussion 226–219. 6. CBTRUS (2005). Statistical Report: Primary Brain Tumors in the United States, 1998–2002. Central Brain Tumor Registry of the United States. Hinsdale, IL, USA. 7. Nakasu S, Hirano A, Shimura T, Llena JF. Incidental meningiomas in autopsy study. Surg Neurol 1987; 27:319–322. 8. Phillips LE, Koepsell TD, van Belle G, al. History of head trauma and risk of intracranial meningioma: population-based casecontrol study. Neurology 2002; 58:1849–1852. 9. Preston-Martin S, Pogoda JM, Schlehofer B, et al. An international case-control study of adult glioma and meningioma: the role of head trauma. Int J Epidemiol 1998; 27:579–586. 10. Krieg P, Scherer G. Cloning of SV40 genomes from human brain tumors. Virology 1984; 138:336–340.

255

11. Custer BS, Koepsell TD, Mueller BA. The association between breast carcinoma and meningioma in women. Cancer 2002; 94: 1626–1635. 12. Modan B, Baidatz D, Mart H, et al. Radiation-induced head and neck tumours. Lancet 1974; 1:277–279. 13. Joachim T, Ram Z, Rappaport ZH, et al. Comparative analysis of the NF2, TP53, PTEN, KRAS, NRAS and HRAS genes in sporadic and radiation-induced human meningiomas. Int J Cancer 2001; 94:218–221. 14. Lieu AS, Howng SL. Intracranial meningiomas and epilepsy: incidence, prognosis and inﬂuencing factors. Epilepsy Res 2000; 38:45–52. 15. DiBiase SJ, Kwok Y, Yovino S, et al. Factors predicting local tumor control after gamma knife stereotactic radiosurgery for benign intracranial meningiomas. Int J Radiat Oncol Biol Phys 2004; 60:1515–1519. 16. Simpson D. The recurrence of intracranial meningiomas after surgical treatment. J Neurol Neurosurg Psychiatry 1957; 20:22– 39. 17. Borovich B, Doron Y. Recurrence of intracranial meningiomas: the role played by regional multicentricity. J Neurosurg 1986; 64:58–63. 18. Borovich B, Doron Y, Braun J, et al. Recurrence of intracranial meningiomas: the role played by regional multicentricity. Part 2: clinical and radiological aspects. J Neurosurg 1986; 65:168– 171.

2 1

Intracranial Meningioma: Fractionated Radiation Therapy Perspective Leland Rogers, Dennis Shrieve, and Arie Perry

Introduction The function of this chapter will be a review of the role of fractionated radiation therapy in the management of intracranial meningiomas. An introductory treatise about meningiomas is found in another chapter of this book. We will reference background information regarding epidemiology and incidence and highlight features that complement the discussion. Topics contrasting fractionated irradiation with radiosurgery will be emphasized.

Epidemiology Except in the setting of neuroﬁbromatosis type 2 (NF2) [1], meningiomas are rare in the pediatric population. The peak occurrence is during the sixth and seventh decades of life, but the age range is broad, with a 5% or greater incidence in all age brackets from the second to ninth decades [2, 3]. Age may have prognostic relevance. More aggressive clinical and histologic features have been noted in children [4] and in the younger adult population. Jung and associates found that the growth rate after subtotal resection was greater in patients younger than 50 years of age (7.51 cm3 per year) than in older patients (0.83 cm3 per year) [5]. A female preponderance is evident, with a female:male ratio of about 2:1 [6–9]. This implicates gender-speciﬁc hormones. Indeed, progesterone receptors have been identiﬁed in more than 70%, as have estrogen receptors in up to 30% [10], although they are reported in meningiomas from both sexes. A heightened risk has been described with obesity, as has an increase in tumor-related symptoms with pregnancy or the luteal phase of menses [10, 11]. Additionally, an increased relative risk of breast cancer has been described in meningioma patients, as has the converse [10]. These endocrine parallels are alluring, but to date anti-progestational and other hormonal therapies have been largely disappointing [12, 13].

Etiology The etiology of meningiomas remains largely unknown. Cushing and others have invoked trauma as potentially causative, but this has not been substantiated in large epidemiologic studies [14, 15]. Meningiomas occur with greater frequency in patients with certain rare genetic conditions, such as NF2 [1, 4], but this explains only a small minority of cases. Ironically, despite its well-established role as an effective therapy for meningiomas, prior exposure to ionizing radiation is also a well-established risk factor for developing meningiomas [11, 16–21]. Again though, postradiation meningiomas account for only a very small fraction of cases. The vast majority of meningiomas arise in the absence of any known predisposing factors.

Radiation-Induced Meningiomas The awareness of meningioma induction after irradiation derives from experience with atomic bomb fallout and with cranial irradiation for tumors or scalp irradiation for maladies such as tinea capitis [11, 16–21]. Strojan and colleagues projected the risk of developing a meningioma after radiation therapy to be 0.53% at 5 years and 8.18% at 25 years [20]. Indeed, radiation-induced meningiomas remain the most common form of radiation-induced neoplasm in the literature [21]. It is anticipated that modern conformal and stereotactic approaches will, by virtue of treating less volume, result in fewer occurrences, but this remains to be conﬁrmed.

Number of Patients Meningiomas are common, comprising approximately 15% to 30% of all primary intracranial tumors and ranking them as the second most common primary intracranial tumor, behind gliomas [6, 22–24]. The reported incidence varies from <1 to >6 per 100,000 depending upon the population and the method of identiﬁcation [22–27], but in composite, an incidence of 2.6 per 100,000 has been calculated [11, 28]. With this inference, there

257

258

l. rogers et al.

are approximately 8000 newly diagnosed meningiomas per year in the United States [11].

Radiologically Diagnosed Meningiomas The estimate of 8000 meningiomas per year is based largely on surgical series and is underrepresentative. Many meningiomas are identiﬁed solely by neuroimaging and are not treated surgically. Investigators from the University of Pittsburgh reported that 35% of their patients had been diagnosed and treated on the basis of neuroimaging alone. In their estimation, an image-based diagnosis was reliable given a 2.3% 10-year rate of ultimately identifying a diagnosis other than meningioma [29]. The proportion of patients diagnosed radiographically was 62% in a recent University of Maryland report [30]. Accepting the smaller University of Pittsburgh ﬁgure, image-diagnosed meningiomas may account for an additional 2800 patients, bringing the total to nearly 11,000 annually. Even this sum is likely an underestimate. Meningiomas have been identiﬁed in as many as 2.3% of autopsies [31], many of which were subclinical.

Histopathology Cell of Origin As originally suggested more than a century ago by Cleland [32] and Schmidt [33], it is now widely accepted that meningiomas arise from the epithelioid cells on the outer surface of arachnoid villi, referred to as arachnoid cap cells or more broadly as meningothelial cells. Arachnoid cap cell derivation lacks unequivocal proof, but several supporting features bolster the premise. The majority of meningiomas are cytologically similar to arachnoid cap cells, and arachnoid villi are most prominent at sites where meningiomas commonly occur. Furthermore, cap cell clusters become more prominent with age, corresponding with the age-related incidence of meningiomas [34].

Tumor Grade Benignity and slow growth are characteristics commonly ascribed to meningiomas. These preconceptions are often supported clinically but by no means uniformly. Some meningiomas, particularly high-grade and recurrent tumors, can be aggressive. The World Health Organization (WHO) updated its grading criteria in 2000 [35], instilling a greater empiricism and reproducibility. Using these criteria, strong associations between grade, relapse-free survival, and overall survival have been demonstrated in large series [36–39]. Outcomes according to histopathologic grade are depicted graphically in Figure 21-1.

Benign (WHO Grade I) Benign (WHO grade I) meningiomas comprise about 70% to 85% of intracranial primaries, a proportion that may be greater if image-diagnosed and incidental meningiomas could be included. With appropriate treatment, approximately 80% of WHO grade I meningiomas remain progression-free at 10 to years [36, 37, 40, 41].

FIGURE 21-1. Recurrence-free survival by grade. Kaplan-Meier estimated recurrence-free survival by WHO tumor grade from a compiled series of 643 patients [36, 37]. The number of patients with each grade is given, as is the percent of patients with each grade. (Courtesy of Christine Lohse, Biostatistics Division, Mayo Clinic.) The grading criteria employed in this study have been incorporated into the current WHO grading scheme.

Atypical (WHO Grade II) Atypical (WHO grade II) meningiomas account for 15% to 25% of patients graded according to WHO 2000 standards [36, 37]. These have greater proliferative capacity and a seven- to eightfold increased recurrence risk at 3 to 5 years [12]. Only about 35% of patients with WHO grade II meningiomas remain disease-free at 10 years, even with curative intent therapy [36, 37] (Fig. 21-1).

Anaplastic/Malignant (WHO Grade III) Only about 1% to 3% of intracranial meningiomas are anaplastic (WHO grade III). These are aggressive malignant tumors, portending a median overall survival of less than 2 years [36, 37]. There is little discrepancy in the recommendations for surgery and radiation therapy, typically entailing larger ﬁeld external beam irradiation. This review will emphasize benign (WHO grade I) meningiomas, although atypical (WHO grade II) meningiomas will also be referenced. We intend no speciﬁc assessment of grade III meningiomas, because the approach toward treatment, poor outcome notwithstanding, is less controversial.

Treatment and Outcome It is germane, as a prelude to any discourse about the treatment of intracranial meningiomas, to emphasize that these tumors can, and with sufﬁcient follow-up often do, recur. Even gross totally resected benign meningiomas may recur, with a fourth to a third recurring at 15 years [2, 7, 40]. Subtotally resected meningiomas have much higher progression rates of approximately 70% to 90% at 15 years [7, 40] and may even be associated with poorer survival. Condra and colleagues found that subtotal resection alone resulted in signiﬁcantly poorer causespeciﬁc survival than subtotal resection and radiation therapy (RT) [40]. Goldsmith and co-investigators determined that

21.

259

intracranial meningioma: fractionated radiation therapy perspective TABLE 21-1. Extent of resection (Simpson’s grade). Resection grade

1 2 3

4 5 FIGURE 21-2. Overall survival. Overall survival after STR and postoperative EBRT for 140 (117 benign and 23 malignant) meningioma patients compared with the age- and sex-matched control population. (Modiﬁed from: Goldsmith B, Wara W, Wilson C, et al. Postoperative irradiation for subtotally resected meningiomas. J Neurosurg 1994; 80:195–201. Used with permission.)

overall survival rates in subtotally resected and irradiated benign meningioma patients were inferior to age- and sexmatched controls (Fig. 21-2) [42]. The characteristically prolonged natural history of meningiomas suggests benignity, though the ultimate high recurrence rates yield considerable morbidity and therapeutic challenges.

Surgery Surgery is the mainstay in the diagnosis and treatment of meningiomas. The completeness of surgical removal has consistently been an important prognostic factor [2, 28, 40, 43]. Resection of the tumor, its involved dura, and any involved soft tissue and bone is accepted procedure [11], and high local control rates can be achieved. Nonetheless, as documented in a classic “surgery series” (8 of 114 patients received RT) by Adegbite and co-workers with prolonged follow-up (Fig. 21-3), even with

Recurrence (%)

Deﬁnition

GTR of tumor, dural attachments, and abnormal bone GTR of tumor, coagulation of dural attachments GTR of tumor without resection or coagulation of dural attachments or extradural extensions (e.g., invaded or hyperostotic bone) Partial resection of tumor Simple decompression (biopsy)

9 19 29

44 —

thorough resection, the likelihood of recurrence is by no means negligible [3].

Simpson “Grades” of Resection The extent of surgical resection was classically deﬁned by Donald Simpson [44]. He related the degree of removal of the tumor, its dural attachments, and any hyperostotic bone to the local recurrence risk. He followed 470 patients over a 26-year span and described 5 “grades of resection,” summarized in Table 21-1 alongside the respective rates of recurrence. Many series have corroborated the correlation between thorough resection and improved outcome, but with extended follow-up, recurrences often occur even with gross totally resected benign meningiomas (Table 21-2).

Grade Zero Resection Kinjo and colleagues characterized a yet more extensive resection as “grade zero.” They reported 37 convexity meningioma patients who underwent gross total resection (GTR) of the tumor, any hyperostotic bone, and all involved dura with a 2-cm dural margin [45] and observed no local recurrences, with more than half of patients followed beyond 5 years. However, apart from convexity primaries, resection to this extent is usually not feasible.

Location and Likelihood of Excision Table 21-3 charts the distribution of intracranial meningioma sites from a Mayo Clinic analysis of 581 patients. The likelihood of gross total resection varies considerably among primary sites [7, 8, 28, 43]. Table 21-4 reviews the likelihood of complete excision by site from a Massachusetts General Hospital surgical

TABLE 21-2. Local recurrence after gross total resection alone. Local recurrence rate (%) Author (institution)

FIGURE 21-3. Recurrence-free survival. Cumulative proportion of patients free of recurrence against the number of years since surgery. (From: Adegbite AB, Kahn MI, Paine KWE, et al. The recurrence of intracranial meningiomas after surgical treatment. J Neurosurg 1983; 58:51–56. Used with permission.)

n

5-year

10-year

15-year

Mirimanoff (MGH) [7] Stafford (Mayo Clinic) [2] Condra (U. of Florida) [40]

145 465 175

7 12 7

20 25 20

32 — 24

Total

785

7–12

20–25

24–32

n, number of patients; MGH, Massachusetts General Hospital; U of Florida, University of Florida.

260

l. rogers et al.

TABLE 21-3. Sites of meningioma in 581 patients. Site

Percent of patients (%)

Parasagittal falx Convexity Multiple sites Sphenoid wing Posterior fossa Parasellar area Anterior visual pathway Clivus Foramen magnum Intraventricular area

18.6 16.8 16.3 15.3 13.6 10.5 2.5 2.4 2.4 1.0

Note: Due to the requirements of rounding, adding the percentages amounts to 99.4%. The item “multiple sites” refers principally to meningiomas that occupy more than one area. Only 3% of patients had multiple lesions. Source: From Stafford S, Perry A, Suman V, et al. Primarily resected meningiomas: outcome and prognostic factors in 581 Mayo Clinic patients, 1978 through 1988. Mayo Clin Proc 1998; 73:936–942. Reprinted with permission of the Mayo Foundation.

series. Overall, at least a third of meningiomas reported in surgical series are not fully resectable [7, 43]. The actual percentage, considering inoperative patients, is almost certainly greater.

Gross Total Resection Alone GTR for benign meningiomas remains the preferred treatment and is generally considered deﬁnitive [2, 7, 40, 43]. Three large series with extended follow-up are listed in Table 21-2. These substantiate remarkably similar rates of local recurrence after GTR: 7% to 12% at 5 years, 20% to 25% at 10 years, and 24% to 32% at 15 years. It has been typical to designate GTR as inclusive of Simpson grades 1 and 2 [43], and even perhaps Simpson grade 3 [40]. A University of Florida series is depicted in Figure 21-4. Among 262 patients, 175 were deemed to have had total excision. With a median follow-up of 8.2 years, they found no signiﬁcant difference in local control or cause-speciﬁc survival between Simpson resection grades 1 through 3 [40]. Intuitively, Simpson TABLE 21-4. Likelihood of complete excision by tumor site. Meningioma location

n

Percent complete excision (%)

Convexity Orbit Spine Olfactory groove Parasagittal area/falx Parasellar region Posterior fossa Sphenoid ridge

47 5 18 22 38 28 31 36

96 80 78 77 76 57 32 28

225

64

Total n, number of patients.

Note: Relative frequency of complete excision in 225 patients. Extent of surgery was determined from the operative report. Source: Modiﬁed from Mirimanoff R, Dosoretz D, Linggood R, et al. Meningioma: analysis of recurrence and progression following neurosurgical resection. J Neurosurg 1985; 62:18–24. Used with permission.

FIGURE 21-4. Local control by resection extent. Local control according to Simpson grade of resection (1 through 3) in 175 meningioma patients deemed to have undergone a total excision. In addition, this study reported local control in 55 patients with Simpson grade 4 resections: 53%, 40%, and 30% at 5, 10, and 15 years, respectively. There were no Simpson grade 5 patients. (Modiﬁed from: Condra K, Buatti J, Mendenhall W, et al. Benign meningiomas: primary treatment selection affects survival. Int J Radiat Oncol Biol Phys 1997; 39:427–436. Used with permission from Elsevier.)

resection grades 1 through 3 would correspond with a lack of any residual disease detectable on postoperative imaging, and the utilization of imaging in conjunction with the surgeon’s appraisal is prevalent to differentiate subtotal from gross total resection.

Subtotal Resection Alone Progression rates after subtotal resection (STR) are substantially higher than after GTR, and STR is often insufﬁcient as a sole modality. Table 21-5 summarizes outcomes after STR alone, largely without radiation therapy, from four large, single institutions with up to 20 years of follow-up. Combining these ﬁndings, 37% to 47% of subtotally resected meningiomas progress locally at 5 years, 55% to 63% by 10 years, and at least 70% by 15 to 20 years [2, 7, 40, 46]. In spite of the substantial local progression risk after STR alone, observation remains commonplace. The Mayo Clinic series detailed outcome in 581 patients, 116 of whom (20%) had STR. Only 10 of the 116 (9%) STR patients received postoperative radiation therapy [2].

TABLE 21-5. Local progression after subtotal resection alone. Local progression rate (%) Author (institution)

n

Wara (UCSF) [46] Mirimanoff (MGH) [7] Condra (U of Florida) [40] Stafford (Mayo Clinic) [2] Total

5-year

10-year

15-year

20-year

58 80

47 37

62 55

— 91

74 —

55

47

60

70

—

116

39

61

—

—

37–47

55–63

70–91

74

n, number of patients; UCSF, University of California San Francisco; MGH, Massachusetts General Hospital; U of Florida, University of Florida.

21.

intracranial meningioma: fractionated radiation therapy perspective

Radiation Therapy RT remains the sole validated nonsurgical treatment for meningiomas. Several factors may be considered in the decision to use RT. Of key importance are the extent of surgical resection [2, 7, 40, 44], tumor grade [36, 37], and histologic subtype, which itself is linked with grade [35]. Other candidate prognostic factors include imaging ﬁndings [6, 47, 48], age [4, 5], and menopausal status [5]. The majority of reports evaluating RT, either as single-fraction radiosurgery or fractionated external beam irradiation, have attested to its inﬂuence in delaying or preventing recurrence. The 40 studies summarized in Tables 21-6 and 21-7 conﬁrm the value of RT after STR, as primary therapy, or as treatment for recurrent meningiomas. For the purpose of comparison, a brief review of radiosurgery for meningiomas is provided. A more thorough treatise is found in a separate chapter of this text. However, identifying expected outcomes and toxicity from radiosurgery will serve as a template for comparison with fractionated therapy.

Radiation Therapy: Stereotactic Radiosurgery In some measure, single-fraction stereotactic radiosurgery (SRS) avails itself of the physical beneﬁts of tight conformality

261

and a steep dose gradient, whereas fractionated external beam radiation therapy (EBRT) exploits the radiobiological advantages of lower doses per fraction, longer treatment course, and higher total dose. The ultimate purpose with either method is to enhance tumor control and limit side effects to healthy tissue near the tumor. SRS is a more recent development than EBRT but has been used with increasing frequency over the past two decades, and is an accepted form of treatment for meningiomas. Local control rates after SRS have been impressive, 75% to 100% at 5 to 10 years, both with linear accelerator–based [49, 50, 80] and Gamma Knife [43, 52, 55, 56, 81–83] systems. Outcomes from multiple radiosurgical series are summarized in Table 21-6. For many patients, radiosurgery has been the primary modality, as indicated in the “No Histology” column of Table 21-6, listing the percentage of SRS patients diagnosed radiographically. SRS is generally considered most appropriate for meningiomas less than 3 to 4 cm in maximum diameter, with distinct margins, and with sufﬁcient distance from critical normal tissues to allow for appropriate dose restrictions. Regarding tumor volume, DiBiase and co-authors reported 5-year disease-free survivals of 91.9% for patients with meningiomas ≤10 cm3 (equivalent diameter 2.7 cm) versus 68% for larger tumors [30]. Regarding dose, many series have used tumor margin doses in the 10- to 18-Gy range. Optimal dosing is the subject of

TABLE 21-6. Stereotactic radiosurgery progression-free survival. Author (year)

n

Follow-up (months)

No histology (%)

Dose (Gy)

Chang (1997) [49] Hakim (1998) [50] Chang (1998) [51] Liscak (1999) [52] Kondziolka (1999) [53] Morita (1999) [54] Roche (2000) [55] Stafford (2001) [56] Shin (2001) [57]

48 31 46 19

42

— 54 — 64 43 44 63 41 30

Nicolato (2002) [58] Lee (2002) [59]

55 127 24 53 99 88 80 168 15 22 111 159

48 35

50 52

18 15 17.7 12 16 16 14 16 10–12 14–18 15 13

Spiegelmann (2002) [60] Pollock (2003) [61] Roche (2003) [55] Iwai (2003) [62] Flickinger (2003) [29] Chuang (2004) [63]

42 62 32 42 219 43

36 64 56 49 29 75

— 46 75 48 100 48

14 17.7 13 11 14 16

DiBiase (2004) [30]

137

54

62

1578

19–75

30–100

Total

35 31

14 10–18

≥5-year PFS (%)

98 89 100 100 93 95 93 93 75 (5- & 10-yr PFS) 100 (5- & 10-yr PFS) 96 93 (97 if SRS sole treatment) 97.5 95 (7-yr PFS) 100 92 93 (5- & 10-yr PFS) 90 (7-yr PFS) (100 if SRS sole treatment) 86.2 (91.9 if <10 cm3) 75–100

n, number of patients; yr, year. Note: Compiled stereotactic radiosurgery with reported 5-year progression-free survival rates. Actuarial intervals other than 5 years are given in parentheses. The “No Histology” column refers to the percentage of patients diagnosed with meningioma on the basis of neuroimaging. Patients in the above reports typically, but not exclusively, had either known or presumed low-grade meningiomas. The follow-up and dose columns list the mean or median ﬁgures.

262

l. rogers et al.

TABLE 21-7. External beam radiation therapy progression-free survival. ≥5-year PFS Author (year)

n

Adegbite (1983) [3] Mirimanoff (1985) [7]

114 225

Barbaro (1987) [64] Taylor (1988) [65]

135 132

Glaholm (1990) [66] Miralbell (1992) [67] Mahmood (1994) [68] Goldsmith (1994) [42] Peele (1996) [69] Condra (1997) [40] Stafford (1998) [2] Nutting (1999) [41] Vendrely (1999) [70] Maguire (1999) [71] Wenkle (2000) [72] Pourel (2001) [73] Dufour (2001) [74] Debus (2001) [75] Uy (2002) [76] Pirzkall (2003) [77] Soyuer (2004) [78] Selch (2004) [79]

117 115 254 117 86 246 581 82 156 28 46 26 31 189 40 20 92 45

60% >60 80 57 61 40 46 98 55 108 40 41 53 30 73 35 30 36 92 36

2887

10–276

Total

STR + EBRT (%)

Follow-up (months)

GTR (%)

STR (%)

10–276

90 93

45 63

82

96 96

60 43

80 85

98

48 54

95 88

52 53 61

65% >60 78

77

38

77–98

48–63

84 88 (8-yr PFS) 89 (98 after 1980) 100 86 92 89 92 (4-yr PFS) 100 95 93 (10-yr PFS) 98 (FSRT) 93 100 91 98 (3-yr PFS) 80–100

n, number of patients; yr, year. Note: Compiled series allowing for comparison in the rates of progression-free survival for patients treated with gross total resection, subtotal resection, or with subtotal resection + external beam radiation therapy. Actuarial intervals other than 5 years are given in parentheses. Patients in the above reports typically, but not exclusively, had either known or presumed low-grade meningiomas.

continued investigation. Ganz and associates noted that a minimum peripheral tumor dose of 10 Gy or less (mean, 7.6 Gy) was linked with a higher risk of failure, whereas ≥12 Gy (mean 19.0 Gy) resulted in improved local control [84]. Morita and colleagues subsequently recommended tumor margin doses of at least 15 to 16 Gy and stressed the importance of at least 5 years of follow-up to assess local control [54]. More recently and with longer follow-up, Stafford and co-authors reported no degradation in local control for benign meningiomas at 5 years with tumor margin doses <16 Gy versus ≥16 Gy [56]. Similarly, Kondziolka and associates found no improvement with marginal doses ≥15 Gy compared with <15 Gy [81]. With further investigation and lengthier evaluation intervals, tumor margin doses from 12 to 16 Gy are now accepted [29, 53, 55, 58–60]. As shown in Table 21-6, SRS has provided excellent 5- to 10-year local control and sets a standard, demanding from longer-course fractionated radiation therapy not only similar control rates but also limited side effects. In this light, a review of the side effects of SRS is relevant.

Stereotactic Radiosurgery: Toxicity The adverse sequelae most commonly attributed to SRS have been cranial nerve deﬁcits and edema, although less common occurrences of radiation necrosis [81, 85, 86], peritumoral cyst

formation [56], carotid stenosis [87], and hypothalamic dysfunction [50] have also been described.

Stereotactic Radiosurgery: Cranial Neuropathy Based on anatomic proximity, cranial nerve deﬁcits are most common with meningiomas of the cranial base. With modern techniques and margin doses ranging from 14 to 16 Gy, four series have identiﬁed new or worsened neurologic deﬁcits in 8% of patients [30, 56, 88, 89]. These deﬁcits are occasionally, but not uniformly, transient and tend to develop within 1 month [56] to 2 or 5 years of treatment [53, 90]. Cranial neuropathies have occurred more frequently with sensory nerves (such as the optic, cochlear, and trigeminal nerves) than with motor nerves. Motor nerves, such as the oculomotor nerves in the cavernous sinus, have tolerated radiosurgery relatively well. Roche and colleagues observed no new oculomotor deﬁcits in 80 cavernous sinus meningioma patients with a mean maximum prescription dose of 28 Gy (range, 12 to 50 Gy), mean prescription isodose 50%, and median follow-up of 30.5 months [83]. Conversely, sensory nerves of the anterior visual pathway have been particularly susceptible. Doses of only 10 Gy or less [54, 90] carry a roughly 1% or 2% rate of radiation optic neuropathy, but the complication curve rises steeply with higher doses [90, 91].

21.

intracranial meningioma: fractionated radiation therapy perspective

263

Stereotactic Radiosurgery: Edema

Radiation Therapy: Fractionated External Beam

Whereas cranial neuropathies are linked with skull base meningiomas, edema is more commonly encountered with nonbasal primaries. Meningiomas produce vasogenic mediators [6, 92– 94], but this alone would not explain the higher rate of edema from nonbasal tumors. One difference may be that nonbasal meningiomas tend to include a broader pial interface [47, 95], perhaps allowing vasogenic substrates greater access to adjacent brain parenchyma. The pial interface permits pial-cortical arterial supply, potentiating vascular steal with nonbasal meningiomas. In contrast, most basal primaries have limited pial involvement [47, 95, 96]. With parasagittal meningiomas, for which edema has been frequently reported [81], bridging vein and/or sagittal sinus occlusion may also contribute. Posttreatment edema has been widely reported subsequent to SRS. Table 21-8 examines edema after radiosurgery in 200 patients from ﬁve series [81, 82, 97–99]. Edema developed in 25% to 78% of patients with nonbasal meningiomas, compared with 0 to 22% of basal meningiomas. In addition to anatomic location, factors associated with an increased risk of edema in these studies were margin dose greater than 15 to 18 Gy, tumor size >3 cm diameter, and the presence of pretreatment edema. Early reports, often with higher doses applied to larger meningiomas, carried higher complication risks. Recent analyses, with more careful patient selection, have encountered less morbidity. Kondziolka and co-authors, with median follow-up 3.5 years, had excellent outcomes (no tumor progression, no management morbidity, and no need for further treatment) in all 41 patients with small meningiomas (<7.5 cm3 or <3 cm diameter) and with no prior surgery. However, patients with larger tumors, prior surgery, and pre-SRS neurologic deﬁcits experienced less favorable results [81]. These data reﬂect the safety and the limits of safety for SRS and should be considered for comparison with fractionated therapy.

Historically, due to infrequent tumor regression after fractionated EBRT, meningiomas were considered resistant to irradiation, and radiation itself was feared to produce considerable side effects [7, 100]. Confounding concerns were voiced regarding malignant degeneration and the relationship between irradiation and the development of meningiomas [16, 17, 19, 20]. There is also apprehension about arachnoid scarring from radiation therapy. These concerns continue to guide patient care and allocate many inoperable or subtotally resected meningioma patients to observation without postoperative therapy [2, 5, 101]. Many series have addressed these issues. Malignant degeneration is more likely the natural history of a subgroup of recurrent meningiomas, and has not been explicitly linked with RT. The risk of developing a meningioma after irradiation has been reviewed by Strojan et al. They reported the actuarial risk after RT to be 0.53% at 5 years and 8.18% at 25 years [20]. This risk may be less with modern image-based and highly conformal radiation therapy.

Primary EBRT Radiation therapy has been used as primary treatment after biopsy or on the basis of imaging ﬁndings alone. Table 21-6 lists the commonality of this tactic with radiosurgery, but this has also been applied to EBRT, with similarly favorable results [40, 71, 77, 79, 102]. An early report from Royal Marsden found 47% disease-free survival at 15 years in 32 patients with EBRT but without resection. This was lower than the 61% rate after partial resection plus RT but surprisingly superior to the 40% 10-year rate after GTR and EBRT [66]. Rather than suggesting inferior outcome after primary EBRT, this likely reﬂects patient selection biases and treatment results from an era predating image-based planning.

TABLE 21-8. Edema after radiosurgery. Author (year)

n

Follow-up

Dose, (range)

Edema, (basal vs. nonbasal)

Comment

Ganz (1996) [97]

34

1–3 yr

15 Gy (12–25 Gy)

21% (7% vs. 71%)

Nakamura (1996) [98]

48

1 yr

15 Gy (mean)

Kondziolka (1998) [81]

203

3.5 yr

15 Gy (9–32 Gy)

10% (all parasagittal)

Ramsey (2002) [99]

23

1.4 yr

14 Gy (9–18 Gy)

39% (22% vs. 78%)

Vermeulen (1999) [82]

95

2.3 yr

17 Gy (8–20 Gy)

32% 22% vs. 41%

86% of patients with edema: Margin dose ≥18 Gy. 12 Gy adequate for LC. Edema with 15–24 Gy. 25% resolved. 3- & 5-yr actuarial edema 16%. More common if >3 cm. Occurred 1–23 mo after SRS. All resolved (median 15 mo). Factors for edema: Volume >4 cm3 (p = 0.005). Parasagittal (0.006). Pre-SRS edema (0.06). 1% basal vs. 4% nonbasal had “deteriorating” edema without tumor growth

1–3.5 yr

8–32 Gy

22% mean (13 vs. 45%)

Total

200

8% (0% vs. 25%)

n, number of patients; yr, year; mo, month; LC, local control. Note: Five series, with 200 total patients, examining new or worsening edema after stereotactic radiosurgery (SRS). The methods of assessing edema varied from clinical (Kondziolka) to imaging (Ganz and Vermeulen) to a combination of both (Nakamura and Ramsey).

264

l. rogers et al.

In a recent series, Debus and associates noted no difference in the outcome comparing patients with primary stereotactic fractionated EBRT (n = 59) to those with surgery plus similar EBRT (n = 130). In fact, there were no recurrences in patients treated by radiotherapy alone. Five- and 10-year recurrencefree probabilities were 100% [75]. Carella and co-authors also discovered uniform local control at 3 to 6 years in 11 patients treated with primary EBRT [102]. Several other publications in the past decade have reported excellent outcomes for patients with primary EBRT [40, 71, 77, 79, 103].

Primary EBRT for Optic Nerve Sheath Meningiomas Optic nerve sheath meningiomas account for only 1% to 2% of meningiomas but are a valuable case in point for primary EBRT. Surgery carries a high risk of visual complications [104– 107] and a high rate of local recurrence [107, 108]. EBRT has thus been used as primary management [103, 108]. In a study by Turbin et al., radiation therapy alone provided more favorable outcome than observation, surgery alone, or even surgery plus EBRT. They used total doses from 40 to 55 Gy and had median follow-up of 8.3 years [108]. Narayan and colleagues, with a median follow-up of 51.3 months, found no radiographic progression in any of 14 optic nerve sheath meningioma patients treated with conformal EBRT, 86% of whom had either improved or stable visual acuity [103].

FIGURE 21-5. Conformal, image-based EBRT. Magnetic resonance image and isodoses for intensity-modulated radiation therapy of a cerebellopontine angle meningioma treated to 54 Gy in 40 fractions, prescribed to the 90% isodose. The yellow line represents the 70% isodose, green 50%, and sky-blue 30%. Note: In this case, the “dural tail” is not included within the planning target volume.

Postoperative EBRT GTR is not always feasible, and EBRT has been used subsequent to STR. As viewed in Table 21-5 and further in Table 21-7 (STR column), the likelihood of progression with incompletely resected meningiomas is well documented. Many retrospective studies now support a role for postoperative EBRT after STR. These have shown improvements in local control (Table 21-7), and possibly even survival [27, 40]. No phase III trials testing this hypothesis have yet been concluded. The European Organisation for Research and Treatment of Cancer (EORTC) began enrolling incompletely resected, newly diagnosed, grade 1 meningioma patients to a randomized trial (EORTC 26021) comparing observation with radiation therapy (either fractionated radiotherapy or radiosurgery). However, the trial was closed due to poor accrual.

EBRT: Technical Factors Technical improvements have favorably impacted both the outcome and the side effect proﬁle of postoperative EBRT. With image-based techniques, treatment is delivered with more precision and conformality (Fig. 21-5), and improved results are to be expected [79, 109]. Indeed, improvements in local control have been documented with computed tomography (CT)- or magnetic resonance imaging (MRI)-based planning. Goldsmith et al. [42] and Milosevic et al. [110] each substantiated improvements in local control accompanying modern imaging. Goldsmith et al. found that CT- and/or MRI-based target deﬁnition, along with appropriate immobilization, resulted in a 21% improvement in progression-free survival (p = 0.002) [8, 42]. Speciﬁcally, patients treated with these advances had a 10-year

progression-free survival of 98% versus 77% without them (p = 0.002) [42].

EBRT: Dose Recommended doses are generally 50 to 55 Gy with fractions of 1.8 to 2.0 Gy [8, 9, 40, 102], but a dose-response has not been unequivocally documented. Goldsmith and associates did report that doses above 52 Gy resulted in an improved 10-year local control of 93% compared with 65% with lower doses (p = 0.04) [42], but this was not retained on multivariate analysis. Among 67 patients, Winkler et al. found no clear dose response from 36 to 79.5 Gy (1.5 to 2.0 Gy per day) [111]. However, a common dose schedule is 54 Gy in 30 fractions of 1.8 Gy each. This results in an acceptable side effect proﬁle, even for meningiomas of the optic sheath or otherwise near the anterior visual pathway, although at these sites, lower total doses in the range of 50 to 52 Gy and even modestly lower doses per fraction may be considered [76, 103, 112–114].

EBRT: Target Volume The recommended planning target volume (PTV) has ranged from gross tumor volume (GTV) with up to a 4-cm margin [66] to GTV plus 2 cm [40, 115], GTV plus 1 cm [8], down to GTV plus 2 mm in a recent series [75]. The latter study included daily stereotactic localization. The literature does not deﬁne the optimal margin breadth with certainty; however, marginal failures have been uncommon [58, 75]. The propriety of targeting the “dural tail” remains contentious. Figure 21-5 depicts a dural tail, which is not included within the deﬁned target. Although some studies have recom-

21.

intracranial meningioma: fractionated radiation therapy perspective

mended treating the dural tail [30], the majority have not expressly included it. It may illuminate the debate to cite that the dural tail is not considered in the characterization of resection extent according to Simpson’s criteria (Table 21-1). In fact, the dural tail is an imaging ﬁnding, typically not visualized at surgery, and Simpson’s deﬁnitions predate CT and MRI. The excellent results from GTR alone (Table 21-2) have been obtained irrespective of removing the dural tail. Furthermore, planning target volumes with SRS have tended to be the contrast-enhanced tumor without a margin [53, 56, 80, 116]. Nicolato and colleagues found no advantage to larger radiosurgery margins. With median follow-up of 4 years, they had 100% tumor growth control with a conformity index (ratio of prescription isodose volume to target volume) of ≤1.5 [58]. Using fractionated radiotherapy with 2-mm margins, Debus et al. noted no marginal failures in 189 patients with 3 years median follow-up [75]. Pathology studies have revealed the dural tail to be composed entirely or almost entirely of hypervascular dura in most cases [117–119]. Microscopic clusters of meningioma cells may occasionally be found, but these are no more likely in the dural tail than within randomly selected dural strips adjacent to the meningioma [120].

EBRT: Toxicity Side effects of EBRT, especially with current methods of treatment planning and delivery, are evidently few, but not negligible. Early publications identiﬁed no deleterious late effects [64–65] but, more recently, a small incidence of clinically important late effects is evident. The largest recent series of fractionated irradiation in the literature, by Debus et al., reported on 189 patients treated with a highly conformal stereotactic approach, using median daily fractions of 1.8 Gy to a mean cumulative dose (at isocenter) of 56.8 Gy. With a median followup of nearly 3 years, they identiﬁed clinically signiﬁcant (grade 3) toxicity in 4 (2.2%) patients, 3 (1.7%) in the absence of a preexisting deﬁcit: reduced vision, a visual-ﬁeld deﬁcit, and trigeminal neuropathy [75]. This is a substantial improvement over the 38% reported by Al-Mefty et al. [121] with older methods of radiation delivery. Goldsmith et al. recognized complications, which they believed may have been attributable to EBRT, in 5 of 140 (3.6%) patients. These were retinopathy in two, optic neuropathy in one, and cerebral necrosis in two patients [42]. In a separate publication, Goldsmith and colleagues constructed a model to predict optic nerve tolerance and recommended a maximum dose of 890 optic ret (e.g., 54 Gy in 30 fractions) [113]. Optic complications have been reported elsewhere and are quite rare with doses below this threshold, particularly if doses per fraction ≤2.0 Gy are used [66, 67, 77]. Uy et al., with a median dose of 50.4 Gy and fractions of 1.7 to 2 Gy, noted no optic pathway toxicity [76]. The beneﬁcial impact of lower doses per fraction on optic tolerance has been recently conﬁrmed in a radiobiologic analysis by Shrieve and colleagues [114]. Nonocular cranial nerve deﬁcits may occur [75] but are uncommon. Selch and colleagues found no treatment-related cranial neuropathies of any kind in 45 cavernous sinus meningioma patients treated with fractions of 1.7 to 1.8 Gy to median total dose of 50.4 Gy [79]. Urie and co-authors also remarked that doses in this range rarely cause cranial neuropathy [122].

265

Brain or brain-stem necrosis is also uncommon but has been observed beyond the above report of Goldsmith et al. [67, 76, 121]. The study by Al-Mefty et al. is especially instructive. They reviewed post-EBRT toxicity in 58 adult patients with a variety of skull base, parasellar, or pineal region tumors. Seventeen (29%) developed late brain parenchymal changes, with latencies of 4 months to 23 years. These changes were encephalomalacia, cerebral atrophy, gliosis and/or necrosis, generalized or involving the temporal lobes in 14 (82%) of the 17 cases [121]. Such occurrences are almost certainly related to antiquated treatment techniques. Integral doses to the temporal lobes and to large volumes of untargeted brain are much lower with current multiﬁeld conformal and intensity modulated methodology. Pituitary dysfunction [41, 67, 77, 121], cerebrovascular events [41, 66, 79], second malignancy [41, 121], orbital ﬁbrosis [71], and other sporadic toxicities have been noted. In an older series, using an array of techniques and doses, Glaholm et al. found that all complications occurred with the former convention of treating only a portion of the ﬁelds daily, and with fractional doses >1.8 Gy (to ﬁnal doses of 50 to 55 Gy) [66]. EDEMA Table 21-7 summarizes data from 22 EBRT series with a total of 2887 patients. Only six (0.2%) patients reportedly developed edema, and two of these six were asymptomatic. This very low rate of edema should be viewed with caution. Seventeen of the 22 studies did not speciﬁcally assess edema, although there was no apparent clinical suspicion of it. Moreover, patients with asymptomatic edema, or with mild and transient symptoms such as headache or nausea, likely escaped detection. Even with this disclosure, Selch and colleagues did evaluate edema and noted none in 45 patients with 3 years median follow-up [79]. It is evident that edema is a much less likely consequence of fractionated EBRT than of single-fraction SRS. COGNITIVE Some retrospective studies, typically with less conformal techniques, have identiﬁed personality changes [76] and memory loss [40, 41, 67, 71] as complications of EBRT. Cognitive outcome has been prospectively evaluated after fractionated stereotactic radiotherapy (FSRT) for skull base meningiomas. Steinvorth and colleagues used a comprehensive battery of neurocognitive tests before, after the ﬁrst fraction, at completion, and 12 months subsequent to FSRT. They observed, after the ﬁrst fraction, a transient decline in memory counterbalanced by an increase in attention. There was no cognitive deterioration with further follow-up [123].

EBRT: Special Circumstances ATYPICAL MENINGIOMAS Atypical (WHO grade II) meningiomas comprise roughly 20% of all meningiomas (Fig. 21-1). However, this is based on relatively newly revised grading criteria, and atypical tumors are commonly underrecognized. For this reason, treatment guidelines are even less uniform than for benign (grade I) meningiomas. Many investigators, recognizing that atypical meningiomas entail a much greater recurrence risk, have recommended irradiation irrespective of resection extent [19, 40, 111].

266

l. rogers et al.

In contrast, others have advocated that gross totally resected atypical meningiomas be observed. Goyal and colleagues identiﬁed atypical features in 6.7% of their patients. Twenty-two atypical meningioma patients were studied, eight of whom received postoperative EBRT, median dose 54 Gy. With a median follow-up of 5.5 years, they found that GTR achieved durable local control in 87% at 5 and 10 years and that EBRT had no signiﬁcant impact upon local control or overall survival [124]. This series highlights some of the difﬁculties in deciphering the literature respective to atypical tumors. The sample size is small, the percentage of grade II patients (6.7%) is too low to be representative of current WHO grading, and the doses used may be too low for grade II tumors [19]. Perry and colleagues reported 108 atypical meningiomas treated with modern surgical techniques, grading, and postoperative imaging, with a 5-year recurrence rate of 40% even after GTR [36]. From a separate Mayo Clinic study, recurrences after either subtotal resection or “radical subtotal resection” were 38.6% at 5 years and 61.4% at 10 years [2]. Regarding dose, Hug and associates found that the local control of atypical meningiomas was signiﬁcantly enhanced by cumulative doses of ≥60 CGE (cobalt Gray equivalent). Doses were discussed as CGE because treatment included protons in about half their patients [19]. Radiosurgery has also been used in the treatment of atypical meningiomas. Stafford and colleagues identiﬁed radiosurgery 5-year actuarial local control (LC) of 93% for grade I meningiomas, compared with 68% for grade II [56]. This is very similar to the 64% (7 of 11 patients) reported by Condra et al. with lower EBRT doses [40] but inferior to the 5-year LC of 90% with ≥60 CGE observed by Hug and co-authors [19]. It is apparent that a subset of atypical meningiomas will fail with GTR alone. Therefore, there is excellent rationale for an aggressive approach toward atypical meningiomas, with relative consensus for irradiation after STR. However, it is currently difﬁcult to predict which grade II patients will beneﬁt from adjuvant therapy after GTR. RECURRENT MENINGIOMAS Recurrent meningiomas exhibit a severalfold increased rate of progression over newly diagnosed tumors [7, 40, 65, 67]. The data in support of radiation therapy for a recurrent meningioma are relatively convincing. Miralbell and co-authors found 78% 8-year progression-free survival in patients treated with surgery and EBRT for recurrent tumors versus 11% with surgery alone [67]. Similarly, Taylor et al. found the respective 5-year progression-free rates to be 88% versus 30% [65]. Additionally, they noted that 5-year overall survival was 90% with surgery plus EBRT versus 45% after surgery alone [65]. These data are striking and strongly support aggressive treatment for recurrent meningiomas.

Comparative Outcome: EBRT Versus SRS The data within Tables 21-6 and 21-7 attest to the relative equivalence in local control between EBRT and SRS. To assert this even further, Table 21-9 summarizes the cumulative data from these many reports and reveals that 5- to 10-year progression-free survival (PFS) rates range from 80% to 100% with EBRT and from 75% to 100% with SRS. This relative equivalence is also evident when viewing recent series, published

TABLE 21-9. External beam radiation therapy versus stereotactic radiosurgery progression-free survival.

SRS combined EBRT combined SRS recent EBRT recent

n

Follow-up (mean or median; months)

PFS (5- to 10-year ; %)

1578 2887 884 480

19–75 30–108 29–75 30–92

75–100 80–100 86–100 91–100

n, number of patients. Note: Compiled series allowing comparison of fractionated EBRT with stereotactic radiosurgery, with 5- to 10-year PFS as the end point. Patients in the above reports typically, but not exclusively, had either known or presumed low-grade meningiomas. “Combined” includes the full range of data from Tables 21–5 and 21–7. “Recent” refers to those reports published within the past 5 years and with reported SRS marginal doses of ≥13 Gy.

within the past 5 years, for which 5- to 10-year PFS rates are 91% to 100% with EBRT and 86% to 100% with SRS. For this latter comparison, SRS patients treated with the lowest doses (≤12 Gy) [57] were excluded. Sibtain and Plowman compared EBRT with SRS for cavernous sinus meningioma patients. They reviewed 13 patients with tumors <3 cm in diameter, most of who received radiosurgery, versus 15 patients with tumors >3 cm, most of who had EBRT. They noted uniform local control at 12 to 83 months [115]. Nakamura et al., addressing technical concerns from 10 irregularly shaped and challenging skull base lesions, found that intensity-modulated radiation therapy (IMRT) provided target coverage and normal tissue sparing analogous to SRS. IMRT offered an improved conformity index at the prescription isodose, although at times with less conformity at lower isodoses [125].

Conclusion Meningiomas are among the most common intracranial neoplasms. The majority of meningiomas are histologically benign, although they can be clinically formidable. As a consequence of their typically slow growth and of a lack of randomized trials, uniform treatment recommendations are difﬁcult to formulate. Small, asymptomatic, stable or slowly progressive meningiomas can be observed. For other patients, GTR remains the gold standard against which other therapeutic strategies are measured. Nevertheless, within constraints of acceptable morbidity, complete removal is often impossible due to involvement of critical neurovascular structures (e.g., cavernous sinus and petroclival meningiomas), and these are among the most challenging of neurosurgical cases [126]. Irradiation, be it fractionated EBRT or single-fraction radiosurgery, plays an integral role in the treatment of patients at high risk of recurrence or progression. As summarized above, a myriad of series documents improvements in local control and in some even overall survival. Those who stand to beneﬁt from irradiation are patients with incompletely resected, higher grade, or recurrent meningiomas. There is also growing experience with irradiation as a primary modality, either after a limited biopsy or based purely on an image-based diagnosis. EBRT and SRS result in very similar local control rates, and either can be recommended for many, but not for all

21.

intracranial meningioma: fractionated radiation therapy perspective

patients. EBRT is suitable for a broader range of patients, whereas excellent outcome with SRS is achieved within deﬁned cohorts. Several studies have recommended diameter or volume restrictions with SRS, both to improve local control and to decrease toxicity. Common relative constraints on diameter have been 3.0 to 3.5 cm, with total volumes of 7.5 to 15 cm3 [61, 81]. Pollock and colleagues found that SRS provides equivalent tumor control to a Simpson grade 1 resection with small- to medium-size meningiomas of <3.5 cm average diameter and <15 cm3 [61]. Kondziolka and co-authors reported excellent outcome within the conﬁnes of 3.0 cm diameter or 7.5 cm3 volume [81]. Speciﬁcally addressing edema risk, Ramsey and associates suggested an even smaller volume restraint of 4 cm3 [99]. Other factors to consider in assessing edema risk from SRS are nonbasal location and the presence of pretreatment edema [99]. An additional constraint on SRS is the anterior visual pathway. Many radiosurgeons limit dose to the optic nerves and chiasm to 8 or 9 Gy, although some suggest that a small volume of optic apparatus can tolerate doses of up to 12 Gy [87]. Even at the higher dose limit, radiosurgery would be excluded for most optic nerve sheath meningiomas, as well as some meningiomas at parasellar and orbit apex sites. Taken together, these data indicate that radiosurgery is most judiciously applied to smaller meningiomas at adequate distance from the optic apparatus and with limited risk and side-effect proﬁles. External beam irradiation does not suffer from these limitations. Within commonly employed dose and fractionation guidelines, EBRT carries only small risks of edema or optic neuropathy. Furthermore, fractionated EBRT can now be delivered in highly conformal and reproducible fashion, with attendant improvements in local control [8, 42, 110], and with minimal risk of adverse neurocognitive sequelae [123]. This promises yet to improve with further developments in dose delivery and image guidance.

References 1. Perry A, Giannini C, Raghavan R, et al. Aggressive phenotypic and genotypic features in pediatric and NF2-associated meningiomas: a clinicopathologic study of 53 cases. J Neuropathol Exp Neurol 2001; 60:994–1003. 2. Stafford S, Perry A, Suman V, et al. Primarily resected meningiomas: outcome and prognostic factors in 581 Mayo Clinic patients, 1978 through 1988. Mayo Clin Proc 1998; 73:936–942. 3. Adegbite AB, Kahn MI, Paine KWE, et al. The recurrence of intracranial meningiomas after surgical treatment. J Neurosurg 1983; 58:51–56. 4. Perry A, Dehner LP. Meningeal tumors of childhood and infancy. An update and literature review. Brain Pathol 2003; 13:386–408. 5. Jung HW, Yoo H, Paek SH, et al. Long-term outcome and growth rate of subtotally resected petroclival meningiomas: experience with 38 cases. Neurosurgery 2000; 46:567–575. 6. McDermott MW, Quinones-Hinojosa A, Fuller GN, et al. Meningiomas. In: Levin VA, ed. Cancer in the Central Nervous System, 2nd ed. New York: Oxford University Press, 2002:69–299. 7. Mirimanoff R, Dosoretz D, Linggood R, et al. Meningioma: analysis of recurrence and progression following neurosurgical resection. J Neurosurg 1985; 62:18–24. 8. Goldsmith B. Meningioma. In: Leibel S, Phillips T, eds. Textbook of Radiation Oncology. Philadelphia: WB Saunders, 1998:324– 340.

267

9. Wara WM, Bauman GS, Sneed PK, et al. Brain, brainstem, and cerebellum. In: Perez CA, Brady LW, eds. Principles and Practice of Radiation Oncology, 3rd ed. Philadelphia: Lippencott-Raven, 1997:818–820. 10. Wahab M, Al-Azzawi F. Meningioma and hormonal inﬂuences. Climacteric 2003; 6:285–292. 11. DeMonte F, Marmor E, Al-Mefty O. Meningiomas. In: Kaye A, Law E, eds. Brain Tumors, 2nd ed. London: Churchill Livingstone, 2001:719–750. 12. Perry, A. Unmasking the secrets of meningiomas: a slow but rewarding journey. Surg Neurol 2004; 61:171–173. 13. Ragel B, Jensen RL. New approaches for the treatment of refractory meningiomas. Cancer Control 2003; 10(2):148–158. 14. Annegers JF, Laws ER, Kurland LT, et al. Head trauma and subsequent brain tumors. Neurosurgery 1979; 4(3):203–206. 15. Inskip PD, Mellemkjaer L, Gridley G, et al. Incidence of intracranial tumors following hospitalization for head injuries (Denmark). Cancer Causes Control (England) 1998; 9(1):109–116. 16. Ron E, Modan B, Boice JD Jr. Mortality after radiotherapy for ringworm of the scalp. Am J Epidemiol 1988; 127:713–725. 17. Ron E, Modan B, Boice JD, et al. Tumors of the brain and nervous system after radiotherapy in childhood. N Engl J Med 1988; 319: 1033–1039. 18. Ghim TT, Seo JJ, O’Brien M, et al. Childhood intracranial meningiomas after high-dose irradiation. Cancer 1993; 71(12):4091– 4095. 19. Hug EB, Devries A, Thornton AF, et al. Management of atypical and malignant meningiomas: role of high-dose, 3D-conformal radiation therapy. J Neurooncol 2000; 48:151–160. 20. Strojan P, Popovic M, Jereb B. Secondary intracranial meningiomas after high-dose cranial irradiation: report of ﬁve cases and review of the literature. Int J Radiat Oncol Biol Phys 2000; 48:65–73. 21. Al-Mefty O, Topsakal C, Pravdenkova S, et al. Radiaiton-induced meningiomas: clinical, pathological, cytokinetic, and cytogenetic characteristics. J Neurosurg 2004; 100:1002–1013. 22. Central Brain Tumor Registry in the United States (CBTRUS). Statistical report: primary brain tumors in the Unites States, 1992– 1997. In: Central Brain Tumor Registry of the United States. Chicago: CBTRUS, 2000:11–26. 23. Kuratsu J, Kochi M, Ushio Y. Incidence and clinical features of asymptomatic meningiomas. J Neurosurg 2000; 92:766–770. 24. Kuratsu J, Ushio Y. Epidemiological study of primary intracranial tumors: a regional survey in Kumamoto prefecture in the southern part of Japan. J Neurosurg 1996; 84:946–950. 25. Jaaskelainen J, Haltia M, Servo A. Atypical and anaplastic meningiomas: radiology, surgery, radiotherapy, and outcome. Surg Neurol 1986; 25:233–242. 26. Longstreth WT, Dennis LK, McGuire VM, et al. Epidemiology of intracranial meningioma. Cancer 1993; 72:639–648. 27. McCarthy BJ, Davis FG, Freels S, et al. Factors associated with survival in patients with meningioma. J Neurosurg 1998; 88:831– 839. 28. DeMonte F, Al-Mefty O. Meningiomas. In: Kaye E, Law E, eds. Brain Tumors. Edinburgh: Churchill Livingstone, 1995:675–704. 29. Flickinger JC, Kondziolka D, Maitz AH, et al. Gamma knife radiosurgery of image-diagnosed intracranial meningioma. Int J Radiat Oncol Biol Phys 2003; 56:801–806. 30. DiBiase SJ, Kwok Y, Yovina S, et al. Factors predicting local tumor control after Gamma Knife stereotactic radiosurgery for benign intracranial meningiomas. Int J Radiat Oncol Biol Phys 2004; 60:1515–1519. 31. Nakasu S, Hirano A, Shimura T, et al. Incidental meningiomas in autopsy study. Surg Neurol 1987; 27:319–322. 32. Cleland, J. Description of two tumours adherent to the deep surface of the dura mater. Glasgow Med J 1864; 11:148–159.

268

l. rogers et al.

33. Schmidt MB. Ueber die Pacchionischen granulationen und ihr verhaeltniss zu den sarcomen und psammomen der dura mater. Virchows Arch 1902; 170:429–464. 34. Perry A. Meningiomas. In: McLendon R, Bigner DD, Rosenblum M, Bruner JM, eds. Russell & Rubinstein’s Pathology of Tumors of the Nervous System, 7th ed. London: Edward Arnold Ltd., 2006:427–474. 35. Louis DN, Scheithauer BW, Budka H. Meningiomas. In: Kleihues P, Cavenee WK, eds. World Health Organization classiﬁcation of tumours. Pathology and genetics of tumours of the nervous system. Lyon: IARC Press, 2000:176–184. 36. Perry A, Stafford SL, Scheithauer BW, et al. Meningioma grading: an analysis of histologic parameters. Am J Surg Pathol 1997; 21:1455–1465. 37. Perry A, Scheithauer BW, Stafford SL, et al. “Malignancy” in meningiomas: a clinicopathologic study of 116 patients. Cancer 1999; 85:2046–2056. 38. Ho DMT, Hsu CY, Ting LT, et al. Histopathology and MIB-1 labeling index predicted recurrence of meningiomas. A proposal of diagnostic criteria for patients with atypical meningioma. Cancer 2002; 94:1538–1547. 39. Korshunov, A. Shishkina, L. Golanov, A. Immunohistochemical analysis of p16INK4A, p14ARF, p18INK4C, p21CIP1, p27KIP1 and p73 expression in 271 meningiomas. Correlation with tumor grade and clinical outcome. Int J Cancer 2003; 104:728–734. 40. Condra K, Buatti J, Mendenhall W, et al. Benign meningiomas: primary treatment selection affects survival. Int J Radiat Oncol Biol Phys 1997; 39:427–436. 41. Nutting C, Brada M, Brazil L, et al. Radiotherapy in the treatment of benign meningioma of the skull base. J Neurosurg 1999; 90:823–827. 42. Goldsmith B, Wara W, Wilson C, et al. Postoperative irradiation for subtotally resected meningiomas. J Neurosurg 1994; 80:195– 201. 43. Pollock BE, Stafford SL, Link MJ. Gamma Knife radiosurgery for skull base meningiomas. Neurosurg Clin N Am 2000; 11:659– 666. 44. Simpson D. The recurrence of intracranial meningiomas after surgical treatment. J Neurol Neurosurg Psychiatry 1957; 20:22–39. 45. Kinjo T, Al-Mefty O, Kanaan I. Grade zero removal of supratentorial convexity meningiomas. Neurosurgery 1993; 33:304–399. 46. Wara W, Sheline G, Newman H, et al. Radiation therapy of meningiomas. Am J Roentgenol Ther Nucl Med 1975; 123:453– 458. 47. Ildan F, Tuna M, Iskender G, et al. Correlation of the relationship of brain-tumor interfaces, magnetic resonance imaging, and angiographic ﬁndings to predict cleavage of meningiomas. J Neurosurg 1999; 19:384–390. 48. Nakamura M, Roser F, Michel J, et al. The natural history of incidental meningiomas. Neurosurgery 2003; 53:62–71. 49. Chang SD, Adler JR Jr. Treatment of cranial base meningiomas with linear accelerator radiosurgery. Neurosurgery 1997; 41:1019– 1025. 50. Hakim R, Alexander E III, Loefﬂer JS, et al. Results of linear accelerator based radiosurgery for intracranial meningiomas. Neurosurgery 1998; 42:446–453. 51. Chang SD, Adler JR, Martin DP. Linac radiosurgery for cavernous sinus meningiomas. Stereotact Funct Neurosurg 1998; 71(1):43–50. 52. Liscak R, Simonova G, Vymazal J, et al. Gamma Knife radiosurgery of meningiomas in the cavernous sinus. Acta Neurochir (Wien) 1999; 141:473–480. 53. Kondziolka D, Niranjan A, Lindsford LD, et al. Stereotactic radiosurgery for meningiomas. Neurosug Clin N Am 1999; 10:317– 325.

54. Morita A, Coffey R, Foote R, et al. Risk of injury to cranial nerves after gamma knife radiosurgery for skull base meningiomas: Experience in 88 patients. J Neurosurg 1999; 90:42–49. 55. Roche PH, Pellet W, Fuentes S, et al. Gamma knife radiosurgical management of petroclival meningiomas: results and indications. Acta Neurochir (Wien) 2003; 145(10):883–888. 56. Stafford S, Pollock B, Foote R, et al. Meningioma radiosurgery: tumor control, outcomes, and complications among 190 consecutive patients. Neurosurgery 2001; 49:1029–1038. 57. Shin M, Kurita H, Sasaki T, et al. Analysis of treatment outcome after stereotactic radiosurgery for cavernous sinus meningiomas. J Neurosurg 2001; 95:435–439. 58. Nicolato A, Foroni R, Alessandrini F, et al. The role of gamma knife radiosurgery in the management of cavernous sinus sinus meningiomas. Int J Radiat Oncol Biol Phys 2002; 53(4):992– 1000. 59. Lee JY, Niranjan A, McInerney J, et al. Stereotactic radiosurgery providing long-term tumor control of cavernous sinus meningiomas. J Neurosurg 2002; 97(1):65–72. 60. Spiegelmann R, Nissim O, Menhel J, et al. Linear accelerator radiosurgery for meningiomas in and around the cavernous sinus. Neurosurgery 2002; 51(6):1373–1379. 61. Pollock BE, Stafford SL, Utter A, et al. 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(4):1000–1005. 62. Iwai Y, Yamanaka K, Ishiguro T. Gamma Knife radiosurgery for the treatment of cavernous sinus meningiomas. Neurosurgery 2003; 52(3):517–524. 63. Chuang CC, Tsang NM, Wei KC, et al. Linear accelerator-based radiosurgery in the management of skull base meningiomas. J Neurooncol 2004; 66:241–249. 64. Barbaro NM, Gutin PH, Wilson CB, et al. Radiation therapy in the treatment of partially resected meningiomas. Neurosurgery 1987; 20(4):525–528. 65. Taylor BW, Marcus RB, Friedman WA, et al. The meningioma controversy: post-operative radiation therapy. Int J Radiat Oncol Biol Phys 1988; 15:299–304. 66. Glaholm J, Bloom HJG, Crow, JH. The role of radiotherapy in the management of intracranial meningiomas: the Royal Marsden Hospital experience with 186 patients. Int J Radiat Oncol Biol Phys 1990; 18(4):755–761. 67. Miralbell R, Linggood RM, De la Monte S, et al. The role of radiotherapy in the treatment of subtotally resected benign meningiomas. J Neurosurg 1992; 13:157–164. 68. Mahmood A, Qureshi NH, Malik GM. Intracranial meningiomas: analysis of recurrence after surgical treatment. Acta Neurochir (Wien) 1994; 126:53–58. 69. Peele KA, Kennerdall JS, Maroon JC, et al. The role of postoperative irradiation in the management of sphenoid wing meningiomas. A preliminary report. Ophthalmology 1996; 103(11):1761–1766. 70. Vendrely V, Maire JP, Darrouzet V, et al. Radiotherapie fractionnee des meningiomes intracraniens: 15 ans d’experience au centre hospitalier universitaire de Bordeaux. Cancer Radiother 1999; 3(4):311–317. 71. Maguire PD, Clough R, Friedman AH, et al. Fractionated external-beam radiation therapy for meningiomas of the cavernous sinus. Int J Radiat Oncol Biol Phys 1999; 44(1):75–79. 72. Wenkel E, Thornton AF, Finkelstein D, et al. Benign meningioma: partially resected, biopsied, and recurrent intracranial tumors treated with combined proton and photon radiotherapy. Int J Radiat Oncol Biol Phys 2000; 48(5):1363–1370. 73. 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.

21.

intracranial meningioma: fractionated radiation therapy perspective

74. Dufour H, Muracciole X, Metellus P, et al. Long-term tumor control and functional outcome in patients with cavernous sinus meningiomas treated by radiotherapy with or without previous surgery: is there an alternative to aggressive tumor removal? Neurosurgery 2001; 48:285–296. 75. Debus J, Wuendrich M, Pirzkall A, et al. High efﬁcacy of fractionated stereotactic radiotherapy of large base-of-skull meningiomas: long-term results. J Clin Oncol 2001; 19:3547–3553. 76. Uy NW, Woo SY, The BS, et al. Intensity modulated radiation therapy (IMRT) for meningioma. Int J Radiat Oncol Biol Phys 2002; 53(5):1265–1270. 77. Pirzkall A, Debus J, Haering P, et al, Wannenmacher M. Intensity modulated radiotherapy (IMRT) for recurrent, residual, or untreated skull-base meningiomas: preliminary clinical experience. Int J Radiat Oncol Biol Phys 2003; 55(2):362–372. 78. Soyuer S, Chang EL, Selek U, et al. Radiotherapy after surgery for benign cerebral meningioma. Radiother Oncol 2004; 71:85– 90. 79. Selch MT, Ahn E, Laskari A, et al. Stereotactic radiotherapy for treatment of cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys 2004; 59(1):101–111. 80. Shafron DH, Friedman WA, Buatti JM, et al. Linac radiosurgery for benign meningiomas. Int J Radiat Oncol Biol Phys 1999; 43:321–327. 81. Kondziolka D, Flickinger J, Perez B, Gamma Knife Study Group: judicious resection and/or radiosurgery for parasagittal meningiomas: outcomes from a multicenter review. Neurosurgery 1998; 43:405–414. 82. Vermeulen S, Young R, Li F, et al. A comparison of single fraction radiosurgery tumor control and toxicity in the treatment of basal and nonbasal meningiomas. Stereotact Funct Neurosurg 1999; 72(Suppl 1):60–66. 83. Roche PH, Regis J, Dufour H, et al. Gamma knife radiosurgery in the management of cavernous sinus meningiomas. J Neurosurg 2000; 93(Suppl 3):68–73. 84. Ganz JC, Backlund EO, Thorsen FA. The results of Gamma Knife surgery for meningiomas, related to size of tumor and dose. Stereotact Funct Neurosurg 1993; 61(Suppl 1):23–29. 85. Engenhart R, Kimmig BN, Hover K, et al. Stereotactic single high dose radiation therapy of benign intracranial meningiomas. Int J Radiat Oncol Biol Phys 1990; 19:1021–1026. 86. Hudgins WR, Barker JL, Schwartz DE, et al. Gamma Knife treatment of 100 consecutive meningiomas. Stereotact Funct Neurosurg 1996; 66:121–128. 87. Stafford SL, Pollock BE, Foote, RL, et al. Stereotactic radiosurgery for meningioma. In: Pollock BE, ed. Contemporary Stereotactic Radiosurgery: Technique and Evaluation. Armonk, NY: Futura Publishing Company, 2002:157–171. 88. Subach BR, Lundsford LD Kondziolka D, et al. Management of petroclival meningiomas by stereotactic radiosurgery. Neurosurgery 1998; 42:437–443. 89. Torres RC, DeSalles AAF, Frighetto L, et al. Long-term follow-up using Linac radiosurgery and stereotactic radiotherapy as a minimally invasive treatment for intracranial meningiomas. In: Kondziolka D, ed. Radiosurgery. Basel: Karger, 2004:115–123. 90. Stafford SL, Pollock BE, Leavitt JA, et al. A study of the radiation tolerance of the optic nerves and chiasm after stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2003; 55(5):1177–1181. 91. 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(1):43–50. 92. Kalkanis SN, Carroll RS, Zhang J, et al. Correlation of vascular endothelial growth factor messenger RNA expression with peritumoral vasogenic cerebral edema in meningiomas. J Neurosurg 1996; 85:1095–1101.

269

93. Provias J, Claffey K, DelAguila L, et al. Meningiomas: role of vascular endothelial growth factor/vascular permeability factor in angiogenesis and peritumoral edema. Neurosurgery 1997; 40(5): 1016–1026. 94. Paek SH, Kim CY, Kim YY, et al. Correlation of clinical and biological parameters with peritumoral edema in meningioma. J Neurooncol 2002; 60(3):235–245. 95. Alvernia JE, Sindou MP. Preoperative neuroimaging ﬁndings as a predictor of the surgical plane of cleavage: prospective study of 100 consecutive cases of intracranial meningioma. J Neurosurg 2004; 100:422–430. 96. Inamura T, Nishio S, Takeshita I, et al. Peritumoral brain edema in meningiomas: inﬂuence of vascular supply on its development. Neurosurgery 1992; 31:179–185. 97. Ganz JC, Schrottner O, Pendl G. Radiation-induced edema after gamma knife treatment for meningiomas. Stereotact Funct Neurosurg 1996; 66(Suppl 1):129–133. 98. Nakamura S, Hiyama H, Arai K, et al. Gamma knife radiosurgery for meningiomas: four cases of radiation-induced edema. Stereotact Funct Neurosurg 1996; 66(Suppl 1):142–145. 99. Ramsey AF, Blurton M, Ekstrand K, et al. Edema following gamma knife radiosurgery for intracranial meningiomas. Int J Radiat Oncol Biol Phys 2002; 54(2 Suppl):146–147 (abstract 250). 100. King D, Chang C, Pool J. Radiotherapy in the management of meningiomas. Acta Radiol Ther Phys Biol 1966; 5:26–33. 101. Akeyson E, McCutcheon I. Management of benign and aggressive intracranial meningiomas. Oncology 1996; 10:747–756. 102. Carella RJ, Ransohoff J, Newall J. Role of radiation therapy in the management of meningioma. Neurosurgery 1982; 10(3):332– 339. 103. Narayan S, Cornblath WT, Sandler HM, et al. Preliminary visual outcomes after three-dimensional conformal radiation therapy for optic nerve sheath meningioma. Int J Radiat Oncol Biol Phys 2003; 56(2):537–543. 104. Kennerdall JS, Maroon JC, Malton M, et al. The management of optic nerve sheath meningiomas. Am J Ophthal 1988; 106:450– 457. 105. Ito M, Ishizawa A, Mikaoka M, et al. Intraorbital meningiomas. Surg Neurol 1988; 29:448–453. 106. Andrews BT, Wilson CB. Suprasellar meningiomas. The effect of tumor location on postoperative visual outcome. J Neurosurg 1988; 69:523–528. 107. Cristante L. Surgical treament of meningiomas of the orbit and optic canal: retrospective study with particular attention to the visual outcome. Acta Neurochir 1994; 126:27–32. 108. Turbin RE, Thompson CS, Kennerdell JS, et al. A long-term visual outcome comparison in patients with optic nerve sheath meningioma managed with observation, surgery, radiotherapy, or surgery and radiotherapy. Ophthalmology 2002; 109:890–899. 109. Shrieve DC, Loefﬂer JS. Radiotherapy in the treatment of meningiomas. In: Black PM, Loefﬂer JS, eds. Cancer of the Nervous System, 2nd ed. Philadelphia: Lippincott Williams & Wilkins, 2005:357–366. 110. Milosevic MF, Frost PJ, Laperriere NJ, et al. Radiotherapy for atypical or malignant intracranial meningioma. Int J Radiat Oncol Biol Phys 1996; 34(4):817–822. 111. Winkler C, Dornfeld S, Schwarz R, et al. The results of radiotherapy in meningiomas with a high risk of recurrence. A retrospective analysis. Strahlenthe Onkol 1998; 174:624–628. 112. Kupersmith MJ, Warren FA, Newall J, et al. Irradiation of meningiomas of the intracranial anterior visual pathway. Ann Neurol 1987; 21:131–137. 113. Goldsmith BJ, Rosenthal SA, Wara WM, et al. Optic neuropathy after irradiation of meningiomas. Radiology 1992; 185:71–76.

270

l. rogers et al.

114. Shrieve DC, Hazard L, Boucher K, et al. Dose fractionation in stereotactic radiotherapy for parasellar meningiomas: radiobiological considerations of efﬁcacy and optic nerve tolerence. J Neurosurg 2004; 101(Suppl 3):390–395. 115. Sibtain A, Plowman PN. Stereotactic radiosurgery. VII. Radiosurgery versus conventionally-fractionated radiotherapy in the treatment of cavernous sinus meningiomas. Br J Neurosurg 1999; 13: 158–166. 116. Kondziolka D, Lundsford LD, Flickinger JC. Stereotactic radiosurgery of meningiomas. In Lundsford LD, Kondziolka D, Flickinger JC, eds. Gamma Knife Brain Surgery. (Prog Neurol Surg, vol. 14.) Basel: Karger, 1998:104–113. 117. Nagele T, Petersen D, Klose U, et al. The “dural tail” adjacent to meningiomas studied by dynamic contrast-enhanced MRI: a comparison with histopathology. Neuroradiology 1994; 36:303–307. 118. Kawahara Y, Niiro M, Yokoyama S, et al. Dural congestion accompanying meningioma invasion into vessels: the dural tail sign. Neuroradiology 2001; 43:462–465. 119. Perry A, Gutmann DH, Reifenberger G. Molecular pathogensis of meningiomas. J Neurooncol 2004; 70:183–202. 120. Borovich B, Doron Y. Recurrence of intracranial meningiomas: the role played by regional multicentricity. J Neurosurg 1986; 64:58–63.

121. Al-Mefty O, Kersh JE, Routh A, et al. The long-term side effects of radiation therapy for benign brain tumors in adults. J Neurosurg 1990; 73:502–512. 122. Urie MM, Fullerton B, Tatsuzaki H, et al. A dose response analysis of injury to cranial nerves and/or nuclei following proton beam radiation therapy. Int J Radiat Oncol Biol Phys 1992; 23: 27–39. 123. Steinvorth S, Welzel G, Fuss M, et al. Neuropsychological outcome after fractionated stereotactic radiotherapy (FSRT) for base of skull meningiomas: a prospective 1-year follow-up. Radiother Oncol 2003; 69(2):177–182. 124. Goyal L, Suh J, Mohan D, et al. Local control and overall survival in atypical meningioma: a retrospective review. Int J Radiat Oncol Biol Phys 2000; 46: 57–61. 125. Nakamura JL, Pirzkall A, Carol MP, et al. 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. 126. Strang RD, Al-Mefty O. Comment on stereotactic radiosurgery for meningioma. In: Pollock BE, ed. Contemporary Stereotactic Radiosurgery: Technique and Evaluation. Armonk, NY: Futura Publishing Company, 2002:172–180.

2 2

Meningioma: Systemic Therapy Perspective Steven Grunberg

eningiomas pose a dilemma for the oncologist. Although meningiomas are indeed tumors (i.e., abnormal tissue growth), most meningiomas are histologically benign. The morbidity of this condition is generally not related to rapid invasive growth or to metastases, the properties that deﬁne malignancies. Instead, symptoms of this condition are primarily due to location. Slow tumor growth in the extremities or trunk can remain asymptomatic and undetected for extended periods of time. Retroperitoneal tumors, such as nonfunctioning adrenocortical carcinoma, can easily reach 10 cm in size before clinical symptoms are apparent [1]. However, there is only a ﬁxed amount of space within the skull. If one structure increases in size, another must give way, usually with the appearance of neurologic symptoms. The proximity of vital neurologic structures can also rapidly lead to complex neurologic deﬁcits while making the physical approach to the tumor mass more difﬁcult. Because meningioma tends to be a localized problem, the use of localized modalities is the logical therapeutic approach, and the use of systemic modalities, which may also result in systemic toxicities, must be adequately justiﬁed. Historically, the favored approach to meningioma is surgical resection. This may be straightforward if the meningioma is found on the cerebral convexity. However, the surgical approach to skull base tumors may be more difﬁcult. Location of the meningioma in the cavernous sinus may preclude an adequate surgical approach. In cases of optic nerve meningioma, a surgical approach, though potentially curative, can lead to the very morbidity (loss of vision) that treatment of the tumor was meant to prevent. In some cases, use of an additional localized modality can enhance the surgical approach. It has been suggested, for example, that presurgical embolization of large meningiomas can decrease operative blood loss, allowing a more aggressive and complete approach [2]. However, if surgery is not possible, other localized modalities should be considered. There is no question that radiotherapy can lengthen the time to recurrence of resected meningioma and the time to progression of unresectable meningioma. However, it is the late toxicities of such therapy that are of concern to the medical oncologist and the internist, who must follow and care for the patient long after the tumor is cured. There is a concept among medical oncologists that late toxicities of therapy are in effect the price of our success [3]. If initial therapy, no matter how aggressive, is not successful, then there will be no extended survival during which late toxicities can occur.

M

One particular area of controversy in radiotherapeutic treatment of brain tumors is the question of treatment-induced neurocognitive deﬁcit. Unfortunately, the prognosis of highgrade astrocytoma, particularly unresectable astrocytoma, remains poor enough that late toxicities of aggressive radiotherapy are seldom an issue, and surviving populations are small enough to be difﬁcult to evaluate. However, most patients with benign meningioma are expected to survive for decades, and the possibility of such complications is a concern. Guidance for this group of patients can be obtained by also considering outcomes in other slow-growing neurologic tumors, such as low-grade glioma. Some series report that neurocognitive deﬁcit is not a signiﬁcant problem at all. Kleinberg [4] reported maintenance of performance status in a series of 30 patients with aggressive astrocytomas treated with cranial radiotherapy. However, median follow-up in this series was only 3.5 years. On the other hand, Surma-aho [5] compared 28 patients with low-grade glioma who had postoperative radiotherapy (whole-brain radiotherapy in 19 patients) at a mean follow-up time of 7 years and 23 patients who did not have postoperative radiotherapy at a mean follow-up time of 10 years and found that patients who received radiotherapy had a greater decrease in memory and performance status. Cull [6] suggested an explanation for these differences by comparing a series of glioma patients for whom at least 4-year follow-up was available. Sixteen of these patients received whole-brain irradiation and 14 received focused irradiation. The patients who received whole-brain irradiation had a greater neuropsychometric deﬁcit in terms of visuospatial organization, visual memory, and complex information processing, suggesting that the amount of brain radiated could be a key variable. This was indirectly supported by Brown [7], who evaluated 203 patients receiving focal radiotherapy for low-grade glioma and found neuropsychiatric deterioration in only 5% at 5 years. In the setting of brain metastases or high-grade brain tumor, whole-brain irradiation must be considered to avoid the possibility of overlapping radiotherapy ﬁelds if there should be further metastasis or tumor extension. However, the geographic localization of meningioma would generally allow focal irradiation. This reasoning could also be extrapolated to suggest a lesser chance of neurocognitive deﬁcit with the highly localized administration of radiosurgery. If meningioma recurs or progresses in spite of local therapy or if local therapy is not technically feasible, then systemic

271

272

s. grunberg

therapy must be considered. Due to the preferential effect on rapidly growing tissue and the possibility of cumulative toxicity after extended treatment, cytotoxic chemotherapy is more suited to the rare malignant meningioma. Occasional success has been seen using regimens modeled on those administered to patients with sarcomas and including agents such as doxorubicin and dacarbazine [8]. A single case report by Travitzky [9] described response of a metastatic meningioma to an extended course of Doxil (Ortho Biotech, Bridgewater, NJ), a liposomal formulation of doxorubicin. Recently temozolomide, an oral chemotherapeutic agent with good CNS penetrance and a modest toxicity proﬁle, has gained popularity in the treatment of a number of different types of brain tumors. However, Chamberlain [10] treated a series of 16 patients with refractory meningioma with low-dose daily oral temozolomide and observed rapid progression of disease with no response. One agent that has maintained a continuing level of modest interest is hydroxyurea, an oral chemotherapeutic agent with mild myelosuppression as its most common toxicity. Schrell [11] originally reported dramatic responses in three patients with recurrent meningioma. This was supported by cell culture studies in 20 meningioma cell lines suggesting apoptotic changes after treatment with hydroxyurea [12]. However, results from later series using hydroxyurea have been more restrained. Mason [13] reported stabilization of tumor (but not shrinkage) in 12 of 20 patients with recurrent or residual benign meningioma, and Newton [14] reported similar results in 18 of 20 patients. Paus [15] reported marked improvement of visual ﬁelds without tumor shrinkage in a patient with an optic nerve meningioma. However, Loven [16] treated 12 patients with meningioma and did not believe that hydroxyurea was efﬁcacious in inducing response or maintaining freedom from progression. Whether this agent really has clinically signiﬁcant activity remains to be seen. Attention has therefore turned to biologics, hormonal therapy, and targeted therapy as better tolerated and perhaps more efﬁcacious systemic alternatives. Interferon-alpha has been shown to inhibit the response of meningioma cell cultures to various growth stimuli [17]. The possibility of a clinically signiﬁcant therapeutic effect was further supported by observation of decreased tracer uptake by responding meningiomas in 12 patients receiving interferon-alpha and being followed with [11C]l-methionine PET scan [18]. However, Kaba [19] reported only stabilization lasting 6 to 14 months in 6 patients with recurrent malignant or unresectable meningiomas treated with interferon-alpha-2B. Meningioma cell lines have been reported to produce interleukin-6, which acts as a growth factor and leads to increased terminal cell density [20]. Cyclooxygenase-2 is expressed by meningioma cells, particularly more atypical or malignant meningiomas [21], and meningioma cell lines are sensitive to cyclooxygenase-2 inhibitors [22]. Meningiomas also commonly express somatostatin receptors, raising the possibility that somatostatin analogues could have a therapeutic role [23]. However, clinical activity of these agents has not yet been demonstrated. One of the more interesting potential systemic interventions for meningioma is based on the relationship of these tumors to sex hormone status. Meningiomas are more common in women [24], are associated with breast cancer [25], and may

wax and wane with menstrual periods or with pregnancy [26]. These epidemiologic observations led to the detection of estrogen and progesterone receptors on meningiomas. However, unlike breast cancer, meningiomas are usually progesterone receptor positive but only occasionally estrogen receptor positive [27]. The logical intervention would therefore be treatment with a progesterone receptor antagonist such as mifepristone, which was shown to have promising activity in preclinical models [28, 29]. Although its common role as an abortifacient requires only a single day of use, long-term mifepristone has been administered to patients with meningioma. A series of 28 patients with unresectable meningioma received daily mifepristone for up to 13 years with good drug tolerance [30]. Eight of these patients had a slight decrease in tumor size or improvement in visual ﬁelds. These encouraging results led to a phase III, double-blind, placebo-controlled trial of mifepristone in patients with documented progressive meningioma [31]. Treatment with mifepristone in this study did not increase time to further progression. However, subset analysis of premenopausal, postmenopausal, and male patients may lead to further insights into potential mechanisms of response. Another possible target for growth modulation of meningiomas is the epidermal growth factor receptor (EGFR). Several investigators [32–34] have detected EGFR on a majority (60% to 100%) of meningiomas assayed for this receptor. The detected receptor is physiologically active [35] and can induce epidermal growth factor (EGF)-stimulated kinase autophosphorylating activity [36]. Camby [37] also found that in some meningioma cell lines, addition of EGF would have a measurable proliferative effect. These ﬁndings are now particularly important due to the clinical development of EGFR inhibitors that target the EGFR receptor (cetuximab) or EGFR tyrosine kinase (geﬁtinib, erlotinib). Clinical trials of these agents in meningioma have not yet been reported but are eagerly awaited. The EGFR pathway may have relevance for other potential therapeutic maneuvers as well. Dirven [38] demonstrated that redirection of the adenovirus type 5 vector from the coxsackievirus and adenovirus receptor to the epidermal growth factor receptor resulted in a threefold enhancement of gene transfer to primary meningioma cell lines. Koper [39], investigating the epidermal growth factor and progesterone receptor pathways, found that progesterone enhanced the sensitivity of meningioma cell lines to the growth stimulatory effects of epidermal growth factor, and addition of the progesterone antagonist mifepristone counteracted this effect. This observation raises the possibility that combined use of an EGFR inhibitor (geﬁtinib, erlotinib, cetuximab) and a progesterone antagonist (mifepristone) might be clinically effective in reducing meningioma growth even in the absence of clinically signiﬁcant activity of either agent alone. Local therapies, such as surgery and radiosurgery, are still the mainstays of treatment for benign meningioma. However, for those patients in whom recurrent or residual tumor remains, strategies for long-term tumor suppression will be necessary. Although cytotoxic chemotherapy has not had a major impact on the treatment of benign meningioma, newer treatment strategies based on biologic therapy, hormonal modulation, or signal transduction targeting may fulﬁll this therapeutic need.

22.

meningioma: systemic therapy perspective

References 1. Macfarlane DA. Cancer of the adrenal cortex. Ann Roy Coll Surg Engl 1958; 23:155–186. 2. Qureshi AI. Endovascular treatment of cerebrovascular diseases and intracranial neoplasms. Lancet 2004; 363:804–813. 3. Grunberg SM, Burris H 3rd, Livingston R. The price of success. J Clin Oncol 1995; 13:797–798. 4. Kleinberg L, Wallner K, Malkin MG. Good performance status of long-term disease-free survivors of intracranial gliomas. Int J Radiat Oncol Biol Phys 1993; 26:129–133. 5. Surma-aho O, Niemelä M, Vilkki J, et al. Adverse long-term effects of brain radiotherapy in adult low-grade glioma patients. Neurology 2001; 56:1285–1290. 6. Gregor A, Cull A, Traynor E, et al. Neuropsychometric evaluation of long-term survivors of adult brain tumours: relationship with tumour and treatment parameters. Radiother Oncol 1996; 41: 55–59. 7. Brown PD, Buckner JC, O’Fallon JR, et al. Effects of radiotherapy on cognitive function in patients with low-grade glioma measured by the Folstein Mini-Mental State Examination. J Clin Oncol 2003; 21:2519–2524. 8. Kyritsis AP. Chemotherapy for meningiomas. J Neurooncol 1996; 29:269–272. 9. Travitzky M, Libson E, Nemirovsky I, et al. Doxil-induced regression of pleuro-pulmonary metastases in a patient with malignant meningioma. Anticancer Drugs 2003; 14:247–250. 10. Chamberlain MC, Tsao-Wei DD, Groshen S. Temozolomide for treatment-resistant recurrent meningioma. Neurology 2004; 62:1210–1212. 11. Schrell UM, Rittig MG, Anders M, et al. Hydroxyurea for treatment of unresectable and recurrent meningiomas. II. Decrease in the size of meningiomas in patients treated with hydroxyurea. J Neurosurg 1997; 86:840–844. 12. Schrell UM, Rittig MG, Anders M, et al. Hydroxyurea for treatment of unresectable and recurrent meningiomas. I. Inhibition of primary human meningioma cells in culture and in meningioma transplants by induction of the apoptotic pathway. J Neurosurg 1997; 86:845–852. 13. Mason WP, Gentili F, Macdonald DR, et al. Stabilization of disease progression by hydroxyurea in patients with recurrent or unresectable meningioma. J Neurosurg 2002; 97:341–346. 14. Newton HB, Scott SR, Volpi C. Hydroxyurea chemotherapy for meningiomas: enlarged cohort with extended follow-up. Br J Neurosurg 2004; 18:495–499. 15. Paus S, Klockgether T, Schlegel U, et al. Meningioma of the optic nerve sheath: treatment with hydroxyurea. J Neurol Neurosurg Psychiatry 2003; 74:1348–1350. 16. Loven D, Hardoff R, Sever ZB, et al. Non-resectable slow-growing meningiomas treated by hydroxyurea. J Neurooncol 2004; 67: 221–226. 17. Koper JW, Zwarthoff EC, Hagemeijer A, et al. Inhibition of the growth of cultured human meningioma cells by recombinant interferon-alpha. Eur J Cancer 1991; 27:416–419. 18. Muhr C, Gudjonsson O, Lilja A, et al. Meningioma treated with interferon-alpha, evaluated with [(11)C]-L-methionine positron emission tomography. Clin Cancer Res 2001; 7:2269–2276. 19. Kaba SE, DeMonte F, Bruner JM, et al. The treatment of recurrent unresectable and malignant meningiomas with interferon alpha-2B. Neurosurgery 1997; 20:271–275. 20. Boyle-Walsh E, Hashim IA, Speirs V, et al. Interleukin-6 (IL-6) production and cell growth of cultured human meningiomas: interactions with interleukin-1 beta (IL-1 beta) and interleukin-4 (IL-4) in vitro. Neurosci Lett 1994; 170:129–132.

273

21. Lin CC, Kenyon L, Hyslop T, et al. Cyclooxygenase-2 (COX-2) expression in human meningioma as a function of tumor grade. Am J Clin Oncol 2003; 26:S98–102. 22. Nathoo N, Barnett GH, Golubic M. The eicosanoid cascade: possible role in gliomas and meningiomas. J Clin Pathol 2004; 57: 6–13. 23. Pistolesi S, Fontanini G, Boldrini L, et al. The role of somatostatin in vasogenic meningioma associated brain edema. Tumori 2003; 89:136–140. 24. Quest DO. Meningiomas: an update. Neurosurgery 1978; 3:219– 225. 25. Schoenberg BS, Christine BW, Wisnant JP. Nervous system neoplasms and primary malignancies of other sites. The unique association between meningiomas and breast cancer. Neurology 1974; 25:705–712. 26. Cushing H, Eisenhardt L. Meningiomas arising from the tuberculum sellae with the syndrome of primary optic atrophy and bitemporal ﬁeld defects combined with a normal sella turcica in a middle-aged person. Arch Ophthalmol 1929; 1:1–41. 27. Grunberg SM. The role of progesterone receptors in meningioma. In: Muggia FM, ed. New drugs, concepts and results in cancer chemotherapy. Boston: Kluwer Academic Publishers, 1992:127– 137. 28. Olsson JJ, Beck DW, Schlechte J, et al. Hormonal manipulation of meningiomas in vitro. J Neurosurg 1986; 65:99–107. 29. Olsson JJ, Beck DW, Schlechte JA, et al. Effect of the antiprogesterone RU-38486 on meningioma implanted into nude mice. J Neurosurg 1987; 66:584–587. 30. Grunberg SM, Weiss MH, Russell CA, et al. Long-term administration of mifepristone (RU486): clinical tolerance during extended treatment of meningioma. Cancer Invest 2006; 24:727– 733. 31. Grunberg SM, Rankin C, Townsend J, et al. Phase III doubleblind randomized placebo-controlled study of mifepristone for the treatment of unresectable meningioma. Proc Am Soc Clin Oncol 2001; 20:56A. 32. Kuratsu JI, Seto H, Kochi M, et al. Expression of PDGF, PDGFreceptor, EGF-receptor and sex hormone receptors on meningioma. Acta Neurochir (Wien) 1994; 131:289–293. 33. Andersson U, Guo D, Malmer B, et al. Epidermal growth factor receptor family (EGFR, ErbB2–4) in gliomas and meningiomas. Acta Neuropathol (Berl) 2004; 108:135–142. 34. Torp SH, Helseth E, Dalen A, et al. Expression of epidermal growth factor receptor in human meningiomas and meningeal tissue. APMIS 1992; 100:797–802. 35. Carroll RS, Black PM, Zhang J, et al. Expression and activation of epidermal growth factor receptors in meningiomas. J Neurosurg 1997; 87:315–323. 36. Sanﬁlippo JS, Rao CV, Guarnaschelli JJ, et al. Detection of epidermal growth factor and transforming growth factor alpha protein in meningiomas and other tumors of the central nervous system in human beings. Surg Gynecol Obstet 1993; 177:488–496. 37. Camby I, Nagy N, Rombaut K, et al. Inﬂuence of epidermal growth factor and gastrin on the cell proliferation of human meningiomas versus astrocytic tumors maintained as ex vivo tissue cultures. Neuropeptides 1997; 31:217–225. 38. Dirven CM, Grill J, Lamfers ML, et al. Gene therapy for meningioma: improved gene delivery with targeted adenoviruses. J Neurosurg 2002; 97:441–449. 39. Koper JW, Lamberts SWJ. Meningiomas, epidermal growth factor and progesterone. Human Reprod 1994; 1(9 Suppl):190–193. 40. Grunberg SM, Weiss MH, Russell CA, et al. Long-term administration of mifepristone (RU486): clinical tolerance during extended treatment of meningioma. Cancer Invest 2006; 24:727–733.

2 3

Acoustic Schwannoma William M. Mendenhall, Robert J. Amdur, Robert S. Malyapa, and William A. Friedman

Introduction Acoustic schwannomas are benign tumors that originate from Schwann cells surrounding the vestibular (eighth) nerve, usually within the internal auditory canal [1]. As the tumor enlarges, it ﬁlls the canal and extends into the cerebellopontine angle. Acoustic schwannomas are relatively uncommon with an incidence of approximately 1 new case per 100,000 [1]. The mean age at diagnosis is approximately 45 to 47 years, and there is a slight female preponderance [2]. Approximately 2% to 4% of all patients with acoustic schwannomas have neuroﬁbromatosis type 2 (NF2), which occurs in 1 per 50,000 of the general population [3]. NF2 is autosomal dominant and is characterized by the development of bilateral acoustic schwannomas, intracranial meningiomas, spinal root and peripheral schwannomas, and presenile lens opacities. Acoustic schwannomas usually grow slowly (approximately 1 mm per year) and patients usually present with the gradual onset of sensorineural hearing loss [1, 2]. Alternative presenting complaints or accompanying symptoms may include tinnitus, dizziness, ataxia, vertigo, and a sensation of fullness in the ear [1]. As the tumor progresses, the patient may experience numbness, headaches, diplopia, loss of coordination, difﬁculty swallowing, and facial numbness [1]. Matthies and Samii [2] reported a series of 962 patients with 1000 acoustic schwannomas treated surgically and observed the following neurologic deﬁcits at presentation: acoustic (95%), vestibular (61%), trigeminal (9%), and facial (6%). The duration of hearing loss before diagnosis is approximately 3 to 4 years [2]. Large tumors with brain-stem compression often present in young patients with a shorter duration of symptoms [2].

Treatment Options The treatment of patients with acoustic schwannomas is controversial. Options include observation, microsurgery, stereotactic radiosurgery, and radiotherapy. The term radiosurgery refers to the delivery of a single fraction of radiotherapy using stereotactic techniques to localize the tumor and align the ﬁelds. Radiosurgery may be administered with a Gamma Knife, a

linear accelerator (linac)-based system, or a proton beam. Radiotherapy refers to the administration of fractionated irradiation and includes stereotactic radiotherapy, intensitymodulated radiation therapy (IMRT), three-dimensional conformal radiotherapy, and conventional radiotherapy techniques such as a “wedge pair” ﬁeld arrangement. The prescription isodose line is at the radiographically deﬁned margin of the tumor for patients treated with radiosurgery, whereas a small margin (approximately 5 mm) of apparently normal surrounding tissue is included within the prescription isodose line for those treated with radiotherapy [4, 5]. Conventional radiotherapy is given at 1.8 to 2.0 Gy per fraction once daily, 5 days a week, to approximately 50 Gy in a continuous course. There is less experience with short hypofractionated courses of radiotherapy and, because the risk of complications increases with dose per fraction, tighter margins should be employed.

End Points The primary outcome measure for the treatment of acoustic schwannomas is local control. The deﬁnition of local control depends on the treatment that is used. Local control after microsurgery implies that the tumor has been completely resected and that there is no evidence of residual tumor on follow-up radiographic studies such as magnetic resonance imaging (MRI). In contrast, local control after either radiosurgery or radiotherapy implies stabilization of tumor growth or regression with no evidence of progression on follow-up evaluations. Benign tumors rarely disappear completely after successful irradiation. As long as they do not progress, they will cause no new symptoms and will require no additional therapy. They are as effectively “locally controlled” as those that are completely resected [6–8]. Because a small subset of acoustic schwannomas may demonstrate growth many years after treatment, it is necessary to have long follow-up and to analyze local control using the Kaplan-Meier product-limit method [9–11]. Secondary outcome measures include treatment-related morbidity, such as facial weakness, and hearing preservation. Survival is not a particularly useful end point to evaluate the efﬁcacy of treatment because very few patients die of disease, irrespective of primary treatment.

275

276

w.m. mendenhall et al.

Rationale for Treatment Alternatives and the Case for Radiosurgery Because acoustic schwannomas are benign, slow-growing tumors, they may initially be observed in asymptomatic patients. However, symptomatic patients as well as those whose lesions exhibit progression on serial MRI require treatment to prevent morbidity associated with tumor growth. The goal of treatment is to prevent the symptoms associated with tumor progression from becoming worse. Although the likelihood of local control exceeds 90% after any of the treatment alternatives, many of the symptoms caused by the tumor may remain stable after successful treatment and some (particularly hearing loss) may worsen.

Observation Patients who are of advanced age or medically inﬁrm with small tumors at presentation may be observed and deﬁnitive treatment withheld until tumor progression is observed on serial imaging studies [12–14]. Measurement of auditory and vestibular function correlates poorly with tumor growth [15, 16]. Strasnick et al. [12] reported a group of 51 patients selected for observation between 1979 and 1992. Patients had a mean tumor size of 1.03 cm (range, 0.2 to 2.8 cm) at initial examination and had follow-up for an average of 2.6 years (range, 6 months to 11 years). Fifty patients had adequate radiographic followup. The mean annual tumor growth rate was 0.11 cm per year (range, 0 to 1.1 cm per year). Eleven patients experienced a tumor growth rate of more than 0.2 cm per year; nine patients eventually required surgery, one was treated with stereotactic radiosurgery, and one refused treatment. In contrast, 39 patients demonstrated a tumor growth rate of less than 0.2 cm per year, and only two subsequently required surgical intervention. Wiet et al. [13] reported 78 patients with acoustic schwannomas who were selected for observation between 1981 and 1994. Patients were excluded from analysis if they had neuroﬁbromatosis, previous treatment, or inadequate radiographic follow-up. The mean tumor size for the 53 patients eligible for analysis was 0.98 cm (range, 0.2 to 3.0 cm). Follow-up averaged 25.8 months (range, 5 to 99 months). The average annual tumor growth rate was 0.16 cm per year (range, 0 to 1.64 cm per year). The authors noted no relationship between tumor size at presentation and growth rate. Fourteen patients (26%) eventually required surgical intervention, four were treated with radiotherapy, and one patient refused recommended therapy. Four additional patients were lost to follow-up. Abaza et al. [3] reported 22 patients with NF2 and acoustic schwannomas who had follow-up with serial MRI for an average of 3.71 years (range, 4 months to 9 years). The average growth rate was 0.30 cm3 per year and was higher in patients age 35 or more and those with concomitant spinal tumors. Rosenberg [15] studied the growth of acoustic schwannomas in 70 patients older than 65 years. Mean follow-up was 4.8 years. Only four (5.7%) patients underwent treatment. In contrast, 18 (26%) patients died from unrelated causes. One must recall that patients selected for observation represent a small subset of the patient population with acoustic schwannomas, generally older and less healthy, and that the growth characteristics of these tumors may not reﬂect the

growth rate in the overall patient population. Additionally, delays in treatment after initial observation may result in loss of useful hearing or make treatment more difﬁcult because of tumor progression. Conversely, observation may obviate the need for intervention in cases of nonneoplastic enhancement of the vestibular nerve (e.g., vestibular neuritis), which has been reported in the range of 3% of acoustic tumors [17].

Surgery Surgery has been considered the gold standard in the treatment of acoustic schwannomas because of its long track record and the low risk of recurrence after an apparent complete resection [18–20]. However, many variables must be considered when evaluating primary and secondary outcomes with microsurgery. First, there are the three standard approaches—suboccipital, middle fossa, and translabyrinthine—and each has its unique advantages and disadvantages. For example, the fundus of the internal auditory canal is incompletely visualized during either suboccipital or middle fossa approach with attempted hearing preservation, this because of the need to preserve the bony capsule of the inner ear. This translates to a higher potential for incomplete resection [21, 22]. The translabyrinthine approach eliminates this risk, but hearing is sacriﬁced. Second, there is a learning curve with microsurgery for acoustic schwannoma [23–25]. The less experienced the surgeon, the lower the rates of success. Third, the larger the tumor, the lower the chance of success, at least without concomitant morbidity, such as facial palsy [25]. This is particularly important when comparing results of microsurgery against radiosurgery because larger tumors are more likely to be treated surgically. Radiosurgery is generally not indicated for tumors larger than 3 cm, which are the most difﬁcult tumors to completely resect and control locally. Finally, technological advancements in nerve monitoring (especially the facial nerve) and surgical optics (e.g., microscopes and endoscopes) have improved outcomes over results from even 10 years earlier [18, 26–28]. Gormley et al. [29] reported 179 patients who underwent surgery at the University of Pittsburgh Medical Center (Pittsburgh, PA) and George Washington University Medical Center (Washington, DC) between 1985 and 1996 and had follow-up from 3 to 171 months (mean, 70 months). Complete resection without evidence of subsequent recurrence was observed in 178 (99%) patients. Wiegand et al. [1] reported 1579 surgically treated patients who were entered into the Acoustic Neuroma Association Registry between 1989 and 1994. Recurrent or persistent tumor was observed on follow-up studies in 7.8% of patients. Gjuric and colleagues [30] reported on the success of the extended middle fossa approach for 735 unilateral acoustic schwannomas, less than 3 cm, treated from 1975 through 1998. Total tumor removal was achieved in 97%, and residual tumor was later identiﬁed in 0.3% of patients thought to have undergone complete resection. The local control rates in some of the surgical series cited are expressed as crude percentages without specifying the number of patients lost to follow-up or using the actuarial product-limit method to specify the probability of local control at a speciﬁc point in time. Additionally, the high rates of local control reported in the surgical literature are usually those of experienced neurootologic surgeons and may not be consistent

23.

acoustic schwannoma

with results obtained in a community setting by less experienced physicians. Samii and Matthies [31] reported on 962 patients who were treated surgically for 1000 acoustic schwannomas between 1978 and 1993. Anatomic preservation of the facial nerve was accomplished in 93% of patients. House-Brackman grade, 2 to 8 weeks postoperatively for those in whom the facial nerve was preserved, was as follows: grade 1, 51%; grade 2, 13%; grade 3, 15%; grade 4, 6%; grade 5, 11%; and grade 6, 4%. Major neurologic complications included tetraparesis (0.1%), hemiparesis (1.0%), and caudal cranial nerve palsies (5.5%). Surgical complications included hematoma (2.2%), cerebrospinal ﬂuid ﬁstulas (9.2%), hydrocephalus (2.3%), bacterial meningitis (1.2%), and wound revisions (1.1%). Eleven (1.1%) patients died 2 to 69 days after surgery. Medical complications (i.e., pneumonia, cardiac complications, etc.), time required for rehabilitation, and interval to return to work were not discussed. Buchman et al. [32] analyzed the ﬁrst 96 patients treated surgically by a new neurootologic team at the University of Pittsburgh School of Medicine. Approximately 60 cases were required before the new team achieved results comparable with those of an experienced team relative to preservation of facial nerve function. The ability to obtain postoperative HouseBrackman grade 1 or 2 facial nerve function was signiﬁcantly related to experience (p < 0.0003). Surgical experience was also related to the likelihood of achieving a complete resection and hearing preservation, as well as reducing the risk of cerebrospinal ﬂuid leaks. Welling and colleagues [24] found similar results, but only 20 cases were required to achieve the requisite experience. A more recent study by Kaylie and colleagues [33] reported grade 1 or 2 facial nerve function in 95% of their smaller tumors treated during the ﬁrst 6 years of a new surgical team’s experience. Sixty-nine of 179 (39%) patients reported by Gormley et al. [29] had functional hearing before surgery. The likelihood of postoperative hearing preservation versus tumor diameter in the cerebellopontine angle was as follows: <2 cm, 48%; 2 to 3.9 cm, 25%; and ≥4 cm, 0%. Sixty of 364 (16%) patients treated surgically by the Chicago Otology Group were candidates for hearing preservation surgery; the average tumor size was 1.4 cm [34]. The rates of hearing preservation were as follows: ≤1.5 cm diameter tumor, 50%; >1.5 cm diameter tumor, 16%; and overall, 37%. Early postoperative preservation of hearing does not imply long-term preservation of hearing. Shelton et al. [23] reported 25 patients who had hearing preservation after resection via the middle cranial fossa approach at the Otologic Medical Group (Los Angeles). The authors noted a signiﬁcant loss of hearing in the ipsilateral ear in 14 of 25 (56%) patients. The mean loss of speech discrimination was 25%, and the mean loss of speech reception threshold was 12 dB. The loss of hearing was proportional to length of follow-up. None had evidence of a local recurrence.

Radiotherapy The rationale for radiotherapy is to reduce the dose per fraction from one large single dose (as is employed with radiosurgery) and administer irradiation in multiple smaller doses protracted over several weeks to reduce the risk of long-term complications. The main risks associated with radiosurgery are damage

277

to the ﬁfth, seventh, and eighth nerves. The probability of ﬁfth or seventh nerve injury after radiosurgery is very low (1% or 2%) with the techniques and doses currently employed. Therefore, the primary potential advantage of radiotherapy compared with radiosurgery is hearing preservation. Radiotherapy is much like cooking, and fractionation schedules may vary widely. Conventional radiotherapy consists of 1.8- to 2.0-Gy once-daily fractions, 5 days a week, to a total dose of 45 to 50 Gy. Although formulae exist to predict local control and complication probabilities after various fractionation schedules, it is risky to employ a schedule for which there is no track record. Shirato et al. [35] reported on 65 patients treated with stereotactic radiotherapy between 1991 and 1999 and had followup for more than 6 months. Patients received 36 to 50 Gy in 20 to 25 fractions speciﬁed at isocenter with the 80% isodose line at the periphery of the tumor. The local control rate at 5 years was 92%. Fifty-one patients were treated with stereotactic radiotherapy at the University of Heidelberg (Germany) between 1989 and 1999 and had follow-up for an average of 42 months [36]. The mean dose was 57.6 Gy at 1.8 to 2.0 Gy per fraction. Ten of 51 (20%) patients had NF2. The 5-year actuarial local control rate was 95%. Andrews et al. [37] reported on 125 patients treated with Gamma Knife radiosurgery (69 patients) or linac-based stereotactic radiotherapy (56 patients) at Thomas Jefferson University between 1994 and 2000. Three patients were lost to follow-up; the remaining patients had follow-up for an average of 119 weeks (radiosurgery group) or 115 weeks (radiotherapy group). The local control rates for radiosurgery and radiotherapy were as follows: sporadic tumors, 98% and 97%; and NF2 patients, 80% and 67% (p = 0.6615), respectively. The differences in the likelihood of local control between these two modalities were not statistically signiﬁcant. Szumacher et al. [5] reported no new trigeminal or facial nerve palsies among 39 patients who received 50 Gy in 25 fractions over 5 weeks. Shirato et al. [35] reported on 65 patients who received 36 to 50 Gy in 20 to 25 fractions at the Hokkaido University School of Medicine (Sapporo, Japan); no patient experienced a permanent trigeminal or facial nerve injury. Thirty patients with 31 acoustic schwannomas received 54 to 60 Gy in 30 to 33 fractions with proton beam radiotherapy at the Loma Linda University Medical Center (California); no transient or permanent trigeminal or facial nerve injuries were observed [38]. Szumacher et al. [5] reported that 19 of 28 (68%) patients who received 50 Gy in 25 fractions retained serviceable hearing after radiotherapy. Sakamoto et al. [39] analyzed the hearing before and after radiotherapy in 21 patients who received 36 to 50 Gy in 20 to 25 fractions. The mean annual rate of hearing loss was lower and was decreasing after radiotherapy compared with the rate of hearing loss before treatment. They also found that a tapered 2-week course of oral prednisone improved hearing loss after radiotherapy [40]. Andrews et al. [37] compared the rates of hearing preservation in patients treated with either radiosurgery or radiotherapy at Thomas Jefferson University. Patients treated with radiosurgery received approximately 12 Gy speciﬁed at the 50% isodose line, and those treated with radiotherapy received 50 Gy in 25 fractions over 5 weeks. There was a 2.5-fold higher likelihood of hearing

278

w.m. mendenhall et al. Treatment Algorithm Asymptomatic

Symptomatic

Re-scan in 6 Months Treat No Progression

Continue Follow-up

Progression Minimum Diameter Maximum Diameter <3 cm ≥ 3 cm Radiosurgery

Microsurgery

FIGURE 23-1. Treatment algorithm.

preservation after radiotherapy. Maintenance of serviceable hearing in patients with sporadic tumors was 81% after radiotherapy versus 33% after radiosurgery (p = 0.0228). The question of whether radiotherapy results in improved hearing compared with radiosurgery remains unresolved. Data sited elsewhere in this chapter indicate that radiosurgery at the current recommended dose levels probably results in comparable hearing preservation rates.

The Case for Radiosurgery Radiosurgery and radiotherapy result in local control rates that are similar to those observed after complete excision. The potential for complications after microsurgery is higher than after either radiosurgery or radiotherapy [41]. Myrseth et al. [42] recently reported on 189 patients treated with either microsurgery (86 patients) or Gamma Knife radiosurgery (103 patients) and had follow-up for an average of 5.9 years. Two standardized questionnaires, the Glasgow Beneﬁt Inventory and Short-Form 36, were sent to 168 living patients, and 83% responded. Posttreatment facial nerve function, hearing, complications rates, and quality of life were all signiﬁcantly better for those patients treated with radiosurgery. The risk of a trigeminal and/or facial nerve injury has decreased signiﬁcantly after radiosurgery at current dose levels (approximately 12.5 Gy). Although some data suggest that patients with serviceable hearing at presentation may have a higher probability of hearing preservation after radiotherapy compared with radiosurgery, other data suggest that the likelihood of hearing preservation is similar. Our current recommendations are to offer the options of microsurgery and radiosurgery to patients in whom treatment

FIGURE 23-2. Dose distribution for a patient treated with radiosurgery at the University of Florida. The patient was treated with a threeisocenter plan with the dose speciﬁed at the 70% line. The isodose lines shown are the 70%, 50%, and 20% lines.

is indicated (Fig. 23-1). Microsurgery is the treatment of choice for larger tumors (≥3 cm) and for salvage of the small subset of patients who experience progression after treatment with irradiation. Although stereotactic radiotherapy may be considered for the small subset of patients who present with serviceable hearing in whom the inconvenience of fractionated radiotherapy may be worth the potential increased likelihood of hearing preservation, in practice, the vast majority of patients with tumors <3 cm are treated with radiosurgery.

Treatment Planning Patients undergo an image fusion magnetic resonance imaging (MRI) and a decision is made whether to proceed with radiosurgery. Thereafter, a stereotactic frame is placed followed by computed tomography (CT), which is then fused with the MRI. Sphere packing is used to create a dose distribution that precisely covers the tumor with additional margin so that no normal tissue is included in the high-dose isodose line. The dose is 12.5 Gy regardless of tumor size with the dose speciﬁed at the 80% isodose line for one isocenter plan and the 70% line for a ≥2 isocenter plan. The vast majority of patients require multiple isocenters (Fig. 23-2). Treatment is administered with 6-MV photons and requires about 15 to 20 minutes per isocenter. After completion of radiosurgery, the frame is removed, and the patient is seen annually for follow-up, which consists of a physical examination and an MRI (Fig. 23-3).

FIGURE 23-3. (A) Patient with a right acoustic schwannoma prior to treatment. The patient received 12.5 Gy to the periphery of the tumor. (B) Follow-up MRI 10 years after radiosurgery shows signiﬁcant regression with no evidence of regrowth.

23.

279

acoustic schwannoma

Outcomes After Radiosurgery Between 1988 and 2005, 390 patients were treated with linear accelerator–based radiosurgery for acoustic schwannomas at the University of Florida [43]. Mean follow-up was 40 months; 63 patients had follow-up for more than 5 years. The 2-, 5-, and 10-year local control rates were 98%, 90%, and 90%, respectively (Table 23-1) [43–45]. Three (1%) patients required surgery because of tumor progression. Seventeen (4.4%) patients experienced seventh nerve paresis, and 14 (3.6%) patients had ﬁfth nerve paresis after treatment. Multivariate analysis of ﬁfth and seventh nerve function after radiosurgery revealed that increasing tumor volume and increasing dose were signiﬁcantly related to an increased risk of paresis. Since 1994, when the dose was lowered to 12.5 Gy, the incidence of ﬁfth and seventh nerve paresis has been 0.7%. Lunsford et al. [46] reported on 829 patients treated with Gamma Knife radiosurgery between 1987 and 2002 at the University of Pittsburgh. Since 1992, 12.5 to 13 Gy speciﬁed at the 50% isodose line has been employed. Tumor control was deﬁned as no additional treatment required and was 97% at 10 years. There was no improvement in local control for doses greater than 13 Gy compared with doses between 12.5 and 13 Gy. No patient experienced a radiation-induced tumor. For the subset of 313 patients treated for unilateral tumors between 1991 and 2001 with doses ranging from 12 to 13 Gy, the 6-year resection-free survival rate was 98.6% [47]. Facial nerve preservation was achieved in 100% of patients, trigeminal nerve function was maintained in 96%, hearing level remained stable in 70%, and preservation of useful hearing was achieved in 79% of patients. Hasagawa et al. [44] reported on 301 patients treated with Gamma Knife radiosurgery between 1991 and 1998 and who had serial follow-up imaging. The 5- and 10-year progression-free survival rates were 93% and 92%, respectively. For patients who received a marginal tumor dose of 13 Gy or less, hearing preservation was achieved in 68%, transient facial nerve injury occurred in 1%, and trigeminal nerve injury was observed in 2% of patients. Weber et al. [45] reported on 88 patients treated with proton beam radiosurgery at the Harvard Cyclotron Laboratory between 1992 and 2000 and had follow-up from 12 to 103 months (median, 39 months). Patients were treated to a median dose of 12 cobalt Gray equivalents (range, 10 to 18 cobalt Gray equivalents) with the dose speciﬁed at the 70% to 108% isodose lines (median, 70%). The 5-year local control rate was 94%. Seven of 21 (33%) patients with Gardner-Robertson grade 1 to 2 pretreatment hearing retained serviceable hearing. The 5-year normal facial and trigeminal nerve preservation rates were 91% and 89%, respectively. No treatment-induced malignancies were observed.

The outcomes after radiosurgery using a linear accelerator–based system, Gamma Knife, or proton beam appear to be similar. Any differences in local control rates, complications, and hearing preservation likely reﬂect tumor extent, dose selection, and the experience of the treatment team rather than equipment. A marginal tumor dose of 12 to 13 Gy appears to offer a high probability of local control, a low risk of facial or trigeminal nerve injury, and a relatively high likelihood of retaining serviceable hearing. Doses in this range result in outcomes similar to those observed after fractionated radiotherapy.

Prognostic Factors Affecting Local Control There is no relationship between the likelihood of local control and tumor size for patients with sporadic unilateral acoustic schwannomas. Patients with NF2 have a lower probability of local control compared with those who are treated for sporadic tumors [48]. Noren et al. [48] reported on 224 acoustic schwannomas treated with Gamma Knife radiosurgery at the Karolinska Hospital (Stockholm) between 1969 and 1991. Follow-up ranged from 1 to 17.2 years. One hundred ninety-three procedures were performed for unilateral tumors and 61 procedures were performed on patients with von Recklinghausen neuroﬁbromatosis. The rates of local control were 94% for unilateral tumors and 84% for those associated with neuroﬁbromatosis.

Complication Avoidance and Management The main complications secondary to radiosurgery of acoustic schwannomas are facial weakness, facial numbness, and hearing loss. The most important parameters inﬂuencing the probability of these complications are radiosurgery dose and tumor volume [43, 49]. Patients who have undergone prior surgery are probably more likely to experience complications [43]. The optimal dose range that appears to offer a high likelihood of local control and a very low risk of complications is 12 to 13 Gy [43, 47, 50]. Reducing the risk of developing a complication is the most effective “management.” Although some cranial neuropathies that are observed after radiosurgery will resolve, there is no effective treatment for those that persist.

Future Directions Advances in the radiosurgical treatment of acoustic schwannomas will likely be due to improvements in treatment planning

TABLE 23-1. Outcomes after radiosurgery. Series

Institution

Friedman et al. [43] Hasegawa et al. [44] Weber et al. [45]

University of Florida Komaki City Hospital, Japan Massachusetts General Hospital

*301 patients with imaging follow-up. †

New permanent nerve deﬁcit.

No. patients

390 317 88

Radiosurgery modality

Local control, (interval)

Linac Gamma Knife Proton beam

90% (10 years) 92%* (10 years) 94% (5 years)

Facial weakness (%)

Facial numbness (%)

4.4 3.2 8.9†

3.6 2.8 10.1†

280

w.m. mendenhall et al.

software and the use of mini-multileaf collimators to improve the speed and “user friendliness” of both treatment planning and delivery.

References 1. Wiegand DA, Ojemann RG, Fickel V. Surgical treatment of acoustic neuroma (vestibular schwannoma) in the United States: report from the Acoustic Neuroma Registry. Laryngoscope 1996; 106:58–66. 2. Matthies C, Samii M. Management of 1000 vestibular schwannomas (acoustic neuromas): Clinical presentation. Neurosurgery 1997; 40:1–10. 3. Abaza MM, Makariou E, Armstrong M, Lalwani AK. Growth rate characteristics of acoustic neuromas associated with neuroﬁbromatosis type 2. Laryngoscope 1996; 106:694–699. 4. Friedman WA, Buatti JM, Bova FJ, Mendenhall WM. Linac Radiosurgery: A Practical Guide. New York: Springer-Verlag, 1998:1–176. 5. Szumacher E, Schwartz ML, Tsao M, et al. 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; 53:987–991. 6. Mendenhall WM, Amdur RJ, Hinerman RW, et al. Radiotherapy and radiosurgery for skull base tumors. Otolaryngol Clin North Am 2001; 34:1065–1077. 7. Condra KS, Buatti JM, Mendenhall WM, et al. Benign meningiomas: primary treatment selection affects survival. Int J Radiat Oncol Biol Phys 1997; 39:427–436. 8. Hinerman RW, Mendenhall WM, Amdur RJ, et al. Deﬁnitive radiotherapy in the management of chemodectomas arising in the temporal bone, carotid body, and glomus vagale. Head Neck 2001; 23:363–371. 9. Parsons JT, McCarty PJ, Rao PV, et al. On the deﬁnition of local control (Editorial). Int J Radiat Oncol Biol Phys 1990; 18: 705–706. 10. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53:457–481. 11. SAS Institute Inc. SAS OnlineDoc(r) version 9. Cary, NC: SAS Institute Inc., 2003. 12. Strasnick B, Glasscock ME III, Haynes D, et al. The natural history of untreated acoustic neuromas. Laryngoscope 1994; 104: 1115–1119. 13. Wiet RJ, Zappia JJ, Hecht CS, O’Connor CA. Conservative management of patients with small acoustic tumors. Laryngoscope 1995; 105:795–800. 14. Deen HG, Ebersold MJ, Harner SG, et al. Conservative management of acoustic neuroma: an outcome study. Neurosurgery 1996; 39:260–266. 15. Rosenberg SI. Natural history of acoustic neuromas. Laryngoscope 2000; 110:497–508. 16. Sakamoto T, Fukuda S, Inuyama Y. Hearing loss and growth rate of acoustic neuromas in follow-up observation policy. Auris Nasus Larynx 2001; 28:S23–S27. 17. Arriaga MA, Carrier D, Houston GD. False-positive magnetic resonance imaging of small internal auditory canal tumors: a clinical, radiologic, and pathologic correlation study. Otolaryngol Head Neck Surg 1995; 113:61–70. 18. Dandy WE. Results of removal of acoustic tumors by the unilateral approach. Arch Surg 1941; 42:1026–1033. 19. Harner SG, Beatty CW, Ebersold MJ. Retrosigmoid removal of acoustic neuroma: experience 1978–1988. Otolaryngol Head Neck Surg 1990; 103:40–45. 20. Rhoton AL Jr, Tedeschi H. Microsurgical anatomy of acoustic neuroma. Otolaryngol Clin North Am 1992; 25:257–294.

21. Roland PS, Meyerhoff WL, Wright CG, Mickey B. Anatomic considerations in the posterior approach to the internal auditory canal. Ann Otol Rhinol Laryngol 1988; 97:621–625. 22. Haberkamp TJ, Meyer GA, Fox M. Surgical exposure of the fundus of the internal auditory canal: Anatomic limits of the middle fossa versus the retrosigmoid transcanal approach. Laryngoscope 1998; 108:1190–1194. 23. Shelton C, Hitselberger WE, House WF, Brackmann DE. Hearing preservation after acoustic tumor removal: Long-term results. Laryngoscope 1990; 100:115–119. 24. Welling DB, Slater PW, Thomas RD, et al. The learning curve in vestibular schwannoma surgery. Am J Otolaryngol 1999; 20:644– 648. 25. Wiet RJ, Mamikoglu B, Odom L, Hoistad DL. Long-term results of the ﬁrst 500 cases of acoustic neuroma surgery. Otolaryngol Head Neck Surg 2001; 124:645–651. 26. Wackym PA, King WA, Poe DS, et al. Adjunctive use of endoscopy during acoustic neuroma surgery. Laryngoscope 1999; 109:1193–1201. 27. Morikawa M, Tamaki N, Nagashima T, Motooka Y. Long-term results of facial nerve function after acoustic neuroma surgery— clinical beneﬁt of intraoperative facial nerve monitoring. Kobe J Med Sci 2000; 46:113–124. 28. Sterkers JM, Morrison GA, Sterkers O, El-Dine MM. Preservation of facial, cochlear, and other nerve functions in acoustic neuroma treatment. Otolaryngol Head Neck Surg 1994; 110:146– 155. 29. Gormley WB, Sekhar LN, Wright DC, et al. Acoustic neuromas: results of current surgical management. Neurosurgery 1997; 41:50–60. 30. Gjuric M, Wigand ME, Wolf SR. Enlarged middle fossa vestibular schwannoma surgery: experience with 735 cases. Otol Neurotol 2001; 22:223–230. 31. Samii M, Matthies C. Management of 1000 vestibular schwannomas (acoustic neuromas): Surgical management and results with an emphasis on complications and how to avoid them. Neurosurgery 1997; 40:11–23. 32. Buchman CA, Chen DA, Flannagan P, et al. The learning curve for acoustic tumor surgery. Laryngoscope 1996; 106:1406– 1411. 33. Kaylie DM, Gilbert E, Horgan MA, et al. Acoustic neuroma surgery outcomes. Otol Neurotol 2001; 22:686–689. 34. Hecht CS, Honrubia VF, Wiet RJ, Sims HS. Hearing preservation after acoustic neuroma resection with tumor size used as a clinical prognosticator. Laryngoscope 1997; 107:1122–1126. 35. Shirato H, Sakamoto T, Takeichi N, et al. Fractionated stereotactic radiotherapy for vestibular schwannona (VS): comparison between cystic-type and solid-type VS. Int J Radiat Oncol Biol Phys 2000; 48:1395–1401. 36. Fuss M, Debus J, Lohr F, et al. Conventionally fractionated stereotactic radiotherapy (FSRT) for acoustic neuromas. Int J Radiat Oncol Biol Phys 2000; 48:1381–1387. 37. 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 Radiat Oncol Biol Phys 2001; 50:1265–1278. 38. Bush DA, Mcallister CJ, Loredo LN, et al. Fractionated proton beam radiotherapy for acoustic neuroma. Neurosurgery 2002; 50: 270–275. 39. Sakamoto T, Shirato H, Takeichi N, et al. Annual rate of hearing loss falls after fractionated stereotactic irradiation for vestibular schwannoma. Radiother Oncol 2001; 60:45–48. 40. Sakamoto T, Shirato H, Takeichi N, et al. Medication for hearing loss after fractionated stereotactic radiotherapy (SRT) for vestibular schwannoma. Int J Radiat Oncol Biol Phys 2001; 50:1295– 1298.

23.

acoustic schwannoma

41. Kaylie DM, Horgan MJ, Delashaw JB, McMenomey SO. A metaanalysis comparing outcomes of microsurgery and gamma knife radiosurgery. Laryngoscope 2000; 110:1850–1856. 42. Myrseth E, Moller P, Pedersen PH, et al. Vestibular schwannomas: clinical results and quality of life after microsurgery or gamma knife radiosurgery. Neurosurgery 2005; 56:927–935. 43. Friedman WA, Bradshaw P, Myers A, Bova FJ. Linear accelerator radiosurgery for vestibular schwannomas. J Neurosurg 2006; 105:657–661. 44. Hasegawa T, Fujitani S, Katsumata S, et al. Stereotactic radiosurgery for vestibular schwannomas: analysis of 317 patients followed more than 5 years. Neurosurgery 2005; 57:257–265. 45. Weber DC, Chan AW, Bussiere MR, et al. Proton beam radiosurgery for vestibular schwannoma: tumor control and cranial nerve toxicity. Neurosurgery 2003; 53:577–586.

281

46. Lunsford LD, Niranjan A, Flickinger JC, et al. Radiosurgery of vestibular schwannomas: summary of experience in 829 cases. J Neurosurg 2005; 102(Suppl):195–199. 47. Flickinger JC, Kondziolka D, Niranjan A, et al. Acoustic neuroma radiosurgery with marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys 2004; 60:225–230. 48. Noren G, Greitz D, Hirsch A, Lax I. Gamma knife surgery in acoustic tumors. Acta Neurochir Suppl (Wien) 1993; 58:104–107. 49. Massager N, Nissim O, Delbrouck C, et al. Role of intracanalicular volumetric and dosimetric parameters on hearing preservation after vestibular schwannoma radiosurgery. Int J Radiat Oncol Biol Phys 2006; 64:1331–1340. 50. Foote KD, Friedman WA, Buatti JM, et al. Analysis of risk factors associated with radiosurgery for vestibular schwannoma. J Neurosurg 2001; 95:440–449.

2 4

Acoustic Neuroma: Surgical Perspective Indro Chakrabarti and Steven L. Giannotta

Introduction For both the physician and the patient, the decision making for treatment options in acoustic neuroma has taken on more complexity. At least in experienced hands, surgical results have become highly commendable and provide a reasonable success rate to those looking for an excisional cure. Radiosurgical tumor control data, again in the best of hands, grows more impressive, now with 10 to 15 years of follow-up. Hypofractionated radiosurgery has a laudable safety record, although longterm efﬁcacy data is lacking. In selected cases, simple observation is also an effective strategy. Clearly, there is not one best way to treat a patient with an acoustic tumor. In the ideal situation, a simple comparison between efﬁcacy and safety data from several different therapeutic options should be sufﬁcient to make a decision. However, besides the various treatment options that exist, other factors come into play, including patient preference, surgeon bias, cost, patient age, and lifestyle issues. Surgical excision of these tumors has been the traditional treatment for these tumors. In our practice, the roles of neurosurgical treatment and/or stereotactic radiosurgical treatment of acoustic neuromas depend on four factors: (1) patient age, (2) tumor size, (3) hearing levels, and (4) recurrence.

Conservative Management Many patients with newly diagnosed acoustic neuromas have minimal symptoms and are aware that this is a benign, slowgrowing process. As a result, a logical strategy is to simply follow patients with surveillance imaging and plot linear growth rates. There are any number of studies, both short-term and long term, that underscore the fact some tumors will change very little over time [1–4]. Depending on the study, up to 85% of follow-up cases may show little or no growth. Several authors have demonstrated shrinkage over time. Annual growth rates have ranged between 0.15 to 4 mm [5–9]. However, most practitioners have witnessed alarming growth in some cases, which highlights the unpredictability of a strategy that relies on nontreatment as an alternative.

Conservative management for young patients is fraught with hazard. Inevitably, their tumors will grow exposing them to greater risk of treatment or eliminating treatment alternatives. We have a large number of patients over the age of 70 who are simply examined on a yearly basis and undergo imaging. Infrequently do they need intervention, usually radiosurgery. Patients in this age group with tumors that are too large to consider radiosurgery are almost always offered operative therapy unless there are substantial comorbidities that preclude general anesthesia.

Surgical Management The gold standard for treatment of any benign tumor is curative total removal. The concept of complete removal without the need for lifelong surveillance is appealing to patients. Thus, when deciding between radiosurgical versus surgical alternatives, the realistic expectation of a safe surgical obliteration is a powerful incentive. Long-term follow-up data are available for selected surgical series documenting the likelihood of longterm freedom from recurrence in experienced surgeons’ hands [9–12]. In 379 surgical cases by the senior author, no recurrences have resulted when the nerve of origin of the tumor is identiﬁed and sacriﬁced. Prior to the advent of radiosurgery, it was the author’s philosophy to offer microscopic total removal to all patients with acoustic tumors. This strategy mandated, in many cases, protracted dissection of a thinned out facial nerve at the porus acusticus. Review of the senior author’s facial nerve results shows 19% of cases with long-term House-Brackman scores of 3 or worse. Given the high rate of radiosurgical control of small lesions, subtotal removal in certain narrow circumstances might be desirable. We have taken to warning the patients about the possibility of a subtotal removal in the circumstance where injury to the facial nerve or brain stem becomes likely. This is most likely to occur in very large or very vascular tumors. In such cases, usually only a small residual is left on the facial nerve. Frequently, no evidence of residual is identiﬁable on postoperative magnetic resonance imaging (MRI).

283

284

i. chakrabarti and s.l. giannotta

Lesions where a greater likelihood of leaving a substantial residual exists include multicystic lesions. Not infrequently, there are substantial adhesions to the brain stem and to the pial vascularity. Judicious use of a subtotal strategy may result in better functional results. Growth of the residual can be treated with radiosurgery [13, 14]. Long-term tumor control or cure is desirable to many patients, but at what price? Among the larger series, the surgical complication rates have become acceptably low. The most common complication is CSF leaks, occurring in 2% to 25% of cases [9–12, 15–17]. Major morbidity such as lower cranial nerve palsy, ataxia, or long-tract dysfunction is also low, occurring 0 to 2% of the time [9–12, 15–17]. Mortality from surgery is rare, occurring in 0.4% to 2%. The two most important factors that relate to safety and efﬁcacy are tumor size and surgeon experience.

Facial Nerve Results Of all potential negative factors, patients are most concerned about facial nerve function. In fact, this is clearly how the most accomplished practitioners of the surgical art keep score. It is important for the surgeon advising a patient to be familiar with state of the art statistics and whether he or she can offer those same results. Most reports deﬁned an “excellent” or “good” outcome as a House-Brackman score of 1 or 2 [18]. A review of the senior author’s results in 379 acoustic tumors of all sizes revealed a 79% House-Brackman 1 or 2 facial nerve function rate. Depending on the size of the tumor, excellent facial nerve functional results have been reported in up to 90% of cases [9–12, 15–17, 19–27]. The Acoustic Neuroma Registry, started in the late 1980s, catalogues the national results of various surgeons through questionnaires and surveys. In a summary of 1579 cases recorded in this registry between 1989 and 1994, Wiegand et al. found that in 92% of cases, surgeons believed the removal to be total with 94% preservation of the facial nerve continuity [17]. Of these, 69% had excellent facial nerve function after follow-up at 1 year. In order to attempt a more direct comparison between surgery and radiosurgery, we sifted through the various publications to identify those who stratiﬁed facial nerve results based on tumor size. Facial nerve outcomes after surgery for tumors less than 3 cm were tabulated as single-fraction radiosurgery only addresses this subgroup of patients. As expected, these facial nerve outcomes are laudable, with favorable facial nerve outcomes in the 80% to 99% range [10–12, 19, 20, 28]. Another way to capture surgical results that can be compared with radiosurgical data is to look at cases where the surgeon attempted to preserve hearing. The majority of theses lesions would be in the size range for single-fraction radiosurgery. Shelton et al. reported that 89% of their patients experienced good facial function in a series of 106 cases performed through the middle fossa approach [29]. Other groups have elevated their surgical techniques to where facial nerve preservation and cochlear nerve preservation are an expected outcome [25, 30–34].

Hearing Preservation The data on hearing is a bit difﬁcult to interpret as patient outcomes are sometimes reported as simply at or near preoperative levels. However, per the Gardner-Robertson scale, a class I or II is considered as serviceable or better hearing [35]. In this scale, pure-tone average of 0 to 30 dB and speech discrimination of 70% to 100% corresponds with class 1 hearing. Class 2 hearing assumes 31 to 50 dB pure-tone average and 50% to 69% speech discrimination. The AAO-HNS guidelines on the evaluation of hearing preservation also developed a system of classiﬁcation with class A or B generally indicating useful hearing [36]. Class A hearing includes <30 dB pure-tone thresholds and >70% speech discrimination. Class B includes 30 to 50 dB pure-tone thresholds and >50% speech discrimination. Most reports use one of these two methods for recording hearing outcomes. The data with regard to hearing preservation among the large series is somewhat variable. Good hearing can range from 24% to 60% depending on the series [9–12, 16, 19, 20, 21, 27, 28, 30]. It is noteworthy that the Acoustic Neuroma Registry data described by Wiegand et al. reports a 22% hearing preservation rate. A cursory look at data from groups who report large series of middle fossa approaches would suggest that this strategy might be more effective in terms of hearing preservation. Most accomplished groups report good hearing rates in the range of 50% [25, 29, 31, 33, 34, 37–39]. However, these series are all biased toward the smallest of tumors.

Surgical Approaches Each of the three traditional approaches to the CP angle have there own putative strategic advantages (Table 24-1). The translabyrinthine approach minimizes retraction, maximizes exposure of the internal auditory canal, and facilitates deﬁning the facial nerve within the canal. The retrosigmoid approach has the advantage of being stereotypical for most neurosurgeons while enabling the preservation of the hearing apparatus (Fig. 24-1). The middle fossa approach, though admittedly for smaller tumors, also facilitates exposure of the lateral-most extent of the internal auditory canal. Proponents of each TABLE 24-1. Approach selection. Retrosigmoid

Middle fossa

Translabyrinthine

+

+++

++

+++ +++

+++ 0

+++

>2.5 cm

++

0

+

Only-hearing ear

+++

0

+

>50/50

+++

+++

+

<50/50

+

+

+++

Recurrent

+

0

+++

Size <1 cm Lateral impact Medial <2.5 cm

++

Hearing

24.

285

acoustic neuroma: surgical perspective

and CyberKnife) are offered. The surgical decision-making is less predicated on available technologies or reliance on a practiced single surgical approach than it is on patient age, tumor size, and need for hearing preservation (Table 24-2).

Age and Size

FIGURE 24-1. (A) A 5-cm acoustic neuroma prior to resection. The patient presented with tinnitus, hearing loss, and signs of brain-stem compression. (B) Postoperative image after a retrosigmoid craniotomy. All preoperative symptoms resolved.

approach are able to post impressive numbers in terms of safety and cure rates. Many studies have looked at outcomes based on approach, and no discernible advantage has been documented for one approach over another [10, 25]. However, for purely intracanalicular tumors, several reports have shown that the middle fossa approach is safer for hearing preservation than the retrosigmoid approach [20, 33, 37]. A cursory review of the literature would suggest no way to resolve controversies that may emerge as to the best approach for a given situation. Hearing preservation, facial nerve function, and incidence of total removal seem to be practitioner- or team-related as opposed to approach-related. This suggests that technical expertise and strategic operative decision-making can negate most disadvantages related to the various approach strategies. A retrospective review of our institution’s past 50 cases of retrosigmoid resections was compared with the past 50 translabyrinthine resections. Tumors were included that were 3 cm or less, and results were categorized in terms of facial nerve outcome, incidence of total resection, and major complications. At discharge, 37 of the retrosigmoid patients had a HouseBrackman score of 1 compared with 28 of the translabyrinthine patients. Combining House-Brackman 1 and 2 scores showed 82% with the retrosigmoid cases and 70% of the translabyrinthine. This did not reach statistical signiﬁcance. There was one major neurologic complication in the retrosigmoid group and none in the other group. Three cases of subtotal removal resided in each group. Long-term follow-up of these patients showed even less of a difference in facial nerve function. This suggests that surgical technique may be more important than approach strategies in terms of outcome for acoustic tumors. Ideally, each practitioner should have a working knowledge of all approaches. This can instill the wary patient with some conﬁdence that decision-making is based on assessment of the patient’s best interests. For those who are experiencing unacceptable results, altering approach strategy may show modest beneﬁts, but doing such will not make up for deﬁciencies in surgical technique as it relates to the brain-stem vascularity and cranial nerves.

For younger patients, the emphasis is on surgical removal. A surveillance strategy in this group is likely to be futile, as inevitably the lesion will cause further symptoms and require treatment. Radiosurgery has a long follow-up period, and the window of vulnerability for recurrence is potentially wide. Reliable data on lifelong tumor control for patients in their thirties or forties is lacking. For those in their ﬁfties and sixties, single-fraction radiosurgery is an attractive alternative. Efﬁcacy and safety statistics are available for this group and are highly acceptable (Fig. 24-2). Patients in older age groups rarely need any therapy unless their tumor is large enough to be threatening. For large lesions greater than 3 cm, single-fraction radiosurgery has no role. For these lesions, surgical removal, or in certain situations subtotal removal, with follow-up radiosurgery is advisable. Subtotal removal for a patient with a single tumor and good hearing in the other ear should be an unusual event. A small residual may be left behind in an effort to avoid a major complication such as facial nerve sacriﬁce or brain-stem injury. For younger patients with large tumors, surgical removal is the preferred strategy. The length of vulnerability for recurrence is too great for younger patients to rely on subtotal removal. Some further therapeutic endeavor will ultimately be necessary, multiplying the potential for complications. Large lesions on older patients can present some strategic problems. This would seem like an ideal situation for hypofractionated radiosurgery. However, with lesions greater than 4 cm or somewhat smaller lesions with associated arachnoid cysts usurping much of the available reserve in the posterior fossa, radiosurgery with its attendant edema formation may produce unacceptable risks 6 to 12 months after treatment. Data is sorely lacking for this modality in larger tumors. Until better long-term studies are available, older patients in good health with large lesions should be offered the option of surgical removal. The decision for total versus subtotal removal is made

TABLE 24-2. Treatment modality. Radiosurgery

Surgery

Observe

Size >2.5 cm 1.5 to 2.5 cm

+

+++

0

++

++

+

<1.5 cm

+++

++

++

Age <40 40 to 60

+

+++

0

++

++

+

Choice of Treatment Method (Radiosurgery Versus Excision)

>60

+++

+

+++

>50/50

+++

++

++

At our institution, the three traditional approaches for surgical excision of acoustic tumors are practiced. Further, three technologies for radiosurgery (Gamma Knife, linear accelerator,

<50/50

++

++

+

Recurrent

+++

+

+

Hearing

286

i. chakrabarti and s.l. giannotta

FIGURE 24-2. (A) A 2.5-cm acoustic neuroma in a patient who presented with slight hearing loss. The fourth ventricle is distorted. This patient elected to have radiosurgery. (B) After Gamma Knife radiosurgical treatment. Central necrosis is present as well as decreased pressure on posterior fossa structures.

at the time of surgery and predicated on the likelihood of complications. Radiosurgery as an adjunct can be offered for any threatening residual. For large tumors in older patients who are poor surgical risks, hypofractionated radiosurgery is a logical option.

Hearing Preservation Hearing preservation in the context of surgical removal can be expected in experienced hands to be successful 50% of the time with intracanalicular tumors and 30% with larger lesions that are generally under 3 cm. This presupposes good functional hearing to begin with. Attempts to save hearing in a marginal or poorly hearing ear will be unrewarding. That ear will be a constant source of distraction to the patient as it picks up unstructured background noise and reduces the overall functionality of the hearing. Thus, compromises in surgical strategy to preserve the function of a poorly hearing ear should be vigorously resisted. Single-fraction radiosurgery is fast growing in popularity and may become the treatment of choice in the absence of an experienced surgeon with a proven track record for the safe and effective removal. For those lesions less than 3 cm, one can expect at least 50% hearing preservation or better assuming accepted proven radiosurgical techniques are utilized. The major drawback in prescribing it for all small acoustic tumors is the lack of long-term efﬁcacy data. For many patients, the need for continued surveillance and the thought of the continued presence of the lesion are negative satisﬁers. If an acoustic neuroma recurs and is deemed in need of treatment, the ﬁrst option should be radiosurgery. This assumes the tumor regrowth has been detected before it has grown too large for radiosurgery. In these cases, a translabyrinthine surgical approach will offer the largest corridor while minimize the need for retraction or dissection of previously scarred brain. A repeat retrosigmoid craniotomy may prove difﬁcult.

Choice of Surgical Approach Our preference is to use the translabyrinthine approach for all tumors where hearing preservation is unlikely (Table 24-1).

Thus, it is used in all large tumors and in those with poor hearing. Certainly, those greater than 3 cm would be treated this way and most with speech discrimination scores of less than 50%. The only time we would favor a retrosigmoid approach in a large tumor would be the case of a large lesion in an onlyhearing ear where a subtotal removal is contemplated. In tumors that protrude from the porus, the retrosigmoid approach is preferred for hearing preservation (Fig. 24-3). In our most recent 80 cases using this approach, functional hearing resulted in 30%. We limit the use of the middle fossa strategy for those small intracanalicular lesions that are affected in the lateral end of the internal auditory canal. Other factors come into play as patients try to make their decisions. Socioeconomic and educational status may complicate decision making in patients who cannot understand a complex set of options. Patient and family biases for or against surgery or radiation may direct the patient’s thinking contrary to the physician’s best judgment. Access to the Internet, inﬂuence from patients who have had one form of therapy or another, and loyalty to a particular institution may be relevant factors in decision making. One can guide the decision making by trying to simplify principles. Explaining away misconceptions is a place to start. Identifying patient and family biases and dealing with them in a forthright way will also help. If it is perceived that this discussion is simply a device to steer the decision making toward the surgeon or radiosurgeon’s bias, confusion and mistrust can develop. If a patient harbors a tumor that may be amenable to either surgical removal or radiosurgery, a simple construct can be presented to the family and patient to facilitate their decision making. Does the patient insist that the tumor be gone? Beneﬁts include diagnostic certainty and the lack of need for longterm surveillance. Ultimately, patients will decide on what treatment option they prefer with some guidance as to the risk/beneﬁt ratio from the physician. The surgical removal of acoustic neuroma has been reﬁned over time to achieve impressive results. However, obvious risk of major morbidity remains. Radiosurgical methods are less invasive and less likely to cause major morbidity. However, large lesions are clearly not amenable to this therapy, and lifelong observation is a requirement even for small lesions. No perfect algorithm exists, and each patient has unique challenges. The ﬁrst question the physician and patient must answer is how important is the removal of the tumor to one’s overall

FIGURE 24-3. (A) Intraoperative view of acoustic neuroma during a left retrosigmoid craniotomy. (B) After tumor removal, the nerves of the porus acusticus are seen. Note the transected superior vestibular nerve from where the tumor originated.

24.

acoustic neuroma: surgical perspective

comfort. If knowing the tumor is still present makes one unable to participate in life activities, then surgery seems logical. Otherwise, the radiosurgical alternatives must be explored. With sound advice and research, the patient will ultimately decide on his own treatment algorithm.

References 1. Rosenberg, S. Natural history of acoustic neuromas. Laryngoscope 2000; 110:497–508. 2. Tshudi DC, Linder TE, Fisch U. Conservative management of unilateral acoustic neuromas. Am J Otol 2000; 21:722–728. 3. Charabi S, Thomsen J, Mantoni M, et al. Acoustic neuroma (vestibular schwannoma): growth and surgical and nonsurgical consequences of the wait-and-see policy. Otolaryngol Head Neck Surg 1995; 113:5–14. 4. Charabi S, Tos M, Thomsen J, et al. Vestibular schwannoma growth-long-term results. Acta Otolaryngol Suppl 2000; 543:7– 10. 5. Deen HG, Ebersold MJ, Harner SG, et al. Conservative management of acoustic neuroma: an outcome study. Neurosurgery 1996; 39:260–266. 6. Hoistad DL, Melnik G, Mamikoglu B, et al. Update on conservative management of acoustic neuroma. Otol Neurotol 2001; 22:682–685. 7. Levo H, Pyykko I, Blomstedt G. Non-surgical treatment of vestibular schwannoma patients. Acta Otolaryngol (Stockh) Suppl 1997; 529:56–58. 8. Leutje CM. Spontaneous involution of acoustic tumors. Am J Otol 2000; 21:393–398. 9. Samii M, Matthies C. Management of 1000 vestibular schwannomas (acoustic neuromas): the facial nerve-preservation and restitution of function. Neurosurgery 1997; 40:684–695. 10. Sampath P, Holiday MJ, Brem H, et al. Facial nerve injury in acoustic neuroma (vestibular schwannoma) surgery: etiology and prevention. J Neurosurg 1997; 87:60–66. 11. Koos WT, Matula C, Kitz K. Microsurgery versus radiosurgery in the treatment of small acoustic neurinomas. Acta Neurochir Suppl 1995; 63:73–80. 12. Ebersold MJ, Harner SG, Beatty CW, et al. Current results of the retrosigmoid approach to acoustic neurinoma. J Neurosurg 1992; 76:901–909. 13. Modugno GC, Pirodda A, Ferri GG, et al. Small acoustic neuromas: monitoring the growth rate by MRI. Acta Neurochir (Wien) 1999; 141:1063–1067. 14. Silverstein H, Rosenberg SL, Flanzer JM, et al. An algorithm for the management of acoustic neuromas regarding age, hearing, tumor size, and symptoms. Otolaryngol Head Neck Surg 1993; 108:1–10. 15. Van Leeuwen JP, Cremers CW, Theunissen EJ, et al. Translabyrinthine and transotic surgery for acoustic neuroma. Clin Otolaryngol 1994; 19:491–495. 16. Gjuric M, Wigand ME, Wolf SR. Enlarged middle fossa vestibular schwannoma surgery: experience with 735 cases. Otol Neurotol 2001; 22:223–231. 17. Wiegand DA, Ojemann RG, Fickel V. Surgical treament of acoustic neuroma (vestibular schwannoma) in the United States: report from the Acoustic Neuroma Registry. Laryngoscope 1996; 106:58–66. 18. House JW, Brackman DE. Facial nerve grading system. Otolaryngol Head Neck Surg 1985; 93:146–147.

287

19. Wiet RJ, Mamikoglu B, Odom L, et al. Long-term results of the ﬁrst 500 cases of acoustic neuroma surgery. Otolaryngol Head Neck Surg 2001; 124:645–651. 20. Sterkers JM, Morrison GAJ, Sterkers O, et al. Preservation of facial, cochlear, and other nerve functions in acoustic neuroma treatment. Otolaryngol Head Neck Surg 1994; 110:146–155. 21. Tonn JC, Schalke HP, Goldbrunner R, et al. Acoustic neuroma surgery as an interdisciplinary approach: a neurosurgical series of 508 patients. J Neurol Neurosurg Psychiatry 2000; 69:161– 166. 22. Ojemann RG. Management of acoustic neuromas (vestibular schwannomas) (honored guest presentation). Clin Neurosurg 1993; 40:408–535. 23. Symon L, Bord LT, Compton JS, et al. Acoustic neuroma: a review of 392 cases. Br J Neurosurg 1989; 3:343–348. 24. Harner SG, Beatty CW, Ebersold MJ. Retrodigmoid removal of acoustic neuroma: experience 1978–1988. Otolaryngol Head Neck Surg 1990; 103:40–45. 25. Arriaga MA, Chen DA, Fukushima T. Individualizing hearing preservation in acoustic neuroma surgery. Laryngoscope 1997; 107:1043–1047. 26. Tos M, Thomsen J, Harmsen A. Results of translabyrinthine removal of 300 acoustic neuromas related to tumour size. Acta Otolaryngol (Stockh) Suppl 1988; 452:38–51. 27. Sugita K, Kobayashi S. Technical and instrumental improvements in the surgical treatment of acoustic neurinomas. J Neurosurg 1982; 57:747–752. 28. Gormley WB, Sekhar LN, Wright DC, et al. Acoustic neuromas: results of current surgical management. Neurosurgery 1997; 41: 50–60. 29. Shelton C, Brackman DE, House WF, et al. Middle fossa acoustic tumor surgery: results in 106 cases. Laryngoscope 1989; 99:405– 408. 30. Nadol JB, Chiong CM, Ojemann RG, et al. Preservation of hearing and facial nerve function in resection of acoustic neuroma. Laryngoscope 1992; 102:1153–1294. 31. Haines SJ, Levine SC. Intracanalicular acoustic neuroma: early surgery for preservation of hearing. J Neurosurg 1993; 79:515– 520. 32. Post KD, Eisenberg MB, Catalano PJ. Hearing preservation in vestibular schwannoma surgery: what factors inﬂuence outcome? J Neurosurg 1995; 83:191–196. 33. Irving RM, Jackler RK, Pitts LH. Hearing preservation in patients undergoing vestibular schwannoma surgery: comparison of middle fossa and retrosigmoid approaches. J Neurosurg 1998; 88:840–845. 34. Slattery WH, Brackman DE, Hitselberger W. Middle fossa approach for hearing preservation with acoustic neuromas. Am J Otol 1997; 18:596–601. 35. Gardner G, Robertson JH. Hearing preservation in unilateral acoustic neuroma surgery. Ann Otol Rhinol Laryngol 1988; 97: 55–66. 36. Committee on Hearing and Equilibrium. Guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma). Otolaryngol Head Neck Surg 1995; 113:179– 180. 37. Rowed DW, Nedzelski JM. Hearing preservation in the removal of intracanalicular acoustic neuromas via the retrosigmoid approach. J Neurosurg 1997; 86:456–461. 38. Glasscock ME, Hays JW, Minor LB, et al. Preservation of hearing in surgery for acoustic neuromas. J Neurosurg 1993; 78:864–870. 39. Brackman DE, Owens RM, Friedman RA, et al. Prognostic factors for hearing preservation in vestibular schwannoma surgery. Am J Otol 2000; 21(3):417–424.

2 5

Acoustic Neuromas and Other Benign Tumors: Fractionated Stereotactic Radiotherapy Perspective David W. Andrews, Greg Bednarz, Beverly Downes, and Maria Werner-Wasik

Introduction Radiosurgery has become an important treatment alternative to surgery for a variety of intracranial lesions. As currently practiced, it has in fact replaced surgery as a standard of care in some instances, complements surgery as a postoperative adjunct in others, and most commonly represents an alternative to surgery or the only treatment option. Radiosurgery techniques have evolved quickly with the development of new technologies enabling more complex yet more efﬁcient treatment plans. As a consequence, these technologies have broadened radiosurgery applications and improved radiosurgery outcomes. Among these newer techniques, treatments involving fractionated stereotactic radiation referred to as fractionated stereotactic radiotherapy, or FSR, have emerged as a consequence of linear accelerators designed for and dedicated to stereotactic techniques. Without the logistical constraints of retroﬁtted general-purpose linear accelerators used in radiation oncology, often available only once or twice a week, dedicated units have enabled the design of treatment paradigms that strive for an ideal treatment based on the radiobiology of the target and dose-limiting contiguous tissues. This chapter will summarize our 12-year experience with the Varian 600SR, initially with the Radionics software more recently modiﬁed to a Novalis shapedbeam radiosurgery unit, and our practice of FSR for a variety of intracranial lesions. Special attention will be devoted to tumors involving or near the special sensory cranial nerves. Given the versatility of the Novalis treatment planning platform, one has the option of comparing different treatment planning solutions at once, including stereotactic intensity-modulated radiation therapy (IMRT). For selected skull base lesions, we have found that stereotactic IMRT yields greater conformality than FSR, and we will therefore include its application among fractionation strategies.

Radiobiological Principles of FSR for Late-Responding Tissues Fractionated radiotherapy has been used to treat patients with skull base tumors during the past 50 years. Stereotactic radiosurgery has made the treatment of skull base tumors far more precise with modern imaging and versatile three-dimensional treatment planning software, which, through dose-volume histograms, maximizes dose to target with high conformality and minimizes dose to contiguous normal structures. Two exceptions are optic nerve sheath meningiomas and acoustic neuromas where these cranial nerves are intrinsic to the target volume. As special sensory cranial nerves, injury to sensory function occurs much more frequently than either sensory or motor function in mixed cranial nerves, reﬂecting a lower threshold for injury. Low daily doses of radiation and a cumulative dose below a threshold value, however, have proved to be safe for the optic nerves and more recently for the cochlear nerve. We will advance arguments based on published data that tumors involving or near special sensory cranial nerves, when treatment is indicated, should be treated with FSR utilizing daily conventional fraction sizes.

Optic Nerve Sheath Meningiomas Single doses of 15 Gy or 54 Gy in 30 daily fractions both result in excellent control of meningiomas [1]. The SRS dose of 15 Gy, however, exceeds the generally accepted single-dose tolerance of the optic nerves (8 to 10 Gy) for treatment of optic nerve sheath meningiomas (ONSMs) and may be associated with a risk of optic neuropathy approaching 78% at ≥15 Gy dose prescriptions [2]. For fractionated radiotherapy, decades of experience have led to guidelines for normal tissue tolerance. For optic nerve and chiasm, guidelines have evolved that have

289

290

d.w. andrews et al.

proved to be extremely safe and are generally based on the optic ret dose of 890 recommended by Goldsmith [3], later corroborated by a large retrospective analysis by Parsons [4] who also established a daily dose of 1.9 Gy as safe. Based on the assumption that 15 Gy in a single fraction and 54 Gy in 30 fractions are biologically equivalent, the α/β ratio for meningiomas may be calculated based on the linear quadratic formula where BED is biologically effective dose: BED (Gy) = D[1 + d/(α/β)] where D is the total dose, d is the dose per fraction, and α/β is the characteristic constant associated with the particular tissue in question. Assuming BED1 = BED2 for tissue of an unknown α/β: D1[1 + d1/(α/β)] = D2[1 + d2/(α/β)] or α/β = (D1 × d1 − D2 × d2)/D2 − D1 For meningiomas using the assumption of equivalent BED for 15 Gy in a single fraction (d1 = 15 Gy) and 54 Gy in 30 fractions (d2 = 1.8 Gy): α/β = (15 × 15) − (54 × 1.8)/(54 − 15) = 3.28 Gy and the corresponding BED is calculated as: BED = 15(1 + 15/3.28) = 54(1 + 1.8/3.28) = 83.6 Gy

scales. The most widely cited scale is the Gardner-Robertson scale [5], but despite this scale there still exist disparities in reported outcomes due to differences, for example, in the derivation of pure tone average. We reviewed published literature and selected only those studies that featured audiometric outcomes utilizing the Gardner-Robertson scale conﬁrmed by audiometric data. Other scales or outcomes based on subjective hearing results were excluded. The most informative reports included the actual audiometric data, but most reports did not. At radiosurgery doses that reﬂected excellent tumor control rates, Table 25-1 [6–23] reﬂects Gamma Knife dose deescalations over time to improve rates of cranial neuropathy. To date, literature does not reﬂect dose iterations for FSR. From Table 25-1, high rates of tumor control have been achieved for both a single dose of 12 to 13 Gy or 45 to 50 Gy in 1.8- to 2-Gy daily fractions, suggesting a dose equivalence for treatment of acoustic neuromas. Unlike the optic apparatus, however, acceptably low rates of hearing loss have not yet been generally agreed upon. In our single institution experience of dose de-escalation for FSR, we have found the lowest dose cohort of 46.8 Gy achieves an excellent tumor control rate. Utilizing a dose equivalence of 12 Gy in a single fraction and 46.8 Gy in 1.8 Gy daily fractions, and once again assuming BED1 = BED2 for tissue of an unknown α/β:

It has been clearly established that the radiation tolerance to the optic nerves and chiasm depends on the total dose of radiation and the dose per fraction. Goldsmith et al. proposed a model predicting the total dose associated with a low risk of optic neuropathy when various doses per fraction were prescribed [3]:

For acoustic neuromas using the assumption of equivalent BED for 12 Gy in a single fraction (d1 = 12 Gy) and 46.8 Gy in 26 fractions (d2 = 1.8 Gy):

Optic ret dose = Dose (cGy)/N 0.53

α/β = (12 × 12) − (46.8 × 1.8) / (46.8 − 12) = 1.72

Shrieve et al. applied these formulas to the treatment of parasellar meningiomas where the optic apparatus is a dose-limiting structure [1]. If the optic ret tolerance of 890 is observed and the α/β is assumed to be 3.28 ± 10%, a range of doses from a minimum of 46 Gy in a minimum of 22 fractions to 54 Gy in 30 fractions is necessary to achieve tumor control and an acceptable rate of optic neuropathy. An assumption in this analysis is that vision is intact at the inception of treatment. The natural history of primary optic nerve sheath meningiomas or meningiomas that encroach upon the optic apparatus reﬂects progressive visual loss to blindness in the affected eye. Assumptions regarding the morbidity of radiation treatment must in these cases be weighed against the morbidity associated with the natural history of the disease. If radiation optic neuropathy, for example, occurred at rates of 5% to 10% in a median range of 4 to 6 years, and the natural history reﬂected rates of visual loss to blindness at 80% to 90% over a broader median range of 6 to 10 years, the therapeutic beneﬁt would outweigh the treatment-related morbidity, even if as high as a 10% rate of optic neuropathy.

Acoustic Neuromas Reports of hearing preservation are nonuniform, ranging from subjective hearing outcomes to standard audiometric outcomes

D1[1 + d1/(α/β)] = D2[1 + d2/(α/β)] or α/β = (D1 × d1 − D2 × d2)/D2 − D1

and the corresponding BED is calculated as: BED = 12(1 + 12/1.72) = 46.8(1 + 1.8/1.72) = 95.7 Gy This model could be used for other dose-fractionation schemes for treatment of acoustic tumor, but its utility for a predictive model of serviceable hearing loss remains unclear because a hearing ret formula does not yet exist. Pan et al. proposed a model relating radiation dose to sensorineural hearing loss when various doses per fraction were prescribed and built into the model other covariates including differences in hearing threshold between the normal and affected ear at baseline and age [24]. Signiﬁcant hearing loss was deﬁned as a >10 dB loss in pure-tone audiometry. Within this study, the authors observed that for almost all cases in which signiﬁcant hearing loss occurred in the affected ear receiving radiation, the dose was ≥45 Gy. Although fewer published data are available, we were able to plot the log of dose versus the log of the number of fractions drawing from Goldsmith’s model [3] for reports with hearing preservation rates of at least 70% based on Gardner-Robertson criteria [10, 14–16, 18–21, 23] (see designated papers marked with double asterisks, Table 25-1). The linear regression gave us a formula for dose/fraction size regimens with a high correlation coefﬁcient (R2 = 0.9705). It may be postulated that dose/ fraction schemes with very high hearing preservation rate can be represented by a parallel line on the same plot, intersecting

25.

acoustic neuromas and other benign tumors: fractionated stereotactic radiotherapy perspective

291

TABLE 25-1. Reported focused radiation series for treatment of acoustic neuromas. Rate of cranial neuropathy (%) V Author/date

N

Gamma Knife radiosurgery Hirsch, 1988 [6] 126 Flickinger, 1991 [7] 85 Foote, 1995 [8] Kondziolka, 1998 [9]

36 162

1988–1998 results Andrews, 2001 [10] Karpinos, 2002 [11]

69 96

Regis, 2002 [12]

104

Iwai Y, 2003 [13]

51

Flickinger, 2004 [14]† Van Eck, 2005 [15]† Hasegawa, 2005 [16]

313 78 74

2001–2005 results

Isodose prescription (Gy)

18–25 14–20 (18 median) 16–20 12–20 (16.6 mean)

Tumor control rate (%)

21 29

15 30

24 (m, 4.7)a 46 (s, 2)b

100 98

59 27

67 21

42 (s, 2)b 47 (s, 5–10)a

96 ± 2 100 91 100

8–12 (12 median) 12–13 13 ≤13

VIII

(m/s/yr*)

91 97

18 median 12 10–24 (14.5 mean) 12–14

13 median

VII

34 ± 8 2 11

33 ± 12 1 4

40 ± 11 33 (s, 0.8)b 44 (s, 4)a

4

2

96

4

0

50 (s, >3)a 56 (s, 5)a

99 98 NS

4 2.5 NS

0 1.2 NS

79 (s, 6)b 83 68 (s, 7)

97 ± 1

5 ± 31

1 ± 12

97

16

8

78 (s, 2)b,d

97

7

2

100

7

4

91 93

14 3.4

4 2.3

81 (s, 1.2)a 71 (s, 5)a,c 85 (s, 1)b 57 (s, 2) b 72 (s, 3.75)b 98 (s, 4)b

10 ± 5

5±3

74 ± 133P

59 ± 18

Conventional dose fractionated stereotactic radiotherapy Kagei, 1999 [17]

39

Andrews, 2001 [10]†

56

Chung, 2004 [18]†

72

Sawamura, 2003 [19]† Combs, 2005 [20]† 1999–2005 results

36 106

36–44 2-Gy/fraction + 4-Gy boost 50.4 2-Gy/fraction 45 1.8-Gy/fraction 48 (median) 57.6 47.5 median

98 ± 2

Higher dose fractionated stereotactic radiation Poen, 1999 [21]†

31

Noren, 2005 [22]†

25

Williams, 2003 [23]†

80

1999–2005 results

21 7-Gy/fraction 15 3-Gy/fraction 25 5-Gy/fraction 21 median

97

16

3

77 (s, 2)b

96

0

0

86 (s, 12–64 mo)a

100

2.5

0

77 (s, 1.1)b

97 ± 2

6 ± 8.7

1 ± 1.7

80 ± 54

*m/s/yr reﬂects measurable or serviceable hearing preservation rates at follow-up intervals in years. †

Data points used for hearing ret formula.

a

Raw hearing preservation rate.

b

Actuarial hearing preservation rate.

c

Current follow-up for this cohort (unpublished observation).

d

Unclear impact of 4-Gy end-boost on hearing preservation.

p = 0.0028 vs. early Gamma Knife cohort, trigeminal neuropathy; 2p = 0.0087 vs. early Gamma Knife cohort, facial neuropathy; 3p = 0.0064, FSR longest follow-up hearing vs. early Gamma Knife cohort hearing; 4p = 0.002, higher dose fractionated stereotactic radiation follow-up hearing vs. early Gamma Knife cohort hearing. 1

292

d.w. andrews et al. LogD vs log # of fractions 4

3.8 3.6 3.4 Log D

3.2 y = 0.4308x + 3.0824

3

R2 = 0.9705

2.8

y intercept = 3.051 2.6 for high rate of hearing preservation

2.4 2.2 2 0

0.5

1

1.5

2

2.5

Log # of fractions FIGURE 25-1. Logarithmic plot of total dose versus number of fractions for cases reporting preservation of serviceable hearing.

a point corresponding with the dosage/fraction scheme documented as safe (Fig. 25-1). Data from Pan et al. [24] suggest that 45 Gy in 25 × 1.8 Gy fractions (or less) is such a point. By substitution to the linear regression equation: intercept = log(4500) − 0.4308 × log(25) with an intercept at 3.051. So the line equation is logD = 0.431 × logN + 3.051 where N is the number of fractions. This can be folded into: D = 1120 × N 0.43 or 1120 hearing ret = D × N −0.43 for dose/fraction schemes with high hearing preservation rates. For one fraction: D = 11.2 Gy. This implies that the radiosurgical dose should not exceed a threshold around 11.2 Gy to obtain high rates of hearing preservation. Based on this hearing ret formula, single fraction doses above 11.2 Gy or a cumulative FSR dose above 45 Gy could result in higher rates of hearing loss. The median value of 13 Gy has, in fact, resulted in serviceable hearing losses ranging from 17% to 67%, and a median FSR dose of 47.5 Gy has resulted in serviceable hearing losses ranging from 29% to 43% (Table 25-1). None of the FSR papers featured in Table 25-1 provide uniform treatment planning data such as MDPD (maximum dose to prescribed dose) ratios or PTV (prescription to tumor volume) ratios. A uniform reporting mechanism for singlefraction treatments should include mean number of isocenters stratiﬁed by tumor sizes, and these data might build a relationship between such variables as tumor size, serviceable hearing loss, and mean number isodose centers as a measure of conformality. When we compared BED plots that were ±5% of the calculated α/β ratio of 1.72 to a hearing ret plot, we noted that a minimum of around 26 fractions delivered with a uniform dose was necessary to achieve a high rate of hearing preservation (Fig. 25-2). As with the optic nerves, these data based on published results reﬂect that an effective dose for acoustic

neuroma tumor control that simultaneously achieves high rates of serviceable hearing preservation is obtained with fractionation, in this case a minimum of approximately 26 fractions. For the treatment of acoustic neuromas, we postulate that FSR is preferable to single-fraction radiosurgery for the following reasons (see Fig. 23-5). Typical isodose prescriptions are 50% for Gamma Knife treatment and 90% for FSR treatments (Fig. 25-3h). Because a 50% isodose prescription falls on the steep portion of a typical dose proﬁle, dose gradients are necessarily greater for the same distance compared with a 90% isodose prescription (Fig. 25-3d, f). A 50% isodose prescription therefore calls for extremely high conformality at the cranial nerve VIII–tumor interface to avoid hearing loss. This has been achieved with the substitution of high-resolution stereotactic magnetic resonance imaging over computed tomography [25, 26]. Coupled with improvements in treatment planning software, large numbers of isocenters can be employed to achieve greater conformality with sharp dose fall off [14]. For less conformal radiosurgery treatment plans, however, it is highly probable that segments of cranial nerve VIII are included inside the standard 50% isodose prescription volume (Fig. 25-3g), resulting in higher dose exposure and increased risk of serviceable hearing loss. Although multiple-shot/high-conformality treatments reﬂect the greatest rates of hearing preservation, an additional concern is the greater the number of shots creating a greater likelihood of dose inhomogeneity [27], which could adversely affect hearing. Other variables at play may be the segment or length of nerve exposed or the inclusion of the cochlea above threshold dose. For FSR treatments, although dose conformality probably plays a role (Fig. 25-3f), the most important variables affecting cranial nerve VIII are most likely cumulative dose and daily fraction dose, as demonstrated with the optic nerves by Parsons et al. [4]. For cumulative dose, as with single-fraction treatments, our preliminary audiometric results after a 10% dose reduction reveal signiﬁcant improvement in pure tone average (see Table 25-2 and “Clinical Outcomes” section below). BED and hearing ret doses 600

500 400

hearing ret alpha/beta=1.63 alpha/beta=1.81

300

200

100 0 20 25 30 35 40 45 # of fractions FIGURE 25-2. Plots of biologically equivalent doses utilizing the linear quadratic formula with α/β ratios of +10%; comparison with a hearing ret plot. 0

5

10

15

25.

acoustic neuromas and other benign tumors: fractionated stereotactic radiotherapy perspective

FIGURE 25-3. Comparison of dose distribution to the cochlear nerve with Gamma Knife and FSR treatments. (a) Axial T1 gadoliniumenhanced MRI scan of right acoustic neuroma. (b) Artist’s rendering of translucent acoustic tumor with cranial nerves VII and VIII 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 inner for right acoustic neuroma. (d) Magniﬁed sagittal view of actual treatment plan in the distal porus acusticus (outer line is 50% isodose prescription line; middle line is 60% isodose line; inner line is tumor surface). Assuming cochlear nerve is around the 7 o’clock position, the nerve is within a 10% dose gradient above isodose prescription. (e) Single-shot Novalis FSR treatment plan with a 1.8-Gy prescription to the 90% isodose line for a right acoustic neuroma (tumor is shaded; outer line is 90% isodose prescription line). (f) Magniﬁed sagittal view of actual treatment plan in the distal porus acusticus (tumor is shaded; outer line is 90% isodose

TABLE 25-2. Comparison of two dose cohorts after FSR treatment of acoustic neuromas.

293

prescription line; inner line is 95% isodose line). Assuming cochlear nerve is around the 7 o’clock position, the nerve is within a 5% dose gradient above isodose prescription. (g) Artist’s rendering of magniﬁed sagittal cross section of intracanalicular portion of tumor and contiguous VIII nerve at 7 o’clock position. The VIII nerve is within the prescribed isodose line and exposed to higher dose gradients. (h) Proﬁle of a focused radiation beam with typical isodose prescriptions at 50% (Gamma Knife) and 90% (FSR). Vertical columns in either scenario could represent narrow location range of cochlear nerve relative to tumor (e.g., within 1 mm of tumor surface). Broad inferior horizontal bar represents potential actual dose range delivered to cochlear nerve with 50% isodose prescription; narrow superior horizontal bar represents potential actual dose range delivered to cochlear nerve with 90% isodose prescription. The potential dose gradient at a 50% isodose prescription is more than three times greater than 90% prescription at the same distance.

Technique

Dose cohort 50.4 Gy (N = 74)

46.8 Gy (N = 38)

Tumor control rate

98%

100%

VII neuropathy rate

1%

2%

V neuropathy rate

1%

0

Outcomes

Hearing preservation rate All measurable Serviceable Corrected for follow-up and initial SNHL*

74% 71% 67%

88% 90% 93%†

SNHL, sensorineural hearing loss; *Pretreatment audiograms limited to patients with PTA between 15 and 50 dB and follow-up limited to 122 weeks. p = 0.06 vs. 50.4 dose cohort, Fisher’s exact test.

†

We commissioned in 1994 the world’s ﬁrst installation of a linear accelerator designed for and dedicated to stereotactic radiosurgery and fractionated stereotactic radiotherapy [28]. Since then, we adopted a treatment technique that incorporates stereotactic technique with conventional daily doses of radiation [10, 29–34]. These dose-fraction protocols are designed to optimize therapeutic effect while minimizing treatment-related morbidities. Our rationale for a conventional fraction FSR paradigm stems from our belief that special sensory cranial nerves, notably the optic nerves and the cochlear nerves, are more sensitive to any therapeutic intervention including radiation therapy. Drawing from previously reported observations documenting injury thresholds for the optic nerves after radiation for head and neck cancers [4], we have adopted a 1.8-Gy daily fraction schedule, and for most

294

d.w. andrews et al.

TABLE 25-3. FSR treatment paradigms for benign intracranial tumors. Target lesion

Acoustic neuroma* Optic nerve sheath meningioma Parasellar lesions Pituitary adenoma Craniopharyngioma Cavernous sinus meningioma Complex skull base Meningioma Chordoma

Daily dose (Gy)

Cumulative dose (Gy)

1.8 1.8

46.8 52

Dynamic arc Dynamic arc

1.8 1.8 1.8

50.4 54 54

Dynamic arc Dynamic arc Dynamic arc

1.8

54

1.8

75

IMRT or conformal arc IMRT

Technique

*Current treatment policy involves a 10% dose de-escalation every 3 years contingent upon unchanged tumor control rate.

lesions, we have achieved high tumor control rates and cranial nerve preservation rates with cumulative doses ranging from 50.4 to 54 Gy (Table 25-3). As one exception, we have achieved high tumor control rates with acoustic neuromas but still document loss of serviceable hearing after treatment. We have therefore adopted a policy of a 10% dose de-escalation every 3 years, attempting to seek the lowest dose yielding the highest hearing preservation rate while maintaining high tumor control rates. Details of our experiences with different intracranial targets are discussed in the following section. Pretreatment patient preparation involves the customized design of a lightweight relocatable frame based on either an upper arch Reprosil dental mold (Gill-Thomas-Cosman frame [35]) or a thermoplast mask of the face. Either template is adapted to the frame and yields a reproducible and accurate frame relocation each time it is applied for treatment. On this day, both magnetic resonance imaging (MRI) and computed tomography (CT) data are obtained. Contrast-enhanced CT data are obtained in the frame with a ﬁducial cage attached. The patient is removed from the frame and a gadoliniumenhanced magnetic resonance (MR) scan is obtained next. Both imaging data sets are electronically transferred to the treatment planning workstation where they are fused into one composite image for treatment planning purposes. Due to the high spatial

ﬁdelity of CT data, the CT data set is an obligatory imaging data set for treatment planning. The patient is discharged home and returns for treatment inception usually a week to 10 days later. The Novalis treatment planning workstation provides a number of planning options ranging from dynamic arc treatment [36] to conformal static arcs or stereotactic IMRT. At our institution, most treatment plans involve a single isocenter treatment with ﬁve non-coplanar arcs utilizing the dynamic arc method. With mini-multileaf collimation, this technique allows for both high target conformality and high dose homogeneity, variables commonly considered to yield the highest therapeutic index. More complex skull base lesions are treated with either conformal arc or, if highly irregular concave surfaces are involved, stereotactic IMRT. Treatment planning is highly effective and efﬁcient due to the software design, which is capable of parallel treatment plans for comparison. Table 25-4 summarizes current dose fractionation protocols at the Jefferson Hospital for Neuroscience.

Clinical Outcomes Optic Nerve Sheath Meningiomas We initiated FSR in the early 1990s with the belief that, for selected intracranial targets, a conventional fraction FSR technique would yield higher cranial nerve preservation rates due to a greater radiobiological advantage. This was thought to be particularly true for special sensory cranial nerves, and the most compelling example to date is the treatment of optic nerve sheath meningiomas (ONSMs). A typical FSR treatment plan is featured in Figure 25-4. Historically, ONSM was either observed or ﬁrst treated surgically and then observed, and in all cases patients lost vision ineluctably to blindness in the affected eye. In the modern MR era, there is no longer any surgical indication to establish a diagnosis as primary ONSM has a classic axial tram-track appearance or coronal bull target appearance in fat-suppressed gadolinium-enhanced MRI scans. Traditional radiation achieved tumor control and stabilization of vision but associated morbidities including hypopituitarism, contralateral visual loss from

TABLE 25-4. Audiometric outcomes for two dose cohorts after FSR treatment of acoustic neuroma. Pretreatment Dose cohort

50.4 Gy Audiometric data All measurable (71) Serviceable (48) Corrected for follow-up and initial SNHL1 (27) 46.8 Gy Audiometric data All measurable (29) Serviceable (25) Corrected for follow-up and initial SNHL1 (14)

Posttreatment

PTA (dB)

SDS (%)

PTA

DPTA

SDS

DSDS

34 24 29

74 89 88

48 40 43

–142 –164 –146

68 69 66

–63 –205 –227

26 22 27

78 83 85

35 30 32a

–98 –810 –512

57 77 76

–219 –611 –913

Pretreatment audiograms limited to patients with PTA between 15 and 50 dB and follow-up limited to 122 weeks; ap = 0.0331 vs. 50.4 dose cohort, unpaired t-test; p < 0.0001 vs. pretreatment PTA, paired t-test; 3p = 0.0002 vs. pretreatment SDS, paired t-test; 4p < 0.0001 vs. pretreatment PTA, paired t-test; 5p = <0.0001 vs. pretreatment SDS, paired t-test; 6p = 0.0002 vs. pretreatment PTA, paired t-test; 7p = 0.0002 vs. pretreatment SDS, paired t-test; 8p = 0.003 vs. pretreatment PTA, paired t-test ; 9p = 0.03 vs. pretreatment SDS, paired t-test; 10p = 0.012 vs. pretreatment PTA, paired t-test; 11p = 0.028 vs. pretreatment SDS, paired t-test ; 12p = 0.039 vs. pretreatment PTA, paired t-test; 13p = 0.08 vs. pretreatment SDS, paired t-test. 1

2

25.

acoustic neuromas and other benign tumors: fractionated stereotactic radiotherapy perspective

295

FIGURE 25-5. Humphrey automated perimetry (24-2 central threshold) of left visual ﬁeld in patient with left ONSM at pretreatment (left) and 46 weeks after FSR (right).

geneity along the axis of the optic nerve with dynamic arc technique (Fig. 25-4).

Other Parasellar Tumors FIGURE 25-4. Axial T1-weighted image of left optic nerve sheath meningioma treated with dynamic arc FSR. Dose prescription is 1.8 Gy to the 91% isodose line for a total dose of 52.2 Gy. Organs at risks (OARs) are optic nerves (shaded area).

radiation-induced optic neuropathy, or loss of cognitive function represented too high a dose to the optic nerve in a single fraction; it is associated with a high risk of visual loss from radiation injury and is therefore not practiced. We initiated conventional fraction FSR in 1995 after we commissioned the ﬁrst Varian 600SR, and, taking advantage of our strong afﬁliation with Wills Eye Hospital, addressed a large population of patients with ONSM. Based on the data published by Parsons et al., we initiated a dose fraction FSR program of 1.8-Gy daily fractions to a cumulative dose of 54 Gy (Table 25-1) and published our initial results involving 33 optic nerves in 30 patients in 2002 [33]. These data reﬂected a 90% actuarial visual preservation rate. Two patients lost vision after treatment both of which were scored as treatment-related morbidities despite lack of adequate follow-up data to determine whether visual loss was actually related to treatment failure. We have not seen late radiation-induced optic neuropathy at a median follow-up of 5 years. When compared with observation or surgery, visual outcomes were signiﬁcantly better, and we concluded that patients with ONSM should be treated upfront solely on radiographic diagnostic criteria [30]. In a more detailed follow-up ophthalmologic analysis of outcomes in a subset of these patients, we have noted improvements in not only visual ﬁelds (Fig. 25-5) but also refracted visual acuity and color perception [30]. Despite these promising results, our initial experience was based on the use of circular collimators and, given the linear shape of ONSM, necessarily involved multiple overlapping isocenters. This created dose inhomogeneity along the axis of the optic nerve without apparent consequence at daily fractions of 1.8 Gy and no excessive incidence of radiation optic neuropathy was observed. With the advent of the Novalis unit, we have since adopted treatments based on a single isocenter, which are much more efﬁcient and which yield much higher dose homo-

We have treated a variety of benign tumors in the parasellar region including craniopharyngiomas (N = 25), nonfunctioning pituitary macroadenomas (N = 68) [31], and parasellar/cavernous sinus meningiomas (N = 150) with FSR. A typical FSR treatment plan for a nonfunctioning pituitary macroadenoma is featured in Figure 25-6. For craniopharyngiomas and nonfunctioning pituitary macroadenomas, we have advocated the integration of surgery with follow-up FSR. For incidentally found lesions (usually nonfunctioning pituitary tumors), we have advocated careful observation with serial MRI scans, visual ﬁelds, and optical coherence tomography [34] at 6-month intervals. The goal of surgery is establishing the diagnosis and decompressing the optic apparatus. If tumor resection is subtotal, our goal is to minimize the target volume for FSR in the postoperative period, thereby minimizing the risk of radiation optic neuropathy. In the case of craniopharyngiomas, we have recommended a transcranial pterional approach to these tumors, although our recent innova-

FIGURE 25-6. Axial T1-weighted image of residual pituitary macroadenoma treated with dynamic arc FSR. Dose prescription is 1.8 Gy to the 92% (innermost) isodose line to a total dose of 50.4 Gy. Organs at risks (OARs) include optic nerves chiasm and brain stem.

296

d.w. andrews et al.

tions in the endoscopic extended transnasal approach has promise to replace this approach as a less-invasive alternative (Evans JJ, unpublished observations). Because craniopharyngiomas are often more posterior in the parasellar region, often extending to the lamina terminalis and into the third ventricle, we prefer the transcranial approach over the transsphenoidal route because the latter often involves traversing the normal gland with potential disruption of pituitary function, a more limited access to the tumor, and potentially a less-effective optic apparatus decompression. For residual disease, we have proceeded with FSR as soon as feasible because residual tumors will often recur with cystic components that compromise vision. In contradistinction to craniopharyngiomas, nonfunctioning pituitary macroadenomas are always initially approached transsphenoidally with an image-guided transnasal endoscopic technique [37]. More recently with image-guided transnasal endoscopic resections, we have achieved more aggressive resections leaving little minimal tumor residuum (Evans JJ, unpublished observations). We and others have found that residual tumor will shrink to a minimum size over the ﬁrst 3 months of the postoperative period, after which if residual tumor remains, we recommend FSR. Our data as well as published Gamma Knife data favor treatment over observation [38]. Any visual compromise attributed to residual tumor that occurs in this 3-month period is addressed with reoperation. If vision remains stable but considerable tumor remains at 3 months, we recommend a transcranial resection to decompress the optic apparatus and achieve tumor volume reduction for a planned FSR treatment thereafter. We do not recommend surgery for parasellar meningiomas unless patients manifest rapidly progressive cranial neuropathies or the tumor is causing edema with shift. As with pituitary tumors, we recommend careful observation for tumors found incidentally until growth is documented on serial MRI scans, usually annually or biannually. We include Humphrey automated perimetry and optical coherence tomography for tumors at baseline and at 6-month intervals if a meningioma is near the optic apparatus. We also stipulate the addition of fatsuppression MRI scans if a meningioma is invading the orbit. As with optic nerve sheath meningiomas, we recommend FSR as the optimal treatment for patients with subacute symptoms associated with their tumor, usually visual loss or diplopia and, less frequently, facial numbness. Collectively for parasellar tumors, we have thus far observed with FSR an overall actuarial visual preservation rate of 93% and a tumor control rate of 95% at a median follow-up of 282 weeks. We noted three cases of post-FSR cyst formation requiring stereotactic placement of an Ommaya reservoir. In all three cases, only one or two aspirations were necessary after which no further accretion of cystic ﬂuid was noted. For the entire parasellar cohort, we have noted only six cases of FSR treatment-related hypopituitarism for a pituitary function preservation rate function of 98%. For the parasellar meningiomas, we have seen no incidence of treatment-related cavernous sinus cranial neuropathies. Fifty-two patients had visual deﬁcits that were potentially reversible with treatment. At the latest follow-up, 11 (21%) patients had a stable visual examination as measured by refraction, near card evaluation, confrontation, or automated perimetry. Twenty-seven (52%) patients had documented

improvement in vision, and three (6%) patients had complete resolution of all visual deﬁcits [39]. Fifty-nine patients had cranial neuropathies, which did not involve vision. Of these, 18 (31%) patients had stable symptoms, 15 (25%) patients had improvement in their deﬁcits, and 11 (19%) patients had complete resolution of all cranial neuropathies. Nine (15%) patients had progressive cranial nerve ﬁndings, ﬁve of which were attributed to progressive disease.

Complex Skull Base Tumors Additional indications for a fractionation technique include large, irregularly shaped skull base lesions, typically meningiomas, or skull base lesions considered to be radioresistant such as clival chordomas or chondrosarcomas. Large skull base meningiomas often have an en plaque component and may invade the skull base with brain-stem compression. Patients with either subacute cranial neuropathies or chronic global headache pain are candidates for FSR. With Novalis treatment planning software, either single isocenter dynamic arc technique or static ﬁeld stereotactic IMRT technique has yielded highly conformal treatment plans with homogeneous dose delivery minimizing treatment morbidity to contiguous dose-limiting brain tissues. A typical FSR treatment plan for a complex cranial base meningioma is featured in Figure 25-7. For treatment of large, irregularly shaped skull base lesions, Nakamura et al. found that IMRT achieved a consistently higher conformality index when compared with Gamma Knife treatment (means of 1.25 and 1.52, respectively) [31]. We have also achieved high-conformality treatment plans with IMRT, which we have found to be superior not only to Gamma Knife but also to dynamic arc FSR. With others [40, 41], our data support IMRT as the treatment of choice for these lesions. Clival chordomas are radioresponsive but only at high doses of radiation therapy [40]. The radiosurgery experience at the Mayo Clinic reﬂects actuarial tumor control rates of 89% and 32% at 2 and 5 years, respectively, at a median isodose prescrip-

FIGURE 25-7. Axial T1-weighted image of medial sphenoid wing meningioma treated with stereotactic IMRT. Dose prescription is 1.8 Gy to the 93% isodose line (innermost) to a total dose of 54 Gy. Dose distribution spares the brain stem.

25.

acoustic neuromas and other benign tumors: fractionated stereotactic radiotherapy perspective

297

tion of 15 Gy to the tumor margin (range, 10 to 20 Gy). Of 29 patients with cranial base chordoma or chondrosarcoma, 7, or 28%, had recurrence after treatment with three in-ﬁeld and four out-of-ﬁeld recurrences. Our experience has been more limited, but 7 of 8 patients have remained free of disease progression at a median follow-up of 4 years after FSR doses of 70 Gy, for a control rate of 88%. We have noted one patient with out-of ﬁeld recurrences ﬁrst at the base of the clivus and subsequently in the upper cervical spine. Our experience has been limited to the use of circular collimators, and with the advent of multileaf collimation or IMRT, higher, more conformal, and more homogeneous doses should be achieved.

Acoustic Neuromas Numerous studies have drawn comparisons between surgical outcomes and single-fraction radiosurgery in the treatment of acoustic neuromas, and radiosurgery has consistently and conclusively demonstrated superior outcomes with higher cranial nerve preservation rates and signiﬁcantly greater patient satisfaction [11, 42]. Radiosurgery results improved in the early to mid-1990s due to dose de-escalation to the current standard of 12 to 12.5 Gy prescribed to the 50% isosurface. Table 25-4 summarizes tumor control rates and cranial nerve outcomes associated with these downward dose iterations. As with ONSM, we anticipated a radiobiological advantage to fractionation when dealing with a special sensory cranial nerve, in this case the cochlear nerve in its close approximation to acoustic neuromas. In a prospective comparison of Gamma Knife treatment and conventional fraction FSR, we have noted comparable radiologic response rates, tumor control rates, and comparably low rates of both facial and trigeminal neuropathies after treatment. The one notable difference in outcomes, however, were hearing preservation rates. FSR yielded signiﬁcantly higher measurable and serviceable hearing preservation rates after treatment utilizing the Gardner-Robertson scale as a standard audiometric grading scale [12]. As a result of this comparison, we have routinely treated patients maintaining measurable but diminished hearing with conventional fraction FSR. Based on a retrospective analysis of cases we have either observed or treated, we are currently recommending observation for newly diagnosed patients with intact and symmetric hearing without associated sensorineural hearing loss (Fig. 25-1). Indications for treatment include any sensorineural hearing loss in the affected ear or documented tumor progression on serial MRI scans. Based on data derived from the single-fraction radiosurgery literature, we initiated a policy of dose de-escalation every 3 years if tumor control rates remained unchanged. Our initial decrease was instituted in November 2002 and results are summarized in Tables 25-3 and 25-4. Raw hearing preservation rates were consistently improved in the lower-dose cohort, which approached statistical signiﬁcance when corrected for comparable pretreatment hearing levels and follow-up period. A typical FSR treatment plan for an acoustic neuroma is featured in Figure 25-8. This was further borne out by the audiometric data, which reﬂected shallower decay rates for both pure tone average (PTA) and speech discrimination score (SDS) in the posttreatment period. When once again corrected for comparable pretreatment hearing levels and follow-up

FIGURE 25-8. Axial T1-weighted image of right acoustic neuroma treated with dynamic arc FSR. Dose prescription is 1.8 Gy to the 90% isodose line (innermost) to a total dose of 46.8 Gy. Organs at risk (OAR) is brain stem (shaded).

period, PTA was signiﬁcantly improved in the lower-dose cohort (p = 0.0331 vs. 50.4 Gy cohort, unpaired t-test), and SDS was not signiﬁcantly different in a paired comparison to the pretreatment value in the lower dose cohort (Table 25-4). These data reﬂect more favorable hearing outcomes in the lower-dose cohort and support a dose-seeking paradigm with further downward dose iterations with concomitant assessments of tumor control rates. Our data conﬂict with a recent report in which the authors reported an almost perfect serviceable hearing preservation rate (98%) at a higher dose of 57.6 Gy [5], but because no actual audiometric data were published, it remains difﬁcult to draw a comparison of hearing outcomes. The ideal FSR dose prescription should achieve both high tumor control rates, judged by serial MRI scans, and the highest hearing preservation rates, judged by serial audiometry.

Conclusion Fractionated stereotactic radiotherapy broadens the treatment armamentarium for a variety of benign intracranial tumors, providing a high therapeutic index often associated with improved or restored cranial nerve function. Particularly for tumors involving or near special sensory cranial nerves, FSR should be considered the primary treatment modality.

References 1. Shrieve DC, Hazard L, Boucher K, et al. Dose fractionation in stereotactic radiotherapy for parasellar meningiomas: radiobiological considerations of efﬁcacy and optic nerve tolerance. J Neurosurg 2004; 101(Suppl 3):390–395. 2. 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. 3. Goldsmith BJ, Rosenthal SA, Wara WM, et al. Optic neuropathy after irradiation of meningioma. Radiology 1992; 185:71–76. 4. Parsons JT, Bova FJ, Fitzgerald CR, et al. Radiation optic neuropathy after megavoltage external-beam irradiation: analysis of time-dose factors. Int J Radiat Oncol Biol Phys 1994; 30:755–763.

298

d.w. andrews et al.

5. Gardner G, Robertson JH. Hearing preservation in unilateral acoustic neuroma surgery. Ann Otol Rhinol Laryngol 1988; 97:55–66. 6. Hirsch A, Noren G. Audiological ﬁndings after stereotactic radiosurgery in acoustic neurinomas. Acta Otolaryngol 1988; 106:244– 251. 7. Flickinger JC, Lunsford LD, Coffey RJ, et al. Radiosurgery of acoustic neurinomas. Cancer 1991; 67:345–353. 8. Foote RL, Coffey RJ, Swanson JW, et al. Stereotactic radiosurgery using the gamma knife for acoustic neuromas. Int J Radiat Oncol Biol Phys 1995; 32:1153–1160. 9. Kondziolka D, Lunsford LD, McLaughlin MR, et al. Long-term outcomes after radiosurgery for acoustic neuromas. N Engl J Med 1998; 339:1426–1433. 10. 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 Radiat Oncol Biol Phys 2001; 50:1265–1278. 11. Karpinos M, Teh BS, Zeck O, et al. Treatment of acoustic neuroma: stereotactic radiosurgery vs. microsurgery. Int J Radiat Oncol Biol Phys 2002; 54:1410–1421. 12. Regis J, Pellet W, Delsanti C, et al. Functional outcome after gamma knife surgery or microsurgery for vestibular schwannomas. J Neurosurg 2002; 97:1091–1100. 13. Iwai Y, Yamanaka K, Shiotani M, et al. Radiosurgery for acoustic neuromas: results of low-dose treatment. Neurosurgery 2003; 53:282–287; discussion 287–288. 14. Flickinger JC, Kondziolka D, Niranjan A, et al. Acoustic neuroma radiosurgery with marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys 2004; 60:225–230. 15. van Eck AT, Horstmann GA. Increased preservation of functional hearing after gamma knife surgery for vestibular schwannoma. J Neurosurg 2005; 102(Suppl):204–206. 16. Hasegawa T, Kida Y, Kobayashi T, et al. Long-term outcomes in patients with vestibular schwannomas treated using gamma knife surgery: 10-year follow up. J Neurosurg 2005; 102:10–16. 17. Kagei K, Shirato H, Suzuki K, et al. Small-ﬁeld fractionated radiotherapy with or without stereotactic boost for vestibular schwannoma. Radiother Oncol 1999; 50:341–347. 18. Chung HT, Ma R, Toyota B, et al. Audiologic and treatment outcomes after linear accelerator-based stereotactic irradiation for acoustic neuroma. Int J Radiat Oncol Biol Phys 2004; 59:1116– 1121. 19. Sawamura Y, Shirato H, Sakamoto T, et al. Management of vestibular schwannoma by fractionated stereotactic radiotherapy and associated cerebrospinal ﬂuid malabsorption. J Neurosurg 2003; 99:685–692. 20. Combs SE, Volk S, Schulz-Ertner D, et al. Management of acoustic neuromas with fractionated stereotactic radiotherapy (FSRT): long-term results in 106 patients treated in a single institution. Int J Radiat Oncol Biol Phys 2005; 63:75–81. 21. Poen JC, Golby AJ, Forster KM, et al. Fractionated stereotactic radiosurgery and preservation of hearing in patients with vestibular schwannoma: a preliminary report. Neurosurgery 1999; 45: 1299–1305; discussion 1305–1307. 22. Cobery S, Remis M, Bradford C, et al. Five session Gamma Knife treatment of acoustic neuromas. Presented at 7th International Stereotactic Radiosurgery Society Congress, Brussels, Belgium, 2005. 23. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas: preservation of function versus size. J Clin Neurosci 2003; 10:48–52. 24. Pan CC, Eisbruch A, Lee JS, et al. Prospective study of inner ear radiation dose and hearing loss in head-and-neck cancer patients. Int J Radiat Oncol Biol Phys 2005; 61:1393–1402.

25. Flickinger JC, Kondziolka D, Pollock BE, et al. Evolution in technique for vestibular schwannoma radiosurgery and effect on outcome. Int J Radiat Oncol Biol Phys 1996; 36:275–280. 26. Flickinger JC, Kondziolka D, Niranjan A, et al. Results of acoustic neuroma radiosurgery: an analysis of 5 years’ experience using current methods. J Neurosurg 2001; 94:1–6. 27. Nedzi LA, Kooy H, Alexander E 3rd, et al. Variables associated with the development of complications from radiosurgery of intracranial tumors. Int J Radiat Oncol Biol Phys 1991; 21:591– 599. 28. Das IJ, Downes MB, Corn BW, et al. Characteristics of a dedicated linear accelerator-based stereotactic radiosurgeryradiotherapy unit. Radiother Oncol 1996; 38:61–68. 29. Andrews DW, Silverman CL, Glass J, et al. Preservation of cranial nerve function after treatment of acoustic neurinomas with fractionated stereotactic radiotherapy. Preliminary observations in 26 patients. Stereotact Funct Neurosurg 1995; 64:165–182. 30. Andrews DW, Faroozan R, Yang BP, et al. Fractionated stereotactic radiotherapy for the treatment of optic nerve sheath meningiomas: preliminary observations of 33 optic nerves in 30 patients with historical comparison to observation with or without prior surgery. Neurosurgery 2002; 51:890–902; discussion 903–904. 31. Behbehani RS, McElveen T, Sergott RC, et al. Fractionated stereotactic radiotherapy for parasellar meningiomas: a preliminary report of visual outcomes. Br J Ophthalmol 2005; 89:130–133. 32. Coke C, Andrews DW, Corn BW, et al. Multiple fractionated stereotactic radiotherapy of residual pituitary macroadenomas: initial experience. Stereotact Funct Neurosurg 1997; 69:183–190. 33. Kalapurakal JA, Silverman CL, Akhtar N, et al. Improved trigeminal and facial nerve tolerance following fractionated stereotactic radiotherapy for large acoustic neuromas. Br J Radiol 1999; 72:1202–1207. 34. Paek SH, Downes MB, Bednarz G, et al. Integration of surgery with fractionated stereotactic radiotherapy for treatment of nonfunctioning pituitary macroadenomas. Int J Radiat Oncol Biol Phys 2005; 61:795–808. 35. Gill SS, Thomas DG, Warrington AP, et al. Relocatable frame for stereotactic external beam radiotherapy. Int J Radiat Oncol Biol Phys 1991; 20:599–603. 36. Solberg TD, Boedeker KL, Fogg R, et al. Dynamic arc radiosurgery ﬁeld shaping: a comparison with static ﬁeld conformal and noncoplanar circular arcs. Int J Radiat Oncol Biol Phys 2001; 49:1481–1491. 37. Kanamori A, Nakamura M, Matsui N, et al. Optical coherence tomography detects characteristic retinal nerve ﬁber layer thickness corresponding to band atrophy of the optic discs. Ophthalmology 2004; 111:2278–2283. 38. Moses RL, Keane WM, Andrews DW, et al. Endoscopic transseptal transsphenoidal hypophysectomy with three-dimensional intraoperative localization technology. Laryngoscope 1999; 109: 509–512. 39. Losa M, Valle M, Mortini P, et al. Gamma knife surgery for treatment of residual nonfunctioning pituitary adenomas after surgical debulking. J Neurosurg 2004; 100:438–444. 40. Nakamura JL, Pirzkall A, Carol MP, et al. Comparison of intensity-modulated radiosurgery with gamma knife radiosurgery for challenging skull base lesions. Int J Radiat Oncol Biol Phys 2003; 55:99–109. 41. Pirzkall A, Debus J, Haering P, et al. Intensity modulated radiotherapy (IMRT) for recurrent, residual, or untreated skull-base meningiomas: preliminary clinical experience. Int J Radiat Oncol Biol Phys 2003; 55:362–372. 42. Fagundes MA, Hug EB, Liebsch NJ, et al. Radiation 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; 33:579–584.

2 6

Pituitary Tumors Kintomo Takakura, Motohiro Hayashi, and Masahiro Izawa

Introduction Pituitary adenomas are tumors that make up a signiﬁcant portion of neurosurgical practice. According to the Japanese Brain Tumor Registry [1], pituitary adenomas (n = 6653) comprise 17.4% of all primary brain tumors (N = 38,273). Their clinical presentation can be divided into two groups: functioning adenomas that secrete excess pituitary hormones resulting in endocrinologic symptoms and signs, and nonfunctioning adenomas that do not secrete biologically active hormones. All adenomas, particularly when they grow and extend outside the sella turcica, will cause neurologic deﬁcit due to compression of surrounding tissues (e.g., bitemporal hemianopsia and hypopituitarism). The goals of treatment are to normalize endocrinologic function and/or to relieve compression to the optic apparatus and the surrounding hypothalamus or normal pituitary gland. Current treatment methods include surgical removal of the adenoma, radiotherapy, hormonally directed medical therapy, and combination treatments. At present, the standard approach for the treatment of pituitary adenomas is surgical resection by transsphenoidal craniotomy. The effectiveness, safety, and comfort to patients of this operation has been fueled by developments in stereotactic guidance, intraoperative magnetic resonance imaging (MRI), and endoscopic techniques. Recently, stereotactic radiosurgery has gained greater acceptance as an adjunctive treatment. Radiosurgery is particularly useful for cavernous sinus extension, postsurgical residual tumor, and persistently elevated hormone levels after surgery. In this chapter, the treatment rationale and results of stereotactic radiosurgery for pituitary adenomas and other parasellar tumors such as craniopharyngioma and meningioma will be presented.

Historical Perspective The characteristic facial features of acromegaly were ﬁrst noted in the Egyptian era. Aldred and Sandison [2], in their study of archeological sculptures, described a limestone portrait from 1365 B.C. that showed Pharaoh Akhenaten with acromegalic facies and eunuchoid obesity. They postulated that the acromegaly developed from excess growth hormone secretion and

the obesity was a result of hypopituitarism from normal pituitary gland compression. In 1889, Sir Victor Horsley ﬁrst performed pituitary tumor surgery using the transfrontal, transcranial approach to remove an adenoma that had caused chiasmal compression and blindness [3]. Before Horsley, Caton and Paul in 1883 treated an acromegalic patient presenting with headaches and blindness by temporal lobe decompression without addressing the pituitary adenoma [4]. In 1907, Schloffer used an incision above and to the left of the nose, which allowed the nose to be ﬂapped downward, to directly access the sphenoid sinus and pituitary [5]. Eiselsberg modiﬁed Schloffer’s approach by making a long bilateral brow incision with a horizontally placed left lateral rhinotomy incision ending at the left alar foot [6]. Those surgical techniques were aesthetically unacceptable and led to development of translabial and transnasal approaches. Halstead developed the sublabial or oronasal incision for the transnasal approach [7], and Cushing, who performed his ﬁrst pituitary operation in 1910, used a sublabial incision with submucosal resection of the septum en route to the sphenoid sinus and the pituitary gland [8]. Between 1910 and 1925, transcranial approaches were rare as both Cushing and Hirsch primarily used the transnasal-transseptal route with a mortality rate of less than 5%. By the late 1920s, however, Cushing abandoned the transnasal approach in favor of the transcranial route with most neurosurgeons following his lead. Cushing felt that craniotomy was safer with better restoration of vision and fewer recurrences. Hirsch continued to champion the transsphenoidal approach and he later reported on 425 procedures treated in this manner [9]. Cushing’s main work in the early years of the 20th century was directed toward the elucidation of the anatomy and functions of the pituitary gland by various analyses and experiments. He gave his celebrated oration in surgery, “The Hypophysis Cerebri; Clinical Aspects of Hyperpituitarism and of Hypopituitarism,” to the Section of Surgery of the American Medical Association at the 16th Annual Session held in Atlantic City in 1909 [10]. This lecture was a historic landmark for the pathophysiology and treatment of pituitary adenomas. Cushing clariﬁed the pathophysiologic evidence of the basophilic pituitary adenomas and their clinical manifestation (pituitary basophilism) [11]. Bishop and Close recognized and reported a case in the same year as “A Case of Basophil Adenoma of the

299

300

k. takakura et al.

Anterior Lobe of the Pituitary: Cushing’s Syndrome.” This was the ﬁrst description of “Cushing’s Syndrome” [12]. In the 1950s and early 1960s, most pituitary tumors were removed by craniotomy. Sir Norman Dott learned the transsphenoidal approach from Cushing and he taught the technique to Gerard Guiot in France. Guiot reported his 495 patients operated on from 1957 to 1972 [13]. Hardy in Montreal, a pupil of Guiot, introduced the concept of selective removal of pituitary microadenomas and reported the results of 434 patients operated on from 1962 to 1972 [14]. Since the early 1970s, the transsphenoidal approach has been favored. In the 1990s, endoscopic surgery techniques were developed, and many neurosurgeons are now performing this operation with endoscopic assistance and image guidance. Radiotherapy for pituitary adenoma started in the early 20th century. Cushing reported that in some cases of pituitary adenomas treated by X-ray irradiation, the results were very encouraging, thereby justifying the efﬁcacy of radiotherapy in improving the therapeutic results of operative measures [15]. In the same period, Gramegna [16], Béclère [17], and Jaugeas [18] showed that in certain forms of pituitary tumors, prolonged radiation had a notable effect in ameliorating local symptoms, most likely due to a deﬁnitive shrinkage of the tumor growth. A widening of the constricted ﬁelds of visions served as a reliable index of any diminution in the size of the pituitary adenomas. By the middle of the 20th century, radiotherapy was commonly used as a postsurgical adjunct treatment for residual tumor after surgical removal. The beneﬁts of fractionated external beam radiotherapy on pituitary adenomas were reported by several investigators. Sheline and Tyrell [19] demonstrated an enlargement of the visual ﬁeld after radiotherapy. They demonstrated that this effect was related to the extent of the pretreatment visual ﬁeld deﬁcit. In those patients with one- to two-quadrant pretreatment loss, 6 of 10 patients improved after the radiotherapy, compared with only 28% after surgery alone, and no patients worsening after radiotherapy. However, where there was greater than two-quadrant visual ﬁeld loss, radiotherapy provided no beneﬁt, even though one third of the patients improved after surgery. Grigsby et al. [20] reported that when the dose given was less than 30 Gy, only 14% of the patients improved when compared with 40% given 40 to 49 Gy and 70% given 50 to 54 Gy. Endocrinologic improvement in functioning pituitary adenomas after radiotherapy has been demonstrated by several investigators. Eastman et al. [21] reported that in acromegalic patients, growth hormone (GH) level normalized in 73% of patients by 5 years, in 81% by 10 years, and in 90% by 15 years. Tsang et al. [22] reported that more than 80% of their patients had shown normalization of GH levels and tumor control was achieved in greater than 90% of the patients. In the cases of Cushing disease, Aristizabal et al. [23] reported that complete regression was noted in 28% of the patients. When the dose given was less than 40 Gy, only 23% of the patients improved or were cured compared with 57% in patients receiving doses greater than 40 Gy. Although the tumor-suppressive effect of fractionated radiation was evident, several complications have been noted. Postradiation optic neuropathy is the most hazardous complication. It should be noted that endocrinologic injury appears

very slowly, often more than 10 years after the completion of radiotherapy. Proton beam treatment for pituitary adenomas was ﬁrst performed in Boston and Berkeley. Kjellberg et al. [24] at Massachusetts General Hospital (MGH) treated 67% of the 510 acromegalic patients with proton beam as the primary treatment. GH levels of less than 5 ng/mL were found in 27.5% of the patients at 2 years, 75% at 10 years, and 92.5% at 20 years. Fabrikant et al. [25] at Berkeley reported similar results. At both institutions, a number of years were required for patients to reach GH values of less than 5 ng/mL, but there was a much more rapid decline in elevated GH levels during the ﬁrst 2 years than in the fractionated series. In patients with Cushing disease at MGH, Kjellberg et al. [24] reported the remission rates of 55% at 2 years and 80% at 5 years. The new concept of stereotactic radiosurgery was introduced by Lars Leksell in 1951 as a method to destroy an intracranial target tissue by using radiation without opening the skull [26]. Leksell was the ﬁrst to treat a patient with pituitary adenoma using Gamma Knife. Because Gamma Knife focuses a tumor target with a high radiation dose at one time, it is able to minimize damage to the surrounding normal structures. Thus, stereotactic radiosurgery has gradually developed into one of the standard adjunct therapeutic methods for treating pituitary adenomas. Gamma Knife radiosurgery has been performed worldwide in 24,604 pituitary adenoma patients as of December 2004. In Japan, 2696 patients with pituitary adenomas have been treated with Gamma Knife during the same period. Finally, in the last decades of the 20th century, pharmacologic research has advanced in the ﬁeld of endocrinologic control of pituitary hormones, especially with the analogous compounds that have the activity to control hormone secretion. It was discovered that dopamine agonist bromocriptine has a strong effect to control prolactin production, and bromocriptine is, today, used routinely to treat prolactin-producing pituitary adenoma (prolactinoma). Similar investigations for each pituitary hormone such as growth hormone and adrenocorticotropic hormone (ACTH) continue actively.

Pathophysiology According to the Japan Brain Tumor Registry, 2000 [1], primary tumors that develop in the region of the pituitary gland (sella turcica) and chiasm are mainly pituitary adenomas (77.1%), followed by craniopharyngiomas (15.8%), germinomas (3.3%), astrocytomas (1.0%), and other tumors (Table 26-1). Seventeen percent of all primary brain tumors consists of pituitary adenomas and develop generally in adults, but the age distribution is different for each type of adenoma. Pituitary adenomas are endocrinologically classiﬁed into two groups: functional adenomas that secrete pituitary hormone and nonfunctioning adenomas that do not secrete any known biologically active hormones. The incidence for each type of adenoma is shown in Table 26-2. Prolactin-producing pituitary adenoma (prolactinoma) is the most common functioning adenoma. It comprises 26% of all pituitary adenomas [1]. The incidence in females is almost three times more than in males. The symptoms of hyperprolac-

26.

pituitary tumors

TABLE 26-1. Primary tumors in the region of the pituitary gland. Tumor

No. of cases

Ratio (%)

Pituitary adenoma Craniopharyngioma Germinoma Astrocytoma Malignant astrocytoma Teratoma Epidermoid Chordoma Dermoid Others

4882 997 211 65 10 19 26 11 21 87

77.1 15.8 3.3 1.0 0.2 0.3 0.4 0.2 0.3 1.4

Total

6329

100.0

tinemia are amenorrhea and galactorrhea (Forbes-Albright syndrome) in women and impotence and infertility in men. The symptoms appear at relatively younger age, mainly from 20 to 45 years. In men, prolactinoma tends to present with visual complaints such as a bitemporal hemianopsia or decreasing visual acuity; therefore, they are more often found as macroadenomas compressing the optic nerves. In women, the hormonal manifestations are more apparent and prolactinomas tend to present earlier as microadenomas. The diagnosis is made from serum prolactin levels and MRI. Medical therapy with dopamine agonists is usually effective in suppressing prolactin hypersecretion and can reduce tumor size. Growth hormone–producing pituitary adenomas (acromegaly, gigantism) make up 23% of all pituitary adenomas [1]. The incidence is the same in men and women. Excess growth hormone causes acromegalic features, hypertension, diabetes, cardiac hypertrophy, cardio- and cerebrovascular disorders. Growth hormone secretion may be controlled by bromocriptine or octreotide, but the results are not as dramatic as those seen in prolactinomas. ACTH-secreting pituitary adenomas (Cushing disease and Nelson syndrome) make up 5% of all pituitary adenomas. The incidence is four times more in women compared with men [1]. The onset of the disease appears from age 25 to 60 years. In the differential diagnosis, ectopic ACTH-producing tumors such as carcinomas of the thymus and pancreas, bronchial carcinoids, and pheochromocytoma must be considered. A PET scan of the whole body can also help to detect ectopic tumors. Nelson syndrome occurs after a bilateral adrenalectomy for Cushing TABLE 26-2. Incidence of each subtype of pituitary adenoma. Type of adenoma

No. of cases

Ratio (%)

Nonfunctioning Growth hormone producing Prolactin producing ACTH producing Other hormone producing Malignant adenoma

2237 1162 1324 255 95 6

44.0 22.9 26.1 5.0 1.9 0.1

Total

5079

100.0

Source: Committee of Brain Tumor Registry of Japan. Report of Brain Tumor Registry of Japan. 10th edition. Neurol Med Chir 2000; 40(Suppl):5–11.

301

disease resulting in the pituitary adenoma secreting an excess of ACTH and resulting in rapid tumor growth. Other pituitary hormone–producing adenomas make up 2% of all pituitary adenomas. Patients with TSH-producing adenoma show mild hyperthyroidism with increased basal metabolic rate, elevated serum thyroxine, and an increase uptake of 131I by the thyroid. Gonadotropin-producing tumors (gonadotropinoma) are quite rare and are distinguished by having increased serum levels of LH or FSH. There are also plurihormonal adenomas that secrete multiple hormones such as GH and prolactin. Two possible mechanisms may explain the abnormal secretion caused by these adenomas. Corenblum et al. [27] have shown that such adenomas might be composed of two cell lines, one secreting GH and the other prolactin. An alternate hypothesis was described by Horvath and Kovacs [28] who found adenomas composed of a single cell type that contains both GH- and prolactin-secreting ability. Polysecretory pituitary adenomas of TSH- and prolactin-producing tumors [29] have also been found. Nonfunctioning pituitary adenomas make up 44% of all pituitary adenomas [1]. There is no sex predilection. Occasionally, nonfunctioning adenomas may secrete a partial gonadotropin such as the α or β subunit of LH or FSH. A null cell adenoma by deﬁnition secretes no hormone. Clinically, nonfunctioning pituitary adenomas are generally diagnosed by tumor growth causing headache or optic nerve compression. Malignant pituitary adenomas (pituitary carcinoma) number only 0.1% of all pituitary adenomas [1]. They invade surrounding brain and optic nerves and may exhibit meningeal spread. They can also metastasize to bone, liver, lung, and lymph nodes. Other rare pituitary tumors include granular cell tumor (choristoma, granular cell myoblastoma), which develops from posterior pituitary or pituitary stalk, and pituicytoma, which develops from pituicytes (a kind of astrocyte) in the posterior lobe of the pituitary gland. Craniopharyngiomas make up 16% of all tumors in the pituitary region (Table 26-1). A craniopharyngioma develops from the epithelial remnants of the Rathke pouch. Most of these tumors are found in the suprasellar region. Other parasellar tumors include meningioma, chordoma, germinoma, teratoma, epidermoid, and astrocytoma (Table 26-1). Metastases originating from breast, lung, and lymphoma are also found in the pituitary region.

Rationale for Treatment The goal of treatment for pituitary adenomas is to remove mass effect on optic nerve, chiasm, and normal pituitary gland and to normalize hormone function. Microsurgical removal of the pituitary adenoma via a transsphenoidal approach is now standard. However, if a large adenoma extends to the supra- or the parasellar region, a transcranial surgical approach may be needed. According to the Japan Brain Tumor Registry [1], the cumulative survival rate of pituitary adenoma is 95.5% (n = 2086) at 5 years after surgery. Even though surgical treatment of pituitary adenoma has become quite safe, there are still considerably high rates of complications: 1% mortality, 3.4% major morbidity (visual loss, ophthalmoplegia, etc.), and 4.6% lesser morbidity (sinusitis, septum perforation, diabetes insipidus, etc.) [30]. Long-term tumor control rates with surgical

302

k. takakura et al.

treatment alone are from 50% to 80% and recurrences are not uncommon [31–33]. Therefore, radiotherapy or radiosurgery after microsurgical removal of a tumor is useful to prevent recurrences. Long-term control of nonfunctioning pituitary adenomas with conventional radiotherapy alone and radiotherapy with surgery were studied by several investigators. Sheline and Tyrell [19] and Halberg and Sheline [34] reported that tumor control rates at 10 years after treatment were 71% by radiotherapy alone and 86% by radiotherapy with surgery. Grigsby et al. [20, 35] also reported similar results, with 80% and 90%, respectively, while Brada et al. [36] reported 92% and 98%, respectively. These authors showed better tumor control rates in the patients treated with both surgical removal and adjuvant radiotherapy when compared with those treated with radiotherapy alone. On the contrary, McColough et al. [37] reported an opposite result where patients treated with radiotherapy alone showed better tumor control: 100% (n = 29) at 10 years after the treatment compared with 92% in those who received radiotherapy and surgery (n = 76). Tumor control was better up to 10 years after treatment but they dropped after 10 years or more. Flickinger et al. [38] reported that tumor control at 5, 10, and 20 years after radiotherapy and/or surgery (n = 112) were 97%, 89%, and 72%, respectively. The standard strategy for the treatment of pituitary adenomas today is microsurgical removal by the transsphenoidal approach together with postsurgical stereotactic radiosurgery with Gamma Knife or some other stereotactic radiosurgical method. For large adenomas extending extensively to the suprasellar or the parasellar region, tumor removal by craniotomy might be needed. If the patient’s medical condition is poor, for example, severe cardiac failure making surgical intervention inappropriate, stereotactic radiosurgery without surgical treatment should be considered. Conventional fractionated radiotherapy might be needed if the residual tumor is extensive or compressing the optic nerve.

Selection of Patients By December 2004, 24,604 pituitary adenomas have been treated with Gamma Knife worldwide. In Japan, Gamma Knife was used to treat 2696 pituitary adenomas, which represented 3.6% of all tumors treated with Gamma Knife. In our Department of Neurosurgery at Tokyo Women’s Medical University, we treated 131 pituitary adenomas with Gamma Knife from 1994 to 2004; these patients formed 3.9% of all the 3360 patients treated with Gamma Knife. The types of pituitary adenoma treated are illustrated in Table 26-3. Among all adenomas, 37 were nonfunctioning adenomas, 57 were growth hormone– secreting adenomas, 23 were prolactinomas, and 14 were ACTH-producing adenomas. We also treated 38 craniopharyngiomas, 7 germ cell tumors, and 24 trigeminal schwannomas in the pituitary region in the same period (Table 26-3). In our department, Gamma Knife treatment for pituitary adenomas was generally performed after a surgical removal of the adenoma or in recurrent tumors. The criteria to select patients with pituitary adenoma for Gamma Knife treatment were • The distance from the tumor border to the optic nerve was more than 5 mm.

TABLE 26-3. Pituitary adenoma and pituitary region tumors treated by Gamma Knife at the Tokyo Women’s Medical University Hospital (1995–2004). Tumor

Pituitary adenoma Nonfunctioning Growth hormone producing Prolactin producing ACTH producing Craniopharyngioma Germ cell tumor Trigeminal schwannoma Total

No. of cases

131 37 57 23 14 38 7 24 200

• The adenoma was present in the cavernous sinus (residual tumor after a surgical removal). • Hormone level in the serum did not normalize after surgery. • The patient’s general condition was inappropriate for general anesthesia. Even if adenoma invaded into the cavernous sinus, but the total size of the adenoma was small, Gamma Knife treatment before surgery was considered. Tumor growth suppression also depended on the size of the adenoma. The average tumor size treated with Gamma Knife in our department was 2.5 mL.

Treatment Plan Among all pituitary adenomas, tumors extending in an anteriorinferior direction, mostly in the case of acromegaly, are the most difﬁcult to treat with Gamma Knife. Because the shape of an acromegalic face is quite different from a normal ﬁgure, there is sometimes no free space to set the frame and thus impossible to perform Gamma Knife treatment. To solve this problem, we made a special frame that has shorter posterior posts and has the anterior portion of the frame set extremely downward when compared with a regular frame. By using this frame, we can set the target position close to the center of the frame. The plane of this frame is almost parallel to the optic nerve. With that, we can avoid unnecessary radiation to the optic nerves. After setting the frame, we connect the box with indicators to the frame and perform image examinations with both computed tomography (CT) scan and MRI. All patients receive three sequences of MRI: (1) T2-weighted coronal image, 2.0-mm slices; (2) FE axial (modiﬁed time-of-ﬂight) 1.0-mm slices; and (3) GdFE axial 1.0-mm slices (early images). By using T2-weighted coronal images, the optic nerve is vividly visualized, and pituitary adenoma and normal pituitary gland can be well differentiated. The optic nerve shows low intensity but the surrounding cerebrospinal ﬂuid shows high-intensity signal. Therefore, the optic nerve is clearly visualized when compared with other methods. Normal pituitary gland and adenoma are well differentiated by using T2-weighted image with FE axial images [39]. The most important problem in a radiosurgical treatment is simulating an appropriate radiation dose distribution to the adenoma and the optic nerve. The effective radiation dose for

26.

pituitary tumors

controlling the growth of pituitary adenoma is relatively high. We are using a marginal dose of 20 Gy for nonfunctioning adenomas and 25 to 30 Gy for functioning adenomas. On the other hand, the dose to the optic nerves must be limited under 8 Gy. Therefore, a dose plan is focused on how to make a sharp dose distribution in the 5-mm space between the adenoma and the optic nerve. For that purpose, we use many small collimeters. The Gamma angle (angle of frame setting) and plugging technique must be employed efﬁciently. We are now also trying to make a more homogeneous dose distribution inside the adenoma. We introduced a new model of Gamma Knife, the C-APS (Automatic Positioning System; Elekta Instrument AB) in November 2002. By using this new type of Gamma Knife, radiation plans with multiple shots are automatically performed in a relatively short period of time. This beneﬁts not only patients but also doctors and the medical staff. The accuracy of C-APS Gamma Knife is 0.2 mm [40]. Before starting a Gamma Knife treatment, all tumor images including three sequences of MRI and thin-sliced CT bone images are incorporated into the GammaPlan (ELEKTA Instrument AB), and fusion images are produced. By displaying the relationship between adenoma and the surrounding tissue, a sharp dose distribution can be made between the tumor border and the optic nerve. We have been able to perform a safe conformal radiosurgery treatment to a patient in whom the distance between the adenoma and the optic nerve was only 2 mm. Now, we can even make a treatment plan for pituitary adenoma with 0.1 mm accuracy.

Clinical Results We have treated 131 cases of pituitary adenomas (Table 26-3), from 1995 to 2004, in which 43 cases were treated by the C-APS Gamma Knife. The treatment plan must be conformal with as

FIGURE 26-1. Therapeutic planning for a pituitary adenoma invading the cavernous sinus. (A) A T2-weighted coronal MRI scan, 2.0 mm thickness. The adenoma and the surrounding tissue (outer wall of cavernous sinus, normal pituitary gland, optic nerve) are well visualized. Images obtained by (B) CT and (C–F) modiﬁed time-ofﬂight with gadolinium enhancement show the accurate morphologic structure of the adenoma. The conformal and selective radiation was possible.

303

high a dose as possible to the tumor tissue itself, with a limited peripheral dose to avoid radiation to the surrounding tissue. Figures 26-1 and 26-2 illustrate examples of a treatment plan. In order to get a conformal radiation, multiple isocenters are used (Fig. 26-1). We treated 88 cases of pituitary adenoma from 1994 to 2001 by using the manual mode Gamma Knife (model B), in which 45 cases were followed up for more than 2 years (average, 36.5 months). The tumor control rate was 91.1% (acromegaly, 100%; prolactinoma, 85%; and Cushing disease, 88%). The endocrinologic improvement was seen in 70% of the patients (acromegaly, 90%; prolactinoma, 69%; and Cushing disease, 50%). Normalization of serum pituitary hormone level alone was seen in 27% of the patients (acromegaly, 41%; prolactinoma, 15%; and Cushing disease, 10%). During the follow-up period, tumor recurrence, complications such as optic neuropathy, or decreasing normal pituitary hormone level in serum were not observed. In 43 cases (21 nonfunctioning and 22 functioning adenomas) treated by C-APS Gamma Knife from 2002, 16 cases were followed up for more than one-half year (average, 17.1 months). The tumor control rate was 100%. The rate of tumor shrinkage was seen in 75% of the patients. Decrease of normal pituitary hormone levels, optic neuropathy, or worsening of any symptoms were not observed. In 10 cases of acromegaly, the tumor control rate was 100% (rate of tumor shrinkage, 50%), and endocrinologic improvement was seen in 70% of all patients. In two cases of prolactinoma, the tumor control rate and the rate of tumor shrinkage were both 100%. In four cases of Cushing disease, the tumor control rate was 100% and the rate of tumor shrinkage was 75%. Endocrinologic improvement was seen in 75%. The normalization of serum pituitary hormone level alone was seen in 50% of the patients. Normalization of serum pituitary hormone level was found more clearly in the patients treated by C-APS than those treated by Gamma Knife model B.

304

k. takakura et al. FIGURE 26-2. Three-dimensional display of pituitary adenoma invading the cavernous sinus appeared on the panel of the GammaPlan.

The shrinkage of pituitary adenoma in an acromegaly patient treated by Gamma Knife (model B) is shown in Figure 26-3. This patient had severe cardiac failure, and general anesthesia was not appropriate. The patient received Gamma Knife treatment without surgical removal of the tumor. Clear tumor shrinkage was seen at 10 months after the radiosurgery. The serum growth hormone level decreased soon after the radiosurgery and normalized within 2 years. The shrinkage of pituitary adenoma in another patient (35-year-old man, acromegaly) treated by Gamma Knife is shown in Figure 26-4. The dose plans of C-APS Gamma Knife treatment are shown in Figures 26-1 and 26-2. An example of a therapeutic plan for a pituitary adenoma invading the cavernous sinus is shown. Small, multiple isocenters are used for conformal and selective radiation to the adenoma and to avoid damage to the optic nerve, the normal pituitary gland, and the hypothalamus. Three-dimensional MRI clearly shows the morphologic detail of the adenoma and the surrounding tissues, such as the optic nerve, the carotid artery, the normal pituitary gland, and the outer wall of cavernous sinus (Fig. 26-2). Many investigators reported the effects of radiotherapy and radiosurgery using Gamma Knife for treating pituitary adenomas. Regarding fractionated radiotherapy, Flickinger et al. [38], Rush et al. [41], and Tsang et al. [42] reported tumor control rates of 76% to 97%. Endocrinologic improvement ranged from 38% to 70% (acromegaly, 55%; prolactinoma, 70%; and Cushing disease, 57%). Complications (decreasing visual acuity, hypopituitarism) occurred in 12% to 100% of the patients [38, 43, 44]. As for radiosurgery with Gamma Knife, the tumor control rate ranged from 93% to 96%, and the tumor shrinkage rate ranged from 46% to 57% [45, 46]. The rate of endocrinologic improvement ranged from 78% to 93% [45–47], and normalization rate of hormonal level ranged from 21% to 52%. Witt et al. [46] stated that acromegaly is most responsive to Gamma Knife radiosurgery (acromegaly, 92%; prolactinoma, 86%; and Cushing disease, 66%). Complications appeared in 12% to 100% of the patients [46].

Sheehan et al. [48] made a collective report from recently published papers on the efﬁcacy of radiosurgery on pituitary adenomas. The data examined a total of 1621 patients. The average marginal dose of radiation was from 14 to 32 Gy and

FIGURE 26-3. Effect of Gamma Knife treatment on MRI and serum growth hormone level. Acromegaly, 45-year-old woman: (A) before treatment; (B) at 10 months after Gamma Knife treatment; (C) serum growth hormone level before and after Gamma Knife treatment.

26.

pituitary tumors

16 M

FIGURE 26-4. Effect of Gamma Knife treatment on MRI. Acromegaly, 35-year-old man: (A) before treatment; (B) at 16 months after Gamma Knife treatment. Marked shrinkage of the adenoma is observed. Serum growth hormone level before treatment was 80 ng/mL and at 16 months after treatment was 15 ng/mL.

the maximal dose inside the adenoma ranged from 19 to 50 Gy. Approximately 90% of patients had tumor growth control. Hormone levels varied mainly due to the different postradiosurgical assessment criteria used. The risks of hypopituitarism, optic neuropathy, cerebrovascular disturbance, or radiation-induced neoplasm were lower when compared with those treated by fractionated radiotherapy. The tumor control rate for pituitary adenoma treated by Gamma Knife in recent years is 96% [48]. Reports also showed visual improvement from the decreasing size of adenomas after Gamma Knife treatments [49–51].

Other Tumors in the Pituitary Region Several other kinds of tumors have been found to develop in the sella turcica or parasellar region. These tumors are important when differentiating tumors from pituitary adenomas and should be considered when making a therapeutic strategy. Craniopharyngioma is the second most common tumor, following pituitary adenoma, in the pituitary region. It forms 15.8% of all primary brain tumors developed in this region (Table 26-1) [1].

FIGURE 26-5. Effect of Gamma Knife treatment on a craniopharyngioma in a 16-year-old boy (MRI): (A) before treatment; (B) 8 months after treatment.

305

Craniopharyngioma should ﬁrst be treated by surgery. Radiosurgery is indicated only when the residual tumor is present after a surgical removal of the tumor. The shrinkage of craniopharyngioma in a patient (16-year-old boy) treated by Gamma Knife is shown in Figure 26-5. Postsurgical radiotherapy has been shown to improve the outcome of surgical treatment. The Japan Brain Tumor Registry [1] showed that the mortality rates of patients with craniopharyngiomas were 5.0% (48/956) among those treated by surgery alone and 0.2% (1/432) among those treated by surgery and radiation (p < 0.01). The 5-year survival rates after surgery were 84.4% in the patients treated by surgery alone and 90.1% in those treated by surgery and radiation (p < 0.01). A combined surgical and stereotactic radiosurgery treatment is recommended to improve the outcome of treatment for craniopharyngiomas. Meningioma is often observed in the sellar and parasellar region. The incidence of meningioma is 7.2% (718/10,089 of all intracranial meningiomas) in the parasellar region and 4.5% (456/10,089) in the medial portion of the sphenoid ridge [1]. The wall of the cavernous sinus is a common site for the development of meningiomas. The initial treatment should be surgery, but a radical operation and its complications should be avoided as stereotactic radiosurgery can be used to treat the residual tumor. The shrinkage of a cavernous sinus meningioma in a patient (45-year-old woman) treated by Gamma Knife is shown in Figure 26-6. Several investigators reported good tumor control rates (90% to 100%) of meningiomas treated by Gamma Knife [52–57]. Germinoma is quite common in Japan and constitutes 3.3% of all tumors developed in the pituitary gland and chiasmal region (211/6329 of all tumors) [1]. Germinomas and other germ cell tumors including choriocarcinomas, embryonal carcinomas, yolk sac tumors, and teratocarcinomas are sensitive to radiation. Stereotactic radiosurgery is generally indicated for these tumors after veriﬁcation of the histologic diagnosis. Teratoma epidermoid, dermoid, and chordoma are sometimes observed in the pituitary gland and the surrounding region. Astrocytomas are also tumors that develop in the pituitary region. Pituicytomas, which develop from the posterior lobe of a pituitary gland, and granular cell tumors (choristoma, granular cell myoblastoma) are rare tumors but are found in this area. A Rathke cyst is not a neoplasm but is often found in the

FIGURE 26-6. Effect of Gamma Knife treatment on a cavernous sinus meningioma in a 45-year-old woman (MRI): (A) before treatment; (B) 19 months after treatment.

306

k. takakura et al.

pituitary gland region and is important for differentiating from a pituitary adenoma. The above-mentioned tumors can be well diagnosed by MRI and CT scan in addition to clinical data, but a histologic diagnosis is generally required. Metastatic tumors are often found in pituitary glands and the surrounding region. The pituitary gland is a common target organ of metastatic tumors. Metastasis to the pituitary gland or its capsule is observed in 6% of the autopsied cancer patients in our surgical series [58]. Abrams et al. [59], Hagerstrand et al. [60], and Kovacs [61] reported a 1.8%, 3.8%, and 1.8% incidence, respectively. Lymphomas, leukemias, and multiple myelomas often affect the pituitary gland and its capsule. Breast cancers are the most frequent metastatic tumors in the pituitary gland (50% of 205 cases with metastasis to the pituitary gland), followed by lung cancer (22%), cancer of the gastrointestinal tract, and the nasopharyngeal cancers. In 526 autopsied cases of breast cancer, pituitary gland involvement was found in 103 cases (20%). Gurling et al. [62], Duchen et al. [63], Hagelstrand and Pribram [64], and Oi et al. [65] reported the frequency of metastases involving the pituitary gland in 9% to 28% of their autopsied cases. The majority of metastatic tumors in the pituitary gland are asymptomatic. Diabetes insipidus is a major symptom, from which we found 6 of 103 patients had pituitary metastases from breast cancer. All patients had metastatic lesions on the posterior lobe of the pituitary gland, and ﬁve patients had extensive lesions that invaded the infundibulum or the hypothalamus. Tecars and Silverman [66] also reported six patients with diabetes insipidus among 88 patients with pituitary gland metastasis. Jones [67], Blotner [68], and Houck et al. [69] reported a 2.4% to 20% incidence of diabetes insipidus among patients with metastatic pituitary tumors. Hormonal disturbances are not detected during the clinical course of these patients. In our 205 patients with metastatic pituitary gland carcinoma, 102 patients had tumors in the posterior lobe, 43 had them in their anterior lobes, and 43 had them in the capsules [52]. On the contrary, in the 96 patients with metastatic lymphomas and leukemias, 84 had tumors in the capsule whereas 13 had them in the posterior lobe. Carcinoma has an obvious predilection for the posterior lobe whereas lymphomas and leukemias tend to involve the capsule. The unique vascular architecture of the pituitary gland probably explains why carcinomas metastasize more frequently into the posterior lobe than the anterior lobe, which has a very limited primary arterial supply that is supplied by the portal system, which enters through the posterior lobe of the pituitary gland [70,71]. Willis [72] reviewed 16 reports of blood-borne metastatic lesions to the pituitary gland, most of them originating from breast carcinoma. Koyama and Takakura [73] and Epstein et al. [74] reported similar cases of blood-borne metastatic tumors to the pituitary gland. The direct invasion of tumor to the pituitary gland is observed in skull base tumors, especially in cancers of the head and neck. Teoh [75] reported a 23% (7 of 31 cases) incidence of pituitary gland invasion from nasopharyngeal carcinomas. Inﬁltration of the pituitary gland by leukemic cells and malignant lymphomas is common. We observed inﬁltration in the pituitary gland in 28% of the patients with leukemia. Masse et al. [76] reported 46% pituitary gland metastases in leukemic patients, most of which were limited to the capsule of the pituitary gland. Although the capsule of the pituitary gland is the only site of involvement in many cases, inﬁltration of tumors

into the pituitary gland is observed in about one ﬁfth of all lymphomas and leukemia with pituitary involvement. Although stereotactic radiosurgery for metastatic tumors to the pituitary gland and the surrounding region is not often indicated, the treatment method should be considered when the general condition of a patient is good and a better quality of life is expected, as Gamma Knife treatment for intracranial metastatic tumor in general showed excellent results.

Complication Avoidance Damage to visual function after radiosurgery must be cautiously avoided. An upper limit of 8 to 10 Gy to the optic nerve is generally accepted [77, 78] for preventing post-radiosurgical complication in visual function. The tolerable level of radiation for the optic nerve and chiasm, however, depends on a number of factors [79]: the presence of optic nerve compression and previous fractionated radiotherapy inﬂuence the degree of damage to visual function. A safe distance from the pituitary adenoma to the optic nerve or chiasm for the stereotactic radiosurgery is considered to be more than 5 mm, but if the radiation dose to the optic nerve can be reduced to less than 10 Gy by accurate dose planning, a closer distance might be permissible. We have safely treated a patient with only 2 mm separation between the adenoma and the optic nerve. Hypopituitarism often occurs after fractionated radiotherapy, but it is less frequently found in patients treated by stereotactic radiosurgery [80, 81]. Fiegel et al. [82] reported that hypopituitarism after radiosurgery correlated with the radiation dose to the pituitary stalk. The incidence of radiation-induced neoplasms by radiosurgery is less than that of fractionated radiotherapy. Sheehan et al. [48] reviewed 35 reports and found no radiation-induced neoplasm in the total of 1621 patients treated by stereotactic radiosurgery. However, six radiation-induced neoplasms were found in patients treated by stereotactic radiosurgery from other studies [83–88]. These patients were among the more than 250,000 patients treated by Gamma Knife in the whole world. The ratio of the incidence is much less than that of patients treated by fractionated radiotherapy. Careful follow-up should be given to patients with pituitary adenoma treated by stereotactic radiosurgery.

Future Directions The much improved technology of imaging diagnosis has reﬁned the accuracy of treatment plans, and the much advanced radiosurgical system has increased treatment effects. With these new technologies, we can expect a total improvement of tumor suppressive effect and a quicker normalization of hormonal disturbance in patients with pituitary adenoma, and fewer complications caused by radiation. Radiosensitizer for radiosurgery of pituitary adenoma cells can be introduced to reduce the radiation dose and to prevent radiation-induced complications. Because we cannot expect a dramatic development of microsurgical techniques to remove adenomas, we should expect new drugs for the management of these tumors. Although not completely satisfactory, bromocriptine can cure 70% to 90% of patients with prolactinomas. For acromegaly and Cushing disease,

26.

pituitary tumors

further endocrinologic and pharmacologic research and the development of new drugs are to be expected. Although a pituitary adenoma is a benign tumor, a quick and safe therapeutic modality to improve visual disturbance and to normalize hormonal abnormalities without any undesirable complications should be investigated, and, when successful, the method should be introduced in the daily medical practice. Acknowledgments. The authors thank Prof. A Teramoto, Department of Neurosurgery, Nippon Medical University, for his great cooperation. This work was supported by the Encouraging Development of Strategic Research Centers, Special Coordination Funds for Promoting Science and Technology, Japanese Ministry of Education, Cultures, Sports, Science and Technology.

References 1. Committee of Brain Tumor Registry of Japan. Report of Brain Tumor Registry of Japan. 10th edition. Neurol Med Chir 2000; 40(Suppl):5–11. 2. Aldred C, Sandison AT. The Pharaoh Akhenaten; a problem in Egyptology and Pathology. Bull Hist Med 1963; 36:293. 3. Horsley V. On the technique of operations on the central nervous system. Br Med J 1906; 2:411–423. 4. Caton R, Paul FI. Notes of a case of acromegaly treated by operation. Br Med J 1893; 2:1421–1423. 5. Schloffer H. Erfolgreiche Operation eines Hypophysentumors auf Nasalem Wege. Wien Klin Wochenschr 1907; 20:621–624. 6. von Eiselsberg F. Operations upon the hypophysis. Trans Am Surg Assoc 1910; 28:55–72. 7. Halstead AE. Remarks on the operative treatment of tumors of the hypophysis; with the report of two cases operated on by an oronasal method. Trans Am Surg Assoc 1910; 28:73–93. 8. Cushing HW. Partial hypophysectomy for acromegaly; With remarks on the function of the hypophysis. Ann Surg 1909; 1: 1002–1017. 9. Hirsch O. Symptoms and treatment of pituitary tumors. Arch Otolaryngol 1952; 55:268. 10. Cushing HW. The hypophysis cerebri; clinical aspects of hyperpituitarism and hypopituitarism. JAMA 1909; 53:249–255. 11. Cushing HW. The basophil adenomas of the pituitary body and their clinical manifestations (pituitary basophilism). Bull Johns Hopkins Hosp 1932; 50:137–195. 12. Bishop PMF, Close HG. A case of basophil adenoma of the anterior lobe of the pituitary “Cushing’s Syndrome.” Guy’s Hosp Rep 1932; 82:143–153. 13. Guiot G. Transsphenoidal approach in surgical treatment of pituitary adenomas, general principles, and indications in nonfunctioning adenomas. Excerpta Medica Int Congr Ser 1973; 303:159. 14. Hardy J. Transsphenoidal surgery of hypersecreting pituitary tumors. Excepta Medica Int Congr Ser 1973; 303:179. 15. Cushing H. The Pituitary Body and Its Disorders. Philadelphia: Lippincott, 1910:321–322. 16. Gramegna A. Un cas d’acromégalie traité par la radiothérapie. Rev Neurol 1909; 17:15–17. 17. Béclère A. The radio-therapeutic treatment of tumors of the hypophysis, gigantism and acromegaly. Arch Roentg Ray 1909– 1910; 14:142–150. 18. Jaugeas F. The X-ray diagnosis of tumors of the hypophysis, Arch Roentg Ray 1910; 15:87–89. 19. Sheline GE, Tyrell JB. Pituitary adenomas. In: Phillips TL, Pisterno PA, eds. Radiation Oncology Annual. New York: Raven Press, 1983:1–35.

307

20. Grigsby PW, Simpson JR, Enami BN, et al. Prognostic factors and results of surgery and postoperative irradiation in the management of pituitary adenomas. Int J Radiat Oncol Biol Phys 1989; 16:1411–1417. 21. Eastman RC, Gorden P, Roth J. Conventional super voltage irradiation is an effective treatment for acromegaly. J Clin Endocrinol Metab 1979; 48:931–940. 22. Tsang RW, Brierley JD, Panzarella T et al. Role of radiation therapy in clinical hormonally-active pituitary adenomas. Radiother Oncol 1996; 41:45–53. 23. Aristizabal S, Caldwell WL, Avila J. The relationship of time-dose fractionation factors to complications in the treatment of pituitary tumors by irradiation. Int J Radiat Oncol Biol Phys 1977; 2:667–673. 24. Kjellberg RN, Masamutsu A. Stereotactic Bragg peak photon beam therapy. In: Lunsford LD, ed. Modern Stereotactic Neurosurgery. Boston: Nijhoff, 1988:463–470. 25. Fabrikant JI, Levy RP, Steinberg GK, et al. Stereotactic chargedparticle radiosurgery: clinical results of treatment of 1200 patients with intracranial arteriovenous malfomations and pituitary disorders. Clin Neurosurg 1992; 38:472–492. 26. Leksell L. The stereotactic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316–319. 27. Corenblum B, Sirek AMT, Horvath E, et al. Human mixed somatotrophic and lactotrophic pituitary adenomas. J Clin Endocrinol 1976; 42:857–863. 28. Horvath, Kovacs K. Ultrastructual diagnosis of pituitary adenomas and hyperplasias. In Loyd RV, ed. Surgical Pathology of the Pituitary Gland. Philadelphia: WB Saunders, 1993:52–84. 29. Horn K, Erhardt F, Fahlbusch R, et al. Recurrent goiter, hyperthyroidism, galactorrhea and amenorrhea due to a thyrotropin and prolactin-producing pituitary tumor. J Clin Endocrinol 1976; 43:137–143. 30. Thapar K, Laws ER Jr. Pituitary tumors. In Kaye AH, Laws ER Jr, eds. Brain Tumors, 2nd ed. London: Churchill Livingstone, 2001:803–854. 31. Friedman RB, Oldﬁeld EH, Nieman LK, et al. Repeat transsphenoidal surgery for Cushing’s disease. J Neurosurg 1989; 71: 520–527. 32. Laws ER Jr, Ebersold MJ, Piepgras DG. The results of transsphenoidal surgery in speciﬁc clinical entities. In: Laws ER, Randall RV, Kern EB, Abboud CF, eds. Management of Pituitary Adenomas and Related Lesions with Emphasis on Transsphenoidal Microsurgery. New York: Appleton-Century-Crofts, 1982:277–305. 33. Laws ER Jr, Vance ML. Radiosurgery for pituitary tumors and craniopharyngiomas. Neurosurg Clin N Am 1999; 10:327–336. 34. Halberg FE, Sheline GE. Radiotherapy of pituitary tumors. Endocrinol Metab Clin North Am 1987; 16:667–684. 35. Grigsby PW. Pituitary adenomas: evolving diagnosis and management. Int J Radiat Oncol Biol Phys 1993; 27:1253–1254. 36. Brada M, Rajan B, Traish P, et al. The longterm efﬁcacy of conservative surgery and radiotherapy in the control of pituitary adenomas. Clin Endocrinol (Oxf) 1993; 38:571–578. 37. McCollough WM, Marcus RB Jr, Rhoton AL Jr, et al. Long-term follow-up of radiotherapy for pituitary adenoma: the absence of late recurrence after greater than or equal to 4500 cGy. Int J Radiat Oncol Biol Phys 1991; 21:607–614. 38. Flickinger JC, Nelson PB, Martinez AS, et al. Radiotherapy of nonfunctional adenomas of the pituitary gland. Results with longterm follow up. Cancer 1989; 63:2409–2414. 39. Hayashi M, Taira T, Chernov M, et al. Pituitary gland and Gamma Knife. Endocrinol Diabetol (Jpn) 2002; 15:296–306. 40. Régis J, Hayashi M, Porcheron D, et al. Impact of the model C and automatic positioning system on gamma knife radiosurgery: an evaluation in vestibular schwannomas. J Neurosurg 2002; 97(Suppl 5):588–591.

308

k. takakura et al.

41. Rush S, Cooper PR. Symptom resolution, tumor control, and side effects following postoperative radiotherapy for pituitary macroadenomas. Int J Radiat Oncol Biol Phys 1997; 37:1031–1034. 42. Tsang RW, Brierley JD, Panzarella T, et al. Radiation therapy for pituitary adenoma: treatment outcome and prognostic factors. Int Radiat Oncol Biol Phys 1994; 30:557–565. 43. Zierhut D, Flentje M, Adolph J, et al. External radiotherapy of pituitary adenomas. Int Radiol Oncol Biol Phys 1995; 33:307– 314. 44. Littley MD, Shalet SM, Beardwell CG, et al. Hypopituitarism following external radiotherapy for pituitary tumors in adults. Am J Med 1989; 79:145–160. 45. 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):30–37. 46. Witt TC, Kondziolka D, Flickinger JC, Lunsford LD. Gamma Knife radiosurgery for pituitary tumors. In: Lunsford LD, Kondziolka D, Flickinger JC, eds. Progress in Neurosurgery; Gamma Knife Brain Surgery. Basel: Karger, 1998:114–127. 47. Park YG, Chang JW, Kim EY, Chung SS. Gamma Knife surgery in pituitary microadenomas. Yonsei Med 1996; 37:165–173. 48. Sheehan JP, Niranjan A, Sheehan JM, et al. Stereotactic radiosurgery for pituitary adenomas: an intermediate review of its safety, efﬁcacy, and role in the neurosurgical treatment armamentarium. J Neurosurg 2005; 102:678–691. 49. Hayashi M, Izawa M, Hiyama S, et al. Gamma Knife radiosurgery for pituitary adenomas. Stereotact Funct Neurosurg 1999; 72 (Suppl 1):111–118. 50. Izawa M, Hayashi M, Nakaya K, et al. Gamma Knife radiosurgery for pituitary adenomas. J Neurosurg 2000; 93(Suppl 3):19–22. 51. Chen JC, Giannotta ST, Yu C, et al. Radiosurgical management of benign cavernous sinus tumors: dose proﬁles and acute complications. Neurosurgery 2001; 48:1022–1032. 52. Kondziolka D, Levy EL, Niranjan A, et al. Long-term outcomes after meningioma radiosurgery: physician and patients perspectives. J Neurosurg 1999; 91:44–50. 53. Iwai Y, Yamanaka K, Nakajima H, et al. Gamma Knife radiosurgery for skull base meningiomas: the treatment results and patient satisfaction expressed in answers to a questionnaire. No-Shinkei Geka 2000; 28:411–415 [Japanese with English summary]. 54. Pendl G, Schrottner O, Eustacchio S, et al. Stereotactic radiosurgery of skull base meningiomas. Minim Invasive Neurosurg 1997; 40:87–90. 55. Pan DH, Guo WY, Chang YC, et al. The effectiveness and factors related to treatment results of Gamma Knife radiosurgery for meningiomas. Stereotact Funct Neurosurg 1998; 70(Suppl 1): 19–32. 56. Liscak R, Simonova G, Vymazal J, et al. Gamma Knife radiosurgery of meningiomas in the cavernous sinus region. Acta Neurochir (Wien) 1999; 141:473–480. 57. Morita A, Coffey RJ, Foote RL, et al. Risk of injury to cranial nerves after Gamma Knife radiosurgery for skull base meningiomas: experience in 88 patients. J Neurosurg 1999; 90:42–49. 58. Takakura K, Sano K, Hojo S, Hirano A. Pituitary gland. In: Takakura K, Sano K, Hojo S, Hirano A, eds. Metastatic Tumors of the Central Nervous System. Igaku Shoin: Tokyo, 1982:101–104. 59. Abramo HL, Spiro R, Goldstein V. Metastasis in carcinoma: analysis of 1,000 autopsied cases. Cancer 1950; 3:74–85. 60. Hagerstrand I, Schonebeck J. Metastasis to pituitary gland. Acta Pathol Microl Scand 1969; 75:14–70. 61. Kovacs A. Metastatic cancer of the pituitary gland. Oncology 1973; 27:533–542. 62. Gurling KJ, Scott GBD, Baron DN. Metastases in pituitary tissue removed at hypophysectomy in women with mammary carcinoma. Br J Cancer 1957; 11:519–523. 63. Duchen LW. Metastatic carcinoma in the pituitary gland and hypothalamus. J Pathol Bacteriol 1966; 91:347–355.

64. Kistler M, Pribram HW. Metastatic disease of the sella turcica. Am J Roentgenol Radium Ther Nucl Med 1975; 123:13–21. 65. Oi S, Ciric I, Mayers TK. Metastatic breast carcinoma in the pituitary gland. Brain Nerve (Tokyo) 1978; 30:69–73 [Japanese with English abstract]. 66. Tecars RJ, Silverman EM. Clinicopathologic review of 88 cases of carcinoma metastatic to the pituitary gland. Cancer 1975; 36: 216–220. 67. Jones GM. Diabetes insipidus. Arch Intern Med 1944; 74:81–93. 68. Plotner H. Primary or idiopathic diabetes insipidus: a system disease. Metabolism 1958; 7:191–200. 69. Houck WA, Olson KB, Horton J. Clinical features of tumor metastasis to the pituitary. Cancer 1970; 26:656–659. 70. Green JD. The comparative anatomy of the hypophysis, with special reference to its blood supply and innervation. Am J Anat 1951; 88:225–311. 71. Page RB, Bergland RM. Pituitary vasculature. In Allen MB Jr, Mahesh VB, eds. The Pituitary. A Current Review. New York: Academic Press, 1977:9–17. 72. Willis RA. The spread of tumors in the human body, 3rd ed. London: Butterworth, 1973. 73. Koyama Y, Takakura K. Intracranial invasion of malignant neoplasms—studies on autopsy cases. Adv Neurol Sci (Tokyo) 1969; 13:188–197 [Japanese with English abstract]. 74. Epstein S, Ranchod M, Goldswain PRT. Pituitary insufﬁciency, inappropriate anti-diuretic hormone (ADH) secretion, and carcinoma of the bronchus. Cancer 1973; 32:476–481. 75. Teoh TB. Epidermoid carcinoma of the nasopharynx among Chinese: a study of 31 necropsies. J Pathol Bacteriol 1957; 72: 451–465. 76. Masse SR, Wolk RW, Conklin RH. Peripituitary gland involvement in acute leukemia in adults. Arch Pathol 1973; 96:141–142. 77. Girkin CA, Comey CH, Lunsford LD, Goodman ML, Kline LB. Radiation optic neuropathy after stereotactic radiosurgery. Ophthalmology 1997; 104:1634–1643. 78. 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; 27:215–221. 79. Lundstrom M, Frisen L. Atrophy of optic nerve ﬁbres in compression of the chiasm. Degree and distribution of ophthalmoscopic changes. Acta Ophthalmol 1976; 54:623–640. 80. Jane JA Jr, Vance ML, Woodburn LJ, Laws ER Jr. Stereotactic radiosurgery for hypersecreting pituitary tumors: part of a multimodality approach. Neurosurg Focus 2003; 14:E12. 81. Pollock BE, Carpenter PC. Stereotactic radiosurgery as an alternative to fractionated radiotherapy for patients with recurrent or residual nonfunctioning pituitary adenomas. Neurosurgery 2003; 53:1086–1094. 82. Fiegl GC, Bonelli CM, Berghold A, Mokry M. Effects of gamma Knife radiosurgery of pituitary adenomas in pituitary function. J Neurosurg 2002; 97(Suppl 5):415–421. 83. Hanabusa K, Morikawa A, Murata T, Taki W. Acoustic neuroma with malignant transformation. Case report. J Neurosurg 2001; 95:519–521. 84. Kaido T, Hoshida T, Uranishi R, et al. Radiosurgery-induced brain tumor. Case report. J Neurosurg 2001; 95:710–713. 85. Shamisa A, Bance M, Nag S, et al. Glioblastoma multiforme occurring in a patient treated with gamma knife surgery. Case report and review of the literature. J Neurosurg 2001; 94:816–821. 86. Comey CH, Mc Laughlin MR, Jho HD, et al. Death from a malignant cerebellopontine angle triton tumor despite stereotactic neurosurgery. Case report. J Neurosurg 1998; 89:653–658. 87. Yu JS, Yong WH, Wilson D, Black KL. Glioblastoma induction after radiosurgery for meningioma. Lancet 2000; 356:1576–1577. 88. Shin M, Ueki K, Kurita H, Kirino T. Malignant transformation of a vestibular schwannoma after Gamma Knife radiosurgery. Lancet 2002; 360:309–310.

2 7

Pituitary Adenomas: Surgery Perspective William T. Couldwell and Martin H. Weiss

Introduction The ﬁrst recorded attempt at surgical resection of a pituitary tumor was a two-stage lateral subtemporal decompression in an acromegalic patient performed by Caton and Paul [1]. They used a temporal approach as suggested by Sir Victor Horsley, who later used both a subfrontal approach and a lateral middle fossa approach in operating on 10 pituitary tumors [2]. The ﬁrst successful removal of a pituitary tumor via a superior nasal transsphenoidal approach was in 1907 (Fig. 27-1) [3–5]. This approach was further modiﬁed to address complications and the disﬁgurement that were associated with the original technique so that the transsphenoidal approach has now become the approach of choice for almost all pituitary tumors. This review will address the indications for surgical resection of pituitary adenomas and describe contemporary surgical techniques applied to pituitary surgery. It will also describe the use of stereotactic radiosurgery as an adjuvant to surgical resection of pituitary adenomas.

Surgical Indications For symptomatic pituitary adenomas that will not respond to medical therapy, the gold standard therapy is surgical removal of the tumor. The goals of surgery for pituitary tumors are elimination of the tumor mass, normalization of hormonal hypersecretion, and preservation of normal pituitary function. Another goal of surgery is to eliminate the potential for recurrence [6]; this may be achieved through adjunctive radiosurgical treatment. A common indication for surgery is progressive visual loss caused by mass effect [7]. Precipitous visual loss associated with headache, cranial neuropathies, and sometimes acute adrenal insufﬁciency (pituitary apoplexy) also indicates immediate steroid replacement and emergent decompression of the optic nerves and chiasm via a transsphenoidal approach [8–10]. Successful removal of growth hormone–secreting adenomas results in a prompt decrease in growth hormone levels in cases of acromegaly [11–14]. Radiation therapy can also serve as an adjunct when a large tumor cannot be “cured” through resection. Failure of prior medical or radiation therapy may warrant surgical intervention to treat recurrent tumor [15]. The transsphenoidal approach is also useful in the removal of sellar

craniopharyngiomas and Rathke cleft cysts, clival chordomas, and occasional meningiomas or metastatic lesions.

Preoperative Radiographic and Endocrinologic Considerations Several characteristics must be considered when evaluating the patient for pituitary surgery. The proximity of the tumor to the intracavernous carotid arteries, as judged on T1-weighted magnetic resonance imaging (MRI), is an important consideration when using a transsphenoidal approach to the tumor. The greater soft-tissue contrast offered by MRI over computed tomography (CT) enables clear visualization of the proximal structures. This is particularly important for identifying deviations of the ectatic carotid arteries, for instance, which may deviate into the midline trajectory and preclude a transsphenoidal approach. The endocrine status of the patient must be evaluated before surgery is performed on lesions near the hypothalamichypophysial axis because of the risk of intra- or postoperative hypopituitarism, which can be dangerous in the perioperative period. Preoperative recognition of hypothyroidism is also important because it can manifest acutely during the early postoperative period. Any patient with a presumptive pituitary tumor should have a comprehensive endocrine evaluation to establish the secretory or nonsecretory status of the tumor, to identify those patients in whom efﬁcacious medical therapy should be initiated (e.g., prolactinoma), and to provide a preoperative evaluation of baseline pituitary function.

Surgical Approaches to the Pituitary The Endonasal Transsphenoidal Approach The transsphenoidal approach to the sella, based on the foundations of Hirsch and Cushing, continues to be the method of choice in treating most sellar lesions. Cushing advocated the sublabial transsphenoidal approach because it offered more extensive access to the lesion. Although the authors originally used the sublabial approach, they have subsequently converted to the endonasal approach advocated by Hirsch. Whereas the

309

310

w.t. couldwell and m.h. weiss

above the heart to encourage venous drainage (Fig. 27-2). Once the sella and dura are opened, the diseased tissue is removed from the sellar region as in any standard transsphenoidal resection (Fig. 27-3) [18, 19]. The authors employ angled endoscopes to inspect areas that are hidden from the microscopic view, particularly residual tumor that is hidden behind redundant folds of suprasellar arachnoid, as well as tumor extending into the cavernous sinus. The angled endoscopes can also be used to resolve the problem of deviation from a strictly midline approach. This “cross-court” view is advantageous for visualizing unilateral sellar tumors that are contralateral to the nostril. Thus, if a tumor projects more to one side, the contralateral nostril is used for the approach. Closure proceeds as described previously unless cerebrospinal ﬂuid leakage is encountered. If a tear in the arachnoid is encountered, sellar reconstruction is performed with autologous fascia lata and fat grafting placed over the dural opening and buttressed in place with fat packed in the sphenoid sinus as recently published [20].

Modiﬁcations of the Transsphenoidal Approach The standard transsphenoidal approach is used in most cases of pituitary tumors with sellar and suprasellar extension; however,

FIGURE 27-1. Schloffer’s transnasal transsphenoidal operation. (A) Incision made along the left nasolabial furrow around the left ala nasi and continued up to the glabella. (B) The incision cuts through the skin, nasal bone, philtrum, and the anterior part of the septum. The whole external nose is reﬂected to the right exposing the remainder of the septum. (C) The rest of the nasal septum has been removed exposing the rostrum of the sphenoid sinus. (D) The anterior wall of the sphenoid sinus is opened, the mucosal lining of the sinus is removed with a sharp spoon, and the ﬂoor of the sella is removed with a small chisel or punch forceps. (From: Cope VZ. The pituitary fossa and the methods of surgical approach thereto. Br J Surg 1916; 4:107–144. Used with permission.)

surgeon gives up some superior trajectory and viewing angle, this is more than offset by the ease of performing the endonasal approach [16]. A transsphenoidal approach is indicated in primary or revision cases of sellar and/or parasellar lesions. The details of the surgical approach have been described previously [17]. The patient is positioned supine with the face parallel to the ceiling and the head elevated approximately 15°

FIGURE 27-2. Patient and operating room positioning for the transsphenoidal approach. The head is elevated 15° above to reduce venous bleeding. The head is tilted toward the left shoulder to facilitate endonasal exposure. (From: Couldwell WT, Weiss MH. The transnasal transsphenoidal approach. In Apuzzo MLJ, ed. Surgery of the Third Ventricle, 2nd ed. Baltimore: Williams & Wilkins, 1998:553–574. Used with permission.)

27.

311

pituitary adenomas: surgery perspective

Diaphragm of sella

Subfrontal arachnoid spaces

Optic Anterior cerebral a. chiasm

Infundibulum

Hypothalamus Tumor Basilar a.

Dura

A

Arachoid

Sphenoid sinus Tumor

Window in sella turcica

B Dura

C Arachnoid Diaphragm

D

E F FIGURE 27-3. Tumor removal. Tumor is readily removed with various curettes and scoops. It is imperative that the integrity of the arachnoid is preserved to minimize the potential for a postoperative cerebrospinal ﬂuid ﬁstula. (From: Couldwell WT, Weiss MH. The

transnasal transsphenoidal approach. In: Apuzzo MLJ, ed. Surgery of the Third Ventricle, 2nd ed. Baltimore: Williams & Wilkins, 1998:553– 574. Used with permission.)

regions of the skull base that were once thought only accessible “from above” are now being approached transfacially. With better knowledge of microsurgical anatomy and modern microinstrumentation, neurosurgeons have modiﬁed the transsphenoidal approach to gain better access to regions such as the cavernous sinus and the suprasellar cisterns. Fraioli et al. [21] treated 11 patients with sellar tumors invading the medial wall of the cavernous sinus with a transmaxillosphenoidal approach, which involves a unilateral, bilateral, or Le Fort maxillary osteotomy and removal of the medial wall of the maxillary sinus in addition to the standard transsphenoidal exposure. It allows direct visualization of the intracavernous carotid artery during tumor resection but is limited by extension of tumor lateral to the carotid artery. Sabit et al.

[22] described a safe, minimally invasive combined transmaxillary transsphenoidal approach to the cavernous sinus that is both extradural and extranasal. This approach provides adequate lateral-to-medial reach in the parasellar and infrasellar regions with visualization of the entire ipsilateral cavernous sinus and the medial aspect of the contralateral cavernous sinus. The extended transsphenoidal approach to tumors of the tuberculum sellae and suprasellar cistern was ﬁrst described by Weiss in 1987 [23]. Suprasellar tumors without sellar enlargement have been successfully resected using a modiﬁed transsphenoidal approach that involves a wide bone exposure of the anterior surface of the sella and removal of the posterior portion of the planum sphenoidale [24, 25]. This novel approach has allowed resection of craniopharyngiomas, CNS hemangioblastomas,

312

w.t. couldwell and m.h. weiss Optic chiasm

Dura of tuberculum Recess of sellae infundibulum

Tumor Arachnoid

A

B

Circular sinus

Dura over pituitary

Dural opening

C

Tumor Tumor seen through arachnoid Pituitary

and ectopic adrenocorticotropic hormone (ACTH)-producing adenomas arising from the pituitary stalk [24, 25]. Additional exposure of the skull base for lesions of the parasellar and clival region can be achieved by extended transsphenoidal approaches [18]. Various portions of the skull base may be exposed and bone resection can be extended by repositioning the patient’s head and the self-retaining speculum. The standard transsphenoidal approach can be extended anteriorly to resect suprasellar lesions, inferiorly to expose clival lesions, and inferolaterally to access cavernous sinus lesions (Figs. 27-4, 27-5, and 27-6) [18]. These variations on the transsphenoidal approach provide a minimally invasive technique that avoids prolonged surgery and brain retraction. Technological advances in the areas of endoscope-assisted microneurosurgery [26, 27] have been applied to the classic transsphenoidal operation in an attempt to further decrease morbidity and mortality risks [27–29]. Jho and Carrau [28] reported encouraging results in a series of 50 patients who underwent endoscopic endonasal transsphenoidal surgery. One of the main advantages of this approach is excellent panoramic visualization of the sellar and suprasellar anatomy with increased illumination and magniﬁcation [30]. Anatomic studies have demonstrated that the endoscope provides a volume of exposure superior to that of the operating microscope [31], but disadvantages include the lack of stereoscopic vision, the lack of adequate instrumentation, the limited space of working through one nostril [26], and a learning curve for using this technique.

Transcranial Surgery for Pituitary Tumors

Circular sinus

D

Dural flap FIGURE 27-4. (A–D) Modiﬁcation of the transsphenoidal approach: exposure of the anterior skull base (extended transsphenoidal). A pure suprasellar tumor may be approached by extending the bony resection anteriorly over the tuberculum sellae, thus exposing the dura mater lying anterior to the circular sinus. An incision is made in the dura anteriorly and inferiorly to the circular sinus. The sinus is then coagulated and transected to gain a direct view of the suprasellar cistern without disturbing the pituitary gland. (From: Couldwell WT, Weiss MH. The transnasal transsphenoidal approach. In: Apuzzo MLJ, ed. Surgery of the Third Ventricle, 2nd ed. Baltimore: Williams & Wilkins, 1998:553–574. Used with permission.)

As the transsphenoidal micro- or endoscopic approach has become the mainstay of pituitary surgery, the indications for transcranial approaches to pituitary tumors have become few. In cases in which a transnasal approach is contraindicated, such as sphenoid sinusitis or ectatic midline (“kissing”) carotid arteries, a transcranial approach may be warranted. Some patients who harbor pituitary macroadenomas with signiﬁcant lateral suprasellar extension that cannot be adequately removed transsphenoidally may beneﬁt from a transcranial approach. The two most widely used transcranial approaches to pituitary tumors are the pterional (frontotemporal) approach and the anterior

Sphenoid Chordoma sinus Rostrum Clivus

Soft palate Pharynx

FIGURE 27-5. Modiﬁcations of the transsphenoidal approach: inferior exposure of the clivus. Exposure of the clivus is facilitated by slight ﬂexion of the patient’s head and repositioning of the nasal selfretaining retractor to point inferiorly. The upper clivus lies directly posterior to the sphenoid sinus, but additional exposure to the mid and lower clivus requires more inferior exposure. (From: Couldwell WT, Weiss MH. The transnasal transsphenoidal approach. In: Apuzzo MLJ, ed. Surgery of the Third Ventricle, 2nd ed. Baltimore: Williams & Wilkins, 1998:553–574. Used with permission.)

27.

Floor of sella

313

pituitary adenomas: surgery perspective

Dura of sella

Floor of sella

Dura over pituitary gland, cavernous sinus, and ICA

Area of bone removal Sphenoid sinus

Pituitary gland

Wall of sphenoid sinus Cavernous sinus

ICA

NII NIII NIV

Dural incision enlarged

NV Sphenoid sinus Tumor exposed Area of bone removal FIGURE 27-6. Modiﬁcation of the transsphenoidal approach: inferolateral exposure of the cavernous sinus. (A) After exposure of the dura overlying the sella, the bone overlying the cavernous sinus, including that overlying the carotid grooves, is carefully removed. This removal deﬁnes the lateral extent of the exposure limited by the cavernous cranial nerves. (B) The dura medial to the internal carotid artery is ﬁrst

subfrontal approach. The pterional approach [32] represents the shortest transcranial trajectory to the suprasellar cistern and should be the method of choice when a transcranial approach is used in a patient with a preﬁxed chiasm because the tumor can be resected beneath the chiasm. The anterior subfrontal approach has the advantage of a straight frontal trajectory with direct visualization to the tumor as it is being removed between the optic nerves. The disadvantages include the potential violation of the frontal sinus and damage to olfactory nerves. This procedure is usually not performed in patients with a preﬁxed chiasm.

Adjuvant Radiosurgery of Pituitary Tumors Involving the Cavernous Sinus The use of radiosurgery for the management of pituitary tumors is discussed in detail in Chapter 26 of this text. We describe the use of micro- or endoscopic removal of pituitary tumors as the primary treatment modality in patients in whom surgery would be tolerated. However, there are speciﬁc instances in which radiosurgical treatment is indicated and preferred to surgical resection. These include recurrent tumors in the presence of hypopituitarism and gross invasion into the cavernous sinus.

incised with a size 11 blade and opened with curved alligator microscissors. Removal of the intracavernous portion of the tumor is carried out with a microcurette. (From: Couldwell WT, Weiss MH. The transnasal transsphenoidal approach. In: Apuzzo MLJ, ed. Surgery of the Third Ventricle, 2nd ed. Baltimore: Williams & Wilkins, 1998:553–574. Used with permission.)

In many cases of pituitary adenoma with gross invasion of the cavernous sinus, planned adjuvant radiosurgery for residual tumor offers an alternative to radical surgical resection of tumor. This can provide excellent tumor control and enables the surgeon to avoid the cranial nerve deﬁcit associated with surgery in the region. Radiosurgical techniques have the potential to deliver high dosages directly to the tumor site while reducing the radiation to the surrounding normal structures. However, the proximity of the pituitary gland to the region of treatment increases the risk of the patient developing hypopituitarism, even with stereotactic radiosurgery (SRS). SRS is being used more frequently in the treatment of pituitary tumors, both functioning and nonfunctioning, for residual tumor within the cavernous sinus [33–41]. The preliminary data for tumor control and normalization of hypersecretory syndromes after radiosurgery appear promising [34–49]. The authors have developed a technique for pituitary transposition (hypophysopexy) in anticipation of postoperative radiosurgical treatment [50, 51]. This technique involves transposing the pituitary gland away from the tumor in the cavernous sinus and interposing a fat graft between the normal gland and the tumor in the cavernous sinus. Treatment of the tumor with SRS is facilitated by the greater distance between the pituitary gland and residual tumor, and the effective biological dose to the normal pituitary gland is reduced. The reduction of

314

w.t. couldwell and m.h. weiss

FIGURE 27-7. Pituitary transposition (hypophysopexy) with planned radiosurgical treatment of residual cavernous sinus tumor. Acromegalic young female with residual tumor within the cavernous sinus after hypophysopexy. (A) Coronal T1-weighted gadolinium-enhanced MR images demonstrating the relationship of the cavernous sinus tumor (T), fat graft (F), optic chiasm (OC), and transposed pituitary gland (P). Isodose curves on coronal image demonstrating doses delivered to

surrounding tissues on (B) coronal and (C) axial images. Note lateral displacement of the gland by the interposed fat graft on coronal image. (D) Diagram representing dosimetry target in relationship to brain stem and optic apparatus with transposed pituitary in green. (From Couldwell WT et al., Hypophysopexy technique for radiosurgical treatment of cavernous sinus pituitary adenoma. Pituitary 2002; 5:169–173 [Fig. 1]. Used with permission.)

radiation exposure to the gland decreases the likelihood of the patient developing hypopituitarism.

dural opening and further fat graft within the sphenoid to buttress the fascia and prevent cerebrospinal ﬂuid ﬁstulae [18, 52, 53]. Necrosis and shrinkage of the graft are expected with time, and after radiosurgical treatment; this factor must be considered with fractionated treatments. If longer time periods to deliver fractionated doses are considered, other more permanent materials, such as contemporary bone substitutes, may be better suited to maintain transposition of the gland.

Hypophysopexy: Pituitary Transposition Hypophysopexy is begun with a standard transsphenoidal approach that is extended laterally toward the side of the cavernous sinus that involves tumor [18, 52]. Maximum tumor removal is performed while leaving residual tumor in the cavernous sinus. The normal pituitary gland is dissected from the adjacent recurrent tumor and displaced laterally away from the involved cavernous sinus. A margin of normal gland is also removed from the tumor interface to ensure that no tumor is adhering to the normal gland before transposition. Fat is then interposed between the gland and the involved cavernous sinus to maintain the transposition (Fig. 27-7). The closure then proceeds in a standard fashion, with fascia lata placed over the

Advantages and Disadvantages of Surgery Compared with Radiosurgery for the Treatment of Pituitary Adenomas Microsurgical or endoscopic resection of a pituitary adenoma by an experienced surgeon has the advantage of extirpating the lesion while producing no collateral damage to the normal

27.

pituitary adenomas: surgery perspective

pituitary gland, cavernous sinus and its contents, and optic apparatus. All of these structures are at some risk with radiosurgery if the tumor is compressing the normal structures, as they will receive some fall-off dose. Radiosurgery, on the other hand, avoids many of the risks associated with surgery, including those associated with general anesthesia, infection, cerebrospinal ﬂuid leak, and injury to the carotid artery or cranial nerves. In addition, the discomfort associated with surgery is avoided, and the procedure may be done without admission to the hospital. Certainly, in the authors’ opinion, radiosurgery is an excellent alternative to open surgery in patients with recurrent tumors (especially in the cavernous sinus). The hypophysopexy technique described above enables treatment of the residual tumor in the cavernous sinus while limiting the dose to the normal pituitary gland. In addition, treatment of recurrent tumor within the sella in a patient who is hypopituitary is an attractive option to avoid the risks of surgical resection. Acknowledgments. Portions of this chapter were derived from Refs. 17, 18, 19, 51, and 52 and from Couldwell WT, Weiss MH: Strategies for the management of non-secreting pituitary adenomas. In: Cooper PR, ed. Neurosurgical Topics: Contemporary Diagnosis and Management of Pituitary Adenomas. Park Ridge, IL: AANS Publications Committee, 1991:29–37. The authors would like to thank Kristin Kraus for her superb editorial assistance.

References 1. Caton R, Paul FT. Notes of a case of acromegaly treated by operation. Br Med J 1983; 2:1421–1423. 2. Cope VZ. The pituitary fossa, and the methods of surgical approach thereto. Br J Surg 1916; 4:107–144. 3. Schloffer H. Zur frage der operationen an der hypophyse. Beitr Klin Chir 1906; 50:767–817. 4. Schloffer H. Erfolgreiche operation eines hypophysentumor auf nasalem wege. Wien Klin Wochenschr 1907; 20:621–624. 5. Schloffer H. Weiterer bericht uber den fall von operiertem hypophysen-tumor. Wien Klin Wochenschr 1907; 20:1075–1078. 6. Laws ER Jr. Pituitary surgery. Endocrinol Clin Metab North Am 1987; 16:647–665. 7. Ebersold MJ, Quast LM, Laws ER Jr, Scheithauer B, Randall RV. Long term results in transsphenoidal removal of nonfunctioning pituitary adneomas. J Neurosurg 1986; 64:713–719. 8. Bills DC, Meyer FB, Laws ER Jr, et al. A retrospective analysis of pituitary apoplexy. Neurosurgery 1993; 33:602–608. 9. Ebersold MJ, Laws ER Jr, Scheithauer BW, et al. Pituitary apoplexy treated by transsphenoidal surgery. J Neurosurg 1983; 58:315. 10. Laws ER Jr. Transsphenoidal decompression of the optic nerve and chiasm: visual results in 62 patients. J Neurosurg 1977; 46:717–722. 11. Davis DH, Laws ER Jr, Ilstrup DM, et al. Results of surgical treatment for growth hormone-secreting pituitary adenomas. J Neurosurg 1993; 79:70–75. 12. Freda PU, Wardlaw SL, Post KD. Long-term endocrinological follow-up evaluation in 115 patients who underwent transsphenoidal surgery for acromegaly. J Neurosurg 1998; 89:353–358. 13. Laws ER Jr, Piepgras DG, Randall RV, et al. Neurosurgical management of acromegaly. Results in 82 patients treated between 1972 and 1977. J Neurosurg 1979; 50:454–4561. 14. Ross DA, Wilson CB. Results of transsphenoidal microsurgery for growth hormone-secreting pituitary adenoma in a series of 214 patients. J Neurosurg 1988; 68:854–867.

315

15. Laws ER Jr, Fode NC, Redmond MJ. Transsphenoidal surgery following unsuccessful prior therapy. J Neurosurg 1985; 63: 823. 16. Spencer WR, Levine JM, Couldwell WT, Brown-Wagner M, Moscatello A. Approaches to the sellar and parasellar region: a retrospective comparison of the endonasal-transsphenoidal and sublabial-transsphenoidal approaches. Otolaryngol Head Neck Surg 2000; 122:367–369. 17. Couldwell WT. Transsphenoidal and transcranial surgery for pituitary adenomas. J Neurooncol 2004; 69:237–256. 18. Couldwell WT, Weiss MH. The transnasal transsphenoidal approach. In: Apuzzo MLJ, ed. Surgery of the Third Ventricle, 2nd ed. Philadelphia: Williams & Wilkins, 1998:553–574. 19. Liu JK, Weiss MH, Couldwell WT. Surgical approaches to the pituitary gland. Neurosurg Clin N Am 2003; 14:93–107. 20. Couldwell WT, Kan P, Weiss M. Simple closure following transsphenoidal surgery. Neurosurg Focus 2006; 20(3):E11. 21. Fraioli B, Esposito V, Santoro A, et al. Transmaxillosphenoidal approach to tumors invading the medial compartment of the cavernous sinus. J Neurosurg 1995; 82:63–69. 22. Sabit I, Schaefer SD, Couldwell WT. Extradural extranasal combined transmaxillary transsphenoidal approach to the cavernous sinus: a minimally invasive microsurgical model. Laryngoscope 2000; 110:286–291. 23. Weiss MH. Transnasal transsphenoidal approach. In: Apuzzo ML, ed. Surgery of the Third Ventricle. Philadelphia: Williams & Wilkins, 1987:476–494. 24. Kouri JG, Chen MY, Watson JC, et al. Resection of suprasellar tumors by using a modiﬁed transsphenoidal approach. Report of four cases. J Neurosurg 2000; 92:1028–1035. 25. Mason RB, Nieman LK, Doppman JL, et al. Selective excision of adenomas originating in or extending into the pituitary stalk with preservation of pituitary function. J Neurosurg 1997; 87:343– 351. 26. Alﬁeri A. Endoscopic endonasal transsphenoidal approach to the sellar region: technical evolution of the methodology and reﬁnement of a dedicated instrumentation. J Neurosurg Sci 1999; 43: 85–92. 27. Cappabianca P, Alﬁeri A, Thermes S, et al. Instruments for endoscopic endonasal transsphenoidal surgery. Neurosurgery 1999; 45:392–396. 28. Jho HD, Carrau RL. Endoscopic endonasal transsphenoidal surgery: experience with 50 patients. J Neurosurg 1997; 87: 44–51. 29. Yaniv E, Rappaport ZH. Endoscopic transseptal transsphenoidal surgery for pituitary tumors. Neurosurgery 1997; 40:944– 946. 30. Jho HD, Carrau RL, Ko Y, et al. Endoscopic pituitary surgery: an early experience. Surg Neurol 1997; 47:213–222. 31. Spencer WR, Das K, Nwagwu C, et al. Approaches to the sellar and parasellar region: anatomic comparison of the microscope vs. endoscope. Laryngoscope 1999; 109:791–794. 32. Yasargil MG, ed. Microneurosurgery. Vol. I. Microsurgical anatomy of the basal cisterns and vessels of the brain, diagnostic studies, general operative techniques, and pathological considerations of the intracranial aneurysms. Stuttgart: Georg Thieme Verlag, 1984. 33. Littley MD, Shalet SM, Beardwell CG, et al. Radiation-induced hypopituitarism is dose-dependent. Clin Endocrinol 1989; 31:363– 373. 34. Kim MS, Lee SI, Sim JH. Gamma knife radiosurgery for functioning pituitary microadenoma. Stereotact Funct Neurosurg 1999; 72(Suppl 1):119–124. 35. Kim SH, Huh R, Chang JW, et al. Gamma knife radiosurgery for functioning pituitary adenomas. Stereotact Funct Neurosurg 1999; 72(Suppl 1):101–110.

316

w.t. couldwell and m.h. weiss

36. Kobayashi T, Kida Y, Mori Y. Gamma knife radiosurgery in the treatment of Cushing’s disease: long-term results. J Neurosurg 2002; 97(Suppl 5):422–428. 37. Landolt AM, Haller D, Lomax N, et al. Stereotactic radiosurgery for recurrent surgically treated acromegaly: comparison with fractionated radiotherapy. J Neurosurg 1998; 88:1002– 1008. 38. Landolt AM, Lomax N. Gamma knife radiosurgery for prolactinomas. J Neurosurg 2000; 93(Suppl 3):13–18. 39. Pan L, Zhang N, Wang EM, et al. Gamma knife radiosurgery as a primary treatment for prolactinomas. J Neurosurg 2000; 93(Suppl 3):10–13. 40. Sheehan JP, Kondziolka D, Flickinger J, et al. Radiosurgery for residual or recurrent nonfunctioning pituitary adenoma. J Neurosurg 2002; 97(Suppl 5):415–421. 41. Shin M, Kurita H, Sasaki T, et al. Stereotactic radiosurgery for pituitary adenoma invading the cavernous sinus. J Neurosurg 2000; 93(Suppl 3):2–5. 42. Boelaert K, Gittoes NJ. Radiotherapy for non-functioning pituitary adenomas. Eur J Endocrinol 2001; 144:569–575. 43. Tsang RW, Brierley JD, Panzararekka T, et al. Radiation therapy for pituitary adenoma: treatment outcome and prognostic factors. Int J Radiat Oncol Biol Phys 1994; 30:557–565. 44. Wowra B, Stummer W. Efﬁcacy of gamma knife radiosurgery for nonfunctioning pituitary adenomas: a quantitative follow up with magnetic resonance imaging-based volumetric analysis. J Neurosurg 2002; 97(Suppl 5):429–432.

45. Izawa M, Hayashi M, Nakaya K, et al. Gamma knife radiosurgery for pituitary adenomas. J Neurosurg 2000; 93(Suppl 3):19–22. 46. Kondziolka D, Flickinger JC, Lunsford LD. Radiation therapy and radiosurgery of pituitary tumors. In: Krisht AF, Tindall GT, eds. Pituitary Disorders: Comprehensive Management. Philadelphia: Lippincott Williams & Wilkins, 1999:407–415. 47. Laws ER Jr, Vance ML. Conventional radiotherapy for pituitary tumors. Neurosurg Clin N Am 2000; 11:617–625. 48. Morange-Ramos I, Regis J, Dufour H, et al. Short-term endocrinological results after gamma knife surgery of pituitary adenomas. Stereotact Funct Neurosurg 1998; 70:127–138. 49. Yoon SC, Suh TS, Jang HS, et al. Clinical results of 24 pituitary macroadenomas with linac-based stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1998; 41:849–853. 50. Liu JK, Schmidt MH, MacDonald JD, Jensen RL, Couldwell WT. Hypophysial transposition (hypophysopexy) for radiosurgical treatment of pituitary tumors involving the cavernous sinus: technical note. Neurosurg Focus 2003; 14(5):E11. 51. Couldwell WT, Rosenow JM, Rovit RL, Benzil DL. Hypophysopexy technique for radiosurgical treatment of cavernous sinus pituitary adenoma. Pituitary 2002; 5:169–173. 52. Liu JK, Das K, Weiss MH, Laws ER, Couldwell WT. The history and evolution of transsphenoidal surgery. J Neurosurg 2001; 95:1083–1096. 53. Couldwell WT, Weiss MH. Pituitary macroadenomas. In: Apuzzo MLJ, ed. Brain Surgery: Complication Avoidance and Management. New York: Churchill Livingstone, 1993:295–312.

2 8

Pituitary and Pituitary Region Tumors: Fractionated Radiation Therapy Perspective Jonathan P.S. Knisely and Paul W. Sperduto

Introduction Fractionated radiotherapy for pituitary tumors and tumors arising in the parasellar region has been used for a century. The multidisciplinary nature of optimal management of tumors of the anterior skull base has been clearly established; the expertise of endocrinologists, neurosurgeons, radiation oncologists, neuroradiologists, neuroophthalmologists, and laboratory medicine specialists are all commonly required to optimally manage tumors that arise in the pituitary or pituitary region. The skull base structures surrounding the pituitary gland are anatomically complex [1]. The proximity of a number of important vascular structures and cranial nerves makes a surgical approach to tumors arising in the pituitary region complex; accordingly, complete resections with acceptable operative morbidity can be difﬁcult to achieve. Iatrogenic damage may occur during the treatment of an anterior skull base tumor that may result in endocrinologic, visual, or other neurologic deﬁcits. Radiation can induce second tumors and cause endocrine insufﬁciencies. It is easy to appreciate the potential of iatrogenic damage to decrease the quality of life for patients that may live for decades after an intervention. Treatment options include the use of microsurgery, medical therapy, and irradiation. Therapeutic goals include the destruction of a tumor or control of its growth, controlling hormonal hypersecretion (if present), and the restoration of vision or lost function without injuring surrounding normal neurovascular tissues or inducing pituitary hypofunction. Radiation may be delivered in a single radiosurgical dose or a variable number of smaller doses. Radiation is frequently employed for incompletely resected tumors, for recurrent tumors, and as a primary treatment for medically inoperable patients and for patients who refuse a surgical intervention. This chapter will primarily review the use of conventionally fractionated radiation therapy (25 to 30 weekday treatments, given over a 5- to 6-week time period, in doses of approximately 1.8 Gy/day) for pituitary adenomas, meningiomas, and cranio-

pharyngiomas, all benign anterior skull base tumors arising in the sella and parasellar tissues. Many of the management principles and concerns are largely the same for other benign or for malignant tumors arising in this area.

Pituitary Adenomas Pituitary adenomas represent approximately 10% to 12% of all adult primary intracranial neoplasms; about half of these pituitary adenomas are hormonally active [2]. Several autopsy series have detected pituitary adenomas in as many as 22% to 25% of cases [3, 4]. Equal numbers of men and women develop macroadenomas, and the incidence peaks in the third and fourth decades of life. Only 10% of cases arise in children or adolescents. Individuals with multiple endocrine neoplasia type I have a hereditarily increased risk of developing a pituitary adenoma. Pituitary adenomas are classiﬁed as either functioning (endocrine-active) or nonfunctioning (endocrine-inactive), with functioning tumors subclassiﬁed by the endocrine axis or axes upon which they act; nonfunctioning tumors are usually classiﬁed as either oncocytic or non-oncocytic (Table 28-1). The detection of these hormones may be possible by testing peripheral blood or by doing immunohistochemical staining of the tumor.

Meningiomas Meningiomas represent approximately 18% to 20% of all primary brain tumors. Twenty-ﬁve percent to 30% of meningiomas arise along the base of the anterior and middle cranial fossae in locations such as the cavernous sinus, tuberculum sellae, Meckel cave, or sphenoid wing [5]. The incidence of intracranial meningioma increases with increasing age, and women are more commonly affected than men by a ratio of approximately 2 : 1. Meningiomas are more common in

317

318

j.p.s. knisely and p.w. sperduto

TABLE 28-1. Pituitary tumor classiﬁcation. Pituitary adenoma cell type

Hormone product

Clinical syndrome

Lactotroph

Prolactin

Somatotroph

Growth hormone

Somatotroph/ lactotroph

Growth hormone and prolactin

Thyrotroph Gonadotroph Mixed cell (non-oncocytic) Oncocytic

Thyrotropin FSH, LH None

Galactorrhea/ amenorrhea/impotence Acromegaly, pituitary giantism Acromegaly and galactorrhea/ amenorrhea Hyperthyroidism Hypopituitarism Hypopituitarism

None

Hypopituitarism

FSH, follicle-stimulating hormone; LH, luteinizing hormone.

individuals with neuroﬁbromatosis type 2, in individuals who have a remote history of cranial irradiation, and in patients with pulmonary lymphangioleiomyomatosis [6]. Histopathologically, resected meningiomas from the skull base should be graded as per the 2000 World Health Organization (WHO) guidelines [7]. If there are ≥4 mitoses per 10 highpowered microscopy ﬁelds (HPF) or at least three features believed to correspond with aggressiveness are seen—small cells, macronucleoli, hypercellularity, necrosis, and sheeting architecture—a meningioma is deemed atypical (grade II). Brain invasion is now an indicator of grade II rather than grade III behavior in the absence of any histologic anaplasia. A grade III meningioma must have frank anaplasia or a mitotic rate of >20/10 HPF. Four variants are considered innately more aggressive by deﬁnition: clear cell (grade II), chordoid (grade II), papillary (grade III), and rhabdoid (grade III). Grade II meningiomas recur at a rate of ∼40% at 5 years time, and grade III meningiomas have a median overall survival of <2 years [8, 9]. Grade I tumors have a short term recurrence rate that is very low, even if bone adjacent to the tumor is invaded.

Craniopharyngiomas Craniopharyngiomas are rare, representing only a few percent of all intracranial primary tumors [10]. Craniopharyngiomas arise from either embryonic remnants of an incompletely involuted hypophyseal-pharyngeal duct or from metaplastic ectodermal cells in the anterior hypophysis [11]. They account for a signiﬁcant portion of the intracranial tumors of childhood but are as common in adults as in children and adolescents [12]. The age distribution shows a bimodal pattern, with a ﬁrst, larger peak between the ages of 5 and 10 years, and a second less prominent peak in the ﬁfth decade, and there is a very slight male preponderance reported in some large series. No recognized hereditary syndromes are associated with craniopharyngiomas. Two major histopathologic categories of craniopharyngioma are recognized. There is evidence that they may have different etiologies [10]. The adamantinomatous variant is more commonly seen in pediatric patients, and papillary squamous tumors are more commonly seen in adults and are apparently

more likely to be durably controlled by surgery alone. There is apparently some occasional intergrading of craniopharyngioma tumor types [12].

Clinical Presentation Pituitary Adenomas Endocrine abnormalities are the most common presentation of pituitary adenomas. Functioning adenomas will manifest themselves by causing one of the classic endocrine syndromes (Table 28-1). Hormonal insufﬁciency may arise from compression of the normal pituitary gland or of the pituitary stalk. Insufﬁciencies may be more subtle than hypersecretory states, but either can develop insidiously and only be discovered after a protracted period of symptoms. Amenorrhea in young women and a decreased libido in men may be the signs and signals that trigger an evaluation of the hypothalamic-pituitary axis. Endocrine-active adenomas are more commonly diagnosed as microadenomas (<1.0 cm in diameter) than as macroadenomas (≥1.0 cm in diameter), but somatotroph or lactotroph macroadenomas are common. Lateral extension of pituitary tumors into the cavernous sinus can cause diplopia or facial sensory symptoms. Superior extension will cause pressure on the optic nerves and chiasm, with a resultant bitemporal superior quadrantanopsia or bitemporal hemianopsia. Frontal headaches may be caused by irritation of meningeal sensory ﬁbers as the adenoma enlarges.

Meningiomas The presenting symptoms of meningiomas may be diverse in these locations but are most commonly from mass effect upon adjacent normal tissues. The exact symptoms will depend upon what structures are compressed. The clinical symptomatology may not be as distinctive as for a functional pituitary adenoma, and high-quality imaging is frequently required to determine whether a patient may have a pituitary adenoma, a meningioma, a schwannoma, or other regional pathology such as an aneurysm, sarcoid, or another condition. Involvement of the cavernous sinus will commonly cause diplopia or facial sensory symptoms; a meningioma arising from the optic nerve sheath will cause unusual visual symptoms.

Craniopharyngiomas These tumors most commonly are diagnosed as a result of their mass effect upon the optic nerves or from mass effect upon normal hypothalamic or pituitary function. As these are generally slow-growing tumors, an insidious loss of function may occur, so that the degree of compensation for loss of vision or other deﬁcits may be remarkable at the time of diagnosis. Increased intracranial pressure may lead to headaches and projectile vomiting. The mass effect in adults usually results in amenorrhea for women and in loss of libido for men, and if the tumor extends far enough into the cranial vault, hypothalamic, frontal, or temporal lobe symptoms may develop. Diabetes insipidus, obesity, and neurocognitive or developmental abnormalities may result from a craniopharyngioma.

28.

pituitary and pituitary region tumors: fractionated radiation therapy perspective

Diagnostic Workup and Staging All patients with tumors arising in this region should be managed with a multidisciplinary approach. Not every specialist needs to see every case, but with a team of specialists who are comfortable with multidisciplinary care and who are aware of available therapeutic options, there will be less of a chance of overlooking tests or therapies that could help develop individualized safe and effective treatment plans (Table 28-2). For tumors that present with mass effect, surgical debulking of the tumor may be needed promptly. A neurosurgeon skilled in skull base and pituitary surgery and knowledgeable about the use of fractionated irradiation and stereotactic radiosurgery should be consulted. This surgeon should be aware of the need to establish a distance between the rostral surface of the tumor and the optic nerves. High-dose glucocorticoids may provide temporary improvement of mass effect while additional evaluations are performed. An endocrinologic evaluation can determine whether there are hormonal abnormalities that need to be addressed before surgery and may also reveal that medical management is possible as an initial therapy. Discussion of each case with a neuroradiologist when imaging studies are being ordered will facilitate obtaining appropriately detailed studies. A dedicated pituitary magnetic resonance imaging (MRI) scan with and without gadolinium contrast administration will show details of the relation of the tumor to the optic nerves, chiasm, and tracts and other critical structures that would affect radiation management options being considered. Pulse sequences that suppress the bright signal from fat can clarify whether or not enhancing tumor extends into the orbit. Computed tomography (CT) imaging may show bone remodeling. A neuroophthalmologic consultation is often warranted to evaluate the patient for subtle funduscopic abnormalities or tumor-related vision changes. Serial perimetry is able to follow visual ﬁelds over time in a way that clinical examination cannot. Clinical evidence of injury to the cranial nerves controlling extraocular motion should be documented by a careful observer; improvement is often seen after irradiation.

319

When a patient appears to have long-standing neurocognitive issues, it is worth the effort to have the patient evaluated with neurocognitive testing appropriate to the clinical situation. This can help identify learning issues that may be addressable with interventions or therapy. A radiation oncologist should also be asked to evaluate the patient so that the strengths and shortcomings of different radiation management strategies may be presented to arrive at a satisfactory management approach. Neurosurgical staging systems for anterior skull base tumors have little applicability to radiation management decisions [13–15]. The critical issues for radiation management decisions for pituitary adenomas and all tumors in the sellar and parasellar area are the proximity of the tumor to the optic apparatus and the size of the tumor. Pituitary adenomas and other tumors that are large or that very closely approach the optic nerves or chiasm are not appropriate for stereotactic radiosurgery. Fractionated irradiation can be used for all tumors in this location. Not surprisingly, tumor volume has been shown to be inversely correlated with long-term control in patients with pituitary adenomas that are treated with fractionated irradiation [16]. For meningiomas, the most common staging system was developed nearly 50 years ago to provide some guidance as to the risk of recurrence after surgical management [15]. Molecular markers may be assessed to help predict the risk for recurrence after surgical resection [17, 18]. Radiotherapy’s value in controlling meningioma growth is broadly recognized. If a meningioma very closely approaches or impinges upon the optic apparatus, there is no role for radiosurgery, although fractionated radiation can be used. There is no accepted staging system for craniopharyngiomas, which are frequently densely adherent to adjacent normal tissues. Aggressive surgical resections may be complicated by intraoperative damage to normal tissues involved with the tumor. Irradiation may also cause morbidity but generally is not as likely to cause an acute loss of function. Radiation options include fractionated irradiation, stereotactic radiosurgery, and instillation of radioisotope colloids into cystic tumors.

TABLE 28-2. Diagnostic evaluation of patient with a sellar or parasellar tumor. History and physical examination with particular attention to signs, symptoms, and stigmata from hormonal hypersecretion or hyposecretion and a detailed cranial nerve examination. Laboratory endocrine evaluation (for appropriate patients with apparent endocrinologic problems, based on clinical assessment, or if imaging reveals a tumor arising in or extending to the pituitary sella or suprasellar-hypothalamic region). Provocative testing of apparently affected axes, with assessment of: • Serum prolactin level • Fasting growth hormone level • IGF-1 level, and growth hormone dynamic testing with insulin tolerance, glucose suppression testing, and thyrotropin-releasing hormone testing • Serum ACTH, 24-hour urinary free cortisol and 17-hydroxycorticosteroid levels, and response to dexamethasone suppression testing • Gonadal function testing with serum lh, fsh levels, and plasma estradiol and testosterone levels • Thyrotropin (TSH), thyroxine, and triiodothyronine levels • Complete blood count (CBC) and coagulation studies Radiological evaluation including an MRI scan with coronal thin cuts through the sphenoid bone with and without gadolinium contrast, and possibly also including CT imaging to evaluate bony changes. Specialized neurointerventional imaging and petrosal venous sinus sampling may be required in cases of Cushing disease with a radiographically occult microadenoma. Neuroophthalmologic examination for patients with tumors that impinge upon the optic apparatus or cause cranial nerve abnormalities affecting vision.

320

j.p.s. knisely and p.w. sperduto

Usual Therapeutic Approaches Radiation therapy is usually deferred until after any operative neurosurgical management has been completed. This is because surgery is able to address both mass effect (if present) as well as hormonal hypersecretion (if the tumor in question is a hormonally active pituitary adenoma). For some pituitary adenomas (prolactinomas, growth hormone–secreting adenomas, and thyroid hormone–secreting adenomas) it is possible to use pharmacologic management with speciﬁc action at the hypothalamic-pituitary axis and avoid or defer surgical therapy or radiation therapy [19–21]. Neither meningiomas nor craniopharyngiomas have systemic management options similar to those for functioning pituitary adenomas. Meningiomas may be detected incidentally and found to be without clinical signs or symptoms. As these tumors are generally slow growing, documenting the rate of growth prior to pursuing deﬁnitive interventions may be useful, as no intervention is without risk. Some meningiomas and most craniopharyngiomas are detected because of clinical signs or symptoms from the tumor’s presence at the skull base. In some of these cases, a craniotomy may be needed to try to alleviate mass effect and bring about a recovery of function, but it may not be appropriate for all cases, when radiation management options exist that are likely to be highly effective. After an initial therapeutic intervention (usually surgery), it is common to repeat many elements of the battery of tests that were performed at the time of diagnosis to document any changes in neurologic, ophthalmologic, or endocrinologic function (Table 28-1). Documentation of improvements and deﬁcits will allow them to be accurately attributed to the effects of the tumor, to the initial intervention, or to subsequent therapy. If there is consideration being given to the use of stereotactic radiosurgery for a tumor that preoperatively was documented to be very close to the optic nerves, volumetric contrast-enhanced MRI must be repeated to ensure that the tumor has been adequately debulked away from the optic apparatus. This geometric separation of several millimeters between the tumor and the optic apparatus is not as critical if fractionated irradiation will be used as a postsurgical therapy. The optic nerves will tolerate fractionated doses of radiation to ∼50 Gy that will durably control most tumors in this area, while a radiation dose >8 to 10 Gy given to the optic nerves in a single fraction will cause vision loss in an unacceptably high number of patients [22–25].

Radiation Therapy The role of radiotherapy in the management of anterior skull base tumors has waxed and waned over the past century. Radiotherapy treatment for pituitary adenomas and parasellar tumors is currently delivered with a very high degree of precision, is extremely well tolerated, and provides excellent outcomes by delivering a cytostatic or cytocidal dose of radiation to the tumor while not exceeding the radiation tolerance of adjacent normal tissues. Current radiation therapy techniques are greatly improved relative to those of earlier eras. Superior resolution on imaging of normal anatomy and pathology is a major component of this,

as tumors may be diagnosed earlier in their natural history. Modern treatment techniques reproducibly immobilize patients to achieve setup accuracies of approximately 2 mm, and radiation ﬁelds may be shaped to match tumor geometries and include less normal tissues. Less normal tissue irradiation will contribute to lower rates of late endocrine hypofunction and other normal tissue damage. Temporal lobe damage has been described as a complication associated with historical radiation therapy techniques [26]. The use of higher-energy photons and multiple treatment portals to treat centrally located tumors should decrease the radiation dose delivered to the temporal lobes. Using more radiation portals may attenuate late neurocognitive sequelae, but the possibility exists that late radiationinduced tumorigenesis may be increased with treatment techniques that increase the volume of normal tissues exposed to low doses of irradiation; useful data regarding incidence rates and dose-volume histograms that might provide clinical guidance regarding this particular complication are essentially nonexistent [27].

Simulation and Radiotherapy Planning Before radiation treatment can commence, preparatory steps must occur, and the exact nature of these steps will vary slightly with the equipment being used to treat the patient. Radiotherapy treatment techniques have evolved from a distant era where the distance between the patient’s right and left temple along the central axis of a pair of opposing X-ray beams aimed at the presumed location of the tumor was the critical factor involved in setting up, planning, and delivering radiation therapy. Current sophisticated radiotherapy techniques use meticulous care in a series of steps before any treatment is given to ensure that treatment will be delivered to the tumor being targeted only as prescribed by the radiation oncologist. Simulation is the process of determining the geometric factors relating to fractionated irradiation delivery, and treatment planning uses information about the radiation beams and these geometric parameters to generate data on radiation doses to the tumor and normal tissues. The ﬁrst step in this process is to fabricate a patient-immobilizing device that will be used for all subsequent treatments. A thermoplastic mask is commonly used for this purpose, together with a baseplate that holds the patient’s head in a position that tucks the chin to the chest. Many centers use baseplates that have the ability to be set at a predetermined angle for reproducibly achieving a high degree of chin tuck. A well-constructed mask and baseplate pair should provide setup accuracies of ∼2 mm. More elaborate immobilization techniques that incorporate custom-molded bite blocks and custom-molded posterior head supports can provide immobilization accuracy that approaches a millimeter or less. Rigid and reproducible immobilization techniques are mandatory for the most conformal radiation therapy delivery techniques. The position of the patient’s head at the time of the fabrication of this device will determine whether simple vertical beams or arcs can be used with the couch in a neutral position. Inattention to this point may reduce the ease of daily treatment or prevent some well characterized and simple treatment techniques from being easily adopted. For pituitary radiation therapy, a traditional simulation method uses a ﬂuoroscopic apparatus with the same geometry

28.

pituitary and pituitary region tumors: fractionated radiation therapy perspective

321

as a linear accelerator to set up an anterior and two lateral radiation portals of approximately 4 × 4 cm, centered just above the pituitary fossa. The immobilizing mask can be marked with the entry points, central axes, and corners for these ﬁelds. A contour of the skull is obtained along the line that joins the central points of these ﬁelds. A dosimetrist can use this contour information and X-ray beam data to calculate radiation parameters with computerized treatment planning software. Obviously, the location of the radiation portals and their size depends upon the size and location of the tumor; the size of smaller portals is frequently limited by a physicist’s ability to accurately measure the beam output through a small and irregular aperture. CT simulation simpliﬁes some aspects of this process. The patient needs to be immobilized in the treatment position, and setup points need to be marked on the mask for later reference. A noncontrast CT scan with ﬁne cuts (1.25–2.5 mm thick, extending from the vertex of the head to the foramen magnum) is obtained. Contouring (identiﬁcation of structures in threedimensional space in the imaging study) needs to be done for all the important tissues, particularly if intensity-modulated radiation therapy (IMRT) techniques will be used to treat the patient. High-resolution treatment planning CT studies generally show the pituitary fossa, optic nerves, and lenses well enough that these structures can be clearly deﬁned. The radiation therapy portals are set up on digitally reconstructed radiographs after the contouring process has been completed, and from this information, a dosimetrist can directly use this digital data to calculate the beam weights and dosimetric parameters. The use of image fusion software to register a high-quality MRI study to the treatment planning CT scan may help in delineation of the tumor and adjacent normal tissues. Image

fusion software rigidly registers the two imaging studies through the use of normalized mutual information, chamfer matching, or other algorithms and then reformats and redisplays the voxels from the MRI in the CT scan’s geometric space. The MRI scan does not need to be done with the patient in the immobilization mask. Structures contoured on the MRI can be transferred to the CT for calculation purposes. Clinicians have realized from more detailed dosimetric analyses that there can be areas of unacceptably high dose (“hot spots”) that may injure normal tissues or areas of incomplete coverage of intracranial target tissue (“cold spots”) that can lead to tumor recurrences [28, 29]. Radiotherapy technology advances that permit more than a few coplanar ﬁelds to be used to treat a tumor have been used by researchers to determine an optimal number of radiation beams for treatment of skull base tumors [30, 31]. Treatments that are three-dimensionally conformal to a radiation target can be routinely delivered with a high degree of precision. This has been termed 3DCRT (threedimensional conformal radiation therapy). The use of rigid, reproducible immobilization and a large number of radiation beams or of arcing beams has led some to term this approach fractionated stereotactic radiation therapy (FSRT), although this terminology is inaccurate in that true stereotaxy (in a neurosurgical sense) is not achieved [32]. An example of an FSRT dosimetric plan for a patient with a pituitary adenoma is shown in Figure 28-1A–C. Recently, software and hardware technology to dynamically control the point-to-point intensity of incident radiotherapy beams has been developed. This innovation is known by the acronym IMRT (intensity-modulated radiation therapy). Radiation therapy plans generated and delivered using IMRT techniques are signiﬁcantly different than plans developed in

FIGURE 28-1. (A) Axial fractionated stereotactic radiation therapy (FSRT) dose distribution. (B) Coronal FSRT dose distribution. (C) Sagittal FSRT dose distribution. (D) Axial IMRT dose distribution. (E) Coronal IMRT dose distribution. (F) Sagittal IMRT dose distribution. It is clear from these dose distributions (shown on MR images) that for this patient with a pituitary adenoma, it was possible to develop

a more conformal FSRT plan than an IMRT plan. Tissues outside the target volume are treated to higher doses in the IMRT plan than in the FSRT plan. Both plans would be acceptable for fractionated irradiation using conventional fractionation, but neither could be accepted for radiosurgical treatment, because the optic nerves tolerance would be exceeded by doses that would control the adenoma.

322

j.p.s. knisely and p.w. sperduto

the past. For an IMRT plan to be generated, a computer must be told what the overall goals of the plan should be. These include limitations of doses to certain structures, acceptable ranges of doses to other structures, and the basic parameters of beam energy and beam entry points. The computer software then selects radiation beam intensities that can achieve the desired therapeutic goals. Better integration of these software and hardware advances and faster computer microprocessors has allowed the use of changes in radiation beam intensity in multiple portals to achieve better radiation dose distributions, with fewer hot or cold spots inside the target volume and no hot spots outside the target volume. These technological advances may decrease treatment-related morbidity and improve treatment results overall, but this has not yet been demonstrated to be the case for intracranial tumors. An example of an IMRT dosimetric plan for a patient with a pituitary adenoma is shown in Figure 28-1D–F.

Radiation Therapy Results Pituitary Macroadenomas Fractionated doses of approximately 45 to 50 Gy in daily doses of 1.8 Gy are currently used for treating pituitary adenomas at most centers because of the knowledge that has been gained over the past decades about the low risk of causing visual complications when these guidelines are followed, the dose response seen for doses lower than this range, and the apparent absence of improved results and increasing rate of visual and other complications when higher total doses and plans with hot spots are used [16, 33–38]. For patients who have undergone a surgical resection of a pituitary adenoma, long-term local control is improved when radiation treatment is administered before a macroscopic recurrence of tumor develops, though a patient with a complete resection of tumor may not require postoperative irradiation [16, 39–41]. The rates of recurrence are not at all trivial in many series, and there is a signiﬁcant potential for progression at 5 to 10 years after surgical resection [42, 43]. Patients must understand the need of close surveillance if radiation therapy is not employed as a routine adjunctive therapy.

Radiotherapy alone achieves control rates of approximately 80% to 95% at 10 years follow-up for recurrent nonfunctional adenomas or for primary treatment of nonfunctional adenomas [16, 33, 43]. Longer-term follow-up has documented continued progression up to >20 years from therapy, documenting the need for lifelong surveillance of patients with pituitary adenomas [44, 45]. A critical factor in the long-term local control after postoperative irradiation of pituitary adenomas may well be the postoperative volume of tumor. This may conceivably range from a small amount of tumor remaining in a cavernous sinus to a macroscopic suprasellar extension into the midbrain. Local control can now be more accurately determined than in the past because of improvements in imaging technology and availability. Side-by-side comparisons can be made between studies obtained a decade or more apart, and subtle changes in size can be detected. In addition to the prevention of adenoma growth, many radiotherapy series have reported an improvement in mass effect and mass effect–related symptoms such as vision, but neurosurgical management is much more effective in promptly relieving mass effect. Several reports on fractionated irradiation for nonfunctioning pituitary adenomas are collected in Table 28-3.

Growth Hormone–Secreting Adenomas (Acromegaly) Acromegaly is still a common indication for fractionated irradiation. These tumors are frequently large, and as a result, surgical debulking of suprasellar extension may be inadequate to permit consideration of stereotactic radiosurgery because the tumor may still be in close juxtaposition to the optic apparatus. There are some who believe acromegaly is associated with a higher risk of radiation optic neuropathy than for nonfunctional adenomas, with a cited risk of as high as 1.4% with conventional treatment techniques [25, 44]. A large tumor mass and high pretreatment growth hormone levels may predict a worse prognosis for patients with acromegaly [45, 49]. Older literature on the effectiveness of radiation therapy for acromegaly cannot provide all the data that is believed necessary to evaluate the effectiveness of the therapy. The deﬁnitions of normal have been lowered with more precise laboratory testing over the past several decades. Current guidelines on the

TABLE 28-3. Results of radiation therapy for nonfunctional adenomas. Results Reference

No. of patients

Treatment

End point

Duration (%)

Breen et al. [43]

120

S + RT, RT

LC

10 y, 88; 20 y, 78; 30 y, 65

Tsang et al. [16]

160

S + RT

LC

10 y, 87

Grigsby et al. [46]

111

S + RT, RT

DFS

13 y, 84

Brada et al. [33]

252

S + RT, RT

PFS

10 y, 97; 20 y, 92

Woollons et al. [39]

50

S + RT

PFS

5 y, 72

Gittoes et al. [47]

66

S + RT

PFS

15 y, 93

Sasaki et al. [48]

65

S + RT

LC

10 y, 98

Grabenbauer et al. [38]

50

S + RT

LC

54 mo, 94%

S, surgery; RT, radiation therapy; DFS, disease-free survival; PFS, progression-free survival; LC, local control.

28.

pituitary and pituitary region tumors: fractionated radiation therapy perspective

323

TABLE 28-4. Results of radiation therapy for growth hormone–secreting adenomas. Results Reference

No. of patients

Treatment

End point

Duration (%)

5 y, 30; 10 y, 53; 15 y, 77; 20 y, 89

Eastman et al. [56]

87

S + RT, RT

GH <5 μg/L

Tsang et al. [34]

52

S + RT, RT

Milker-Zabel et al. [60]

20

S + RT, RT

GH <10 μg/L “Normalized GH level”; “Normalized IGF-1 level”

Biermasz et al. [58]

40

S + RT

GH <5 μg/L; “Normalized IGF-1 level”

26 mo, 80; 25 mo, 45 5 y, 75; 10 y, 76; 15 y, 87 Last follow-up, 73

Thalassinos et al. [59]

46

RT

GH <5 μg/L

5 y, 30; 10 y, 32; >10 y, 55

Dowsett et al. [49]

15

S + RT

GH <5 μg/L

3–5 y, 25; 5–10 y, 29; >10 y, 100%

10 y, 46

S, surgery; RT, radiation therapy; GH, growth hormone; IGF, insulin-like growth factor 1.

deﬁnition of cure include a basal growth hormone <4 μg/L, or <1 μg/L with a glucose load, and insulin-like growth factor 1 (IGF-1) levels that are within age-adjusted normative ranges [50]. There is data that the IGF-1 levels after therapy may correlate better with cure than GH values [51]. Modern multidisciplinary management for many patients will include medical management with growth hormone receptor antagonists such as pegvisomant or somatostatin analogues such as octreotide or lanreotide [52, 53]. The use of such medical management strategies may alter the rate at which conventionally delivered radiation can achieve durable control of hormonal excess and may provide additional opportunities for improving the therapeutic ratio by effecting regressions in tumor bulk prior to commencing deﬁnitive irradiation [54]. Radiotherapy should be considered for all patients with a postoperative basal GH level of >5 μg/L, with medical management possibly used to help decrease tumor bulk before commencing irradiation and also to help suppress hormonal excess after irradiation while waiting for the eventual declines that are expected. As there is some data that hormonal suppression prior to radiosurgery may depress the ultimate control rates for some functional pituitary adenomas, these agents may need to be evaluated for any protective effect in the setting of conventionally fractionated and delivered radiotherapy [55]. Adequate follow-up has been obtained in several series of patients with growth hormone–secreting pituitary adenomas that document a progressive response over time [34, 56–59]. The completeness of surgical resection appears to contribute to the long-term results achievable with radiation therapy [57, 59]. The results of FSRT are only now becoming available; further follow-up will document whether or not superior results are

possible with the technological advances that have been used to treat these patients [60]. The results of several series of acromegalic patients treated with fractionated irradiation are listed in Table 28-4.

Prolactin-Secreting Adenomas Prolactinomas are less commonly treated with irradiation at present, and the role of surgical intervention in this disorder may also be declining with the availability of reasonably well tolerated medications that can be taken for protracted periods of time. Historical series show that for patients treated with irradiation, there is a decreased rate of durable control without the addition of medical therapy, but that over time, the proportion of patients with control without medical management does increase, with perhaps 50% of patients achieving a serum level <500 mU/L at 10 years after irradiation [34, 61–63]. The results of several series of prolactinomas treated with fractionated irradiation are summarized in Table 28-5.

Adrenocorticotropic Hormone–Secreting Adenomas (Cushing’s Disease) Adrenocorticotropic hormone (ACTH)-hypersecreting pituitary adenomas should be irradiated if neurosurgical treatment is unsuccessful [66–70]. Fractionated irradiation has historically been used as primary therapy for Cushing’s disease as well, but in current practice, only a rare case will be referred for irradiation without a pituitary operation [70, 71]. In patients medically unable to undergo microsurgery, medical therapy can suppress cortisol excess until the effects of irradiation on hypercortisolemia can be achieved, commonly seen in a year or two

TABLE 28-5. Results of radiation therapy for prolactinomas. Results Reference

No. of patients

Treatment

End point

Duration (%)

Ozgen et al. [62]

106

S + RT

PRL <30 μg/L

Not stated, 58

Tsang et al. [34]

64

S + RT

Wallace & Holdaway [64]

25

S + RT RT

PRL <20 μg/L Normal PRL

4 y, 33

Tsagarakis et al. [65]

36

Williams et al. [61] Littley et al. [63]

28 58

PRL, prolactin; S, surgery; RT, radiation therapy.

RT

PRL <360 mU/L Normal PRL

S + RT, RT

PRL <500 mU/L

10 y, 25 8 y, 50 Not stated, 29 10 y, 50% (predicted)

324

j.p.s. knisely and p.w. sperduto

TABLE 28-6. Results of radiation therapy for Cushing’s disease. Results Reference

No. of patients

Treatment

End point

Duration (%)

9.5 y, 57 10 y, 53

Howlett et al. [72] Tsang et al. [34]

21 29

RT

Normal mean serum cortisol

S + RT, RT

Urinary free cortisol (UFC) <220 nmol/day

Vicente et al. [66]

14 86

UFC <303.5 nmol/24 h Low-dose dexamethasone suppression test (serum cortisol <80 nmol/L)

2 y, 70

Nagesser et al. [73]

S + RT RT and unilateral adrenalectomy

Estrada et al. [75]

30

S + RT

UFC <303.5 nmol/24 h; low-dose dexamethasone suppression test normal

42 mo, 83%

21 y, 64

S, surgery; RT, radiation therapy.

[72–74]. The results of several radiotherapy series for Cushing’s disease are detailed in Table 28-6.

Meningiomas Radiotherapy has been shown to be highly effective at controlling meningiomas in numerous series [29, 76–80]. Large, recurrent, and histologically aggressive tumors are more likely to progress after treatment [29, 78]. The morbidity associated with fractionated irradiation is acceptably low, particularly when compared with surgical series, but can certainly be minimized for individual patients by not using radiotherapy plans that have dosimetric hot spots or cold spots that may lead to radiation injury to normal tissues or early recurrences [78–81].

Craniopharyngiomas Radiotherapy has been used in the treatment of craniopharyngiomas for many decades. Detailed evaluations of surgical outcomes and of radiation therapy outcomes have been compiled [82, 83]. Aggressive surgery has been associated with higher rates of visual loss and diabetes insipidus, and larger tumors are recognized to be more likely to recur [82]. Novel radiation delivery technologies have been used to treat craniopharyngiomas recently [84–87]. These treatment techniques may limit the dose of irradiation given to contiguous normal brain and pituitary, but care must be taken to avoid radiation injury to the optic apparatus [84] or not irradiating the entire tumor [85, 86].

Conclusion Numerous advances in imaging and treatment delivery have made fractionated irradiation for pituitary adenomas and other anterior skull base tumors an extremely good therapeutic option. Evaluation of radiation management approaches for individual patients with tumors in the pituitary and parasellar region should bear in mind the potential for signiﬁcant complications from the attempt to use the most advanced technology available. Not every patient should be treated with radiosurgery or IMRT; extremely good and predictable results are possible with highly conformal fractionated irradiation, an option that should be nearly universally available.

References 1. Rhoton AL Jr. The anterior and middle cranial base. Neurosurgery 2002; 51:S273–S302.

2. Laws ER Jr, Thapar K. Brain tumors. CA Cancer J Clin 1993; 43: 263–271. 3. Costello RT. Subclinical adenoma of the pituitary gland. Am J Pathol 1936; 12:205–216. 4. Kernohan JW, Sayre GP. Tumors of the pituitary gland and infundibulum. In: Atlas of Tumor Pathology. Washington, DC: Armed Forces Institute of Pathology, Section X, Fascicle 36, 1956. 5. DeMonte F, Al-Mefty O. Meningiomas. In: Kaye AH, Laws ER, eds. Brain Tumors. New York: Churchill Livingstone, 1995:675– 704. 6. Moss J, DeCastro R, Patronas NJ, et al. Meningiomas in lymphangioleiomyomatosis. JAMA 2001; 286:1879–1881. 7. Louis DN, Scheithauer BW, Budka H, et al. Meningiomas. In: Kleihues P, Cavenee WK, eds. World Health Organization Classiﬁcation of Tumours: Pathology and Genetics of Tumours of the Nervous System. Lyon: IARC Press, 2000:176–184. 8. Perry A, Stafford SL, Scheithauer BW, et al. Meningioma grading: an analysis of histologic parameters. Am J Surg Pathol 1997; 21: 1455–1465. 9. Perry A, Scheithauer BW, Stafford SL, et al. “Malignancy” in meningiomas: a clinicopathologic study of 116 patients. Cancer 1999; 85:2046–2056. 10. Kahn EA, Gosch HH, Seeger JF, et al. Forty-ﬁve years experience with the craniopharyngiomas. Surg Neurol 1973; 1:5–12. 11. Yasargil MG, Curcic M, Kis M, et al. Total removal of craniopharyngiomas. Approaches and long-term results in 144 patients. J Neurosurg 1990; 73:3–11. 12. Adamson TE, Wiestler OD, Kleihues P, et al. Correlation of clinical and pathological features in surgically treated craniopharyngiomas. J Neurosurg 1990; 73:12–17. 13. Hardy J. Transphenoidal microsurgery of the normal and pathological pituitary. Clin Neurosurg 1969; 16:185–217. 14. Leavens ME, McCutcheon IF, Samaan NA. Management of pituitary adenomas. Oncology (Williston Park). 1992; 6(6):69–79; discussion 79–80. 15. Simpson D. The recurrence of intracranial meningiomas after surgical treatment. J Neurol Neurosurg Psychiatry 1957; 20:22–39. 16. Tsang RW, Brierley JD, Panzarella T, et al. Radiation therapy for pituitary adenoma: treatment outcome and prognostic factors. Int J Radiat Oncol Biol Phys 1994; 30:557–565. 17. Perry A, Gutmann DH, Reifenberger G. Molecular pathogenesis of meningiomas. J Neurooncol 2004; 70:183–202. 18. Perry A. Unmasking the secrets of meningioma: a slow but rewarding journey. Surg Neurol 2004; 61:171–173. 19. Molitch ME. Medical management of prolactin-secreting pituitary adenomas. Pituitary 2002; 5:55–65. 20. Racine MS, Barkan AL. Medical management of growth hormonesecreting pituitary adenomas. Pituitary 2002; 5:67–76. 21. Beck-Peccoz P, Persani L. Medical management of thyrotropinsecreting pituitary adenomas. Pituitary 2002; 5:83–88.

28.

pituitary and pituitary region tumors: fractionated radiation therapy perspective

22. 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; 27:215–221. 23. 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. 24. Stafford SL, Pollock BE, Leavitt JA, et al. A study on the radiation tolerance of the optic nerves and chiasm after stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2003; 55:1177–1181. 25. van den Bergh AC, Schoorl MA, Dullaart RP, et al. Lack of radiation optic neuropathy in 72 patients treated for pituitary adenoma. J Neuroophthalmol 2004; 24:200–205. 26. al-Mefty O, Kersh JE, Routh A, et al. The long-term side effects of radiation therapy for benign brain tumors in adults. J Neurosurg 1990; 73:502–512. 27. Hall EJ, Wuu CS. Radiation-induced second cancers: the impact of 3D-CRT and IMRT. Int J Radiat Oncol Biol Phys 2003; 56:83– 88. 28. Marks JE, Baglan RJ, Prassad SC, et al. Cerebral radionecrosis: incidence and risk in relation to dose, time, fractionation and volume. Int J Radiat Oncol Biol Phys 1981; 7:243–252. 29. Goldsmith BJ, Larson DA. Conventional radiation therapy for skull base meningiomas. Neurosurg Clin N Am 2000; 11:605– 615. 30. Graham JD, Nahum AE, Brada M. A comparison of techniques for stereotactic radiotherapy by linear accelerator based on 3-dimensional dose distributions. Radiother Oncol 1991; 22:29– 35. 31. Perks JR, Jalali R, Cosgrove VP, et al. Optimization of stereotactically-guided conformal treatment planning of sellar and parasellar tumors, based on normal brain dose volume histograms. Int J Radiat Oncol Biol Phys 1999; 45:507–513. 32. Kondziolka D, Lunsford LD, Loefﬂer JS, et al. Radiosurgery and radiotherapy: observations and clariﬁcations. J Neurosurg 2004; 101:585–589. 33. 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; 38:571–578. 34. Tsang RW, Brierley JD, Panzarella T, et al. Role of radiation therapy in clinical hormonally-active pituitary adenomas. Radiother Oncol 1996; 41:45–53. 35. McCord MW, Buatti JM, Fennell EM, et al. Radiotherapy for pituitary adenoma: long-term outcome and sequelae. Int J Radiat Oncol Biol Phys 1997; 39:437–444. 36. Zierhut D, Flentje M, Adolph J, et al. External radiotherapy of pituitary adenomas. Int J Radiat Oncol Biol Phys 1995; 33:307– 314. 37. Milker-Zabel S, Debus J, Thilmann C, et al. Fractionated stereotactically guided radiotherapy and radiosurgery in the treatment of functional and nonfunctional adenomas of the pituitary gland. Int J Radiat Oncol Biol Phys 2001; 50:1279–1286. 38. Grabenbauer GG, Fietkau R, Buchfelder M, et al. Hormonally inactive hypophyseal adenomas: the results and late sequelae after surgery and radiotherapy. Strahlenther Onkol 1996; 172:193– 197. 39. Woollons AC, Hunn MK, Rajapakse YR, et al. Non-functioning pituitary adenomas: indications for postoperative radiotherapy. Clin Endocrinol (Oxf) 2000; 53:713–717. 40. Lillehei KO, Kirschman DL, Kleinschmidt-DeMasters BK, et al. Reassessment of the role of radiation therapy in the treatment of endocrine-inactive pituitary macroadenomas. Neurosurgery 1998; 43:432–438. 41. Greenman Y, Ouaknine G, Veshchev I, et al. Postoperative surveillance of clinically nonfunctioning pituitary macroadenomas: markers of tumour quiescence and regrowth. Clin Endocrinol (Oxf) 2003; 58:763–769.

325

42. Grigsby PW, Simpson JR, Fineberg B. Late regrowth of pituitary adenomas after irradiation and/or surgery. Hazard function analysis. Cancer 1989; 63:1308–1312. 43. Breen P, Flickinger JC, Kondziolka D, et al. Radiotherapy for nonfunctional pituitary adenoma: analysis of long-term tumor control. J Neurosurg 1998; 89:933–938. 44. van den Bergh AC, Hoving MA, Links TP, et al. Radiation optic neuropathy after external beam radiation therapy for acromegaly: report of two cases. Radiother Oncol 2003; 68:101–103. 45. Goffman TE, Dewan R, Arakaki R, et al. Persistent or recurrent acromegaly. Long-term endocrinologic efﬁcacy and neurologic safety of postsurgical radiation therapy. Cancer 1992; 69:271– 275. 46. Grigsby PW, Stokes S, Marks JE, et al. Prognostic factors and results of radiotherapy alone in the management of pituitary adenomas. Int J Radiat Oncol Biol Phys 1988; 15:1103–1110. 47. Gittoes NJ, Bates AS, Tse W, et al. Radiotherapy for non-function pituitary tumours. Clin Endocrinol (Oxf) 1998; 48:331–337. 48. Sasaki R, Murakami M, Okamoto Y, et al. The efﬁcacy of conventional radiation therapy in the management of pituitary adenoma. Int J Radiat Oncol Biol Phys 2000; 47:1337–1345. 49. Dowsett RJ, Fowble B, Sergott RC, et al. Results of radiotherapy in the treatment of acromegaly: lack of ophthalmologic complications. Int J Radiat Oncol Biol Phys 1990; 19:453–459. 50. Giustina A, Barkan A, Casanueva FF, et al. Criteria for cure of acromegaly: a consensus statement. J Clin Endocrinol Metab 2000; 85:526–529. 51. Gullu S, Keles H, Delibasi T, et al. Remission criteria for the follow-up of patients with acromegaly. Eur J Endocrinol 2004; 150:465–471. 52. Clemmons DR, Chihara K, Freda PU, et al. Optimizing control of acromegaly: integrating a growth hormone receptor antagonist into the treatment algorithm. J Clin Endocrinol Metab 2003; 88:4759–4767. 53. Melmed S, Vance ML, Barkan AL, et al. Current status and future opportunities for controlling acromegaly. Pituitary 2002; 5:185– 196. 54. Melmed S, Sternberg R, Cook D, et al. A critical analysis of pituitary tumor shrinkage during primary medical therapy in acromegaly. J Clin Endocrinol Metab 2005; 90:4405–4410. 55. Landolt AM, Haller D, Lomax N, et al. Octreotide may act as a radioprotective agent in acromegaly. J Clin Endocrinol Metab 2000; 85:1287–1289. 56. Eastman RC, Gorden P, Glatstein E, et al. Radiation therapy of acromegaly. Endocrinol Metab Clin North Am 1992; 21:693–712. 57. Minniti G, Jaffrain-Rea ML, Osti M, et al. The long-term efﬁcacy of conventional radiotherapy in patients with GH-secreting pituitary adenomas. Clin Endocrinol (Oxf) 2005; 62:210–216. 58. Biermasz NR, van Dulken H Roelfsema F. Long-term follow-up results of postoperative radiotherapy in 36 patients with acromegaly. J Clin Endocrinol Metab 2000; 85:2476–2482. 59. Thalassinos NC, Tsagarakis S, Ioannides G, et al. Megavoltage pituitary irradiation lowers but seldom leads to safe GH levels in acromegaly: a long-term follow-up study. Eur J Endocrinol 1998; 138:160–163. 60. Milker-Zabel S, Zabel A, Huber P, et al. Stereotactic conformal radiotherapy in patients with growth hormone-secreting pituitary adenoma. Int J Radiat Oncol Biol Phys 2004; 59:1088–1096. 61. Williams M, van Seters AP, Hermans J, et al. Evaluation of the effects of radiotherapy on macroprolactinomas using the decline rate of serum prolactin levels as a dynamic parameter. Clin Oncol (R Coll Radiol) 1994; 6:102–109. 62. Ozgen T, Oruckaptan HH, Ozcan OE, et al. Prolactin secreting pituitary adenomas: analysis of 429 surgically treated patients, effect of adjuvant treatment modalities and review of the literature. Acta Neurochir (Wien) 1999; 141:1287–1294.

326

j.p.s. knisely and p.w. sperduto

63. Littley MD, Shalet SM, Reid H, et al. The effect of external pituitary irradiation on elevated serum prolactin levels in patients with pituitary macroadenomas. Q J Med 1991; 81:985–998. 64. Wallace EA, Holdaway IM. Treatment of macroprolactinomas at Auckland Hospital 1975–91. N Z Med J 1995; 108:50–52. 65. Tsagarakis S, Grossman A, Plowman PN, et al. Megavoltage pituitary irradiation in the management of prolactinomas: long-term follow-up. Clin Endocrinol (Oxf) 1991; 34:399–406. 66. Vicente A, Estrada J, de la Cuerda C, et al. Results of external pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing’s disease. Acta Endocrinol (Copenh) 1991; 125:470–474. 67. Plowman PN. Pituitary adenoma radiotherapy—when, who and how? Clin Endocrinol (Oxf) 1999; 51:265–271. 68. Becker G, Kocher M, Kortmann RD, et al. Radiation therapy in the multimodal treatment approach of pituitary adenoma. Strahlenther Onkol 2002; 178:173–186. 69. Mahmoud-Ahmed AS, Suh JH. Radiation therapy for Cushing’s disease: a review. Pituitary 2002; 5:175–180. 70. Melby JC. Therapy of Cushing [sic] disease: a consensus for pituitary microsurgery. Ann Int Med 1988; 109:445–446. 71. Mampalam TJ, Tyrrell JB, Wilson CB. Transsphenoidal microsurgery for Cushing [sic)] disease. A report of 216 cases. Ann Intern Med 1988; 109:487–493. 72. Howlett TA, Plowman PN, Wass JA, et al. Megavoltage pituitary irradiation in the management of Cushing’s disease and Nelson’s syndrome: long-term follow-up. Clin Endocrinol (Oxf) 1989; 31: 309–323. 73. Nagesser SK, van Seters AP, Kievit J, et al. Treatment of pituitarydependent Cushing’s syndrome: long-term results of unilateral adrenalectomy followed by external pituitary irradiation compared to transsphenoidal pituitary surgery. Clin Endocrinol (Oxf) 2000; 52:427–435. 74. Miller JW, Crapo L. The medical treatment of Cushing’s syndrome. Endocr Rev 1993; 14:443–458. 75. Estrada J, Boronat M, Mielgo M, et al. The long-term outcome of pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing’s disease. N Engl J Med 1997; 336:172–177. 76. Selch MT, Ahn E, Laskari A, Lee SP, et al. Stereotactic radiotherapy for treatment of cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys 2004; 59:101–111.

77. Dufour H, Muracciole X, Metellus P, et al. Long-term tumor control and functional outcome in patients with cavernous sinus meningiomas treated by radiotherapy with or without previous surgery: is there an alternative to aggressive tumor removal? Neurosurgery 2001; 48:285–294. 78. Milker-Zabel S, Zabel A, Schulz-Ertner D, et al. Fractionated stereotactic radiotherapy in patients with benign or atypical intracranial meningioma: long-term experience and prognostic factors. Int J Radiat Oncol Biol Phys 2005; 61:809–816. 79. Jalali R, Loughrey C, Baumert B, et al. High precision focused irradiation in the form of fractionated stereotactic conformal radiotherapy (SCRT) for benign meningiomas predominantly in the skull base location. Clin Oncol (R Coll Radiol) 2002; 14:103– 109. 80. Debus J, Wuendrich M, Pirzkall A, et al. High efﬁcacy of fractionated stereotactic radiotherapy of large base-of-skull meningiomas: long-term results. J Clin Oncol 2001; 19:3547–3553. 81. Uy NW, Woo SY, Teh BS, et al. Intensity-modulated radiation therapy (IMRT) for meningioma. Int J Radiat Oncol Biol Phys 2002; 53:1265–1270. 82. Hetelekidis S, Barnes PD, Tao ML, et al. 20-year experience in childhood craniopharyngioma. Int J Radiat Oncol Biol Phys 1993; 27(2):189–195. 83. Kalapurakal JA. Radiation therapy in the management of pediatric craniopharyngiomas-a review. Childs Nerv Syst 2005; 21(8–9): 808–816. 84. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. Management of cystic craniopharyngiomas with phosphorus-32 intracavitary irradiation. Neurosurgery 2004; 54:813–820. 85. Glod J, Koch B, Myseros J, Breneman J, et al. Issues concerning the treatment of a child with a craniopharyngioma. Med Pediatr Oncol 2002; 38:360–367. 86. Freeman CR, Patrocinio H, Farmer JP. Highly conformal radiotherapy for craniopharyngioma: (potentially) throwing the baby out with the bathwater. Med Pediatr Oncol 2003; 40:340– 341. 87. Selch MT, DeSalles AA, Wade M, et al. Initial clinical results of stereotactic radiotherapy for the treatment of craniopharyngiomas. Technol Cancer Res Treat 2002; 1:51–59.

2 9

Pituitary and Pituitary Region Tumors: Medical Therapy Perspective Mansur E. Shomali

Introduction The management of tumors in or around the pituitary region requires specialized attention to both issues of mass effect on the adjacent critical structures and to the diagnosis and management of endocrine dysfunction. Careful assessment of endocrine function is necessary before therapy for a number of reasons. First, the nature of the pituitary lesion needs to be determined. Tumors of pituitary origin have the potential to respond to medical therapy. In fact, medical therapy is preferred as a primary therapy over surgery and radiation for some tumor subtypes such as prolactinomas. Tumors not of pituitary origin do not respond to medical therapy and should be treated with the other modalities. Second, syndromes of pituitary hormone excess may occur and may be an important factor in the choice of therapy. Third, these patients need to be treated for the hypopituitarism that may occur as a result of the mass lesion or of the treatments thereof.

Diagnosis and Classiﬁcation Pituitary tumors may come to clinical attention in a number of ways. Small tumors (microadenomas; deﬁned as less than 1 cm) that are not hypersecreting a pituitary hormone may be discovered incidentally when an imaging study is performed for an unrelated reason. Microadenomas that hypersecrete a pituitary hormone present with syndromes of hormonal excess; for example, Cushing syndrome or acromegaly. In addition, large tumors, or macroadenomas, may present with headaches, mass effect on the optic chiasm, and dysfunction of the cranial nerves running through the cavernous sinus. The diagnosis of the mass lesion in the region of the pituitary may be aided by the characteristics of the imaging study and by biochemical testing. Often, the radiologic features can clearly distinguish a tumor of pituitary origin from that of another mass lesion. Meningiomas, hamartomas, and craniopharyngiomas have classic radiologic characteristics. Inﬂammatory and inﬁltrative processes such as sarcoidosis or lymphocytic hypophysitis also have typical appearances although it is not unusual for them to appear tumor-like. Mass lesions arising

from the sella turcica are most commonly of pituitary origin and are discussed in detail later. Pituitary neoplasms may arise from any pituitary cell type and thus have the potential to respond to medical therapy (Table 29-1).

Pituitary Tumor Subtypes Prolactinomas Diagnosis Prolactinomas are the most common type of functioning pituitary tumor. The diagnosis is usually made when patients present with signs and symptoms of hyperprolactinemia. In women, this would be amenorrhea and galactorrhea. Men present with hypogonadism. When the circulating prolactin concentration is more than 150 to 200 ng/mL, the diagnosis of a prolactinoma is usually clear [1]. At lower concentrations of hyperprolactinemia, the differential diagnosis includes medications that raise prolactin and the pituitary stalk effect. The latter point merits further discussion. Pituitary tumors that are not prolactinomas or non–pituitary tumors in the region of the pituitary may interrupt the access of hypothalamic dopamine to its receptors on the normal pituitary gland. Because prolactin is normally tonically inhibited by dopamine, the stalk effect results in hyperprolactinemia. Treating the patient with medications as if the tumor were a prolactinoma would decrease the circulating concentrations of prolactin but would not result in tumor shrinkage. Large pituitary tumors associated with hyperprolactinemia of less than 100 or 150 ng/mL are usually not prolactinomas but rather are raising the prolactin due to a stalk compression effect.

Management In large neurosurgical series, prolactinomas used to account for the greatest percentage of cases. Over the past 25 years, advances in the understanding of neuroendocrine physiology has led to the development of medications that have made surgery for prolactinomas generally unnecessary. Dopamine agonists such as bromocriptine and cabergoline effectively

327

328

m.e. shomali

TABLE 29-1. Pituitary tumor cell types and potential to respond to medical therapy. Name

Hormone secreted

Prolactinomas

Prolactin

Somatotroph adenomas Corticotroph adenomas Thyrotroph adenomas Gonadotroph adenomas

GH

*Medical therapy prolactinomas.

ACTH TSH LH, FSH, alpha-subunit is

considered

Clinical syndrome

Response to medical therapy

Hypogonadism, galactorrhea Gigantism, acromegaly Cushing syndrome Hyperthyroidism

Good

None

Poor

the

primary

concern here is that a small, incidental pituitary tumor, which is noted on an imaging study, is not the source of the ACTH hypersecretion, in which case pituitary surgery or radiation would not be effective at treating the hypercortisolism.

Excellent*

Management Good Poor

treatment

modality

for

lower prolactin and cause tumor shrinkage. These drugs may be used even in cases of signiﬁcant mass effect because tumor shrinkage occurs quite rapidly. Special cases where surgery or radiation may be considered are (1) patients who are intolerant of the medications or for whom the medications are not completely effective, (2) cystic prolactinomas, which may not shrink enough with medical therapy to eliminate the risk of mass effect, (3) aggressive prolactinomas or prolactin-producing pituitary carcinomas, which will need to be treated with multiple modalities, and (4) pituitary apoplexy with neurologic compromise due to hemorrhage within a prolactinoma. Given the cost and inconvenience of lifelong medications, stereotactic radiosurgery has been proposed as a primary therapeutic modality in these patients. Gamma Knife radiosurgery has been administered to patients with functional pituitary tumors as primary therapy with some success [2], but, at this point, the standard of care remains pharmacotherapy except for the reasons indicated above.

Corticotroph Adenomas Diagnosis In 1917, Harvey Cushing described a syndrome of cortisol excess associated with basophilic adenomas of the pituitary gland [3]. The clinical features of Cushing syndrome are usually striking but may be subtle and hard to distinguish from generalized obesity as the incidence of obesity increases in the general population. The diagnosis is made by documenting hypercortisolism usually by measuring free cortisol in multiple 24-hour urine collections. The differential diagnosis of hypercortisolism includes adrenal tumors, pituitary corticotroph adenomas, and ectopic secretion of corticotropin (ACTH). In the case of adrenal tumors that cause the Cushing syndrome, the circulating concentration of ACTH should be low. When pituitary tumors are identiﬁed by imaging studies in patients with hypercortisolism in the setting of normal or high ACTH, the diagnosis of a corticotroph adenoma may be made. If there is doubt, for example, a patient with a questionable pituitary lesion that may not be a corticotroph adenoma, excess ACTH production by the pituitary should be documented by selective sampling of the petrosal venous sinuses by interventional radiologists or neurovascular specialists. The

The goal of therapy for the patient with a corticotroph adenoma is the prompt elimination of hypercortisolism. Neurosurgery is the primary treatment for corticotroph adenomas. Conventional radiotherapy and stereotactic radiotherapy are not used as primary therapy due to the slow biochemical response, which puts the patients at risk for prolonged hypercortisolism [4]. A number of medical therapies that attempt to decrease the secretion of ACTH or cortisol are available [5]. Due to intolerable side effects and lack of efﬁcacy, these therapies are usually used as a last resort when surgery and radiation cannot eliminate the source of hypercortisolism. Bilateral adrenalectomy is a potential therapy that would eliminate the hypercortisolism when the source of ACTH overproduction cannot be identiﬁed or cannot be eliminated. Caution is advised with adrenalectomy because enlargement of the corticotroph adenoma may occur with cortisol disinhibition. This phenomenon is known as Nelson syndrome. Limited data suggest that stereotactic radiosurgery may control tumor growth after adrenalectomy [6].

Thyrotroph Adenomas Diagnosis Thyrotroph adenomas are uncommon, accounting for approximately 1% of pituitary tumors. The diagnosis is made by identifying patients with hyperthyroidism and inappropriately normal or elevated concentrations of thyrotropin (TSH). Most patients will also have elevations of the free alpha-subunit or an increase in the alpha-subunit to TSH ratio. Magnetic resonance imaging (MRI) can usually identify a tumor, as the majority are macroadenomas. The differential diagnosis of TSH-dependent hyperthyroidism is the rare syndrome of pituitary-speciﬁc thyroid hormone resistance. In this case, the pituitary gland does not detect appropriate concentrations of thyroid hormone and thus secretes excessive TSH. These patients do not have pituitary tumors and will have lower alpha-subunit to TSH ratios.

Management The goal of therapy is the correction of hyperthyroidism and shrinkage of the tumor to reduce mass effect. Most patients with thyrotroph adenomas will require therapy with multiple modalities. Surgery is indicated as the primary therapy. Radiation therapy is considered when there is signiﬁcant tumor residual and when hyperthyroidism is not controlled. Thyroidectomy is not a good option to control the hyperthyroidism because it may result in aggressive growth of the thyrotroph adenomas. In large series of patients with thyrotroph adenomas, complete tumor excision and normalization of thyrotropin levels occurs in about one third of patients [7], which suggests an important role for adjuvant therapies such as radiosurgery and pharmacotherapy. Surgery with adjuvant conventional radiation is able to control another one third of patients [7]. There are no data yet on the effectiveness of stereotactic radiosurgery on this rare pituitary tumor subtype. Most of the remaining

29.

329

pituitary and pituitary region tumors: medical therapy perspective

FT4, pmol/L

TSH, mU/L 120.0 80.0 60.0 40.0

20.0

10.0 8.0 6.0 4.0 2.0 1.0 0.8 0.6 0.4

0.2

80

40

30

15

0

* * *

*

**

FIGURE 29-1. Correction of hyperthyroidism in 33 patients with thyrotroph adenomas treated with octreotide. The upper panel shows TSH levels, and the lower panel shows free T4 levels. Solid circles refer to baseline measurements, and open circles represent values obtained after octreotide therapy. (From: Chanson P, Weintraub BD, Harris AG. Octreotide therapy for thyroid-stimulating hormone-secreting pituitary adenomas. A follow-up of 52 patients. Ann Intern Med 1993; 119:236–240. Used with permission.)

third of patients not controlled with surgery and adjuvant radiation will normalize their TSH levels when treated with the somatostatin analogue octreotide (Fig. 29-1) [8].

Somatotroph Adenomas Diagnosis Somatotroph adenomas are tumors that release excess growth hormone (GH), which results in gigantism in children and acromegaly in adults. These tumors may also cosecrete prolactin as both somatotroph and lactotroph cells are thought to be derived from a common stem cell. The diagnosis cannot be made based on a single measurement of GH but by means of serial measurements during a glucose tolerance test. Patients with GH excess usually have elevations of insulin-like growth factor 1 (IGF-1), the concentrations of which are used to monitor response to therapy.

Management As in the case of the other functional pituitary neoplasms, the goal of therapy of somatotroph adenomas is both tumor shrinkage and elimination of GH hypersecretion. Most somatotroph adenomas are macroadenomas, whose surgical cure rate is about 50%. The cure rate for microadenomas is higher at 50% to 90% [9]. Only about one third of these patients will have a signiﬁcant response to adjuvant radiation therapy, and there may not be a higher response rate with Gamma Knife radiation [9].

Although surgery and radiation have been shown to be useful, GH oversecretion as documented by IGF-1 concentrations is not normalized in as many as 50% to 70% of patients treated with both surgery and radiation [10, 11], which leaves an important role for medical therapy. Medical therapies for gonadotroph adenomas include dopamine agonists like cabergoline, somatostatin analogues like octreotide, and the new GH-receptor antagonist pegvisomant. Cabergoline may be given orally once or twice weekly. It is generally better tolerated than the other dopamine agonists such as bromocriptine. The response to this drug with regard to tumor shrinkage and lower IGF-1 concentrations is good only in a minority of patients, but given the relative ease of administration and the tolerability, a trial of cabergoline may be reasonable in patients with mild elevations of IGF-1. Somatostatin analogues are available in the form of regular octreotide, which is administered two to three times daily by means of subcutaneous injections, and long-acting octreotide, which is given intramuscularly once monthly. These drugs are more effective than cabergoline and may result in effective control of disease activity in two thirds of patients. Future somatostatin receptor subtype–speciﬁc analogues are being developed and may be more effective [12]. Patients who do not achieve normal IGF-1 concentrations despite surgery, radiation, and somatostatin analogues may respond to pegvisomant, the novel growth hormone receptor antagonist. This drug must be given by subcutaneous injection once daily. This drug improves clinical features of GH excess by blocking the effects of GH on the peripheral tissues. GH levels actually increase, and no tumor shrinkage occurs. In fact, tumors may increase in size, making this medication inappropriate monotherapy for patients with large tumors that have the potential for causing mass effect. A comparison of the medications used to treat acromegaly are summarized in Table 29-2 [13].

Gonadotroph and Clinically Nonfunctional Adenomas Diagnosis Tumors may originate from pituitary cell types without oversecreting any pituitary hormones. This may be due to altered gene expression or abnormal or inefﬁcient hormone production and/ or secretion by these neoplastic cells. Most of these clinically nonfunctional tumors are composed of neoplastic cells of gonadotropin lineage. In the absence of clinical syndromes, TABLE 29-2. Comparison of medications used to treat acromegaly.

Effectiveness Tumor shrinkage Administration

Usage

Cabergoline

Octreotide

Pegvisomant

30% Yes

65% Yes

100% No

Once or twice weekly (oral) Mild disease only

Once or twice monthly (intramuscular)

Daily (subcutaneous) When IGF-1 is not normalized with octreotide

Source: Adapted from van der Lely AJ, Lamberts SWJ. Medical therapy for acromegaly. In: Wass J, ed. Handbook of Acromegaly. Bristol: BioScientiﬁca Ltd, 2001:49–64. Used with permission.

330

m.e. shomali

these tumors are either discovered incidentally when imaging studies are done for other reasons or when the tumors grow large enough to cause mass effect symptoms. In the biochemical evaluation, circulating levels of intact luteinizing hormone (LH) and follicle-stimulating hormone (FSH) are uncommon. The free alpha-subunit component of these hormones may be measured, is often elevated, and may be used as a tumor marker. The deﬁnitive diagnosis of a gonadotroph adenoma is usually made after the tumor is resected by immunohistochemical identiﬁcation of LH, FSH, or the free alpha-subunit. Tumors with no immunostaining may express mRNA for the gonadotroph hormone subunits by Northern analysis or in situ hybridization, tests that are not performed in clinical practice but will attest to the gonadotroph origin of the tumors [14].

tion therapy modalities, as discussed in other sections of the text, are useful as adjuvant therapies, particularly when there is signiﬁcant tumor residual after surgery or if there is a high risk of tumor recurrence. Adjuvant medical therapies are useful in treating thyrotroph and somatotroph adenomas, with regard to tumor shrinkage and control of hormonal hypersecretion. Caution is advised when monitoring patients after radiation therapy because endocrine dysfunction may occur years after the treatment. Hormone replacement with glucocorticoids, thyroid hormone, sex steroids, and growth hormone is not only life-sustaining but also increases quality of life. Even fertility can be restored with appropriate hormone regimens in both men and women. Future advances in molecular medicine may lead to pharmacologic and radiologic “magic bullets” that may reduce or even eliminate the need for traditional neurosurgery.

Management Clinically nonfunctioning microadenomas may not need surgery as long as there is no endocrine dysfunction and as long as they are too small to cause mass effect. Surgery is the primary treatment modality for macroadenomas. Although hormonal hypersecretion is not a concern with the nonfunctioning tumors, hypopituitarism is. Careful endocrine evaluation should be done prior to surgery looking for endocrine dysfunction. After surgery, endocrine function should be monitored carefully for two reasons: (1) some deﬁcits may have normalized if the mass effect on the pituitary stalk is relieved; (2) additional deﬁcits may occur postoperatively due to intraoperative injury to the normal pituitary gland and/or pituitary stalk. As discussed in other sections of the text, patients with incomplete resection of the pituitary tumor or patients at high risk of recurrence may be candidates for adjuvant radiation. Patients who do receive pituitary radiation, either conventional or stereotactic, need long-term monitoring for endocrine dysfunction. A decline in endocrine function may occur months to years from the date of radiation. This may be due to slow destruction of tissue in the region of the hypothalamus or pituitary or perhaps due to a delayed radiation-induced apoptosis of normal endocrine cells. Future studies are required to assess if stereotactic techniques result in less posttreatment hypopituitarism than conventional radiation therapy [15]. Medical therapy for gonadotropin and nonfunctional pituitary tumors has been attempted in a number of small studies [14]. In principle, manipulation of dopamine, somatostatin, and gonadotropin-releasing hormone receptors, which are usually present on these tumors, is expected to have antiproliferative effects. Unfortunately, tumor shrinkage with medical therapy occurred infrequently in the majority of published reports [14]. Medical therapy should be reserved as a last resort when surgery and radiation are not effective or not possible.

Conclusion The effectiveness of multiple modalities for treating tumors in the region of the pituitary makes treating these disorders rewarding for the interdisciplinary team of physicians. Neurosurgery is the primary modality for the majority of these disorders. The only exception is the prolactinoma, where the effectiveness and good tolerability of dopamine agonist drugs make medical therapy the preferred ﬁrst-line treatment. Radia-

References 1. Balagura S, Frantz AG, Housepian EM, Carmel PW. The speciﬁcity of serum prolactin as a diagnostic indicator of pituitary adenoma. J Neurosurg 1979; 51:42–46. 2. Zhang N, Pan L, Dai J, et al. Gamma Knife radiosurgery as a primary surgical treatment for hypersecreting pituitary adenomas. Stereotact Funct Neurosurg 2000; 75:123–128. 3. Cushing H. The basophil adenomas of the pituitary body and their clinical manifestations (pituitary basophilism). Bull Johns Hopkins Hospital 1932; L: 137–195. 4. Hoybye C, Grenback E, Rahn T, et al. Adrenocorticotropic hormone-producing pituitary tumors: 12- to 22-year follow-up after treatment with stereotactic radiosurgery. Neurosurgery 2001; 49:284–292. 5. Sonino N, Boscaro M, Fallo F. Pharmacologic management of Cushing syndrome: new targets for therapy. Treat Endocrinol 2005; 4:87–94. 6. Pollock BE, Young WF. Stereotactic radiosurgery for patients with ACTH-producing pituitary adenomas after prior adrenalectomy. Int J Radiat Oncol Biol Phys 2002; 54:839–841. 7. Beck-Peccoz P, Brucker-Davis F, Persani L, et al. Thyrotropinsecreting pituitary tumors. Endocr Rev 1996; 7:610–638. 8. Chanson P, Weintraub BD, Harris AG. Octreotide therapy for thyroid-stimulating hormone-secreting pituitary adenomas: a follow-up of 52 patients. Ann Intern Med 1993; 119:236–240. 9. Melmed S, Vance M, Barkan A, et al. Current status and future opportunities for controlling acromegaly. Pituitary 2002; 5:185–196. 10. Gutt B, Wowra B, Alexandrov R, et al. Gamma-knife surgery is effective in normalising plasma insulin-like growth factor I in patients with acromegaly. Exp Clin Endocrinol Diabetes 2005; 113:219–224. 11. Attanasio R, Epaminonda P, Motti E, et al. Gamma-knife radiosurgery in acromegaly: a 4-year follow-up study. J Clin Endocrinol Metab 2003; 88:3105–3112. 12. Cap J, Marekova M, Cerman J, et al. Inhibition of hormone secretion in GH-secreting pituitary adenomas by receptor-subtype speciﬁc somatostatin analogues in vitro. Gen Physiol Biophys 2003; 22:201–212. 13. van der Lely AJ, Lamberts SWJ. Medical therapy for acromegaly. In: Wass J, ed. Handbook of Acromegaly. Bristol: BioScientiﬁca Ltd, 2001:49–64. 14. Shomali ME, Katznelson L. Medical therapy of gonadotropinproducing and nonfunctioning pituitary adenomas. Pituitary 2002; 5:89–98. 15. Vladyka V, Liscak R, Novotny J, et al. Radiation tolerance of functioning pituitary tissue in gamma knife surgery for pituitary adenomas. Neurosurgery 2003; 52:309–317.

3 0

Pediatric Radiosurgery Andrew Reisner, Nicholas J. Szerlip, and Lawrence S. Chin

Introduction Over the past ﬁve decades, extensive experience has been obtained with stereotactic radiosurgical procedures. This experience, and research, is primarily based on studies of the adult population. Increasingly, stereotactic radiosurgical treatments are being applied to treat a variety of neurologic disorders in children [1–3]. Children are especially sensitive to the potential side effects of brain irradiation [4]. Thus, the capability of neurosurgical technology to deliver high radiation doses to a deﬁned area and minimize radiation doses to surrounding brain tissue makes this treatment modality a very attractive option in the pediatric population. Currently, there are several available technologies, but new technologies are constantly being developed. The most widely studied are the Gamma Knife and linear accelerator (linac) systems. Each technology has inherent advantages and limitations. At the author’s respective institutions, the Gamma Knife system is the primary tool, thus forming the basis of this review. The literature of other systems, although far less extensive, is also included. The two most common indications for Gamma Knife surgery in children are to treat brain tumors and arteriovenous malformations (AVMs). In this chapter, an outline of contemporary management of pediatric brain tumors and AVMs is presented. In addition, stereotactic radiosurgery is discussed with regard to the unique aspects of this treatment as it pertains to children.

Gamma Knife Radiosurgery The goal of radiosurgery is to deliver high doses of radiation to a discrete target while minimizing radiation to surrounding tissue. The Gamma Knife system accomplishes this by directing multiple (201) beams of gamma rays to a single target. A large amount of energy is deposited at the target while minimal radiation is delivered to the tissue traversed by any single beam. The source of the ionizing radiation is cobalt-60. Cobalt-60 decays, emitting an electron and two gamma rays. The electron is absorbed by the source and has no therapeutic effect. The gamma rays are directed to the targets in beams by collimators

that focus the beams of radiation from the source to the target. The beams converge at a common focal point called the isocenter. The accuracy of the Gamma Knife system is such that the beams of energy converge within 0.3 mm of the target. The Gamma Knife is designed so that the patient’s head is placed at the center of the system, which coincides with the isocenter. The position of the head can be varied in all three planes (x, y, and z) by a fraction of a millimeter so that any speciﬁc part of the brain (the target) can be exactly positioned in the isocenter. The head is afﬁxed to a stereotactic head frame and the precise positioning of the head frame within the Gamma Knife is accomplished using stereotactic techniques. The radiation dose delivered by the Gamma Knife is adjusted in two ways: (a) the amount of energy deposited at the isocenter can be increased by exposing the target of the converging 201 beams of radiation for longer periods of time (the longer the exposure, the more the radiation), and (b) the size of the isocenter can be varied by changing the diameter of the beams, which is a function of the collimator. The Gamma Knife has four different collimators (4, 8, 14, and 18 mm) that adjust the radiation beam diameter. The radiation delivered can also be varied according to the shape of the target. The uniform delivery of the high doses of focused irradiation to an irregularly shaped target is accomplished with multiple shots of radiation. Perfect spherical targets require only one radiation exposure to the 201 converging beams of radiation. More complex shapes require more complicated treatment plans. Calculating the treatment plan is done using a high-speed computer and specially designed software. In general, more complex lesion geometry means a more complicated plan.

Unique Aspects of Radiosurgery in Children Although there are many similarities, stereotactic radiosurgery in children differs from that in the adult population in a number of important respects. First, the pathologies encountered are different (as highlighted in the section “Pediatric Brain Tumors” below). Thus, the treatment options including the indications for Gamma Knife vary [1–3]. Second, children have a higher sensitivity of brain tissue to ionizing radiation. This is especially true for children younger

331

332

a. reisner et al.

than 3 years of age [4]. Whole-brain irradiation (as opposed to radiosurgery) can have devastating consequences for the developing brain of a young child. Although these effects are limited by focal radiation as delivered by stereotactic radiation, these issues must be taken into consideration while formulating a treatment decision or plan. Third, in children less than 24 months old, the lack of skull thickness complicates the ﬁxation of a stereotactic localization frame to the skull [5]. It is the placement and maintenance of the localizing frame to the skull that is important to achieve and maintain accurate intracranial targeting (the targeting error with Gamma Knife stereotaxis is less than 1 mm). For obvious reasons, the thin skull of a child under 2 years of age is problematic. An inappropriately applied head frame pin may easily penetrate the skull. To avoid this complication, we have developed a special torque wrench that allows accurate measurement of torque force applied to a child’s skull during head frame placement, which should be no more than 2 to 4 inch-pounds of torque for younger children and slightly more for a child over 3. Another potential complication of a child’s thin skull is that it is deformable. This may result in a loss of precision that is unacceptable for radiosurgery treatment. Despite these issues, we have treated infants less than 2 years of age with Gamma Knife surgery without compromising precision or skull penetration. In younger children (2 to 10 years of age), precision is maintained by using special posts of the head frame that are tailored to the curve of a small child’s head. Often, longer pins are used because of their small head circumference, making placement of the frame more critical than in adolescents and adults. Because of the need to use long pins, special care must be used when moving the child during the procedure. Most adult patients have Gamma Knife surgery under local anesthesia with light sedation. Almost all children less than 12 to 14 years undergo radiosurgery using general anesthesia [6, 7]. This requires anesthesiologists who are comfortable with all aspects of pediatric anesthesiology as well as imaging patients under anesthesia (requires special magnetic resonance– compatible anesthesia equipment) and Gamma Knife surgery. Appropriate preoperative workup and teaching are conducted. Preoperative medications, including a mild oral sedation to calm the child, are given prior to intravenous line placement. Once general anesthesia is established, full cardiorespiratory monitoring is performed during the entire procedure. Often, the anesthetized child is moved multiple times during the treatment (with the head frame in place), from stretcher to angiography table to magnetic resonance imaging (MRI) unit to the Gamma Knife itself. The Gamma Knife treatment of a child entails a great deal of teamwork between multiple pediatric subspecialists who are experienced in all aspects of the treatment [8].

Pediatric Brain Tumors Brain tumors are the most common solid tumors occurring in childhood. Overall, malignancies of the central nervous system (CNS; brain and spinal cord) are second in frequency only to the hematologic tumors (leukemias and lymphomas) in children. CNS tumors represent 20% of all childhood malignancies. In contrast, brain tumors represent only 1% to 2% of all new

cancers in adults [9]. Approximately 1800 new cases of pediatric brain tumors are diagnosed per year in the United States, yielding an incidence of brain tumors of 2.5 per 100,000 children under the age of 15. Of concern is the fact that the incidence of brain tumors in children appears to be increasing. The reasons for this trend are currently the subject of intense research with investigation focusing on genetic factors, environmental factors, or both. It is postulated that certain individuals may have a genetic predisposition due to an alteration or susceptibility of the cellular DNA to tumor development when exposed to certain environmental factors.

Unique Features of Pediatric Brain Tumors Children’s Brain Tumors Are Different from Those of Adults Although there are many similarities, brain tumors in children differ from those in adults in several key respects. First, different types of brain tumors are found in children compared with adults. Generally, less aggressive tumors such as low-grade glial tumors, germ cell neoplasms, and craniopharyngiomas occur in childhood. Effective treatment of these tumors may result in a cure. For example, some low-grade astrocytomas may be cured by surgery alone, whereas germ cell tumors are sensitive to both chemotherapy and radiation therapy [10, 11]. In contrast, the majority of adult brain tumors are malignant at presentation. The two most common forms of adult brain tumors are highgrade glial tumors and metastatic tumors. Both tend to recur despite treatment. Second, children with brain tumors have an overall better prognosis than their adult counterparts. Whereas 60% of all children with brain tumors can expect to survive into adulthood, the 5-year survival rate of adults with primary brain tumors of all types is approximately 30%. This makes the longterm effects of treatment therapies important in the provider’s decision-making process. When radiation is used in the treatment of pediatric brain tumors, higher dose is often associated with improved tumor control. In children with medulloblastomas, posterior fossa doses >50 Gy are associated with fewer recurrences in this area [12]. Similar ﬁndings have been reported in children with ependymomas [13]. The use of nonconformal radiotherapy techniques makes radiation dose escalation difﬁcult due to the volume of normal brain that also receives a higher dose and the particular problem of neurocognitive deﬁcits and second malignancy seen in the pediatric population.

Children’s Brain Tumors Are Different from Other Childhood Tumors Brain tumors differ from other childhood malignancies in a number of important respects. First, brain tumors in children are a varied group of tumors. This is not surprising when one considers that the developing brain contains many different types of cells—from immature precursor to fully differentiated cells. Any one of these cells (it is estimated that there are 10 billion cells in the brain) may form a tumor. There are more than 40 brain tumor subtypes in children, each with a unique natural history (typical behavior), set of appropriate treatment

30.

pediatric radiosurgery

options, and prognosis. Other common tumors of childhood, such as lymphomas and leukemias (tumors of blood elements), sarcomas (soft tissue tumors), and nephromas (kidney tumors), tend to be more homogeneous (of similar cell types). Second, brain tumors tend to recur at the original site or within the central nervous system, and thus control at the primary site is most important. Other childhood tumors are more likely to metastasize (spread to other organs) so that treatment must be directed throughout the body. Third, brain tissue is essential for normal life, and its regenerative capacity is limited. Childhood tumors that occur in areas of the body that are not essential for life (e.g., a limb) and in tissue that regenerates (e.g., the liver) permit aggressive removal of the tumor including a tumor-free border to ensure that the tumor is completely removed. The critical and nonregenerative nature of brain tissue precludes removal of anything but tumor-bearing tissue. Wide-margin removal of brain tumors is therefore not feasible. Finally, the blood-brain barrier protects the brain and spinal cord. This barrier protects the brain and spine from potentially dangerous chemicals that may circulate in the blood. However, it also prevents certain therapeutic drugs from entering the brain, limiting the chemotherapies that can treat brain tumors.

The Pathologic Spectrum of Pediatric Brain Tumors Brain tumors in children are a heterogeneous group of tumors. The pathologic spectrum of brain tumors varies by age. In the extremes of childhood (young infants and older adolescents), cerebral tumors are more common than cerebellar tumors; however, overall, children are more apt to develop cerebellar tumors than are adults. Different areas of the brain have unique functions, which is largely due to the cellular makeup of that speciﬁc area. Thus tumors derived from a particular cell line are more likely to be found in certain areas of the brain. The common brain tumors encountered in childhood include medulloblastoma (25%), low-grade cerebral astrocytoma (23%), cerebellar astrocytoma (13%), high-grade astrocytoma (11%), brain-stem glioma (10%) and ependymoma (9%).

Benign and Malignant Pediatric Brain Tumors Confusion often occurs regarding the terminology benign and malignant when applied to brain tumors. In most tissues, other than the brain, a benign tumor is one that remains localized, grows slowly, and has a better prognosis; a malignant tumor invades adjacent tissue and/or metastasizes (spreads to distant tissue), responds poorly to treatment, and has an unfavorable prognosis. Most malignant non-CNS tumors are fatal because of metastases to the brain. Therefore, the terms benign (low grade) and malignant (high grade) are relative terms in the context of brain tumors and are used to describe relative degrees of aggressiveness, cellular differentiation, and prognosis. Unlike benign tumors in tissues other than the brain, low-grade brain tumors, despite a slow growth pattern, may be potentially lethal because of a critical location near vital brain structures. A benign or malignant brain tumor may “push” on other healthy areas of the brain causing problems with reduced eyesight, seizures, weakness in the arms or legs, or personality problems. It

333

may be these symptoms that guide pediatricians to rule out a brain tumor with proper scanning.

Initial Management and Overview of Treatment Options Most commonly, the brain tumor is suspected by a primary care physician who orders an imaging study, which conﬁrms the diagnosis. Referral is typically made to a pediatric neurosurgeon. As discussed, it is imperative that a diagnosis be established because pediatric brain tumors are a heterogeneous group of tumors, each of which has a different natural history, treatment protocol, and prognosis. Therefore, in almost all cases, the patient will undergo surgery with the dual goals of obtaining a tissue sample to establish a pathologic diagnosis and to debulk (remove) the tumor, if appropriate. Exceptions do occur. These would include a child with an obviously malignant, inﬁltrating brain-stem tumor or a child with a benign tumor of a visual pathway. There is no role for surgery in these cases, unless hydrocephalus is present. Hydrocephalus is treated with either a ventriculostomy or shunt placement. Once the pathologic diagnosis has been established, a multidisciplinary approach is employed with evaluation by a panel of subspecialists, including a pediatric neurosurgeon, neurooncologist, radiation therapist, endocrinologist, radiologist, ophthalmologist, dietitian, physiotherapist, psychologist, and social worker. The multidisciplinary approach is invaluable in that it not only offers convenience to patients by allowing them to be seen by numerous physicians at one time, but it also allows for vigorous debate among the various subspecialists as to what is the most appropriate treatment option. Further, this approach allows us to address all the needs of the patient and family, rather than only issues related to the tumor in isolation. All aspects of the child’s physical, emotional, and social well-being are addressed. We believe that this approach not only allows for more effective tumor treatment and monitoring but also more compassionate care to the patient and family [8]. Postoperatively, the multidisciplinary team evaluates the child’s neurologic status, pre- and postoperative radiographs, and tumor pathology. All treatment options, including simple tumor surveillance with serial scans, further surgery, chemotherapy and/or Gamma Knife surgery, are considered. The indications for postoperative treatment are largely based on the tumor pathology. Other factors include the patient’s age, the degree of resection (how much tumor was removed) and any associated condition (such as neuroﬁbromatosis) are also factored in.

Low-Grade Tumors Total resection of a low-grade tumor should be the surgical goal. However, tumors in certain locations within the brain make this goal unobtainable [2]. For example, tumors within the basal ganglia or brain stem are not resectable without signiﬁcant morbidity. These tumors may be excellent candidates for Gamma Knife surgery [14]. In some cases of benign brain tumors, an expectant policy of observation (“watch” with repeated scans only—no treatment) may be appropriate. These patients are usually followed with serial scans and subjected to

334

a. reisner et al.

surgery only if they become symptomatic or there is an increase in tumor size based on the radiographic studies. Increasingly, the approach to children with brain tumors is tempered by the natural history of the malignancy and the effectiveness and side effects of treatment options. For example, optic pathway tumors and hypothalamic tumors are known to be associated with a slow growth potential [15]. Current management of such tumors is expectant until they become symptomatic or there is evidence of growth on neuroimaging studies. In the case of optic pathway tumors, treatment is considered in the face of any visual loss or when growth is seen on imaging studies (usually MRI). For patients with low-grade astrocytomas, the current management is a debulking surgery (if possible) with no adjuvant treatment. Adjuvant chemotherapy or radiation therapy is reserved for recurrence (regrowth). If recurrence is localized and in an area of the brain that precludes a safe resection, Gamma Knife surgery is an option. An example of this is tectal plate gliomas [2]. The area where radiosurgery in young children needs to be tempered is in the region of the optic apparatus and brain stem. The radiosurgery tolerance doses of the optic chiasm and brain stem are of the order 8 to 9 Gy and 12 to 14 Gy, respectively [16]. Astrocytic tumors that grow directly within critical structures such as the brain stem or optic chiasm may not easily be excluded from the treatment volume during radiosurgery. However, some intra- and parasellar (in the area of the pituitary gland) tumors are amenable to Gamma Knife surgery if there is an appropriate distance between the tumor and the optic pathways. A “safe” clearance between lesion and optic pathway is >3 mm [16]. Closer proximity to the optic pathway is often the dose-limiting factor in this location. With pituitary tumors, results similar to those seen in adults are expected in children. Pituitary tumors that are not cured by surgery or medical management are potential candidates for Gamma Knife [17]. Lim and Leem treated two patients with residual optic nerve gliomas with stereotactic irradiation after subtotal surgical resection [18]. Both were diagnosed with ﬁbrillary astrocytomas. In their report, they used a marginal isodose that was at the 40% isodose line. Doses of 12 Gy and 14.4 Gy were given to a volume of 14.4 cm3 and 12.3 cm3, respectively. At 24+ month follow-up, both patients showed a marked decrease in tumor volume and improved visual symptoms. Craniopharyngiomas are common low-grade tumors of childhood that occur near the pituitary stalk (Fig. 30-1). They arise from epithelial remnants of the Rathke pouch. Craniopharyngiomas are slow growing and are typiﬁed by a relentless progressive course. All require treatment. Traditional treatments are primary total resection and subtotal resection followed by radiation therapy. The side effects of these two treatments vary. There is a higher incidence of endocrinopathies after total resection that often requires lifelong multiple hormonal replacement for panhypopituitarism. There is considerable debate among pediatric neurosurgeons and neurooncologists as to which is the best treatment [16, 19]. Although the form of treatment is controversial, all agree that treatment, be it surgery or radiotherapy, should be instituted. Our practice is to operate on these lesions initially with

FIGURE 30-1. A T1 enhanced axial MRI scan shows a recurrent craniopharyngioma in the suprasellar space.

the dual goals of establishing the diagnosis and, if possible, achieving a gross total resection. If a gross total resection is achieved, no further treatment is instituted, and we follow these patients with serial MRIs with monitoring of pituitary and visual functions. On occasion, a complete resection is not possible. Small residual craniopharyngiomas are observed with serial MRIs. If the residual lesion remains static and unassociated with clinical problems, we simply adopt a “wait and watch” approach. Enlarging residual or recurrent craniopharyngiomas, particularly intrasellar recurrences, are treated with stereotactic irradiation [16]. It is usually possible to treat the residual lesion without toxic doses to the optic nerves and chiasm. The minimum prescribed doses range from 9 to 20 Gy [20]. The results are encouraging. Eder et al. achieved tumor control in 6 of 7 patients with residual/recurrent craniopharyngiomas treated with stereotactic irradiation [2]. In that series, no patient had new endocrinopathy or optic neuropathy. Hypothalamic hamartomas are rare and curious benign congenital tumors located near the ﬂoor of the third ventricle. These tumors are associated with precocious puberty and intractable seizures, usually of the gelastic and tonic-clonic form [21, 22]. Stereotactic irradiation and surgical resection are both treatment options. The seizure disorder is often ameliorated after either treatment [21]. In our practice, we have had three patients with hypothalamic hamartomas, all of whose intractable seizures improved after treatment. Two were treated surgically and one with stereotactic irradiation of the lesion. Arita et al. reported a case of a patient with seizures secondary to a hypothalamic hamartoma that became seizure free [15]. The patient was treated with 18 Gy (50% isodose line) to the margin. A postirradiation MRI revealed complete disappearance of the lesion 1 year later. The antiepileptic effect of radiation is not known [21].

30.

pediatric radiosurgery

We have been encouraged with the results of Gamma Knife surgery in the treatment of recurrent ependymomas. Survival is improved with radiosurgery in cases of both residual and recurrent ependymomas [12, 23]. Children with neuroﬁbromatosis type 1 and especially neuroﬁbromatosis type 2 are excellent candidates for radiosurgery. Neuroﬁbromatosis type 2 is associated with bilateral acoustic neuromas and intracranial meningiomas (Fig. 30-2). The chance for deafness in bilateral acoustic neuromas is very high. Treatment with Gamma Knife at our institution has achieved 90% or greater tumor control rates, with better hearing and facial nerve preservation than with standard surgical resection [24]. Thus, young patients with neuroﬁbromatosis type 2 should be evaluated for Gamma Knife radiosurgery as a primary treatment at the time of diagnosis and when a new tumor develops. Meningiomas tend to occur along the skull base and, as in adults; the cavernous sinus is a frequent site. Surgical resection of cavernous sinus tumors is likely to impart signiﬁcant cranial nerve defects, and complete removal of the tumor is unusual [25]. Even with the most careful microsurgical technique, complications such as cranial nerve deﬁcit or vascular injury may occur. External beam fractionated radiation treatments, even with advanced planning programs, is likely to deliver a signiﬁcant dose to the temporal lobes with potential for memory and development of moya-moya phenomenon, a condition of slow constriction of skull base arteries. Many of these complications are obviated by employing stereotactic radiosurgical techniques [26].

FIGURE 30-2. A coronal T1 MRI scan shows bilateral vestibular schwannomas and part of a falx meningioma.

335

Similarly, patients with other hereditary neurocutaneous disorders should be evaluated for radiosurgery. Patients with tuberous sclerosis develop giant cell astrocytomas within the ventricular system. The pathology of these tumors is rarely in question, and a biopsy is not beneﬁcial. Radiosurgery is highly effective for this vascular benign tumor, and no invasive surgery may be required. Patients with von Hippel–Lindau disorder are prone to develop multiple highly vascular tumors called hemangioblastomas. Careful radiosurgery with Gamma Knife can control the majority of these tumors without microsurgery or the need for external beam radiation or embolization procedures [27]. But the natural history of this disease makes the multiple use of radiosurgery controversial and should be reserved only for hemangioblastomas that are deemed inoperable.

High-Grade Tumors High-grade (malignant) tumors such as glioblastomas, primitive neuroectodermal tumors, and ependymomas represent a daunting challenge and may necessitate multiple forms of treatments [2, 23]. Increasingly, untoward side effects have been documented in patients treated with aggressive radiation and chemotherapy protocols [28]. These side effects consist of leukoencephalopathy (changes in the brain’s white matter), dementia, learning disabilities, secondary induced tumors, and vascular and endocrine pathologies. The major complication rate may be as high as 50% in patients treated with aggressive protocols. Newer treatments are being sought to increase the sensitivity of the agents and circumvent some of the side effects. Given the potentially devastating complications associated with brain and spinal irradiation in young children, chemotherapy alone is being used with increasing frequency in children less than 3 years of age with malignant brain tumors. For high-grade tumors that do not respond to standard treatment, “novel” radiation approaches consisting of stereotactic radiosurgery, hyperfractionation, and brachytherapy (radioactive materials placed in direct contact with the tumor) are increasingly applied [2]. In children, primitive neuroectodermal tumors such as medulloblastoma, ependymoblastoma, and pineoblastoma are common. Although they have a propensity to disseminate within the spinal ﬂuid spaces, the main site of treatment failure is the primary tumor site. In a series of 12 children with malignant brain tumors, Eder et al. achieved local control using Gamma Knife surgery in six cases [2]. This series consisted of eight children with medulloblastomas and four with ependymoblastomas. The mean followup was 32 months (range, 8–66 months). Factors that appear to correspond with a favorable response of high-grade tumors to stereotactic irradiation are small tumor volumes, young age, and a lack of distant metastases [2]. Protocols using radiosurgery boosts to the primary site are under evaluation. For the present, radiosurgery has been reserved for the treatment of local recurrences in high-grade brain tumors. Generally, Gamma Knife surgery as the primary treatment for these tumors is not an option, but it has promise as an adjunct to surgery, multisession (fractionated) external beam radiation treatments, and chemotherapy in these patients [29] (Case Study 30-1).

336

a. reisner et al.

Case Study 30-1 A 17-month-old boy presented with unsteady gait and macrocephaly. The workup included a CT and a MRI scan that revealed a large left hemispheric cystic mass with mesial enhancement. He underwent a gross total resection, and the pathology returned as ependymoma. Eight months later, tumor recurrence was noted on routine follow-up MRI, and he underwent frameless stereotactic-guided gross total resection. He received postoperative chemotherapy (Baby POG1

[Pediatric Oncology Group protocol 1]: VCR [vincristine], cisplatin, Cytoxan, VP-16). Fourteen months later, he underwent his third craniotomy for gross total resection of recurrent tumor nodules followed by focal radiation therapy (5580 Gy via 3D conformal technique). Eight months later, multiple new nodules presented in the tumor bed that enlarged over serial studies and were treated with Gamma Knife radiosurgery to a dose of 8 Gy at the 50% line (Fig. 303). The tumors remain controlled 6 years after radiosurgery (Fig. 30-4).

FIGURE 30-3. The GammaPlan for a child with recurrent ependymomas.

Pediatric Arteriovenous Malformations

FIGURE 30-4. The enhanced axial MRI scan 6 years after radiosurgery showing control of ependymomas.

Cerebral vascular malformations include a wide variety of structural abnormalities of the brain and spinal cord. They are best described as a region of abnormal tangles of blood vessels. The widespread use of MRI over the past two decades has given us a better understanding of the epidemiology and natural history of vascular malformations of the brain. Distinct entities in this group include AVMs, cavernous malformations, and venous angiomas [37]. These lesions are usually of congenital origin and have a natural history typiﬁed by recurrent intracranial hemorrhages and/or seizures. Of the three types, AVMs have the most hazardous clinical course and prognosis. Typically, AVMs present in an abrupt fashion. Features include sudden onset of severe headaches, seizures, focal neurologic deﬁcits, and raised intracranial pressure (ICP). Approximately 20% of AVMs are symptomatic in childhood. On occasion, they are diagnosed incidentally before they become symptomatic (e.g., when a head CT or MRI is obtained after a

30.

pediatric radiosurgery

head injury). Mortality resulting from initial ruptures is approximately 10% and increases with each bleeding episode. Although the natural history is becoming more clearly deﬁned with modern imaging techniques, it is not completely known. However, it appears that the risk of rebleeding is 2% to 3% per year. Most studies suggest the risk of rebleed in children with AVMs is signiﬁcantly greater than in adults. Additionally, brainstem AVMs carry a high morbidity and mortality in children. Thus, there is little doubt that most AVMs should be treated. However, there is considerable controversy over the most appropriate treatment option. The major treatment options are embolization, surgical resection, and radiation. All have inherent advantages and disadvantages. In all cases, the treatment must be individualized for the patient. Typically, if the AVM lends itself to surgical resection (by size and location), we recommend surgery as the primary option [30]. Gamma Knife surgery, either alone or occasionally after interventional radiology (embolization) techniques, is a very attractive option in children with AVMs. The irradiated AVM undergoes a slow obliteration over approximately 1 to 5 years after the treatment [31]. This obliteration is the result of radiation-induced changes to the vessel wall, especially a thick proliferation of endothelial (lining) cells of the vessel wall. This process continues until the interior of the vessel narrows so much that there is no blood ﬂow though the AVM. The disadvantage of radiosurgery (compared with surgical excision) is that rebleeding may occur during this time, although at a reduced rate. Further, the child is exposed to radiation, which, although focused on a limited area of the brain, may have lifelong inherent risks [4]. The advantages of radiosurgery include a slow, steady adaptation of the cerebral circulation to the incremental reduction of blood ﬂow through the AVM. A rapid, abrupt decrease in ﬂow through the AVM (as may occur with surgical excision) can result in swelling of the surrounding brain—a phenomenon known as perfusion pressure breakthrough. Additional advantages of radiosurgery are the avoidance of conventional surgery and its attendant potential complications. Typically, radiosurgery allows a rapid return to most school and sports activities. Radiosurgery success is inversely related to the size of the AVM and directly proportional to the radiation dose. The usual time to resolution and occlusion in adults is 2 to 5 years for AVMs that are approximately 3 cm in average diameter. For lesions 3 cm or less, the rate of complete occlusion approaches 80% with less than 1% mortality and less than 3% morbidity. Tanaka et al. compared the results of AVMs treated with stereotactic radiosurgery in adult and pediatric populations and found that the obliteration rate was greater among children [32]. Complete nidus occlusion was obtained in 95% of pediatric cases at 2-year follow-up versus 81% in the adult group. The mean volume of the nidus was slightly larger among the children in that series (4.8 vs. 4.2 cm3). The mean marginal dose was 20 Gy in adults and 20.5 Gy in children. No complications or re-hemorrhages were noted in the pediatric group. For larger AVMs, combination treatment is an option. These lesions are treated with a combination of surgery, embolization, and/or Gamma Knife radiosurgery [33, 34]. Staged Gamma Knife surgery, spaced 3 to 6 months apart, is also an option for larger AVMs not suitable for surgery [2]. Regardless of the treatment, it is usually necessary to conﬁrm complete obliteration (cure) with angiography. In chil-

337

FIGURE 30-5. A T2 MRI shows a cavernous malformation in the pons.

dren, this is usually done 3 years after Gamma Knife surgery [35]. AVMs in the pediatric population tend to recur after treatment. There have been two cases in our practice where AVMs recurred after apparent complete surgical resection. Immediate postoperative cerebral angiograms in both cases conﬁrmed the apparent resection. Years later, the recurrent AVM manifested with a recurrent hemorrhage in one cases and was found incidentally in another. Rodriquez et al. describe a case of recurrence of an AVM after successful radiosurgery [36]. The authors suggest that children who undergo radiosurgery for treatment of AVM undergo angiography until early adulthood to exclude AVM recurrence. After obliteration has been conﬁrmed with angiography, we follow the children with MRI and MR angiograms every 2 years. On rare occasions, other forms of vascular malformations may require radiosurgery. Cavernous malformations are usually treated with surgical resection if located near the brain surface. If a cavernous malformation is located in an area of the brain that does not allow safe surgical resection (Fig. 30-5), Gamma Knife surgery should be considered, but this has yet to be systematically studied [37] (Case Study 30-2).

Case Study 30-2 An 8-year-old boy presented with complex partial seizures and was found by imaging studies to have a left temporoparietal AVM (Fig. 30-6). Because the lesion was in eloquent cortex, the patient underwent Gamma Knife radiosurgery to a dose of 16 Gy at the 50% line (Fig. 30-7). At 4 years after treatment, he remains seizure-free and the lesion has disappeared on MRI (Fig. 30-8).

338

a. reisner et al.

FIGURE 30-8. Follow-up MRI 4 years after treatment indicating obliteration of the AVM.

FIGURE 30-6. Flow voids on T1 MRI are seen in the left temporoparietal region indicating an AVM.

FIGURE 30-7. GammaPlan for treatment of an AVM.

30.

pediatric radiosurgery

Conclusion Radiosurgery is playing an increasing role in the management of children with neurosurgical disorders. The majority of treated pathology involves children with brain tumors and AVMs. With further experience, it is likely the indications for Gamma Knife surgery in the management of these disorders will continue to expand. For example, radiosurgery is now used to treat nasopharyngeal tumors (such as juvenile nasopharyngeal angioﬁbromas) and skull base tumors (osteosarcomas). Future uses in young children with brain tumors may include combination treatments with gene therapy, immunotherapy, and a variety of newly developed radiosensitizing agents. Future uses in functional disorders such as epilepsy, obsessive-compulsive disorder, and rare movement disorders in childhood are currently under investigation [38]. Radiosurgery is now an established part of the armamentarium used to treat children with brain tumors and vascular malformations in pediatric neurosurgery. The future of this neurosurgical technology remains exciting, especially in children where limiting radiation to normal tissue remains a critical issue. As with any operation on a child, gentleness, foresight, and extreme care are the key ingredients to a successful outcome.

References 1. Baumann GS, Wara WM, Larson DA, et al. Gamma knife radiosurgery in children. Pediatr Neurosurg 1996; 24:193–201. 2. Eder HG, Leber KA, Eustacchio S, Pendl G. The role of gamma knife radiosurgery in children. Childs Nerv Syst 2001; 17:341–346; discussion 347. 3. Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery in children and adolescents. Pediatr Neurosurg 1990; 16: 219–221. 4. Duffner PK, Cohen ME, Thomas PR, Lansky SB. The long-term effects of cranial irradiation on the central nervous system. Cancer 1985; 56:1841–1846. 5. Copley LA, Pepe MD, Tan V, et al. A comparison of various angles of halo pin insertion in an immature skull model. Spine 1999; 24:1777–1780. 6. Bauman GS, Brett CM, Ciricillo SF, et al. Anesthesia for pediatric stereotactic radiosurgery. Anesthesiology 1998; 89:255–257. 7. Newﬁeld P, Hamid RK. Pediatric neuroanesthesia. Arteriovenous malformations. Anesthesiol Clin North Am 2001; 19:229–235. 8. Messing-Junger AM, Janssen G, Pape H, et al. Interdisciplinary treatment in pediatric patients with malignant CNS tumors. Childs Nerv Syst 2000; 16:742–750. 9. Surawicz TS, McCarthy BJ, Kupelian V, et al. Descriptive epidemiology of primary brain and CNS tumors: results from the Central Brain Tumor Registry of the United States, 1990–1994. Neurooncology 1999; 1:14–25. 10. Kellie SJ, Wong CK, Pozza LD, et al. Activity of postoperative carboplatin, etoposide, and high-dose methotrexate in pediatric CNS embryonal tumors: results of a phase II study in newly diagnosed children. Med Pediatr Oncol 2002; 39:168–174. 11. Zissiadis Y, Dutton S, Kieran M, et al. Stereotactic radiotherapy for pediatric intracranial germ cell tumors. Int J Radiat Oncol Biol Phys 2001; 51:108–112. 12. Hodgson DC, Goumnerova LC, Loefﬂer JS, et al. Radiosurgery in the management of pediatric brain tumors. Int J Radiat Oncol Biol Phys 2001; 50:929–935.

339

13. Colombo F, Benedetti A, Casentini L, et al. Stereotactic radiosurgery of intracranial tumors in childhood. J Neurosurg Sci 1985; 29:233–237. 14. Freeman CR, Souhami L, Caron JL, et al. Stereotactic external beam irradiation in previously untreated brain tumors in children and adolescents. Med Pediatr Oncol 1994; 22:173–180. 15. Arita K, Ikawa F, Kurisu K, et al. The relationship between magnetic resonance imaging ﬁndings and clinical manifestations of hypothalamic hamartoma. J Neurosurg 1999; 91:212–220. 16. Chiou SM, Lunsford LD, Niranjan A, et al. Stereotactic radiosurgery of residual or recurrent craniopharyngioma, after surgery, with or without radiation therapy. Neurooncology 2001; 3:159– 166. 17. Hentschel SJ, McCutcheon IE. Stereotactic radiosurgery for Cushing disease. Neurosurg Focus 2004; 16:E5. 18. Lim YJ, Leem W. Two cases of Gamma Knife radiosurgery for low-grade optic chiasm glioma. Stereotact Funct Neurosurg 1996; 66(Suppl 1):174–183. 19. Chakrabarti I, Amar AP, Couldwell W, Weiss MH. Long-term neurological, visual, and endocrine outcomes following transnasal resection of craniopharyngioma. J Neurosurg 2005; 102:650–657. 20. Buatti JM, Meeks SL, Marcus RB Jr, Mendenhall NP. Radiotherapy for pediatric brain tumors. Semin Pediatr Neurol 1997; 4:304–319. 21. Arita K, Kurisu K, Iida K, et al. Subsidence of seizure induced by stereotactic radiation in a patient with hypothalamic hamartoma. Case report. J Neurosurg 1998; 89:645–648. 22. Unger F, Schrottner O, Haselsberger K, et al. Gamma knife radiosurgery for hypothalamic hamartomas in patients with medically intractable epilepsy and precocious puberty. Report of two cases. J Neurosurg 2000; 92:726–731. 23. Aggarwal R, Yeung D, Kumar P, et al. Efﬁcacy and feasibility of stereotactic radiosurgery in the primary management of unfavorable pediatric ependymoma. Radiother Oncol 1997; 43:269– 273. 24. Kondziolka D, Lunsford LD, McLaughlin MR, Flickinger JC. Long-term outcomes after radiosurgery for acoustic neuromas. N Engl J Med 1998; 339:1426–1433. 25. Larson JJ, van Loveren HR, Balko MG, Tew JM Jr. Evidence of meningioma inﬁltration into cranial nerves: clinical implications for cavernous sinus meningiomas. J Neurosurg 1995; 83:596–599. 26. Brell M, Villa S, Teixidor P, et al. Fractionated stereotactic radiotherapy in the treatment of exclusive cavernous sinus meningioma: functional outcome, local control, and tolerance. Surg Neurol 2006; 65:28–33; discussion 33–34. 27. Thompson TP, Lunsford LD, Flickinger JC. Radiosurgery for hemangiomas of the cavernous sinus and orbit: technical case report. Neurosurgery 2000; 47:778–783. 28. Duffner PK, Horowitz ME, Krischer JP, et al. Postoperative chemotherapy and delayed radiation in children less than three years of age with malignant brain tumors. N Engl J Med 1993; 328: 1725–1731. 29. Tsao MN, Mehta MP, Whelan TJ, 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. 30. Kurokawa T, Matsuzaki A, Hasuo K, et al. Cerebral arteriovenous malformations in children. Brain Dev 1985; 7:408–413. 31. Nicolato A, Lupidi F, Sandri MF, et al. Gamma Knife radiosurgery for cerebral arteriovenous malformations in children/ adolescents and adults. Part II: differences in obliteration rates, treatment-obliteration intervals, and prognostic factors. Int J Radiat Oncol Biol Phys 2006; 64:914–921. 32. Tanaka T, Kobayashi T, Kida Y, et al. [The comparison between adult and pediatric AVMs treated by gamma knife radiosurgery]. No Shinkei Geka 1995; 23:773–777.

340

a. reisner et al.

33. Nicolato A, Foroni R, Seghedoni A, et al. Leksell gamma knife radiosurgery for cerebral arteriovenous malformations in pediatric patients. Childs Nerv Syst 2005; 21:301–307; discussion 308. 34. Smyth MD, Sneed PK, Ciricillo SF, et al. Stereotactic radiosurgery for pediatric intracranial arteriovenous malformations: the University of California at San Francisco experience. J Neurosurg 2002; 97:48–55. 35. Tranchida JV, Mehall CJ, Slovis TL, Lis-Planells M. Imaging of arteriovenous malformation following stereotactic radiosurgery. Pediatr Radiol 1997; 27:299–304.

36. Rodriguez-Arias C, Martinez R, Rey G, Bravo G. Recurrence in a different location of a cerebral arteriovenous malformation in a child after radiosurgery. Childs Nerv Syst 2000; 16:363– 365. 37. Mottolese C, Hermier M, Stan H, et al. Central nervous system cavernomas in the pediatric age group. Neurosurg Rev 2001; 24:55–71; discussion 72–73. 38. Grabenbauer GG, Ernst-Stecken A, Ganslandt O, Stefan H. Gamma knife surgery in mesial temporal lobe epilepsy. Epilepsia 2005; 46:457; author reply 457–459.

3 1

Pediatric Brain Tumors: Conformal Radiation Therapy Perspective Thomas E. Merchant

Introduction Childhood cancer is the most common cause of disease related death among children ages 0 to 19 years. It has not been deﬁned as a public health problem because fewer than 12,500 cases are diagnosed each year. CNS tumors represent 21% of childhood cancers or about 2700 cases per year [1]. This rare group of tumors is composed of astrocytoma, 56% (n = 1510); embryonal tumors including medulloblastoma, supratentorial primitive neuroectodermal tumor, and pineoblastoma, 20% (n = 550); ependymoma, 7% (n = 190); craniopharyngioma, 3% (n = 100); germ cell tumors, 3% (n = 100); and tumors involving the spinal cord, pituitary, and choroid plexus, 11% (n = 300) (Fig. 31-1). As recent as 10 years ago, radiotherapy avoidance was the primary objective of clinical trials developed for CNS tumors in children. Most treatment morbidity was attributed to radiation therapy, and there was a major focus on cognitive effects and second malignancies. These concerns were justiﬁed based on the results of the time and the available treatment. After decades of pursuing radiotherapy avoidance, however, investigators found that side effects remained despite the omission of radiation therapy, and follow-up reports demonstrated that disease control was inferior without radiation therapy as a component of frontline management. Additionally, functional outcomes were relatively unchanged among long-term survivors: patients would pursue radiotherapy avoidance, experience side effects associated with those treatments, suffer progression of disease and eventually require further surgery, radiation therapy, or both. Based on published reports demonstrating major reductions in acute side effects associated with radiation therapy for adult cancers, three-dimensional radiation has been introduced incrementally to the pediatric population. For pediatric patients with CNS tumors, age has been the major driving force behind the inclusion or exclusion of radiation therapy and the sequencing of treatment. The age minimum of 3 years has often been used to determine whether a child should receive radiation therapy as part of their frontline treatment. The embryonal tumors and ependymoma serve as poignant examples because children with these tumors are generally young. For other

tumors, such as low-grade astrocytoma, the age at which radiation therapy has been considered acceptable has varied from 3 years to 10 years of age. For CNS germ cell tumors, craniopharyngioma, and high-grade glioma, age has been considered an important factor although most patients with these diagnoses tend to present at an older age. With the objective of reintroducing radiation therapy for the treatment of young children and improving outcomes for all patients, investigators have looked at ways in which the total prescribed dose might be reduced, the prescribed volume of radiation might be limited, or how newer methods of focal radiation delivery might be implemented to increase conformity of the high dose-volume. To achieve the aforementioned goals, clinical trials have been designed to test the safety of lower doses and reduced target volumes for a variety of tumors and to study radiation-related treatment effects focusing on objective measures including neurologic, endocrine, and cognitive function. Because newer methods of radiation therapy planning yield information about the three-dimensional distribution of dose in normal tissues, the long-term goal of many investigations has been to correlate treatment dosimetry with functional outcomes and to model the effects of radiation dose on normal tissues.

Contrasting Fractionated Irradiation and Radiosurgery Although this chapter is meant to focus on the applications of fractionated irradiation, the use of radiosurgery in pediatric CNS tumors requires comment. Radiosurgery is often proposed as an alternative to fractionated irradiation for primary pediatric CNS tumors based on the assumption that it would be less toxic. Most investigators would agree that indications exist for the use of radiosurgery in the frontline adjuvant treatment of craniopharyngioma and certain low-grade astrocytomas, supplemental boost treatment in conjunction with fractionated irradiation of pineal region tumors, and recurrent or metastatic disease. The usual caveats regarding tolerance of critical normal tissue structures apply, namely, tolerance of optic chiasm,

341

342

t.e. merchant

1800 1648 1600 1400 1200 1000 N 800 600

484

400 200

200

131

100

92

th

m

/O

io

ita tu Pi

op C

ra

ni

ry

ng ry ha

el C G

er

m

Ep

er

a

s or lT

m en

dy

br Em

um

om

na yo

m lio G

a

l

a

0

FIGURE 31-1. Estimated incidence of pediatric CNS tumors in the United States.

nerves, brain stem, and diencephalic structures to high-dose single-fraction irradiation. The true value of radiosurgery in the treatment of pediatric patients is unknown because of a lack of prospective experience and long-term outcomes. The potential beneﬁts of radiosurgery are often overshadowed by poor patient selection or when long-term therapeutic goals or the natural history of the tumor is not considered. For the treatment of pediatric CNS tumors, radiosurgery has not been proved to be equivalent to surgery or fractionated irradiation in terms of outcomes and side effects. Questionable applications of radiosurgery include the following: low dose (<12 Gy) single-fraction irradiation in lieu of a full-dose fractionated treatment course regardless of tumor size; irradiation of a tumor bed without an identiﬁable target; targeting macroscopic residual tumor within a tumor bed known to contain microscopic disease with no plan for fractionated irradiation of the tumor bed; primary radiosurgery in lieu of craniospinal irradiation for seeding tumors; radiosurgery in lieu of relatively simple surgery when feasible; radiosurgery to the solid component of a cystic/solid tumor complex; boost treatment with radiosurgery with subtherapeutic dose attenuation of the fractionated component. Finally, the pediatric caregiver requires advance notice that imaging changes are

prone to occur after radiosurgery that should not be interpreted as progressive disease. Practical considerations regarding fractionated irradiation and radiosurgery are given in Table 31-1.

Medulloblastoma Historically, the treatment for medulloblastoma has consisted of surgery, postoperative radiation therapy, and, during the past 15 years, systematic postirradiation chemotherapy for those older than the age of 3 years at the time of diagnosis. Surgery followed by chemotherapy, in an effort to delay or avoid radiation therapy, has been the path followed by most children under the age of 3 years. For decades, radiation therapy for older children consisted of craniospinal irradiation to 36 Gy followed by “boost” treatment of the anatomic posterior fossa to 54 Gy. These guidelines were developed in an era where adjuvant therapy consisted of radiation therapy alone, and they remained the standard until it was eventually demonstrated that the craniospinal dose could be reduced for low-risk patients (i.e., no evidence of neuraxis dissemination and residual tumor measuring <1.5 cm2) with the

31.

pediatric brain tumors: conformal radiation therapy perspective

343

TABLE 31-1. Practical comparison of fractionated irradiation and radiosurgery. Feature

Fractionated Irradiation

Radiosurgery

Safety

Proven track-record of safety in clinical trials and with long-term empirical data; late effects are well understood in terms of incidence and severity Disease control and functional outcomes documented for the wide variety of pediatric CNS tumors

Special considerations required to avoid devastating complications when used for newly diagnosed and recurrent tumors Lack of prospective clinical trials data limits understanding of efﬁcacy and results in underreporting of side effects Convenient treatment regardless of anesthesia usage and expense

Efﬁcacy

Convenience

Cost-effectiveness

Daily outpatient treatment is relatively inconvenient especially in the setting of a young patient who requires daily anesthesia More expensive overall considering costs associated with parental supervision and loss of employment or productivity

addition of multiagent chemotherapy [2]. Patients with neuraxis dissemination or substantial residual tumor continue to receive craniospinal irradiation to dose levels ≥36 Gy despite the addition of multiagent chemotherapy. Fear of craniospinal irradiation and its effects on cognitive function have always been and remain justiﬁable. The expectation for a child who receives craniospinal irradiation at an early age is a loss of IQ at a rate of approximately 3.9 points per year [3] without plateau over the time period of the reported data, about 6 years (Fig. 31-2). These data were derived from the St. Jude experience that included children initially treated postoperatively with chemotherapy that eventually required conventional-dose craniospinal irradiation. Craniospinal doses of 35.2 Gy and posterior fossa doses of 54 to 55.8 Gy were prescribed for these patients with a median age of approximately 3.2 years. Until this article by Walter et al. [3], the data in the available literature was not modeled, included few subjects, did not include acceptable follow-up, and suffered from signiﬁcant attrition. Working on the assumption that the craniospinal dose and treatment volume were

100 90

–2.4 pts/year

80

–3.9 pts/year

IQ Score

70 60

–4.3 pts/year

50 40 30

SJMB96 1996–2003 (Mulhern et al. J Clin Oncol 2005)

20

St Jude 1984–1995 (Walter et al. J Clin Oncol 1999)

10

CCG-9892 1990–1996 (Ris et al. J Clin Oncol 2001)

0 0

1

2

3 4 5 6 Time (Years) FIGURE 31-2. Linear models of IQ outcomes after craniospinal irradiation for medulloblastoma patients under the age of 7 years at the time of irradiation. Series include Walter et al. [3] (high-dose craniospinal irradiation), Ris et al. [4] (reduced-dose craniospinal irradiation), and Mulhern et al. [6] (reduced-dose craniospinal irradiation and reduced-volume primary site boost).

Overall costs should be similar to fractionated irradiation with advantage given to radiosurgery when anesthesia is required

largely responsible for the effects observed in these children, investigators sought ways to reduce the dose of craniospinal irradiation from 36 Gy to 23.4 Gy for selected average-risk patients using adjuvant chemotherapy. In the largest study reported for that time, and using longitudinal psychology data from the CCG-9892 trial, Ris et al. [4] showed a decline in IQ of 4.3 points per year among patients between the ages of 3 and 7 years treated with craniospinal irradiation to 23.4 Gy, posterior fossa irradiation to 55.8 Gy, and adjuvant chemotherapy. In other words, reducing the craniospinal dose to 23.4 Gy did not improve cognitive outcomes in young patients with average-risk medulloblastoma (Fig. 31-2). This lack of improvement should have come as no surprise given the volume of normal brain subtended by the treatment of the entire posterior fossa. It has been estimated, depending on the individual patient, that nearly 40% of the entire brain receives the prescription dose of 54 to 55.8 Gy when the posterior fossa is the target and conventional radiation therapy planning is used. The more recent focus for investigators has been the target volume for the “boost” component of treatment after craniospinal irradiation in an attempt to alter treatment from the entire posterior fossa to the primary site with a limited margin. Although there are a number of single-institution reports that appear to have accomplished this goal, the largest study of its kind was the SJMB96 trial conducted at St. Jude Children’s Research Hospital and collaborating centers at Texas Children’s Hospital, the Royal Children’s Hospital in Melbourne, and the New Children’s Hospital in Sydney from 1996 through 2003. This study included patients with a variety of embryonal tumors and presentations including average and high-risk medulloblastoma, supratentorial PNET, and atypical teratoid rhabdoid tumor. All patients received similar postirradiation high-dose chemotherapy with peripheral stem cell support, and all patients received postoperative craniospinal irradiation that was uniquely risk adapted. For example, patients classiﬁed as average risk (no evidence of metastases, residual tumor <1.5 cm2), regardless of embryonal tumor type or location, received 23.4 Gy craniospinal irradiation, 36 Gy posterior fossa irradiation (infratentorial tumors only), and primary site irradiation using an anatomically conﬁned 2-cm clinical target volume margin surrounding the postoperative tumor bed and

344

t.e. merchant

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Average-Risk MB 3Yr EFS 83% ± 3%

23.4 Gy

36.0 Gy Volume (cc)

Probability

residual tumor and 0.3- to 0.5-cm planning target volume margin. Patients with high-risk tumors received 36 to 39.6 Gy craniospinal irradiation and primary site irradiation using similarly deﬁned clinical and planning target volume margins. This small and stepwise reduction in primary site targeting did not affect the rate or patterns of failure for patients with average risk medulloblastoma. The 3-year event-free survival was reported at 83 ± 3% [5]. Although the gross tumor volume was signiﬁcantly smaller than the anatomic posterior fossa in these patients, with the additional 2 cm clinical target volume margin and 0.3- to 0.5-cm planning target volume margin, the ﬁnal target approached, and in some cases exceeded, the anatomic posterior fossa for these same patients. The volumetric details from this study provided some insight into the effect of the actual postoperative primary site volume, location, and targeting guidelines on our ability to further reduce the volume of irradiation after the craniospinal component (Fig. 31-3). Despite the small reduction in the volume of irradiation, the decline in IQ for patients with average-risk medulloblastoma, ages 3 to 7 years, was signiﬁcantly improved to approximately −2.4 points per year [6]. Indeed, the psychology data from this study forms the largest data set for patients treated with craniospinal irradiation for medulloblastoma and provides data from which parents and caregivers may estimate the effects of craniospinal irradiation (Fig. 31-2). In summary, eliminating posterior fossa irradiation and treating the primary site with a limited margin after craniospinal irradiation does improve cognitive outcomes for patients with medulloblastoma. The next step taken by the St. Jude consortium and separately by the Children’s Oncology Group is further dose and volume reduction. The current St. Jude trial known as SJMB03 includes identical risk-based craniospinal dosing and postirradiation chemotherapy as the SJMB96 trial with newer targeting guidelines for the primary site boost. The clinical target volume margin is an anatomically conﬁned margin of 1 cm surrounding the postoperative tumor bed and residual tumor. The planning target volume margins remain 0.3 to 0.5 cm. The

250 200 150 100 150 0 GTV CTV PTV

PF

55.8 Gy 0

1

2

3

4

5

6

7

8

Years On Study FIGURE 31-3. Event-free survival after reduced-dose craniospinal irradiation and reduced-volume primary site boost. Primary site target volumes (±SD) for medulloblastoma demonstrating the range of volumes for gross tumor volume (GTV), clinical target volume (CTV = GTV + 2 cm), planning target volume (PTV = CTV + 0.5 cm), and infratentorial volume.

Children’s Oncology Groups is conducting separate medulloblastoma trials for patients 3 to 8 years of age (ACNS 0331) and 8 to 21 years of age (ACNS 0332) [7]. The trial for the younger children involves an ambitious four-arm trial with two steps of randomization. Postoperatively, patients are randomized to 23.4 Gy craniospinal irradiation with concurrent vincristine versus 18 Gy craniospinal irradiation and boost treatment of the anatomic posterior fossa to 5.4 Gy (total 23.4 Gy posterior fossa) and concurrent vincristine. Patients are further randomized to primary site treatment that may include 55.8 Gy to the conventional posterior fossa volume or 55.8 Gy conformal primary site irradiation using a 1.5-cm clinical target volume margin. All patients receive postirradiation chemotherapy consisting of cisplatin, CCNU, cyclophosphamide, and vincristine every 6 weeks for a total of nine cycles. The chemotherapy regimen was designed based on the equivocal results from the A9961 intergroup study that compared platinum, CCNU, and vincristine to cyclophosphamide, cisplatin, and vincristine. The trial for the older patients includes 23.4 Gy craniospinal irradiation and concurrent vincristine followed by randomized to 55.8 Gy to the conventional posterior fossa volume or 55.8 Gy conformal treatment of the primary site using a 1.5-cm clinical target volume margin. The older patients will receive postirradiation chemotherapy identical to the younger patients. There are a number of important questions embedded in the Children’s Oncology Group study: (1) will patients and caregivers accept randomization to receive treatment to the entire anatomic posterior fossa; (2) will 18 Gy craniospinal irradiation prove to be as effective (similar disease control) as 23.4 Gy craniospinal irradiation; (3) are the volumetric differences between the different treatment arms sufﬁciently large to detect a difference in cognitive or audiometric outcomes? Figure 31-4 shows a graphical representation of the dose distribution for the total brain volume that would be expected for a typical patient planned and treated using conformal radiation therapy on the Children’s Oncology Group trial for the younger children. There are several regions of these curves that are notable moving from lowest to highest doses. The total brain is expected to receive the entire craniospinal dose as prescribed, 18 or 23.4 Gy. The dose to the total brain falls steeply and the curves separate according to those who receive conventional posterior fossa irradiation (parallel opposed portals in the example) or those who receive conformal treatment of the primary site—who tend to have an increase in the volume that receives the intermediate doses. The curves come together at the point where 30% of the total brain receives approximately 40 Gy, and for the remainder of the high-dose region, the curves diverge based on conventional targeting and treatment of the posterior fossa versus conformal treatment of the primary site. These curves demonstrate the volumetric effects of craniospinal dose reduction, the volumetric effects of conformal irradiation, and ﬁnally volumetric effects of the reduction in the targeted volume from the anatomic posterior fossa to the primary site with the prescribed clinical target volume margin of 1.5 cm. Based on the improvement shown for the SJMB96 trial over conventional radiation therapy, one would anticipate an improved outcome for those treated with 18 Gy craniospinal irradiation and perhaps those treated on the other arms.

COG Average-Risk MB Trial 18.0 Gy CSI + 5.4 Gy PF + CRT 18.0 Gy CSI + Conventional 23.4 Gy CSI + CRT 23.4 Gy CSI + Conventional 100 90 CSI Dose Reduction

80 % Volume

70 60 50 40 30

CRT Effect

20 10

Primary Site Volume Reduction

0 3000 4000 5000 6000 7000 Dose (cGy) FIGURE 31-4. Total brain cumulative dose-volume histograms based on targeting guidelines for the current Children’s Oncology Group medulloblastoma study, ages 3 to 8 years. 0

1000

345

pediatric brain tumors: conformal radiation therapy perspective

2000

Reintroducing radiation therapy to very young children under the age of 3 with embryonal tumors has meant omitting craniospinal irradiation from their treatment regimen. The obvious advantages of CSI elimination combined with conformal treatment of the posterior fossa and/or primary site are apparent in Figure 31-5. The current trial sponsored by the Children’s Oncology Group (A9934) [8] includes a 20-week course of induction chemotherapy after initial surgery, second surgery for patients with identiﬁable and resectable residual tumor, radiation therapy, and postirradiation chemotherapy for 8 months. Age-, risk-, and response-adapted conformal radiation includes treatment of the posterior fossa to 18 or 23.4 Gy (age adapted) and primary site irradiation to 50.4 or 54 Gy (age and response adapted). Children under the age of 24 months at the time of irradiation will receive 18 Gy to the anatomic posterior fossa followed by boost treatment of the primary site using a 1.5-cm clinical target volume margin. Patients in complete response receive 50.4 Gy; those with residual tumor will receive 54 Gy. Patients older than age 24 months at the time of irradiation will receive 23.4 Gy to the anatomic posterior fossa followed by primary site irradiation to 54 Gy using a 1.5-cm clinical target margin.

the tumor bed in addition to the residual tumor. In the ICRU50 report, the tumor bed is not meant as part of the gross tumor volume, and the clinical target volume margin is the margin surrounding the GTV in consideration of microscopic extension. Because margins are not assessed for brain tumors, all margins of the tumor bed are considered to be at risk and likely to harbor microscopic disease with extension into surrounding normal tissues. Until the day when neurosurgeons biopsy the operative cavity and the pathologists determine margins or the neuroradiologists have the means to identify subclinical microscopic disease, the margin of the tumor bed will most often be considered to be the basic volume at risk and deﬁned as the gross tumor volume. The governing principals of target volume deﬁnitions and beam’s-eye view treatment planning, whether forward or inverse, are similar in children and adults. After the target volume deﬁnition, the next step in the planning process is the deﬁnition of normal tissue volumes or so-called critical structures. These volumes take on additional meaning in the treatment of younger patients where there is an incentive to minimize dose to normal tissues and spare these patients from the late effects of radiation on the central nervous system. For children destined to receive central nervous system irradiation, treatment planners should have, at minimum, dose-volume data for the entire brain, supratentorial brain or temporal lobes, hypothalamic-pituitary unit, cochleae, optic chiasm, cervical spinal cord, and brain stem. These patients are at risk for radiation-related CNS effects including abnormalities in neurovascular function affecting hearing, vision, and gross neurologic function; cognitive effects involving learning, memory, attention, behavior, academic achievement, and global IQ; endocrine function affecting growth hormone secretion, thyroid hormone secretion, adrenocorticotropic hormone (ACTH) secretion, and the gonadotropins. Patients who receive hypothalamic

100 90 80 % Volume

31.

70 60 50 40 30 20 10 0

Primary Brain Tumors That Require Focal Irradiation The ICRU-50 Report Guidelines published in 1993 [9] and more recently the ICRU-62 Report Guidelines published in 1999 [10] have been taken to heart by those involved in the design and conduct of brain tumor trials for children. The deﬁnitions of the gross tumor volume (GTV), clinical target volume (CTV), and planning target volume (PTV) follow closely those of the ICRU with one notable exception. The GTV includes

0

1000

2000

3000 4000 Dose (cGy)

5000

6000

FIGURE 31-5. Total brain cumulative dose-volume histograms based on targeting guidelines for the current Children’s Oncology Group medulloblastoma study, ages 8 to 36 months.

346

t.e. merchant

irradiation are at risk for hypothalamic obesity. Patients treated with large-volume cranial, superﬁcial cranial, or craniospinal irradiation are at risk for abnormalities in growth and development including deformity of the calvarium, facial bones, and spine [11]. In consideration of radiation-related CNS effects, four important points should be understood: (1) preirradiation morbidity has not been systematically studied; (2) most reports describing radiation-related side effects are outdated; (3) most studies include small numbers of subjects and have large rates of attrition; (4) radiation has been treated as a categorical variable in most analyses. The governing hypothesis for current clinical trials involving children with CNS tumors that require only focal irradiation is that irradiation of smaller volumes will reduce radiation-related side effects without affecting tumor control. In support of this hypothesis, current studies include statistical monitoring of rate and patterns of failure and comparison of disease control rates to those achieved for patients treated using historic guidelines. In support of this hypothesis, most studies include longitudinal assessment to evaluate patients for CNS effects and include baseline assessments and some prospect of comparing results with historical controls although, as mentioned earlier, there are limited data for such a comparison. In a landmark study conducted at St. Jude Children’s Research Hospital from 1997 to 2003, patients with localized primary brain tumors were treated on a phase II study to estimate the local control and patterns of failure for pediatric patients with localized primary CNS tumors treated with threedimensional conformal and intensity-modulated radiation therapy. The study also sought to quantify radiation-related CNS effects looking at preirradiation morbidity and estimating the time to onset of CNS effects, speciﬁcally evaluating the effects of radiation dose and volume on longitudinal functional outcomes. This study, known as RT-1, included patients as young as 12 months of age and up to age 25 years, tumors that required only focal irradiation, biopsy-proven tumors except for optic pathway glioma, no evidence of dissemination (patients requiring craniospinal radiation were not eligible), no prior fractionated external beam irradiation history, and prior chemotherapy was not exclusionary. Patients were required to have a minimum performance status, and parents were required to sign a protocol consent indicating that they understood the risks of target-volume reduction including an increase in the rate of failure. Notably, patients with brain-stem glioma because of their poor prognosis and patients treated with craniospinal irradiation were excluded because the whole-brain component of this treatment precludes a high degree of normal tissue sparing. The RT-1 protocol had two treatment strata. The ﬁrst included a 1-cm clinical target-volume margin for ependymoma, WHO grade I and grade II astrocytoma, and craniopharyngioma. The second included a 2-cm clinical target-volume margin for grade III and IV astrocytoma. Planning targetvolume margins for all tumors were 0.3 to 0.5 cm based on the methods of immobilization of the time that included vacuumbag and thermoplastic technology, relocatable stereotactic head frames, and frameless stereotaxy using radiocamera systems. During the time period of the study, additional developments included the uniform use of intravenously adminis-

tered general anesthesia instead of conscious sedation, routine use of prone position general anesthesia, spiral computed tomography (CT), non-coplanar multiﬁeld conformal irradiation, routine magnetic resonance (MR) registration, linear accelerator treatment automation with independent gantry and couch movement, microleaf collimation (<1 cm), widely available inverse planning, and intensity-modulated radiation therapy. One of the key success factors in radiation planning has been the introduction of MR registration into the planning process: treatment standards should include multisequence MR registration. Dedicated MR systems for treatment planning have been implemented. Compared with prior treatment eras, dose prescriptions for patients with pediatric CNS tumors are now similar or identical to those used for adults. One example is ependymoma: postoperatively, 59.4Gy is prescribed with dose reductions to 54 Gy only for children under the age of 18 months who have undergone gross total resection. This dose reduction is not based on experimental data, rather concerns about late effects in very young children who have not previously been irradiated as a part of their frontline management. Grades I and II astrocytoma and craniopharyngioma are prescribed 54 Gy and grades III and IV astrocytoma 59.4 Gy. The data showing no difference in dose for adults with low-grade glioma [12] does not apply to children because pediatric patients are more likely to have juvenile pilocytic astrocytoma (WHO grade I) compared with adults who are more likely to have WHO grade II astrocytoma. One important aspect of dose prescription for localized brain tumor protocols in children is the tolerance doses for the spinal cord and optic chiasm. At our own center, we routinely prescribe 54 Gy to the upper cervical spinal cord, which is meant to include the upper 10 cm3 of spinal cord deﬁned as 30 × 2 mm CT slices multiplied by the area of the cross section of the spinal cord ranging from 3 to 5 cm2 (Fig. 31-6). No neurologic deﬁcits ascribed to the spinal cord have been identiﬁed in patients treated in this manner. Similarly, dose levels of 54 to 55.8 Gy to

FIGURE 31-6. Spinal cord outlined for treatment planning.

31.

pediatric brain tumors: conformal radiation therapy perspective

FIGURE 31-7. Cochlea outlined for treatment planning.

the optic chiasm have also been prescribed in patients who have tumors adjacent to or involving this critical normal tissue structure. Again, no delayed vision loss has been observed in this patient population. It is important to note that while we limit doses to the tolerance levels as described, regardless of the method of volume reduction or “cone down,” these structures will continue to receive additional dose. In our own practice, we have limited the cone-down dose objective to be less than 70% of the daily prescription dose. Although a point for more lengthy discussion, the variances in PTV or normal tissue (unspeciﬁed tissue) inhomogeneity, though often included in cooperative group or institutional guidelines, are unsupported by data. There are a number of critical aspects to planning (pretreatment), treatment (on treatment), and posttreatment imaging. At the present time, thin-slice CT (≤2 mm) is the fundamental data set. CT is used for radiation dose calculation as well as to deﬁne cochlea and spinal cord. As posterior fossa tumors are the most common site for CNS tumors in children, the cervical spinal cord is often subtended by the irradiated volume and needs to be precisely deﬁned. The cochlea, an adjacent and important structure, is easily deﬁned within the temporal bone and needs to be precisely deﬁned (Fig. 31-7). Because these structures are small and prone to the errors of registration (e.g., ﬂexion or extension errors for the spinal cord when registering diagnostic MR and treatment planning CT), it is our suggestion, while CT remains the fundamental data set, to deﬁne the cochleae within the temporal bone and the cervical spinal cord beginning below the foramen magnum on CT. Magnetic resonance imaging (MRI) is of course used to deﬁne target volumes and other normal tissue structures. Gadolinium-enhanced T1weighted MRI works well for most tumor systems to identify

347

postoperative tumor bed, enhancing residual tumor, and normal tissue structures. The postcontrast T1-weighted imaging sequence should be acquired three-dimensionally to improve spatial resolution and to minimize errors in registration. Certain types of low-grade astrocytoma, most cases of high-grade astrocytoma, and nearly all cases of brain-stem glioma are best seen on T2-weighted and ﬂuid-attenuated inversion recover (FLAIR) imaging. Unfortunately, even at 1.5 T, the T2-weighted and FLAIR sequences must be acquired two-dimensionally and have low resolution in order to achieve adequate signal. MRI at 3 T may overcome some of these problems and allow acquisition of thinly sliced, three-dimensionally acquired, T2-weighted and FLAIR imaging speciﬁcally for patients with low-grade astrocytoma. In developing targeting guidelines, the imaging must have resolution lower than the prescribed margins. Imaging during treatment is a concept that requires careful study and discussion. Imaging during treatment may be critical for patients with craniopharyngioma and cystic low-grade astrocytoma. These patients are prone to cyst expansion during radiation therapy and in the setting of highly focused small margin irradiation have the potential to expand outside of the prescribed isodose volume. On the St. Jude RT-1 protocol, imaging was performed at weeks 3 and 5 of irradiation therapy in order to acquire experimental data for quantitative measure of the MR T1 value in normal tissues. This maneuver proved to be critical for patients with craniopharyngioma, the majority of which required cyst aspiration during radiation therapy either stereotactically or through embedded catheter and reservoir systems. Despite aggressive aspiration in the statistical majority of patients, 15% required replanning. In addition, several cases of expansive low-grade astrocytoma were noted during treatment. Given the standard to perform MRI at some point in time prior to the initiation of radiation therapy, often 2 to 3 weeks, followed by 6 weeks of radiation therapy and the usual 6-week follow-up after radiation, nearly 4 months may elapse between preplanning MRI and posttreatment MRI. Posttreatment imaging on the RT-1 protocol includes scans every 3 months for 2 years, every 6 months for an additional 3 years, followed by yearly imaging. Patients with potentially seeding tumors such as ependymoma underwent MRI of the spine every 12 months. Over a 5.5-year period of time, from July 1997 to January 2003, 202 patients with localized CNS tumors were enrolled into the RT-1 protocol including 88 children with ependymoma, 49 children with low-grade astrocytoma, 35 patients with highgrade astrocytoma, and 28 patients with craniopharyngioma. There were two additional patients eligible for study, one with a CNS germ cell tumor and the other with a choroid plexus tumor. The rates and patterns of failure for all these patient groups have been previously reported or presented and fell within the expected range of failure noted for patients treated in the conventional radiation therapy era.

Ependymoma Focusing on children with ependymoma treated on the RT-1 protocol, and as previously reported, there were 88 children treated from July 1997 through January 2003. The age at the

348

t.e. merchant

time of irradiation was 2.85 years (median) with a standard deviation of 4.47 years. Seventy-four of the patients underwent gross total resection, six underwent near total resection, and eight subtotal resection. Thirty-two of the patients underwent second surgery prior to the initiation of radiation therapy. Sixteen of the patients had preirradiation chemotherapy, and 20 of the patients had supratentorial tumor location. As reported in the Journal of Clinical Oncology [13], the 3-year event-free survival was 75 ± 6% with a cumulative incidence of local failure of 14 ± 4%. The median time to failure was 16 months ranging from 6 to 26 months. These data compared favorably with the CCG9942 study [14], which has only been reported in abstract form. That study reported a 3-year event-free survival of 59 ± 6%. Through the RT-1 trial, the volume of irradiation for ependymoma was shown to be reducible to a 1-cm clinical target volume margin without effecting tumor control. Based on the RT-1 study, the Children’s Oncology Group initiated a study (ACNS0121) [15] in August 2003. The improved tumor control rate was attributed to the high rate of gross total resection, improved targeting through MR and CT registration, the relatively high total dose 59.4 Gy, and the relatively high tolerance dose for the cervical spinal cord (54 Gy). Patients on the RT-1 protocol submitted to a battery of psychology testing including measures of intelligence, academic achievement, attention, impulsivity and reaction time, memory, auditory learning, problem and adaptive behavior. Testing was performed at baseline and serially at 6, 12, 24, 36, 48, and 60 months. As previously reported [13], children under the age of 3 years had a statistically lower baseline IQ. Both young and older patients showed no change in IQ with time after irradiation. These data are an improvement compared with historic St. Jude data where much older children had a decline in IQ of approximately 2.4 points per year. Academic achievement, memory, verbal learning, and adaptive behavior have been stable in these same patients. Patients on the RT-1 study also underwent provocative endocrine testing to measure growth hormone secretion, thyroid hormone secretion, ACTH reserve, and gonadotropin secretion in appropriate patients. These tests were performed at baseline and 6, 12, 36, and 60 months after the initiation of irradiation. In the ﬁrst report of its kind, 46 of 68 initial patients were found to have preexisting hormone deﬁciencies including 16 of 32 patients with posterior fossa tumors [16]. Among these patients, speciﬁc deﬁciencies included growth hormone secretion abnormality in 38%, thyroid hormone secretion abnormality in 43%, ACTH secretion abnormality in 22% and GNRH abnormality in 18%. There was one patient with hyperprolactinemia. Among children with ependymoma, the baseline rate of growth hormone deﬁciency using the ATT/l-dopa test was 29%, thyroid hormone abnormality using the TSH surge 27% or TRH stimulation test 10%, ACTH secretion abnormality using the 1 μg ACTH test was 12%, and metyrapone test was 12%. The true value of past, present, and future threedimensional treatment trials will be the correlation of threedimensional radiation dosimetry with functional outcomes. Three-dimensional radiation planning has greatly improved our understanding of radiation dose distributions in tumor and normal tissues. The most practical way for those involved in treatment planning to view dose-volume data is with the twodimensional dose-volume histogram, which is a reduction of the three-dimensional dose-volume information based on the

assumption that all elements of the brain volume in question are functionally similar. In other words, the dose-volume histogram lacks spatial information. The cumulative dose-volume histogram graphs the percent volume of a normal tissue structure receiving a speciﬁed dose in a cumulative manner. Although this is helpful to the planner, the most important graph is the differential dose-volume histogram, which represents the percent volume of the normal tissue structure (y-axis) receiving the given dose on the x-axis. Integrating these data, one can determine the distribution or burden of radiation therapy to the normal tissue volume in question, and when using advanced modeling procedures, one may integrate dose over speciﬁed intervals and treat these data as clinical variables in a multivariate analysis. Most modeling procedures have used speciﬁc doses chosen along the dose-volume curves and entered these values into one of the many models meant to determine the probability of a speciﬁc effect.

Cognitive Effects In the ﬁrst study of its kind, DVH data for the total brain, supratentorial brain, and left and right temporal lobes were acquired for 86 of the 88 ependymoma patients included on the RT-1 protocol. These data were partitioned into three intervals representing the percent volume receiving less than 20 Gy, the percent volume receiving 20 to 40 Gy, and the percent volume receiving 40 to 60 Gy. The DVH data were combined with clinical data to form a linear aggression model estimating IQ as a function of time [17]. Initially, the correlation involved dosimetry and the IQ data for all patients. A total brain dose-volume estimating equation was produced that included in the intercept age and in the slope time and dose-volume. No other clinical variables were included in the equation. A similar equation was developed for the supratentorial brain, however no correlation could be developed for the temporal lobes or with any of the clinical covariants. It is important to note in the total brain volume dosimetry estimating equation that the low-dose coefﬁcient (v 0 to 25 Gy) was negative and small, the high-dose coefﬁcient (v 45 to 65 Gy) was negative and small, and the intermediate dose-volume term v 20 to 40 or v −25 to 45 was similar in weight to the high-dose coefﬁcient but had a positive effect. To increase the speciﬁcity of our model, we performed a correlation only with infratentorial patients and developed three equations; one for the total brain, another for the supratentorial brain, and one for the left temporal lobe. The dose-volume equations have similar coefﬁcients, negative low dose, positive intermediate dose, and negative high-dose terms. The low-dose terms were not statistically signiﬁcant, and in the total brain or left temporal lobe equations, age continued to affect the intercept and time affected the slope.

Endocrine Effects Similar preliminary results have been obtained from the study of dosimetry of the hypothalamic-pituitary unit [18]. Change in peak growth hormone levels after ATT/l-dopa have been studied in children from baseline through 12 months after irradiation. In one series, the decline in peak growth hormone

31.

pediatric brain tumors: conformal radiation therapy perspective

levels for 25 patients without deﬁcit prior to irradiation was shown to be exponential. A population-based estimating equation was developed for which all dose terms (low, intermediate, and high) were statistically signiﬁcant and negative. The effect of low doses was smallest and the effects of intermediate and high doses were similar. Similar changes in peak growth hormone determined after ATT/l-dopa were developed speciﬁcally for children with ependymoma. These data demonstrate that even the lowest doses are likely to result in growth hormone deﬁciency. What remains unclear is the incidence and time to onset of thyroid hormone, ACTH, and gonadotropin deﬁciency. These deﬁcits, if not acquired early as a result of tumor or surgery, appear years after fractionated irradiation even if the hypothalamic-pituitary axis receives the highest doses.

6.

7.

8.

Conclusion For childhood brain tumors, the study of focal irradiation is currently centered on volume reduction in an effort to limit normal tissue irradiation. The results of recent trials for localized brain tumors have made radiation therapy more acceptable for young children with brain tumors. Little is known about the relationship between radiation dose and volume; therefore, most studies recently completed or under way will provide baseline data for comparison with future trial results. In addition to disease control, critical to the successful dissemination of focal irradiation techniques will be the assessment of functional outcomes. There are a number of objective measures of CNS effects that may be studied in children, and because side effects are in part related to tumor location and therefore the spatial aspect of radiation dosimetry, the goal of research should be to increase the speciﬁcity of acquired data stratifying patients according to tumor type, tumor location, and other clinical factors. Implementing new treatment methods should be limited to clinical trials for children. They have the most to gain and lose by advances in treatment technology. Our mission should be stepwise improvements in disease control and functional outcomes using the power of clinical follow-up evaluations of neurologic, endocrine, and cognitive function to guide our protocols.

9.

10.

11.

12.

13.

14.

References 1. Ries LAG, Smith MA, Gurney JG, Linet M, Tamra T, Young JL, Bunin GR, eds. Cancer Incidence and Survival among Children and Adolescents: United States SEER Program 1975–1995. National Cancer Institute, SEER Program. NIH Publication No. 99–4649. Bethesda, MD: NIH, 1999. 2. Packer RJ, Goldwein J, Nicholson HS, et al. Treatment of children with medulloblastomas with reduced-dose craniospinal radiation therapy and adjuvant chemotherapy: a Children’s Cancer Group Study. J Clin Oncol 1999; 17:2127–2136. 3. Walter AW, Mulhern RK, Gajjar A, et al. Survival and neurodevelopmental outcome of young children with medulloblastoma at St. Jude Children’s Research Hospital. J Clin Oncol 1999; 17: 3720–3728. 4. Ris MD, Packer R, Goldwein J, et al. Intellectual outcome after reduced-dose radiation therapy plus adjuvant chemotherapy for medulloblastoma: a Children’s Cancer Group study. J Clin Oncol 2001; 19:3470–3476. 5. Merchant TE, Kun LE, Krasin MJ, et al. A multi-institution prospective trial of reduced-dose craniospinal irradiation (23.4 Gy)

15.

16.

17.

18.

349

followed by conformal posterior fossa (36 Gy) and primary site irradiation (55.8 Gy) and dose-intensive chemotherapy for average-risk medulloblastoma. Int J Radiat Oncol Biol Phys 2003; 57:S194–195. Mulhern RK, Palmer SL, Merchant TE, et al. Neurocognitive consequences of risk-adapted therapy for childhood medulloblastoma. J Clin Oncol 2005; 23:5511–5519. National Cancer Institute. Phase III randomized adjuvant study of standard-dose versus reduced-dose craniospinal radiotherapy and posterior fossa boost versus tumor bed boost radiotherapy in combination with chemotherapy comprising vincristine, cisplatin, lomustine, and cyclophosphamide in pediatric patients with newly diagnosed standard-risk medulloblastoma. Available at http:// www.cancer.gov/search/ViewClinicalTrials.aspx?cdrid=365506& protocolsearchid=1695050&version=healthprofessional. Accessed August 10, 2006. National Cancer Institute. Phase III study of cyclophosphamide, vincristine, cisplatin, and etoposide followed by second-look surgery and focal conformal radiotherapy in children with nonmetastatic medulloblastoma or posterior fossa primitive neuroectodermal tumor. Available at http://www.cancer.gov/search/View ClinicalTrials.aspx?cdrid=68269&version=HealthProfessional& protocolsearchid=1695054. Accessed August 10, 2006. ICRU Report 50. Dose speciﬁcation for reporting external beam therapy with photons and electrons. No. 9. Washington, DC: International Commission on Radiation Units and Measurements, 1993. ICRU Report 62. Dose speciﬁcation for reporting external beam therapy with photons and electrons. Washington, DC: International Commission on Radiation Units and Measurements, 1999. Leung W, Rose SR, Merchant TE. In Schwartz C, Hobbie W, Constine L, Ruccione K, eds. Survivors of Childhood Cancer: Assessment and Management, 2nd ed. Heidelberg: SpringerVerlag, 2005:51–80. Shaw EG, Tatter SB, Lesser GJ, et al. Current controversies in the radiotherapeutic management of adult low-grade glioma. Semin Oncol 2004; 5:653–658. Merchant TE, Mulhern RK, Krasin MJ, et al. Preliminary results from a phase II trial of conformal radiation therapy and the evaluation of radiation-related CNS effects for pediatric patients with localized ependymoma. J Clin Oncol 2004; 15:3156–3162. Garvin J, Sposto R, Stanley P, et al. Childhood ependymoma: improved survival for patients with incompletely resected tumors with the use of pre-irradiation chemotherapy. 11th International Symposium on Pediatric Neuro-Oncology, Boston, 2004 (abstract THER 29). Available at http://home.comcast.net/~turne038/ abstracts.htm. Accessed August 10, 2006. National Cancer Institute. Phase II study of conformal radiotherapy for pediatric patients with localized ependymoma, chemotherapy prior to second surgery in pediatric patients with incompletely resected ependymoma, and observation only in pediatric patients with completely resected differentiated, supratentorial ependymoma. Available at http://www.cancer.gov/search/ ViewClinicalTrials.aspx?cdrid=69086&protocolsearchid=1694995 &version=healthprofessional. Accessed August 10, 2006. Merchant TE, Williams T, Smith JM, et al. Pre-irradiation endocrinopathies in pediatric brain tumor patients determined by dynamic tests of endocrine function. Int J Radiat Oncol Biol Phys 2002; 54:45–50. Merchant TE, Kiehna EN, Li C, et al. Radiation dosimetry predicts IQ after conformal radiation therapy in pediatric patients with localized ependymoma. Int J Radiat Oncol Biol Phys 2005; 63:1546–1554. Merchant TE, Goloubeva O, Pritchard DL, et al. Radiation dosevolume effects on growth hormone secretion. Int J Radiat Oncol Biol Phys 2002; 52:1264–1270.

3 2

Pediatric Brain Tumors: Chemotherapy Perspective Amar Gajjar

Introduction Advances in the diagnosis and therapy of children diagnosed with brain tumors have led to an improved survival over that of the past decade [1]. Surgery, radiation therapy, and chemotherapy are the three treatment modalities that are used to treat brain tumors in the pediatric age group. With improved survival rate, there is an increased emphasis to use therapeutic modalities that reduce the neurocognitive deﬁcits that are seen in the long-term survivors [2]. Stereotactic radiosurgery (SRS) offers the advantage of treating residual tumor with higher doses of radiation therapy without exposing large areas of the developing brain to potentially damaging radiation therapy. SRS has been used in treating patients with ependymoma, low-grade glioma, medulloblastoma/primitive neuroectodermal tumor (PNET), and craniopharyngioma. SRS has been used for these various histologies at the time of either primary therapy or at the time of recurrence. This chapter will discuss the perspective of chemotherapy and the role of SRS in the therapy of pediatric brain tumors.

Ependymoma Ependymomas account for approximately 200 new cases of pediatric brain tumors annually in the United States [3]. The most common location of these tumors is the posterior fossa, with a smaller percentage of these tumors arising in the supratentorial hemispheres and in the spinal cord. Cure rates for ependymoma in published series range from 50% to 60% at 5 years [4–7]. Factors that affect outcome are the extent of surgical resection, tumor histology, and presence of metastatic disease at presentation [8]. Preliminary results form a trial in which the majority of the tumors were gross total resected and the primary tumor bed was treated with three-dimensional conformal radiation to a dose of 59.4 Gy reported a 3-year event free survival (EFS) at 75 ± 6% [9]. In infants diagnosed with ependymoma, chemotherapy is used ﬁrst in order to

postpone the use of radiation therapy [5, 10]. Even though ependymoma does respond to chemotherapy and there is a small subset of patients that can be cured with chemotherapy alone, the duration of tumor control is limited and most of the patients eventually need radiation therapy. Chemotherapy agents that are commonly used in infants are cisplatin, carboplatin, cyclophosphamide, vincristine, etoposide, and ifosphamide [8]. In older children, the role of chemotherapy is still being debated. Several series have failed to show a survival advantage when chemotherapy is used as adjuvant therapy after radiation therapy [11–15]. Current protocols use chemotherapy in patients who have residual tumors after surgical resection so as to facilitate a second attempt at surgical resection. Since almost all large series document that gross total resection is an essential prognostic factor that determines long term disease control, aggressive surgical resection is warranted in all newly diagnosed patients. Despite attempts at resecting all visible tumor, approximately 20% to 25% of the patients will have residual disease on postoperative scans. The role of SRS has been documented to enhance tumor control in these patients. In a series of four patients with ependymoma who were treated with SRS within a month of completing radiation therapy, all are free of progression at a median follow-up of 24 months (range, 15 to 40 months) [16]. The magnetic resonance (MR) scans show no residual tumor based on enhancement and mass effect. Each patient received a single-fraction SRS boost on a modiﬁed 6-MV linear accelerator. All doses were speciﬁed at the 80% isodose volume. The median dose delivered was 10 Gy (range, 9 to 15 Gy) [16]. Similarly, as Mansur et al. documented in a series of nine patients with ependymoma that were treated with SRS, four treated at the time of initial therapy had excellent disease control with no failures at a median follow-up of 28 months (range, 11 to 92 months). Of the ﬁve patients that received SRS as part of salvage therapy, three patients were dead of disease, two patients were alive with disease, and one had no evidence of disease. The median dose delivered was 16 Gy (range, 14 to 20 Gy) [17]. Hence, SRS does have an important role in ependymoma patients who have residual disease at the primary tumor site

351

352

a. gajjar

despite attempts at gross total resection after therapy with chemotherapy. The evidence clearly demonstrates that the use of SRS as part of initial therapy results in durable local disease control compared with when it is used as part of a salvage therapy.

SRT technique combines the precision of SRS with the radiobiologic advantages of fractionation. Though there is a paucity of published data for SRS in low-grade gliomas, it is conceivable that SRS could augment SRT in select patients with high-risk histologies (i.e., pilomyxoid astrocytoma).

Low-Grade Glioma

Medulloblastoma/Primitive Neuroectodermal Tumor

Low-grade gliomas encompass tumors of different histologies and are the most common tumors in the pediatric age group. The chronic nature of the disease often causes patients to suffer from severe morbidity, which can ultimately lead to their demise [18]. They occur either in the midline (hypothalamus, optic chiasm, tectal plate) where complete surgical resection of the tumor is not easily achieved without signiﬁcant morbidity, or they occur in the cerebral and cerebellar hemispheres where completed surgical resection is readily accomplished and no further therapy is indicated [19]. Indications for therapy in patients with chiasmatic/hypothalamic tumors are presence of progressive disease, visual decline, precocious puberty, diencephalic syndrome and other endocrine deﬁciencies. Chemotherapy is currently used as the therapeutic modality of choice for these patients as it signiﬁcantly delays the use of deﬁnitive radiation therapy. Several chemotherapy agents have demonstrated efﬁcacy against low-grade gliomas, and chemotherapy can achieve durable disease control in a majority of the patients. Carboplatin, vincristine, etoposide, and temozolomide either alone or in combination have documented response rates ranging from 55% to 85%. A combination of ﬁve drugs, namely 6-thioguanine, vincristine, CCNU, procarbazine, and dibromodulcitol, has also demonstrated efﬁcacy against these tumors. Unfortunately, despite high response rates, the duration of disease control varies, but most patients with residual disease need radiation therapy for durable disease control. Surgical resection of these tumors is often done with a view to reducing the tumor mass so as to reduce the ﬁeld of high-dose radiation therapy, particularly for chiasmatic/hypothalamic tumors [18]. Stereotactic radiotherapy (SRT) has been used effectively to achieve local control. Marcus et al. have published the largest series of low-grade gliomas treated with SRT [20]. In their series, 50 patients (26 male and 24 female), median age 9 years (range, 2 to 26 years) with low-grade gliomas were treated with SRT. Patients had either progressed after chemotherapy or surgery alone. SRT was delivered using a dedicated 6-MV linear accelerator. The target tumor volume included the preoperative tumor volume plus a 2-mm margin for the planning target volume. The median collimator size was 47.25 mm (range, 30 to 60 mm). Three to nine arcs were used to deliver a mean total dose of 52.2 Gy in 1.8-Gy daily fractions. With a median follow-up of 6.9 years (range, 0.9 to 10.2 years), the progressionfree survival rate was 82.5% at 5 years and 65% at 8 years. Five patients developed disseminated disease; two patients who had local progression had malignant transformation of their pathology to anaplastic astrocytoma; one patient developed a radiation-induced primitive neuroectodermal tumor. There were no marginal failures.

Embryonal tumors are the most common malignant brain tumors in the pediatric age group. When these tumors occur in the posterior fossa, they are called medulloblastoma. Therapy for medulloblastoma/PNET has evolved over the past three decades [21]. For patients with gross total resection and no metastatic disease at diagnosis, initial therapy consists of craniospinal irradiation (23.4 Gy) to the neuraxis with an irradiation boost to the primary tumor bed delivered by threedimensional conformal technique to a total dose of 55.8 Gy. Six weeks after completion of irradiation, patients usually get a year’s duration of chemotherapy (cisplatin, CCNU or cyclophosphamide, vincristine). Eighty percent to 85% of these patients are now long-term survivors. Patients with high-risk disease presentations (presence of metastatic disease or >1.5 cm2 of residual disease on the postoperative scan) are treated with high-dose (36 to 39.6 Gy) craniospinal irradiation along with either neoadjuvant or adjuvant chemotherapy. Survival of patients with high-risk disease is 40% to 60% at 5 years. Published experience with SRS in newly diagnosed medulloblastoma is limited. SRS was used in conjunction with upfront radiation in patients with residual disease at the primary site or in patients with discrete metastatic recurrent disease in conjunction with salvage chemotherapy [22–24]. It is conceivable in the future that as novel targeted therapies are developed against medulloblastoma, SRS may be incorporated in the therapy of newly diagnosed patients.

Craniopharyngioma Craniopharyngiomas are histologically benign tumors that most often arise in the sella or suprasellar region. The incidence of craniopharyngioma is highest among the pediatric age group: craniopharyngiomas represent 3.4% of all brain tumors in patients 0 to 14 years of age [3]. Craniopharyngioma is thought to arise from either embryonic remnants of an incompletely involuted craniopharyngeal duct or transformed cells from the lower infundibulum or anterior pituitary [25]. Due to the close proximity of this neoplasm with the pituitary, infundibulum, hypothalamus, optic chiasm, and vascular structures, diagnosis and treatment options pose unique challenges. Among the factors that inﬂuence decisions regarding therapy are the nature of the tumor (solid vs. cystic), presenting symptoms, endocrine function at diagnosis, potential side effects based on treatment, and the experience of the treating specialist involved.

32.

pediatric brain tumors: chemotherapy perspective

Therapy either involves surgical resection alone or biopsy followed by three-dimensional conformal radiation therapy. Both treatment approaches have produced durable disease control [26–28]. The morbidity of the disease is due to the location of the tumor to the structures mentioned previously. SRS has most commonly been used to treat patients who have experienced progressive disease after fractionated radiotherapy [29]. SRS as primary treatment has had mixed results for disease control and subsequent side effects. Long-term results (median follow-up period 16.8 years) for 21 patients who underwent SRS documented progression in 14 (67%) patients with a higher frequency in children [30]. Additionally, those patients treated with low marginal doses had a higher risk of failure compared with patients treated with higher doses. Eight (38%) patients reported deterioration of visual function. Among the factors that need to be considered when considering SRS as a therapeutic modality are constraints imposed by volume limitations [31], the potential of visual deterioration when the anterior optic pathway receives more than 8 Gy [32], and treatment that includes only the solid portion of the tumor. Despite the above restrictions, SRS is indicated in a select group of patients with craniopharyngioma where treatment options are limited and after careful consideration of the risk/beneﬁt ratio. SRS does have a role in the therapy of pediatric brain tumors. The indications for using this therapeutic modality are very speciﬁc; it is most efﬁcacious when used as an adjuvant to upfront radiation therapy. In a relapse setting, SRS in combination with chemotherapy offers an excellent choice for palliative care and relief of symptoms caused by locally progressive or metastatic disease.

10.

11.

12.

13.

14.

15.

16.

17.

18.

References 1. Packer R. Progress and challenges in childhood brain tumors. J Neurooncol 2005; 75:239–242. 2. Mulhern RK, Merchant TE, Gajjar A, et al. Late neurocognitive sequelae among survivors of pediatric brain tumors. Lancet Oncol 2004; 5(7):399–408. 3. CBTRUS US Statistical Report. Primary brain tumors in the United States 1995–1999. Central Brain Tumor Registry of the United States, Hinsdale, IL; 2002. 4. Horn B, Heideman R, Geyer R et al. A multi-institutional retrospective study of intracranial ependymoma in children: identiﬁcation of risk factors. J Pediatr Hematol Oncol 1999; 21:203–211. 5. Geyer JR, Zeltzer PM, Boyett JM, et al. Survival of infants with primitive neuroectodermal tumors or malignant ependymomas of the CNS with 8 drugs in 1 day: a report from the Children’s Cancer Group. J Clin Oncol 1994; 12:1607–1615. 6. Oya N, Shibamoto Y, Nagata Y, et al. Post operative radiotherapy for intracranial ependymoma: analysis of prognostic factors and patterns of failure. J Neurooncol 2002; 56:87–96. 7. Pollack IF, Gerszten RC, Martinez AJ, et al. Intracranial ependymomas of childhood: long term outcome and prognostic factors. Neurosurgery 1995; 37:655–666. 8. Merchant TE, Fouladi M. Ependymoma: new therapeutic approaches including radiation and chemotherapy. J Neurooncol 2005; 75:287–209. 9. Merchant TE, Mulhern RK, Krasin MJ, et al. Preliminary results from a phase II trial of conformal radiation therapy and evaluation

19.

20.

21.

22.

23.

24.

25.

26.

27.

353

of radiation-related CNS effects for pediatric patients with localized ependymoma. J Clin Oncol 2004; 15:3156–3162. Duffner PK, Horowitz ME, Krischer JP, et al. Postoperative chemotherapy and delayed radiation in children less than three years of age with malignant brain tumors N Engl J Med 1993; 328: 1725–1731. Robertson P, Zeltzer P, Boyett JM, et al. Survival and prognostic factors following radiation therapy and chemotherapy for ependymoma in children: a report of the Children’s Cancer Study Group. J Neurosurg 1998; 88:695–703. Timmerman B, Kortmann RD, Kuhl J, et al. Combined post operative irradiation and chemotherapy for anaplastic ependymoma: result of the German prospective trials HIT 88/89 and HIT 91. Int J Radiat Oncol Bio Phys 2000; 46:287–295. Evans AE, Anderson JR, Lefkowitz-Boudreaux IB, et al. Adjuvant chemotherapy of childhood posterior fossa ependymoma: craniospinal irradiation with or without adjuvant CCNU, vincristine and prednisone. A Children’s Cancer Group Study. Med Pediatr Oncol 1996; 27:8–14. Needle MN, Goldwein JW, Grass J, et al. Adjuvant chemotherapy for the treatment of ependymomas of childhood. Cancer 1997; 80:341–347. Massimino M, Gandola L, Giangaspero F, et al. AIEOP Pediatric Neuro Oncology Group: hyperfractionated radiotherapy and chemotherapy for childhood ependymomas: ﬁnal results of the ﬁrst prospective AIEOP study. Int J Radiat Oncol Biol Phys 2004; 58:1336–1345. Aggarwal R, Yeung D, Kumar P, et al. Efﬁcacy and feasibility of stereotactic radiosurgery in the primary management of unfavorable pediatric ependymoma. Radiother Oncol 1997; 43:269–273. Mansur DB, Drzymala RE, Rich KM, et al. The efﬁcacy of stereotactic radiosurgery in the management of intracranial ependymoma. J Neurooncol 2004; 66:187–190. Perilongo G. Considerations on the role of chemotherapy and modern radiotherapy in the treatment of childhood low grade gliomas. J Neurooncol 2005; 75:301–307. Gajjar A, Sanford RA, Heideman RL, et al. Low grade astrocytomas: a decade of experience at St. Jude Children’s Research Hospital. J Clin Oncol 1997; 15:2792–2799. Marcus KJ, Goumnerova L, Billet A, et al. Stereotactic radiotherapy for localized low grade gliomas in children: ﬁnal results of a prospective trial. Int J Radiat Oncol Biol Phys 2005; 61(2): 374–379. Ellison D, Clifford SC, Gajjar A, Gilbertson RJ. What’s new in neuro-oncology? Recent advances in medulloblastoma. Eur J Pediatr Neurol 2003; 7(2):53–66. Giller CA, Berger B, Pistenmaa DA, et al. Robotically guided radiosurgery in children. Pediatr Blood Cancer 2005; 45:304– 310. Hirth A, Pedersen PH, Baarsden R, et al. Gamma-Knife radiosurgery in pediatric cerebral and skull based tumors. Med Pediatr Oncol 2003; 40:90–103. Hodgson DC, Goumnerova L, Loefﬂer J, et al. Radiosurgery in the management of pediatric brain tumors. Int J Radiat Oncol Biol Phys 2001; 50(4):929–935. Zhang YQ, Wang CC, Ma ZY. Pediatric craniopharyngoimas: clinicomorphological study of 189 cases. Pediatr Neurosurg 2002; 36:80–84. Stripp DCH, Maity A, Janss AJ, et al. Surgery with or without radiation therapy in the management of craniopharyngiomas in children and young adults. Int J Radiat Oncol Biol Phys 2004; 58(3):714–720. Fitzer MM, Linggood RM, Adams J, et al. Combined proton and photon irradiation for craniopharyngioma: long-term results of the

354

a. gajjar

early cohort of patients treated at Harvard cyclotron laboratory and Massachusetts General Hospital. Int J Radiat Oncol Biol Phys 2006; 64(5):1348–1354. 28. Merchant TE, Kiehna EN, Kun LE, et al. Phase II trial of conformal radiation therapy for pediatric patients with craniopharyngioma and correlation of surgical factors and radiation dosimetry with change in cognitive function. J Neurosurg 2006; 104 (2 Suppl Pediatrics):94–102. 29. Kondziolka D, Nathoo N, Flickinger JC, et al. Long-term results after radiosurgery for benign intracranial tumors. Neurosurgery 2003; 53:815–821.

30. Ulfarsson E, Lindquist C, Roberts M, et al. Gamma knife radiosurgery for craniopharyngiomas: long term results in the ﬁrst Swedish patients. J Neurosurg 2002; 97(5 Suppl):613–622. 31. Chung WY, Pan DH, Shiau CY, et al. Gamma knife radiosurgery for craniopharyngioma. J Neurosurg 2000; 93(3 Suppl):47– 56. 32. Stafford SL, Pollock BE, Leavitt JA, et al. A study on the radiation tolerance of the optic nerves and chiasm after stereotactic radiosurgery. Int J Radation Oncol Biol Phys 2003; 55:1177– 1181.

3 3

Pineal Region Tumors Gregory P. Lekovic and Andrew G. Shetter

Introduction The pineal gland is situated at the geographic center of the brain, between the posterior and habenular commissures. Its central location, as well as its solitary nature in an organ composed of paired structures, led early anatomists and physiologists to impart to it singular importance (e.g., famously asserting that the pineal was the seat of the human soul). As cited by Baumgartner [1], the ﬁrst description of a lesion in the pineal region is attributed to Virchow in 1865. By the early 20th century, neurosurgical pioneers had devised approaches to the pineal region, variants of which are still in use today, including the transcallosal approach of Dandy and Fedor Krause’s supracerebellar infratentorial approach. These early experiences with surgery for pineal regions were highly morbid, a fact that is not surprising given the technological limitations of the time [2]. At the same time, the radiosensitivity of germinomas, the most common pineal region tumor, was beginning to be appreciated. Gradually, surgical approaches were given up in lieu of empiric treatment with radiotherapy. With the advent of the era of microneurosurgery, interest in the surgical management of pineal region tumors remerged, and empiric treatments with radiotherapy are now obsolete. Most recently, the development of chemotherapy and radiosurgery have added additional variables to the treatment algorithm of these tumors. Chemotherapy has been used successfully to lower radiation doses used to treat patients with germinoma. This is especially important in the treatment of children, for whom whole-brain irradiation carries signiﬁcant morbidity. Gamma Knife radiosurgery (GKRS) offers a similar advantage in providing highly conformal targeting with maximization of delivered dose to the therapeutic target and concomitant minimization of dose exposure to neighboring structures. Although the role of radiosurgery in the management of pineal region tumors is still evolving, it is safe to say that radiosurgery plays an important role in the multimodality management of these lesions.

Anatomy The pineal region is broadly deﬁned as the region lying between the splenium of the corpus callosum and tela choroidea dorsally, the quadrigeminal plate and tectum ventrally, and between

the posterior third ventricle rostrally and cerebellar vermis caudally. The pineal gland itself is situated between the habenula dorsally and posterior commissure anterior-inferiorly (Fig. 33-1). The pineal region is notable in that it is a region rich in vasculature, especially the deep cerebral veins. The vascular supply to the pineal gland proper is predominately from the medial posterior choroidal arteries, with contributions from the lateral posterior choroidal and quadrigeminal arteries as well [3]. However, the pineal is surrounded by the deep venous drainage of the brain on all sides: the velum interpositum, containing the internal cerebral veins, lies rostral and dorsal to the pineal gland (to which it is often attached to pineal region tumors by thick arachnoid adhesions) [4]; the basal veins of Rosenthal are located at either ﬂank; and the vein of Galen lies posterior and superior to it. For this reason, stereotactic biopsy of the pineal region is viewed with trepidation by some. The cellular constituents of the pineal region include pinealocytes, astrocytes, and sympathetic nerves. The sympathetic innervation comes from the superior cervical ganglion. The physiologic role of the pineal is thought to be concerned with the regulation of circadian rhythms. Ontologically and embryologically, the pineal is related to photoreceptor organs. Indeed, this role persists in certain reptiles where the pineal is known as the “parietal eye.” Because of this shared evolutionary background, pineoblastomas are sometimes seen in association with retinoblastoma (so-called trilateral retinoblastoma).

Pineal Region Tumor Presentation and Natural History Although the physiologic role of the pineal in humans is still somewhat obscure, it is an important locus of pathology. Broadly speaking, there are four categories of lesions occurring in the pineal region: germinoma (including pure germinoma, teratoma, nongerminomatous germ cell tumors), pineal parenchymal tumors (pineocytoma and pineoblastoma), glial tumors, and miscellaneous tumors (including metastases, meningioma, ependymoma, etc.). Although pineal region tumors are relatively rare (accounting for less than 1% of all intracranial neoplasms), they are much more common in Asia [5]. It is well-known that there is a higher incidence of germ cell tumors in Asian populations

355

356

g.p. lekovic and a.g. shetter

Pathology Germ Cell Tumors Pure germinoma, the most common central nervous system (CNS) germ cell tumor (GCT), consists of large, undifferentiated cells arranged in monomorphous sheets. Nuclei are round and typically centrally positioned amidst abundant clear cytoplasm. Mitoses are common and necrosis uncommon. Some tumors may incite an inﬂammatory response that is evident on histology by the presence of either many lymphocytes or occasionally ﬁbrous tissue. Immunohistochemical positivity for placental alkaline phosphatase is typical.

Nongerminomatous Germ Cell Tumors

FIGURE 33-1. Midsagittal section demonstrating the anatomy of the pineal region. Labeled structures include the pineal body (PB), the subforniceal organ (SFO), choroid plexus (CP), area postrema (AP), neurohypophysis (NH), median eminence (ME), and organum vasculorum of the lamina terminalis (OVLT). (Reproduced with permission from Barrow Neurological Institute.)

than in North American or European populations. Germ cell tumors typically account for the majority of pineal region tumors, with pineal parenchymal tumors second and miscellaneous tumors third. In Western populations, germ cell tumors account for 0.3% to 0.5% of all primary intracranial neoplasms but comprise approximately 3.0% of those encountered in children. In contrast, in Asia, germ cell tumors comprise at least 2.0%, and up to 9% to 15%, of all intracranial neoplasms and pediatric neoplasms, respectively. In Japanese and Korean populations, 80% of pineal region tumors in patients 15 to 35 years of age are germinoma [6]. In contrast, out of 370 French patients undergoing stereotactic biopsy or pineal region tumors, only 51% of patients under age of 30 had radiosensitive tumors (germinoma + pineoblastoma) [7]. The most common presentation of pineal region tumors is from symptoms attributable to local mass effect, principally increased intracranial pressure related to aqueductal compression and/or to mass effect on the quadrigeminal plate (e.g., Parinaud syndrome). The growth pattern of pineal region tumors tends to reﬂect the underlying histology, with benign, well-encapsulated tumors ﬁlling the posterior third ventricle, quadrigeminal cistern, and displacing the anterior cerebellum [4]. Malignant tumors are more likely to diffusely inﬁltrate the midbrain, thalamus, or other neighboring structures. Importantly, there is tremendous variation in the natural history of pineal region tumors depending on their histology. The management of tumors of this region, therefore, is extremely dependent on an accurate diagnosis. In most cases, this necessitates a tissue diagnosis, whether through stereotactic, endoscopic, or open biopsy. This issue will be discussed in greater detail below.

The group of tumors that together comprise the nongerminomatous germ cell tumors (NGGCTs) include teratoma (mature and immature), yolk sac tumors, embryonal carcinoma, and choriocarcinoma. Only teratoma is commonly encountered as a pure tumor type, with most NGGCTs having regional variations in histology suggestive of multiple tumor subtypes. By deﬁnition, teratomas recapitulate somatic elements from endodermal, mesodermal, and ectodermal lines. Differentiated teratomas include well-formed, fully differentiated elements such as rests of skin, hair, teeth, or cysts of mucosal epithelia. Immature teratoma are much more common and contain incompletely differentiated tissue. Occasionally, a carcinoma or sarcoma can arise from within a teratoma (so-called teratoma with malignant transformation, classically a sarcoma). Yolk sac tumors are characterized by a reticular epithelium set in a myxoid matrix; classically, these may form papillae known as Schiller-Duvall bodies. Embryonal carcinoma is composed of sheets of large cells with prominent nucleoli, many mitoses, and abundant cytoplasm. Necrosis is common. Trophoblastic elements, such as syncytiotrophoblastic cells (which may be seen scattered with other CNS GCTs) indicate tumor trophoblastic differentiation and hence choriocarcinoma. The immunohistochemistry of GCTs is helpful in differentiating tumor types. Classically, cell membranes and/or cytoplasm are positive for placental alkaline phosphatase germinoma; β-HCG (human chorionic gonadotropin) and human placental lactogen in choriocarcinoma; and α-fetoprotein in yolk sac tumor and teratoma.

Pineal Parenchymal Tumor Pineocytoma (also known as pinealocytoma) are slow-growing neoplasms of the pineal gland composed of small, uniform-sized cells. Characteristically, the cells exhibit cytoplasmic processes that are said to resemble club-like endings. Unlike pineoblastomas, there is some preservation of the lobular architecture of the normal pineal gland. Pineoblastomas are members of the family of small blue cell primitive neuroectodermal tumors, similar to medulloblastoma or retinoblastoma. Like these tumors, they are highly cellular neoplasms with a high nucleus to cytoplasm ratio, characterized by the presence of either Homer-Wright or FlexnerWintersteiner rosettes, the latter being a mark of retinoblastic differentiation. True to the oncologic commonality between

33.

pineal region tumors

the pineal and photoreceptor organs, pineal parenchymal tumors stain positively for interphotoreceptor retinoid binding protein. True intermediate tumors classiﬁcation is relatively rare, with only about 8% of pineal parenchymal tumors belonging to this category (WHO). These tumors are highly cellular with numerous mitoses and nuclear atypia but without pineocytomatous rosettes. Even more rarely, a pineal parenchymal tumor may have within it areas of both pineocytomatous and pineoblastomatous differentiation. Because many tumors of the pineal region are histologically heterogeneous (especially the NGGCTs), the ability of the surgeon to avoid sampling errors in obtaining tissue for histopathologic examination is paramount. In general, this need for accurate histologic diagnosis of disparate regions of tumor has been commonly cited as support for open or endoscopic biopsies, during which regions of tumor appearing grossly dissimilar can be biopsied separately.

Imaging Magnetic resonance imaging (MRI) with and without gadolinium contrast enhancement is the diagnostic imaging modality of choice. Pineocytomas are typically hypointense on T1weighted images and hyperintense on T2-weighted sequences. Pineoblastomas are heterogeneous and can be either hypo- to isointense on T1-weighted MRI. GCTs are typically well circumscribed, iso- to hyperdense relative to gray matter, and intensely enhancing. However, no certain method exists for differentiating pineal region tumors on the basis of imaging studies alone. Tumors of the pineal region enhance variably, as does the normal pineal gland due to the absence of a blood-brain barrier. Incidental calciﬁcation of the pineal is common, although such calciﬁcations are also associated with “brain sand” seen in pineocytomas. The diagnosis of GCT is suggested if the neoplasm appears to surround or engulf the normal pineal gland calciﬁcations, whereas pineal parenchymal tumors are said to cause an “explosion” of pineal calciﬁcations to the periphery of the lesion as the mass expands. On the other hand, hemorrhage (pineal apoplexy) is suggestive of choriocarcinoma. In general, tumors arising from the collicular plate are more likely to be of glial origin [1], and the presence of fat signal is characteristic of lipoma, mature teratoma, or dermoid tumors. The presence of mature teratomatous elements such as hair or teeth may naturally also be evident on imaging studies.

Management of Pineal Region Tumors Because of the varied natural history of pineal region tumors and the lack of diagnostic speciﬁcity of imaging studies, the importance of obtaining a tissue diagnosis cannot be underemphasized. Therefore, a biopsy, whether stereotactic, endoscopic, or open (see discussion below), should be obtained as the initial step in the management of tumors of this region in all cases, except where cerebrospinal ﬂuid (CSF) is positive for markers of malignant germinoma (i.e., β-HCG or α-fetoprotein), in the case of new metastatic lesions (where the tissue diagnosis is

357

already known), or perhaps in patients who are too frail to undergo even a biopsy. As is discussed in more detail below, there is no role for empiric radiation therapy in the management of these tumors, at least not in occidental populations. Whether it is better to perform an open biopsy, coincident with attempted gross total resection or cytoreductive surgery, or to perform a minimally invasive (i.e., stereotactic or endoscopic) biopsy is controversial. Proponents of open biopsy cite the relative safety and low morbidity of approaches to the pineal region given modern instrumentation and microsurgical technique. Because of the histopathologic heterogeneity common to tumors of this region, the ability to minimize sampling errors with open biopsy may yield more accurate specimens for diagnosis. Benign tumors may be completely resectable; therefore, an open approach offers the potential for cure of these lesions. Moreover, even malignant tumors and radiationsensitive tumors may beneﬁt from cytoreduction. On the other hand, minimally invasive approaches avoid the risks of craniotomy and have high diagnostic yield. Because of the pineal region’s proximity to the deep cerebral veins, some authors argue that stereotactic biopsy of this region is a relatively higher risk. However, Regis et al. [7] reviewed 370 stereotactic biopsies of the pineal region from 15 French neurosurgical centers. They reported a 1.3% mortality and 0.8% severe neurologic morbidity attributable to stereotactic biopsy. This was believed not to reﬂect an increased risk of death or neurologic injury over other intracranial sites. Overall, diagnostic tissue was obtained in 94% of patients. However, 2.3% of biopsied patients underwent repeat biopsy or subsequent open surgery and were found to have been initially misdiagnosed. Sampling errors were believed to contribute to about half of these cases (i.e., about 1% diagnostic error due to sampling errors). The endoscopic approach to the pineal region for the purposes of biopsy was ﬁrst described by Fukushima [8, 9]. The advantages of an endoscopic approach include visualization of biopsy site—this may lead to greater diagnostic accuracy as sampling errors may be minimized with visualization [10]. In addition, because most patients with pineal region tumors either present with, or are at high risk of, hydrocephalus, most require a CSF diversionary procedure. An endoscopic biopsy affords the possibility of simultaneously performing an endoscopic third ventriculostomy at the same operative sitting. Oi et al. reported their experience with 20 consecutive patients followed prospectively who underwent endoscopic biopsy in the initial management of their pineal region tumors [11]. Importantly, they reported the endoscopic detection of spread of tumor not visualized in preoperative neuroimaging—this suggests that endoscopy may have beneﬁt as a diagnostic tool per se. Nevertheless, the proper role of the endoscope in the management of pineal region lesions remains controversial.

The Role of Surgery Pineal region tumors can be successfully resected with microsurgical techniques at tolerable levels of morbidity and mortality. Approaches to the pineal region include both supratentorial (interhemispheric transcallosal and occipital transtentorial) and infratentorial (supracerebellar-infratentorial approach as seen in Fig. 33-2). Bruce and Stein [12] described their experience

358

g.p. lekovic and a.g. shetter

FIGURE 33-2. Surgical approaches to the pineal region. Midsagittal section of the human brain demonstrating the surgical corridors to the pineal region: the transcallosal, occipital-transtentorial, and the supracerebellar infratentorial. (Reproduced with permission from Barrow Neurological Institute.)

with 160 pineal region tumors. They achieved a gross total resection in 46/53 benign pineal region tumors, with overall mortality and morbidity of 4% and 3%, respectively. Their results are comparable with other major surgical series in which mortality rates range from 2% to 11% and serious morbidity from 3% to nearly 30% (for a recent review, see Bruce and Ogden [13]). The role of surgery for maximal debulking and/or resection is best established for benign lesions, for which gross total resection can afford a permanent cure. Importantly, these are also lesions that are arguably best suited to radiosurgical treatment. The argument for surgery for benign pineal region tumors is that gross total resection is usually possible, and that this affords both a permanent cure and may obviate the need for CSF diversion. In addition, benign pineal region tumors may be indolent and slow growing, as well as biologically relatively radiosensitive. Whether surgical resection or debulking plays a signiﬁcant role in the management of malignant tumors is much more hotly debated. Aggressive malignancies, especially those that on preoperative imaging are seen to invade the brain stem or other neighboring structures, are unlikely to be amenable to resection. Nevertheless, some studies have shown beneﬁt of cytoreduction even in these cases, although these data have not reached statistical signiﬁcance [14, 15]. Finally, many centers advocate for “second-look” surgery for treatment of persistent radiographic abnormalities after treatment with either chemotherapy or radiation. The majority of these surgeries ﬁnd either necrosis or, in the case of mixed NGGCTs, rests of mature teratoma.

Indeed, radiation remains the mainstay of treatment of most pineal region tumors, whether as primary treatment modality or as adjuvant therapy. Nevertheless, radiation has signiﬁcant morbidity. Efforts at reducing radiation exposure have included avoiding whole-brain and prophylactic craniospinal irradiation, chemotherapy, and radiosurgery. Traditionally, germinoma was treated with craniospinal irradiation, given estimates of spinal cord metastasis or seeding after surgery range of 13% to 40%. However, as the sequelae of prophylactic craniospinal irradiation have become more appreciated, efforts have been made to delay or avoid spinal irradiation and to limit cranial irradiation from whole brain to local ﬁeld. The effectiveness of partial brain ﬁelds comprising tumor plus a 2-cm margin was shown initially by Dattoli and Newall in 1990 [17]. In this retrospective study, 12 patients were treated with partial brain irradiation, consisting of either ﬁelds comprising tumor plus a 2-cm margin (n = 10) or ﬁelds comprising the ventricular system with a boost to the tumor (n = 2). Nine of 10 patients treated with partial brain irradiation were complete responders; however, one patient relapsed and ultimately succumbed to the disease. This patient, however, also received a reduced dose of less than 40 Gy. Thus, the authors concluded that whole neuraxis radiation could be safely avoided in the majority of patients, provided that sufﬁcient tumoricidal doses of radiation were delivered. Because gonadal germ cell tumors respond well to chemotherapy, chemotherapy has been advocated as a means to further reduce the amount of radiation necessary to obtain control of the tumor. In this regard, the rationale for chemotherapy anticipates the argument made for radiosurgery. Balmaceda et al. [18] enrolled 71 patients with germ cell tumors (including both pure and NGGCTs) in an international cooperative study comprising 31 institutions in 6 countries, demonstrating that chemotherapy could not be used alone for the treatment of germ cell tumors. They employed a high-dose regimen of carboplatin, etoposide, and bleomycin. Although 41 of the 71 patients were successfully managed with chemotherapy alone initially, 35 patients showed evidence of recurrence or progression. Furthermore, 7 of 71 patients treated died of chemotherapy-related toxicity. In contrast, Buckner et al. [19] reported the results of a phase II trial of primary chemotherapy followed by reduceddose radiation for the treatment of CNS germ cell tumor in which the partial brain dose was reduced to 30 Gy from 54 Gy. All patients were alive without progression at 51-month mean follow-up. One patient relapsed distally (spinal cord) and was salvaged with spinal irradiation. Thus, the prevailing paradigm at the present time is to use chemotherapy as an adjuvant therapy in combination with reduced-dose radiation. Spinal irradiation should be reserved for those cases presenting with disseminated disease or in the case of relapse.

The Role of Radiosurgery The Role of Radiation It has been known for decades that radiation is curative for intracranial germinoma [16], with 5-year survival rates of up to 95% reported for pure germinoma and up to 76% for NGGCTs.

The role of radiosurgery in the management of pineal region tumors remains controversial. Traditionally, pineal region tumors have been approached either surgically with the goal of gross total resection and cure (if benign) or with conventional radiation if the tumor is of a histology known to be

33.

pineal region tumors

Case Study 33-1 A 12-year-old boy presenting with progressive headache, nausea and vomiting for 5 days underwent computed tomography (CT) of the head revealing a pineal region mass with hydrocephalus at an outside institution and was subsequently transferred to our institution for deﬁnitive care. The patient had a prior history of strabismus but was otherwise previously healthy. On admission examination, the patient was awake, alert, and fully oriented. He complained of diplopia but his extraocular movements were intact to examination except for some divergence of gaze looking upwards. There were no other cranial nerve ﬁndings, and motor examination revealed 5/5 motor strength throughout without drift. Imaging revealed a pineal region tumor with hydrocephalus. Serum markers were positive for an elevated α-fetoprotein of 213, although CSF markers were absent. The patient underwent stereotactic wand-guided endoscopic biopsy and third ventriculostomy. However, endoscopic biopsy material was nondiagnostic, and so the patient subsequently underwent a suboccipital craniotomy for supracerebellar infratentorial approach for biopsy and tumor debulking 2 days later. Despite third ventriculostomy, the patient developed recurrent symptoms of hydrocephalus and eventually required ventriculoperitoneal shunting. Pathologic examination was consistent with a malignant mixed germ cell tumor. After surgery, the patient was begun on induction chemotherapy with a 3-day course of VePesid, carboplatin, cyto-

radiosensitive and/or malignant. Radiosurgery, on the other hand, has the potential to be used either as a “boost” modality in conjunction with conventional radiotherapy or even as an alternative to radiotherapy and/or surgery altogether (Case Study 33-1). To date, there are nine reports in the English language literature on the use of stereotactic radiosurgery in the treatment of pineal region tumors (Table 33-1). Several factors, however, make it difﬁcult to draw easy comparisons or conclusions from these studies, including small sample sizes (2 to 32 patients) and

359

taxin, and bleomycin followed by high-dose Neupogen and Procrit. He returned for subsequent cycles of chemotherapy every 3 weeks for a total of four cycles. Repeat imaging demonstrated persistent disease. The patient then underwent craniospinal irradiation with planned Gamma Knife boost to the pineal region. The brain was treated with shaped-opposed 6-MV photon beams with dose calculated at midplane. The brain was treated to 3600 cGy in 20 fractions. The spine was treated with two separate ﬁelds, one including the cervical and thoracic spines primarily, and the other comprising the lumbar region. Both ﬁelds were treated to a depth of 4.5 cm to 3600 cGy. The patient tolerated both chemotherapy and craniospinal irradiation well and was believed to be an excellent candidate for GKRS boost to the pineal. Two days after the completion of his course of craniospinal irradiation, the patient underwent GKRS. The prescription dose was 15 Gy prescribed at the 50% isodose line. The target volume was 1.2 cm3 and was covered in a single matrix with two 4-mm collimator isocenters and 6 8-mm isocenters. Follow-up imaging studies demonstrated the complete response of the tumor to treatment at 1 year of follow-up. Six years after GKRS, the patient’s MRI scan continues to show no evidence of tumor (Fig. 33-3). He is performing in high school at his grade level, and he has been off treatment for 5 years, although he continues to require pituitary replacement therapy.

because of the heterogeneity of diagnoses included within individual series. Backlund et al. [20] ﬁrst reported the treatment of two cases of pineocytoma with GKRS in 1974. The patients were treated with peak dose of 50 Gy; at 13 and 36 months follow-up, neither patient had displayed evidence of tumor progression. Subach et al. [23] included eight patients with pineocytoma (out of 14), with three tumors demonstrating a complete response to GKRS, three a partial response, and two no change. GKRS mean marginal doses for all tumor types treated was 15.4 Gy, and all

FIGURE 33-3. A 12-year-old boy presented with hydrocephalus secondary to pineal region NGGCT. (A) Axial and (B) sagittal gadoliniumenhanced MRI scans demonstrating complete response of pineal region NGGCT 6 years after GKRS boost to craniospinal irradiation.

5 OR 5 yes

4 NGGCT

5 pineocytoma

6 germinoma

Hasegawa et al. [26]*

Deshmukh et al. [27] Casentini et al. [28] Mean 27.6 (1–58)

14.6

25

52

NA

Mean imaging follow-up 21

12.3 (2–34)

20.7 (3–32)

13 and 36

Mean follow-up (months)

“Inverse boost” paradigm with extended ﬁled radiotherapy: mean 31 Gy, range 24–36

5 OR

2 OR

4 OR 4 EBRT (2 as boost and 2 after recurrence)

Germinoma: 8 chemo, 5 OR Malignant GCT: 10 OR, 10 chemo, 8 EBRT

1 55 Gy EBRT 1 subtotal OR 2 chemo

1 OR

4 primary 5 OR 4 with previous WBRT 1 chemotherapy

None

Prior treatment

4/5 partial response, 100% local control Complete response

4 complete response, 8 partial, 2 no change, 2 no imaging follow-up, local control = 100% 2 partial, 1 no change, 1 progression

3 complete responses (1 pinealoblastoma, 2 germinoma), 5 partial, 2 no change, 1 insufﬁcient follow-up 4 compete responses (3 pineocytoma and 1 pineoblastoma), 6 partial (3 pineocytoma, 1 pineoblastoma, and 2 germinoma) 8 complete response overall response rate = 73.3% For germinoma and pineocytoma, 100% control rate

6 partial or complete responses, 3 no change (all meningioma)

Regression of tumor

Radiologic results

GKRS mean marginal dose of 16.6 Gy (10–20 Gy) at 50%–60% isodose line GKRS mean marginal dose = 9.3 Gy (6–20), 40%–50% isodose line GKRS marginal doses 15.4 Gy (12–20 Gy), all treated to 50% isodose line, mean volume 6.4 cm3 GKRS marginal dose = 16.8 Gy for germinoma, 13.4 Gy for malignant GCT, 17.5 Gy for PPT and others GKRS mean marginal dose = 15.3 Gy at 50% isodose line 12–16 Gy marginal dose to 50% isodose line GKRS marginal dose 14–16 Gy Varian linac (marginal dose 10–11 Gy, peak doses 10–12.5 Gy)

GKRS 50 Gy

Dosimetry

None

None

1 (death)

4 distant treatment failure (1 pineocytoma, 3 pineoblastomas)

8

1 (embryonal carcinoma)

None

None

None

Progression

1 periprocedural death (unrelated)

None

None

2 new deﬁcits 5 deaths (4 attributable to progression)

7 deaths attributable to progression

1 new Parinaud syndrome, 3 deaths (only 1 attributed to progression)

None

3 new neurologic deﬁcits

None

Complication

Source: Reproduced with permission from Barrow Neurological Institute.

*Patients partially included in previous reports.

GKRS, Gamma Knife radiosurgery; NR, not reported; NGGCT, nongerminomatous germ cell tumor; SB, stereotactic biopsy; OB, open biopsy (craniotomy); STGC, germinoma with syncytiotrophoblastic giant cells; OR, surgery; EBRT, external beam radiation therapy; WBRT, whole-brain radiation therapy; PPT, pineal parenchymal tumor; bx, biopsy; NO, no; NA, not applicable.

1 SB 2 OB 1 CSF

10 SB 4 OB

10 pineocytoma 2 mixed histology 4 pineoblastomas

Hasegawa et al. [25]*

11 NO

8 germinoma 4 STGC 13 malignant GCT 3 pineocytoma 2 pineoblastoma 2 unknown

9 SB 2 OB 3 NO

Kobayashi et al. [24]

Subach et al. [23]

7 SB 4 NO

4 SB 5 OB

4 meningioma 2 anaplastic astrocytoma 1 ependymoma 1 craniopharyngioma 1 pineocytoma 1 pineocytoma 1 astrocytoma 2 germinoma 2 pinealoblastoma 3 meningioma 8 pineocytoma 2 pineoblastoma 2 germinoma 2 NGGCT

Manera et al. [22]

2 SB

2 pineocytoma

Backlund et al. [20] Dempsey and Lunsford [21]

Bx?

No. and diagnosis

Reference

TABLE 33-1. Summary of pineal region tumors in literature.

360 g.p. lekovic and a.g. shetter

33.

pineal region tumors

361

Case Study 33-2

Case Study 33-3

A 52-year-old woman, a nursing home resident with severe psychiatric disease, was found to have a diminishing level of consciousness and incontinence of bowel and bladder. She denied headache, and her neurologic exam was nonfocal. Imaging studies revealed a large (>5 cm) contrastenhancing lesion in the pineal region. The patient underwent endoscopic third ventriculostomy and biopsy. Pathologic examination demonstrated a low-grade appearing lesion with preserved pineal glandular architecture; the MIB labeling index was less than 0.1%. Because of the benign pathologic appearance of the tumor, the patient was taken back to the operating room for resection of pineocytoma via a suboccipital craniotomy and supracerebellar infratentorial approach. In addition, she eventually required ventriculoperitoneal shunting after failure of her third ventriculostomy. After surgery, the patient was evaluated for GKRS to residual tumor. The deﬁned target volume was 7.33 cm3. The prescription dose was 14 Gy to the 50% isodose line. The target dose-volume histogram volume was 7.22 cm3; the treatment plan included ﬁve 8-mm and ﬁve 14-mm collimator isocenters. The patient tolerated the treatment well, and there were no treatment-related complications. After treatment, the patient returned to her care facility in her usual state of health. On 25-month follow-up, MRI demonstrated slight shrinkage of the tumor (Fig. 33-4).

A 20-year-old man presented with 3 weeks of persistent headache, dizziness, and diplopia. CT demonstrated an enhancing pineal region mass. The patient underwent a suboccipital craniotomy for supracerebellar infratentorial approach for resection of the mass. Pathologic examination was consistent with pineoblastoma. The patient subsequently underwent ventriculoperitoneal shunting. The patient was begun on emergent whole-brain radiotherapy, and an initial response was found after seven treatments. The patient was therefore administered craniospinal irradiation, with 3600 cGy prescribed to the brain and spine with an intensity modulated radiation therapy (IMRT) boost to the pineal, for a total of 5580 cGy to this location. In addition, the patient underwent GKRS to the residual tumor, with a prescription dose of 14 Gy at the 50% isodose line. The deﬁned target volume was 23.00 cm3, and 22.00 cm3 was covered within the dose matrix with a combination of three 8-mm collimator isocenters and twelve 18-mm collimator isocenters. The patient tolerated craniospinal irradiation and GKRS well. There were no treatment-related complications. The patient has been followed with serial imaging, which at 4.5 years after GKRS demonstrates stable tumor with loss of central enhancement and without evidence of growth (Fig. 33-5). Clinically, the patient has a Karnofsky performance score (KPS) of 100 and has returned to work full-time.

tumors were treated to the 50% isodose line. Hasegawa et al. [25], however, reported a distant recurrence of pineocytoma after GKRS, although 100% local control was achieved (Case Study 33-2). A complete response of pineoblastoma to GKRS was demonstrated by Manera et al. [22] in one patient and a partial response in another (Case Study 33-3). However, Hasegawa et al. [14] experienced three distant failures after GKRS for pineoblastoma treated with a mean marginal dose of 15.3 Gy to the 50% isodose line.

Although germinoma is highly curable by external beam irradiation, efforts to reduce the dose of external beam radiation therapy (EBRT) radiation have gained momentum because of concerns of the toxicity of even partial brain irradiation (especially in children). In 1990, Casentini et al. [28] reported their results using Varian linac radiosurgery as an “inverse boost” paradigm for the treatment of germinoma, with all six patients so treated demonstrating a complete response. A case report of early successful use of GKRS alone for the treatment

FIGURE 33-4. A 52-year-old woman presented with a large (>5 cm) pineocytoma. (A) The patient underwent subtotal resection with treatment of 7.3 cm3 of residual tumor with GKRS. (B) Two years after

treatment, axial MRI with contrast through the pineal demonstrates stable appearance of the lesion.

362

g.p. lekovic and a.g. shetter

FIGURE 33-5. A 20-year-old man presented with pineoblastoma, which was treated with GKRS boost after craniospinal irradiation. Sagittal gadolinium-enhanced MRI at the time of GKRS (A) compared

with that on follow-up imaging (B) demonstrates stable tumor size with a prominent loss of central enhancement.

of germinoma was published in 2001 [29]. Kobayashi et al. [24] included eight patients with the diagnosis of germinoma who were treated with chemotherapy prior to GKRS. They obtained a 100% control rate for these tumors with a marginal dose of 16.8 Gy. There are a total of 19 cases of nongerminomatous germ cell tumors treated with GKRS in the literature, 2 in the series reported by Subach et al. [23], 13 in the series reported by Kobayashi et al. [24], and 4 in the series by Hasegawa et al. [26]. Only the latter series is exclusively composed of NGGCT. Again, due to the heterogeneity of diagnoses in the reported series, it is difﬁcult to draw comparisons. Nevertheless, it is clear that NGGCTs are less responsive to GKRS than are germinoma or pineocytoma. Neither of the tumors reported by Subach et al. [23] demonstrated a response to GKRS; although Hasegawa et al. [26] treated four NGGCTs with between 12 and 16 Gy at the 50% isodose line and at a mean follow-up of 25 months, one tumor progressed, one demonstrated no change, and two tumors partially responded. Kobayashi et al. [24] reported the results of 13 malignant GCTs treated with GKRS. Follow-up data were obtained in 12 patients, of whom 3 demonstrated a complete response, 3, a partial response, and 6 had progressed. Five patients were deceased at a mean follow-up of 12.6 months.

Dosimetry

Barrow Neurological Institute GKRS Experience Between 1997 and 2007, patients with nonmetastatic tumors of the pineal region have undergone GKRS (Lekovic et al., [30]). Diagnoses included eight pineocytomas, one pineoblastoma, one mixed germ cell tumor, one malignant teratoma, one primitive neuroectodermal tumor (PNET), two astrocytomas, one neurocytoma, one pineal tumor of intermediate differentiation, and one choroid plexus papilloma. Six patients were female and nine were male (mean age, 43.6 years). A tissue diagnosis was obtained via supracerebellar-infratentorial approach in all patients except for one, in whom an endoscopic biopsy was performed. One patient died 6 days after GKRS and is, therefore, excluded from further analysis.

All patients were treated using Leksell GammaPlan treatment planning software (versions 4.12 to 5.34). Mean prescription dose was 14.06 Gy (range, 12–18 to 16 Gy). All doses were prescribed to the 50% isodose line. Mean target volume was 7.42 cm3 (range, 1.2 to 32.5 cm3). Between 7 and 27 isocenters (mean, 11.4) were used in treatment planning, taking advantage of automated positioning system (APS) whenever possible, to obtain highly conformal treatment plans. Fourteen patients in this study were treated with GKRS as either the primary radiation modality or as a salvage treatment after recurrence after conventional EBRT. Two patients (one with NGGCT, one with anaplastic astrocytoma) were administered GKRS boost to the pineal region.

Results Clinical follow-up was obtained by chart review of the patient’s last follow-up visit, or in some cases by telephone, and ranged from 2 months to 95 months, except for one patient who died 6 days after GKRS. All follow-up imaging studies were interpreted by independent neuroradiologists for evidence of progression of disease. There were no complications attributable to GKRS. There were three mortalities after GKRS. One patient, already mentioned, died 6 days after radiosurgery and is excluded from further analysis, one died 2 months after radiosurgery, and the third died after developing widespread metastatic disease. These latter two patients demonstrated local control of tumor, despite their clinical progression. Therefore, upon latest followup imaging, local control was established in 100% (16/16) of the patients. Two patients (one with anaplastic astrocytoma, the other with PNET) developed leptomeningeal and spinal spread of tumor despite control of the pineal lesion. The patient with anaplastic astrocytoma underwent craniospinal irradiation and chemotherapy but continued to progress clinically and died secondary to progressive disseminated disease 8 months after radiosurgery. The patient with PNET underwent repeat GKRS for seven ventricular and suprasellar metastases and CyberKnife stereotactic radiosurgery for a chiasmatic lesion. At last followup, all of these lesions have responded well to irradiation (Fig. 33-6).

33.

pineal region tumors

FIGURE 33-6. Distant spread of tumor treated with repeat radiosurgery. A 55-year-old man presenting with a pineal region PNET was treated with GKRS. Thirty-ﬁve months after initial treatment, the patient presented with spread of disease within the lateral ventricle (A), suprasellar, and chiasmatic regions with no evidence of local recurrence (B). Six months after repeat GKRS for leptomeningeal metastatic disease (METS) and CyberKnife radiosurgery to the chiasmatic lesion, repeat imaging demonstrates response of the metastatic lesions to repeat treatment (C, D).

Conclusion Although tumors in this region are relatively rare, the pineal region is an important site of a wide range of pathologic processes. Because of the wide variation in natural history of these tumors depending on this histologic heterogeneity, obtaining a tissue diagnosis prior to instituting deﬁnitive therapy is paramount. The only exceptions to this rule may be in metastatic disease (where the presumptive tissue diagnosis is already known) and in patients that are truly too poor surgical candidates. In our experience, open craniotomy with attempted resection or at least debulking of tumor with biopsy has been the favored approach to obtaining a diagnosis. Endoscopic biopsy potentially spares the patient the risk of craniotomy if the tumor is found to be radiosensitive; however, in our experience only one of our patients did not undergo craniotomy prior to radiosurgery. In addition, endoscopy does afford the theoretical advantage of being able to perform an endoscopic third ventriculostomy at the same surgery as that for obtaining a biopsy; however, in our series all patients’ third ventriculosto-

363

mies eventually failed, ultimately requiring ventriculoperitoneal shunting. The role of radiosurgery in the management of pineal region tumors remains controversial. We have found it helpful to dichotomize our approach to the appropriateness of pineal region tumors based on the grade of tumor involved. For benign lesions (i.e., those with an indolent natural history) that are relatively radioresistant and curable with gross total surgical resection, radiosurgery competes with conventional surgery in the treatment options. Despite advances in microsurgical approaches to the pineal region, surgical extirpation of tumors occurring in this region is often impossible. Moreover, the sensitivity of pineal region tumors to radiation tends to weigh against surgical aggressiveness. Nevertheless, because surgery does offer at least the potential for cure, we believe radiosurgery may be reserved in these cases to poor surgical candidates or for the treatment of residual or recurrent disease. For malignant lesions, the role of radiosurgery is perhaps less well deﬁned. Many malignant lesions are surgically incurable and likely treatable with radiotherapy and chemotherapy alone. Nevertheless, like surgery, radiotherapy has signiﬁcant morbidity that may be ameliorated by using radiosurgery as either a boost modality or as an alternative treatment altogether. The rationale for radiosurgery for radiation-sensitive tumors is therefore analogous to that of chemotherapy (i.e., to avoid, delay, or at a minimum to reduce the dose of conventional EBRT). However, enthusiasm for the role of radiosurgery in malignant disease must be tempered, in our experience, by the fact that we have found that Gamma Knife as a primary radiation modality in the treatment of malignant disease may lead to distal failure. Of course, in these cases, the possibility of salvage with either repeat radiosurgery or radiotherapy remains. In such case, radiosurgery may still be of value in delaying the exposure to whole-brain EBRT.

References 1. Baumgartner JE, Edwards MS. Pineal tumors. Neurosurg Clin N Am 1992; 3:853–862. 2. Bruce JN. Pineal tumors. Historical perspective. In: Winn HR, ed. Youman’s Neurological Surgery. Philadelphia: Elsevier, 2005: 1–6. 3. Yamamoto I, Kageyama N. Microsurgical anatomy of the pineal region. J Neurosurg 1980; 53:205–221. 4. Stein BM. Supracerebellar-infratentorial approach to pineal tumors. Surg Neurol 1979; 11:331–337. 5. Knierim DS, Yamada S. Pineal tumors and associated lesions: the effect of ethnicity on tumor type and treatment. Pediatr Neurosurg 2003; 38:307–323. 6. Oi S, Matsumoto S. Controversy pertaining to therapeutic modalities for tumors of the pineal region: a worldwide survey of different patient populations. Childs Nerv Syst 1992; 8:332–336. 7. Regis J, Bouillot P, Rouby-Volot F, et al. Pineal region tumors and the role of stereotactic biopsy: review of the mortality, morbidity, and diagnostic rates in 370 cases. Neurosurgery 1996; 39: 907–912. 8. Fukushima T. Endoscopic biopsy of intraventricular tumors with the use of a ventriculoﬁberscope. Neurosurgery 1978; 2:110–113. 9. Fukushima T, Ishijima B, Hirakawa K, et al. Ventriculoﬁberscope: a new technique for endoscopic diagnosis and operation. Technical note. J Neurosurg 1973; 38:251–256.

364

g.p. lekovic and a.g. shetter

10. Robinson S, Cohen AR. The role of neuroendoscopy in the treatment of pineal region tumors. Surg Neurol 1997; 48:360–365. 11. Oi S, Shibata M, Tominaga J, et al. Efﬁcacy of neuroendoscopic procedures in minimally invasive preferential management of pineal region tumors: a prospective study. J Neurosurg 2000; 93: 245–253. 12. Bruce JN, Stein BM. Surgical management of pineal region tumors. Acta Neurochir (Wien) 1995; 134:130–135. 13. Bruce JN, Ogden AT. Surgical strategies for treating patients with pineal region tumors. J Neurooncol 2004; 69:221–236. 14. Matsutani M, Sano K, Takakura K, al. Primary intracranial germ cell tumors: a clinical analysis of 153 histologically veriﬁed cases. J Neurosurg 1997; 86:446–455. 15. Weiner HL, Finlay JL. Surgery in the management of primary intracranial germ cell tumors. Childs Nerv Syst 1999; 15:770–773. 16. Sung DI, Harisliadis L, Chang CH. Midline pineal tumors and suprasellar germinomas: highly curable by irradiation. Radiology 1978; 128:745–751. 17. Dattoli MJ, Newall J. Radiation therapy for intracranial germinoma: the case for limited volume treatment. Int J Radiat Oncol Biol Phys 1990; 19:429–433s. 18. Balmaceda C, Heller G, Rosenblum M, et al. Chemotherapy without irradiation—a novel approach for newly diagnosed CNS germ cell tumors: results of an international cooperative trial. The First International Central Nervous System Germ Cell Tumor Study. J Clin Oncol 1996; 14:2908–2915. 19. Buckner JC, Peethambaram PP, Smithson WA, et al. Phase II trial of primary chemotherapy followed by reduced-dose radiation for CNS germ cell tumors. J Clin Oncol 1999; 17:933–940.

20. Backlund EO, Rahn T, Sarby B. Treatment of pinealomas by stereotaxic radiation surgery. Acta Radiol Ther Phys Biol 1974; 13:368–376. 21. Dempsey PK, Lunsford LD. Stereotactic radiosurgery for pineal region tumors. Neurosurg Clin N Am 1992; 3:245–253. 22. Manera L, Regis J, Chinot O, et al. Pineal region tumors: the role of stereotactic radiosurgery. Stereotact Funct Neurosurg 66 Suppl 1996; 6(1):164–173. 23. Subach BR, Lunsford LD, Kondziolka D. Stereotactic radiosurgery in the treatment of pineal region tumors. Prog Neurol Surg 1998; 14:175–194. 24. Kobayashi T, Kida Y, Mori Y. Stereotactic gamma radiosurgery for pineal and related tumors. J Neurooncol 2001; 54:301–309. 25. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. The role of radiosurgery for the treatment of pineal parenchymal tumors. Neurosurgery 2002; 51:880–889. 26. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. Stereotactic radiosurgery for CNS nongerminomatous germ cell tumors. Report of four cases. Pediatr Neurosurg 2003; 38:329–333. 27. Deshmukh VR, Smith KA, Rekate HL, et al. Diagnosis and management of pineocytomas. Neurosurgery 2004; 55:349–355. 28. Casentini L, Colombo F, Pozza F, Benedetti A. Combined radiosurgery and external radiotherapy of intracranial germinomas. Surg Neurol 1990; 34:79–86. 29. Regine WF, Hodes JE, Patchell RY. Intracranial germinoma: treatment with radiosurgery alone. J Neurooncol 1998; 37(1):75–77. 30. Lekovic G, Gonzalez F, Shetter A, et al. The role of Gamma Knife radiosurgery in the management of pineal region tumors. Neurosurgical Focus 2007; 23(6).

3 4

Pineal Region Tumors: Surgery Perspective Alfred T. Ogden and Jeffrey N. Bruce

Introduction Since the seminal articles that described safe and effective approaches to the pineal region in the early 1970s [1, 2], numerous surgical series have established open surgery as a critical tool in the treatment of pineal region lesions [3–13]. Although the difﬁculties and perils of pineal surgery were considered prohibitive during the nascent years of neurosurgery [14–16], the advent of the operating microscope, microneurosurgical techniques, magnetic resonance imaging (MRI), and the development of neurologic intensive care as a specialty have all helped to develop pineal region surgery within a reasonable degree of safety. Open surgery is the treatment of choice for benign tumors. Depending on speciﬁc histology, open surgery has a variable role in the treatment of malignant tumors and tumors with mixed elements, either as a means of cytoreduction prior to chemotherapy and/or radiation [17, 18] or as the best option to eradicate benign elements in mixed tumors after chemotherapy/radiation [4, 19]. Additionally, through immediate removal of obstructing lesions, open surgery can obviate the need for permanent cerebrospinal ﬂuid (CSF) diversion. Open surgery also offers comprehensive tissue sampling for a group of tumors that can be of mixed histology and are notoriously difﬁcult to diagnose. Stereotactic radiosurgery has an unproven role as an adjunct or even an alternative to open surgery. The great variety of tumors that occur in the pineal region and the need to direct adjuvant therapy based on histopathologic diagnosis mandate that, with few exceptions, treatment, including stereotactic radiosurgery, should not be pursued without a tissue diagnosis. Stereotactic biopsy followed by stereotactic radiosurgery may prove to be a reasonable alternative to open surgery in selected cases, but this strategy is currently most appropriate for patients with signiﬁcant medical contraindications to open surgery. Stereotactic radiosurgery may also have an additional role as an alternative to whole-brain radiation in the treatment of select radiosensitive tumors.

Surgical Management: An Overview Initial management of a pineal region mass entails evaluation for hydrocephalus, which is present in most patients. Although ventriculoperitoneal shunting is acceptable, endoscopic third

ventriculostomy is preferred, achieving the same end without exposing a patient to the potential problems of shunt malfunction, shunt infection, and abdominal seeding of a malignancy. After CSF diversion, a procedure to obtain tissue is indicated, either via a stereotactic, an endoscopic, or an open approach. Either diverting procedure provides a convenient opportunity to assay CSF for malignant germ cell tumor markers (α-fetoprotein or β-HCG) and cytology, which can in very rare cases obviate the need for tissue sampling. If these markers are elevated in either CSF or serum, then by deﬁnition a malignant nongerminomatous germ cell tumor is present, and chemotherapy and radiation therapy can proceed without the need for histologic conﬁrmation.

Stereotactic Biopsy Although safe, stereotactic biopsy in the pineal region requires more care and planning than biopsies of most other areas of the brain. An anterolateral approach is favored with a precoronal entry point just behind the nexus of the superior temporal line and the hairline. Using image guidance, the trajectory can be adjusted to enter a gyrus at the pial surface, to avoid the lateral ventricle, and to steer clear of any obvious vasculature that may be distorted into the path of the biopsy needle by the tumor itself. Close proximity of the target to the deep venous system and a lack of adjacent tissue turgor should theoretically increase the likelihood of a signiﬁcant biopsy-induced hemorrhage. In fact, such hemorrhages appear to be mildly more common in the pineal region, but they are rarely of any clinical signiﬁcance [20, 21]. Some series have suggested a higher morbidity associated with pineal region biopsies compared with routine biopsies, whereas others do not. In any case, overall complication rates are low and are almost always a function of transient exacerbations of existing symptoms [22–25]. The major drawback to biopsies is the limited tissue sampling that makes deﬁnitive diagnosis difﬁcult when dealing with these widely diverse and often heterogeneous lesions.

Endoscopic Biopsy Some authors have advocated obtaining tissue specimens endoscopically during the same operation as third ventriculostomy.

365

366

a.t. ogden and j.n. bruce

TABLE 34-1. Results of large microsurgical series for pineal region tumors.

Authors

Year

No. of cases

Approach

Patient population

Hoffman et al. [7] Neuwelt et al. [8] Lapras et al. [9] Edwards et al. [10] Pluchino et al. [11] Luo SQ et al. [12] Vaquero et al. [13] Herrmann et al. [27] Bruce and Stein [3] Chandy et al. [28]

1983 1985 1987 1988 1989 1989 1992 1992 1995 1998

61 13 86 36 40 64 29 49 160 48

TC/ITSC OTT TC/OTT TT/OTT/ITSC ITSC OTT TC/ITSC/OTT IHTC/ITSC ITSC/TC/OTT ITSC/OTT

Pediatric Adult/pediatric Adult/pediatric Pediatric Adult/pediatric Adult/pediatric Adult/pediatric Adult/pediatric Adult/pediatric Adult/pediatric

1998 1998 2003

16 21 201

Adult/pediatric Pediatric/adult Adult/pediatric

2004

81

OTT/ITSC/TC OTT OTT(54%) ITSC(34%) ITSC/TC/OTT

All All All All All All All All All “Benign lesions” All All All

Kang et al. [29] Shin et al. [30] Konovalov et al. [5] Bruce [31]

Adult/pediatric

All

Pathology

Percent of patients with GTR (%)

Mortality (%)

NA 60 65 ? 25 21 NA NA 45 55 37.5 54.5

47

Major morbidity (%)

Permanent minor morbidity (%)

20* 0 5.8† 0 5 10 11 8 4 0

NA 0 5.8† 3.3 NA

NA 20 28 3.3 NA NA NA NA 19 NA

0 0 10‡

0 0 NA

19 5

2

NA

1

NA NA 3 NA

>20

*All except one mortality prior to 1975. †Combined major morbidity/mortality reduced to 2.8% in last 40 patients. ‡Mortality rate of 1.8% in the 168 resections after 1990.

Although a potentially elegant one-step solution to CSF diversion and tissue diagnosis, such a strategy necessarily biases tissue sampling toward one quadrant of the tumor capsule. Also, unless a ﬂexible endoscope is used, endoscopic biopsy and third ventriculostomy must be performed through two separate burr holes and as such offers no less “invasiveness” than stereotactic biopsy and third ventriculostomy.

Craniotomy for Open Resection Although diagnostic biopsies are a suitable ﬁrst option in some cases, an open approach is generally preferred. The advantages of an open resection are numerous, including generous tissue sampling, potential obviation of a shunt, and the ability to proceed with a radical resection if indicated after intraoperative pathology consultation. The relative advantages of pursuing radical resection depend on an accurate histopathologic diagnosis and must be weighed against rates of morbidity and mortality from radical surgery. Complications from open surgery can be serious and even catastrophic typically resulting from postoperative hematoma, venous infarction from vein sacriﬁce, thalamic injury from dissection of the anterior tumor capsule, and visual deﬁcits if the occipital transtentorial approach is used. However, when complete tumor removal can be achieved, the risk of postoperative hemorrhage into a subtotally resected tumor is reduced. Minor complications include infection and exacerbation of existing ataxia or Parinaud phenomena. Pineal region surgery is certainly not trivial, but within the past 25 years, major morbidity and mortality from published surgical

series have improved dramatically and over the past quarter century has dropped to 0 to 2% [26] (Table 34-1).

Surgical Results by Tumor Histology The ﬁeld of oncology is predicated on the belief that any treatment modality must be evaluated on the basis of efﬁcacy in treating groups of patients bound by a common diagnosis. This diagnosis is almost always made on the basis of tissue histology. Pineal region tumors can arise from any of the myriad cell types that normally occur in and around the pineal gland, as well as from the range of ectopic tissues that can be trapped near the pineal gland during embryogenesis. Because few places in the body can match such cellular diversity and biopsies of pineal lesions have been demonstrated to be safe, histologic diagnosis of pineal region tumors is mandatory unless tumor markers are positive. Magnetic resonance imaging (MRI) is excellent at deﬁning the relationship between tumor and adjacent anatomic structures but is not able to distinguish between tumors of different histologies (Fig. 34-1). Tumors in the pineal region can be categorized as benign or malignant. Alternatively, they can be classiﬁed into one of four broad histologic categories: (1) pineal cell; (2) germ cell; (3) glial cell; and (4) miscellaneous. Although surgical results signiﬁcantly inﬂuence the outcome for these patients, individual prognosis is heavily dependent upon tumor type. The rarity of individual histopathologic types within the pineal region makes it difﬁcult to generate sufﬁcient numbers of patients treated to evaluate individual therapies, however some generalizations can be made.

34.

pineal region tumors: surgery perspective

367

irradiation after biopsy [33, 34]. The “true” pineal astrocytomas are often cystic, encapsulated, and resemble pilocytic astrocytomas on histology [35]. These can be completely resected and have excellent long-term results. Surgical outcomes from ependymomas are binary, depending on the degree of anaplasia. Pineal region ependymomas with low cellularity and few mitoses have excellent long-term outcomes although they may recur more readily than ependymomas associated with the lateral ventricles [36].

Pineal Parenchymal Tumors

FIGURE 34-1. Similar imaging characteristics of pineal lesions with different histologies. T1-weighted MRIs, precontrast (left panels) and postcontrast (right panels), of three patients with pineal region tumors: (A) ependymoma, (B) pineocytoma, (C) germinoma.

Benign Pineal Region Tumors Benign tumors account for around one third of the masses found in pineal region [3]. These are composed mainly of welldifferentiated ependymomas, meningiomas, teratomas, pineocytomas, and rare pilocytic astrocytomas. In each case, radical resection is the standard of care if it can be performed within a reasonable degree of safety. Gross total resection provides the best chance for a cure or extended remission but a goal that must be weighed against the risks associated with radical resection. Surgical series demonstrate good outcomes over a range of pathologies, although there is a need in the literature for surgical outcomes to be analyzed according to speciﬁc histology. There are only two small surgical series dedicated to benign pineal lesions of a single histologic type [6, 32].

Glial Tumors Glial neoplasms encountered in the pineal region fall into three categories: (1) brain-stem astrocytomas, usually tectal lesions that extend rostrally, (2) “true” pineal astrocytomas that arise from the supporting astrocytes of the pineal gland itself, and (3) ependymomas associated with the third ventricle. Brain-stem astrocytomas that grow into the pineal region are solid and, although they may be “low grade,” are generally invasive. These are not amenable to aggressive resection and require

Arising from the melatonin-producing cells of the pineal gland, pineal parenchymal tumors exist along a histopathologic continuum from benign and indolent pineocytomas to malignant and aggressive pineoblastomas. Tumors of intermediate grade are referred to as mixed pineal parenchymal tumors, and various classiﬁcation schemes that are tied to prognosis have been proposed [32, 37–41]. Making an accurate pathologic diagnosis is difﬁcult even with generous tissue sampling. The rarity of these tumors makes the aggregation of sufﬁcient numbers of patients for an instructive clinical trial difﬁcult. Further complicating matters is the variable behavior of tumors with similar histopathologies depending upon age of presentation. Amidst the paucity of data to direct clinical decisions, certain general treatment guidelines are apparent. The goal for well-differentiated pineocytomas should be a cure, at least in adults. There is some evidence that this standard is unrealistic in pediatric cases where pineocytomas may behave more aggressively [42]. The standard of care for pineocytoma has been set through clinical experience of low recurrence rates with gross total resection, with the only surgical series dedicated to pineocytomas reporting no recurrences in ﬁve patients after 2 to 8 years of follow-up [32]. Outcomes from radiosurgical treatment of pineocytoma have been favorable as well [43–45], although of the 14 adult cases of radiosurgically treated pineocytoma in the literature, some of the follow-up time is short [45] and there was at least one adult treatment failure that resulted in death from CNS metastasis [44]. The one radiosurgical series reporting pediatric pineocytoma outcomes reported four of seven treatment failures resulting in death during a follow-up of 3 months to 4 years [46]. At the other end of the neoplastic spectrum are pineoblastomas. Although pineoblastomas often appear identical to pineocytomas on MRI, they are histologically indistinguishable from PNETs and behave in a similar clinical fashion. Like PNETs, they tend to be more aggressive in children than in adults [47], and within the pediatric population they are increasingly aggressive with decreasing age of presentation [48–55]. Although there is some indication of increased survival in both adults [17] and children [18] undergoing open surgery, no statistically rigorous study has examined the impact of extent of resection on survival or disease progression. In the absence of such studies, it seems logical to follow the standard of care for PNETs of the posterior fossa for which signiﬁcant survival beneﬁts are apparent after reduction of tumor mass under a cubic centimeter. There are 10 published cases of stereotactic radiosurgery for pineoblastoma and the results are poor [43–46], likely reﬂecting the malignant nature of these tumors and their

368

a.t. ogden and j.n. bruce

potential to spread through CSF. The relative utility of radiosurgery cannot be ascertained based on a few published cases except to say that radiosurgery does not appear to be a “magic bullet” for these lesions, and it does not address the metastatic potential of these tumors. Ultimately, prognosis for pineoblastoma is most predicted by age of presentation and disease dissemination at the time of diagnosis. Although far from comprehensive, evidence points toward surgical cytoreduction as having a beneﬁcial impact on survival. Intermediate-grade pineal parenchymal tumors behave in an unpredictable fashion. Treatment successes and failures have been reported from a range of treatment strategies, and there is little helpful data to direct clinical decisions [17, 40, 41]. Although patients with pineoblastoma are routinely irradiated after surgery, it is unclear whether this is universally required in cases of intermediate-grade pineal parenchymal tumors. Clinical studies that tie adjuvant therapy to tumor grade and outcome are required with long-term outcome as tumors can recur more than 5 years from diagnosis [17]. On the whole, pineal parenchymal tumors present a formidable challenge to clinicians. Treatment optimization will only be possible with systematic and expert pathologic diagnosis that is not restricted by availability of tissue. Ideally, this tissue should be sought via open surgery as radical resection loosely correlates with long-term survival.

Germ Cell Tumors Germ cell tumors are considered and studied in two separate groups, germinomas and nongerminomatous germ cell tumors (NGGCTs), comprised of endodermal sinus tumors, choriocarcinomas, embryonal carcinomas, mature teratomas, and immature teratomas. The role of open surgery for these lesions is perhaps as well deﬁned as for any pineal region tumors. Germinomas are the most common of pineal region tumors, especially in adolescent boys and young men. In Japan and Korea, for reasons that are not understood, germ cell tumors in general and germinomas in particular are more prevalent than in any other part of the world including neighboring China [51]. Because they do not secrete a speciﬁc tumor marker and they cannot be distinguished radiographically from other types of tumors that call for different treatment paradigms, diagnosis should be made with tissue conﬁrmation regardless of age of presentation. Cytoreduction from open surgery has not been shown to improve the excellent outcomes with radiation alone [52] and thus the overwhelming majority of germinomas are diagnosed by open or stereotactic biopsy and treated with whole-brain radiation. Historically, the relatively high prevalence of germinomas, their exquisite radiosensitivity, and the perceived dangers of pineal region biopsy all contributed to the practice of up-front radiation without tissue diagnosis for all pineal region tumors. This approach was still commonly pursued in the Far East as recently as 1992 [53, 54]. Although it is difﬁcult to take issue with cure rates that today are over 90% [55–59], improvements in the treatment of intracranial germinoma can still be made. A small percentage of patients fail radiation and suffer CSF dissemination that is ultimately fatal. Treatment failure is much more likely when syncytiotrophoblastic giant cells (STGCs) are mixed within the usual histologic features,

and failure rates within this histologic subtype have been reported as high as 40% [60–62]. Because these are the rare germinomas that secrete β-HCG to produce mildly elevated serum levels, their diagnosis should be straightforward with adequate biopsy tissue and standard assays of serum markers. As such, they should be considered in a separate category from “pure germinomas” and the subject of separate clinical trials to reexamine all treatment modalities, including radiosurgery and surgical cytoreduction. Because many patients with intracranial germinoma have long-term remission, the extended sequelae of whole-brain radiation, particularly in the pediatric population, are another major concern. Strides in chemotherapy have been made in trials composed of children too young to receive full doses of whole-brain radiation [63, 64]. Radiosurgery is a theoretically attractive alternative because of relatively limited radiation exposure to adjacent brain tissue. Reports of successful treatment of germinoma with stereotactic radiosurgery exist with very short follow-up [43, 45]. Although stereotactic radiosurgery like conventional radiotherapy may not be effective for the germinomas with STGCs [43], it may serve as an effective adjunct for pure germinomas that could reduce cumulative radiation doses. This hypothesis could be tested in an ethically responsible manner in the young pediatric population. NGGCTs have historically had a much worse prognosis that germinomas. Because the individual types are so rare and they are frequently of mixed histology, they have been lumped together in retrospective analyses and clinical trials. These are the only pineal region tumors that should be treated without a tissue diagnosis as elevated levels of markers in the serum and/ or CSF are pathognomonic for speciﬁc histopathologies. Although open surgery in a preadjuvant therapy cytoreductive role has been examined with variable results [63, 65, 66], the best results seem to occur when radiation and/or chemotherapy is followed by “second look” surgery when a persistent radiographic lesion exists [4, 63, 64, 67]. Using this approach, only residual teratomatous elements or scar tissue has been found, and 5-year survival rates have improved dramatically to >90%. Stereotactic radiosurgery has an undeﬁned role for NGGCTs. One report [68] of four patients who received radiosurgery along with fractionated radiation and chemotherapy showed tumor regression in three patients after a follow-up of 2 years.

Conclusion Despite sharing a common anatomic location and similar imaging characteristics, pineal region tumors are extremely heterogenous with respect to histopathology, natural history, and response to therapy. Except for cases of marker-secreting germ cell tumors, a tissue diagnosis is required to direct therapy. In the absence of elevated markers, in most cases, open surgery is the best initial therapeutic option. Benign lesions can be cured with gross total resection, and in experienced hands gross total resection can be achieved with low complication rates. Stereotactic biopsy followed by stereotactic radiosurgery is a potential alternative that is most appropriate

34.

pineal region tumors: surgery perspective

in patients with tumor seeding or medical contraindications to open surgery. The most effective treatments for malignant lesions will likely require a combination of modalities that will be directed by histopathologic diagnosis. The merit of this approach is apparent in the example of NGGCTs whose prognosis has vastly improved after numerous clinical trials that were predicated on cohorts of patients with speciﬁc diagnoses. Further progress will depend on consistent histologic conﬁrmation of these rare tumors rather than empiric up-front conventional radiation or stereotactic radiosurgery as some authors have proposed [46].

References 1. Stein BM. The infratentorial supracerebellar approach to pineal lesions. J Neurosurg 1971; 35:197–202. 2. Jamieson KG. Excision of pineal tumors. J Neurosurg 1971; 35: 550–553. 3. Bruce JN, Stein BM. Surgical management of pineal region tumors. Acta Neurochir (Wien) 1995; 134:130–135. 4. Weiner HL, Lichtenbaum RA, Wisoff JH, et al. Delayed surgical resection of central nervous system germ cell tumors. Neurosurgery 2002; 50:727–733; discussion 733–734. 5. Konovalov AN, Pitskhelauri DI. Principles of treatment of the pineal region tumors. Surg Neurol 2003; 59:250–268. 6. Konovalov AN, Spallone A, Pitzkhelauri DI. Meningioma of the pineal region: a surgical series of 10 cases. J Neurosurg 1996; 85: 586–590. 7. Hoffman HJ, Yoshida M, Becker LE, et al. Pineal region tumors in childhood. Experience at the Hospital for Sick Children. 1983. Pediatr Neurosurg 1994; 21:91–103; discussion 104. 8. Neuwelt EA. An update on the surgical treatment of malignant pineal region tumors. Clin Neurosurg 1985; 32:397–428. 9. Lapras C, Patet JD, Mottolese C, Lapras C Jr. Direct surgery for pineal tumors: occipital-transtentorial approach. Prog Exp Tumor Res 1987; 30:268–280. 10. Edwards MS, Hudgins RJ, Wilson CB, et al. Pineal region tumors in children. J Neurosurg 1988; 68:689–697. 11. Pluchino F, Broggi G, Fornari M, et al. Surgical approach to pineal tumours. Acta Neurochir (Wien) 1989: 96:26–31. 12. Luo, SQ, Li DZ, Zhang MZ, et al. Occipital transtentorial approach for removal of pineal region tumors: report of 64 consecutive cases. Surg Neurol 1989; 32:36–39. 13. Vaquero J, Ramiro J, Martinez R, Bravo G. Neurosurgical experience with tumours of the pineal region at Clinica Puerta de Hierro. Acta Neurochir (Wien) 1992; 116:23–32. 14. Dandy W. Operative experience in cases of pineal tumor. Arch Surg 1936; 33:19–46. 15. Horrax G. Treatment of tumors of the pineal body. Experience in a series of twenty-two cases. Arch Neurol Psychiatry 1950; 64:227– 242. 16. Wagenen V. A surgical appraoch for the removal of certain pineal tumors. Surg Gynecol Obstet 1931; 37:216–220. 17. Lutterbach J, Fauchon F, Schild SE, et al. Malignant pineal parenchymal tumors in adult patients: patterns of care and prognostic factors. Neurosurgery 2002; 51:44–55; discussion 55–56. 18. Reddy AT, Janss AJ, Phillips PC, et al. Outcome for children with supratentorial primitive neuroectodermal tumors treated with surgery, radiation, and chemotherapy. Cancer 2000; 88:2189–2193. 19. Weiner HL, Finlay JL. Surgery in the management of primary intracranial germ cell tumors. Childs Nerv Syst 1999; 15:770–773. 20. Sawin PD, Hitchon PW, Follett KA, Torner JC. Computed imaging-assisted stereotactic brain biopsy: a risk analysis of 225 consecutive cases. Surg Neurol 1998; 49:640–649.

369

21. Field M, Witham TF, Flickinger JC, et al. Comprehensive assessment of hemorrhage risks and outcomes after stereotactic brain biopsy. J Neurosurg 2001; 94:545–551. 22. Kreth FW, Schatz CR, Pagenstecher A, et al. Stereotactic management of lesions of the pineal region. Neurosurgery 1996; 39:280– 289; discussion 289–291. 23. Regis J, Bouillot P, Rouby-Volot F, et al. Pineal region tumors and the role of stereotactic biopsy: review of the mortality, morbidity, and diagnostic rates in 370 cases. Neurosurgery 1996; 39:907–912; discussion 912–914. 24. Dempsey PK, Kondziolka D, Lunsford LD. Stereotactic diagnosis and treatment of pineal region tumours and vascular malformations. Acta Neurochir (Wien) 1992; 116:14–22. 25. Apuzzo ML, Chandrasoma PT, Cohen D, et al. Computed imaging stereotaxy: experience and perspective related to 500 procedures applied to brain masses. Neurosurgery 1987; 20:930–937. 26. Bruce JN, Ogden AT. Surgical strategies for treating patients with pineal region tumors. J Neurooncol 2004; 69:221–236. 27. Herrmann HD, Winkler D, Westphal M. Treatment of tumours of the pineal region and posterior part of the third ventricle. Acta Neurochir (Wien) 1992; 116:137–146. 28. Chandy MJ, Damaraju SC. Benign tumours of the pineal region: a prospective study from 1983 to 1997. Br J Neurosurg 1998; 12: 228–233. 29. Kang JK, Jeun SS, Hong YK et al. Experience with pineal region tumors. Childs Nerv Syst 1998; 14:63–68. 30. Shin HJ, Cho BK, Jung HW, Wang KC. Pediatric pineal tumors: need for a direct surgical approach and complications of the occipital transtentorial approach. Childs Nerv Syst 1998; 14:174– 178. 31. Bruce JN. Pineal tumors. In: Winn H, ed. Youman’s Neurological Surgery, vol. 1. Philadelphia: WB Saunders, 2004:1011–1029. 32. Vaquero J, Ramiro J, Martinez R, et al. Clinicopathological experience with pineocytomas: report of ﬁve surgically treated cases. Neurosurgery 1990; 27:612–8; discussion 618–619. 33. Barnett DW, Olson JJ, Thomas WG, Hunter SB. Low-grade astrocytomas arising from the pineal gland. Surg Neurol 1995; 43:70–75; discussion 75–76. 34. Stein BM, Bruce JN. Surgical management of pineal region tumors (honored guest lecture). Clin Neurosurg 1992; 39:509–532. 35. DeGirolami U, Armbrustmacher V.. Juvenile pilocytic astrocytoma of the pineal region: report of a case. Cancer 1982; 50:1185– 1188. 36. Schwartz TH, Kim S, Glick RS, et al. Supratentorial ependymomas in adult patients. Neurosurgery 1999; 44:721–731. 37. Borit A, Blackwood W, Mair WG. The separation of pineocytoma from pineoblastoma. 1980; Cancer 45:1408–1418. 38. Cho BK, Wang KC, Nam DH, et al. Pineal tumors: experience with 48 cases over 10 years. Childs Nerv Syst 1998; 14:53–58. 39. Coca S, Vaquero J, Escandon J, et al. Immunohistochemical characterization of pineocytomas. Clin Neuropathol 1992; 11:298– 303. 40. Fauchon F, Jouvet A, Paquis P, et al. Parenchymal pineal tumors: a clinicopathological study of 76 cases. Int J Radiat Oncol Biol Phys 2000; 46:959–968. 41. Jouvet A, Saint-Pierre G, Fauchon F, et al. Pineal parenchymal tumors: a correlation of histological features with prognosis in 66 cases. Brain Pathol 2000; 10:49–60. 42. D’Andrea AD, Packer RJ, Rorke LB, et al. Pineocytomas of childhood. A reappraisal of natural history and response to therapy. Cancer 1987; 59:1353–1357. 43. Kobayashi T, Kida Y, Mori Y. Stereotactic gamma radiosurgery for pineal and related tumors. J Neurooncol 2001; 54:301–309. 44. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. The role of radiosurgery for the treatment of pineal parenchymal tumors. Neurosurgery 2002; 51:880–889.

370

a.t. ogden and j.n. bruce

45. Manera L, Regis J, Chinot O, et al. Pineal region tumors: the role of stereotactic radiosurgery. Stereotact Funct Neurosurg 1996; 66: 164–173. 46. Raco A, Raimondi AJ, D’Alonzo A, et al. Radiosurgery in the management of pediatric brain tumors. Childs Nerv Syst 2000; 16:287–295. 47. Chang SM, Lillis-Hearne PK, Larson DA, et al. Pineoblastoma in adults. Neurosurgery 1995; 37:383–390; discussion 390–391. 48. Duffner PK, Cohen ME, Sanford RA, et al. Lack of efﬁcacy of postoperative chemotherapy and delayed radiation in very young children with pineoblastoma. Pediatric Oncology Group. Med Pediatr Oncol 1995; 25:38–44. 49. Abay EO 2nd, Laws ER Jr, Grado GL, et al. Pineal tumors in children and adolescents. Treatment by CSF shunting and radiotherapy. J Neurosurg 1981; 55:889–895. 50. Jakacki RI, Zeltzer PM, Boyett JM, et al. Survival and prognostic factors following radiation and/or chemotherapy for primitive neuroectodermal tumors of the pineal region in infants and children: a report of the Childrens Cancer Group. J Clin Oncol 1995; 13:1377–1383. 51. Oi S, Matsuzawa K, Choi JU, et al. Identical characteristics of the patient populations with pineal region tumors in Japan and in Korea and therapeutic modalities. Childs Nerv Syst 1998; 14:36– 40. 52. Sawamura Y, de Tribolet N, Ishii N, Abe H. Management of primary intracranial germinomas: diagnostic surgery or radical resection? J Neurosurg 1997; 87:262–266. 53. Oi S. Recent advances and racial differences in therapeutic strategy to the pineal region tumor. Childs Nerv Syst 1998; 14:33–35. 54. Oi S, Matsumoto S. Controversy pertaining to therapeutic modalities for tumors of the pineal region: a worldwide survey of different patient populations. Childs Nerv Syst 1992; 8:332–336. 55. Jenkin RD, Simpson WJ, Keen CW. Pineal and suprasellar germinomas. Results of radiation treatment. J Neurosurg 1978; 48: 99–107. 56. Rich TA, Cassady JR, Strand RD, Winston KR. Radiation therapy for pineal and suprasellar germ cell tumors. Cancer 1985; 55:932– 940. 57. Shirato H, Nishio M, Sawamura Y, et al. Analysis of long-term treatment of intracranial germinoma. Int J Radiat Oncol Biol Phys 1997; 37:511–515.

58. Sung DI, Harisliadis L, Chang CH. Midline pineal tumors and suprasellar germinomas: highly curable by irradiation. Radiology 1978; 128:745–751. 59. Wara WM, Fellows CF, Sheline GE, et al. Radiation therapy for pineal tumors and suprasellar germinomas. Radiology 1977; 124: 221–223. 60. Uematsu Y, Tsuura Y, Miyamoto K, et al. The recurrence of primary intracranial germinomas. Special reference to germinoma with STGC (syncytiotrophoblastic giant cell). J Neurooncol 1992; 13:247–256. 61. Utsuki S, Kawano N, Okra H, et al. Cerebral germinoma with syncytiotrophoblastic giant cells: feasibility of predicting prognosis using the serum hCG level. Acta Neurochir (Wien) 1999; 141:975– 977; discussion 977–978. 62. Utsuki S, Oka H, Tanaka S, et al. Long-term outcome of intracranial germinoma with hCG elevation in cerebrospinal ﬂuid but not in serum. Acta Neurochir (Wien) 2002; 144:1151–1154; discussion 1154–1155. 63. Balmaceda C, Heller G, Rosenblum M, et al. Chemotherapy without irradiation—a novel approach for newly diagnosed CNS germ cell tumors: results of an international cooperative trial. The First International Central Nervous System Germ Cell Tumor Study. J Clin Oncol 1996; 14:2908–2915. 64. Kochi M, Itoyama Y, Shiraishi S, et al. Successful treatment of intracranial nongerminomatous malignant germ cell tumors by administering neoadjuvant chemotherapy and radiotherapy before excision of residual tumors. J Neurosurg 2003; 99:106– 114. 65. Matsutani M, Sano K, Takakura K, et al. Primary intracranial germ cell tumors: a clinical analysis of 153 histologically veriﬁed cases. J Neurosurg 1997; 86:446–455. 66. Robertson PL, DaRosso RC, Allen JC. Improved prognosis of intracranial non-germinoma germ cell tumors with multimodality therapy. J Neurooncol 1997; 32:71–80. 67. Friedman JA, Lynch JJ, Buckner JC, et al. Management of malignant pineal germ cell tumors with residual mature teratoma. Neurosurgery 2001; 48:518–522; discussion 522–523. 68. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. Stereotactic radiosurgery for CNS nongerminomatous germ cell tumors. Report of four cases. Pediatr Neurosurg 2003; 38:329–333.

3 5

Pineal Tumors: Fractionated Radiation Therapy Perspective Steven E. Schild

Introduction Approximately 18,400 primary brain tumors were diagnosed in the United States during 2004, and, of these, approximately 12,690 resulted in death [1]. Pineal tumors make up only 0.4% to 1% of all brain tumors [2]. The pineal gland is composed of pineal parenchymal cells or pineocytes. The pineocyte is a specialized neuron surrounded by blood vessels and astrocytes. The pineocytes produce the hormone melatonin, which affects circadian rhythms, inhibiting the release of both luteinizing and follicle-stimulating hormones. Serum concentrations of melatonin are low during exposure to light and increase in the dark. Melatonin stimulates sleep and is sold over the counter to treat insomnia. The pineal gland region is unusual in the wide variety of primary tumors that arise from this site [3–8]. Benign cysts can also occur in this site as do ordinary brain tumors such as astrocytomas or meningiomas. However, these are not as common as the germ cell tumors, which make up about two thirds of the malignancies of the pineal gland. It is assumed that germ cells migrate during embryogenesis and can lodge at this ectopic site, becoming malignant germ cell tumors later in life. The most common of these are the germinomas, which are histologically indistinguishable from testicular seminomas and ovarian dysgerminomas. Nongerminomatous germ cell tumors (NGGCTs) also occur in pure and mixed forms. Histologic varieties of nongerminomatous germ cell tumors include the teratoma, embryonal carcinoma, yolk sac tumor (endodermal sinus tumor), choriocarcinoma, or combinations referred to as mixed germ cell tumors. Teratomas typically include cell types derived from all three germ cell layers: endoderm, mesoderm, and ectoderm. Mature teratomas contain fully differentiated cells. Immature teratomas consist of cells and tissues resembling those of the developing fetus. By convention, tumors in which mature and immature components coexist are classiﬁed as immature. Other teratomas with malignant epithelial or mesenchymal elements are considered to be teratomas with malignant transformation. Embryonal carcinomas contain large, primitive-appearing epithelial cells that form sheets, cords, and gland-like arrangements. In rare instances, they contain plate-

like miniature embryos or embryoid bodies. Yolk sac tumors are epithelial tumors that contain compact sheets, ribbons, cords, or papillae. Schiller-Duval bodies are diagnostic of yolk sac tumors. They contain tufted epithelium-covered vessels projecting into clear spaces. In addition to Schiller-Duval bodies, yolk sac tumors often have a loose-knit vitelline pattern. Choriocarcinomas contain a bilaminar arrangement of both syncytiotrophoblasts and cytotrophoblasts. Combinations of any of these tumor types are referred to as mixed germ cell tumors. One speciﬁc mixed germ cell tumor, the teratocarcinoma, contains elements of both embryonal carcinoma and teratoma [3–8]. Some of these tumors produce markers such as AFP (α-fetoprotein) or β-HCG (β-human chorionic gonadotropin). Yolk sac tumors (endodermal sinus tumors) can produce AFP, choriocarcinomas can produce β-HCG, and embryonal tumors can produce both markers. Pure germinomas can produce βHCG but usually not in very high levels. These markers are generally found in higher concentration in the cerebrospinal ﬂuid (CSF) than in the serum and can aid in identifying the tumor type and monitoring the effects of therapy. Tumors may also arise in the pineocytes forming a group of tumors called pineal parenchymal tumors, which compose 15% to 30% of all pineal tumors. These can be classiﬁed into four groups: pineocytomas, mixed pineal parenchymal tumors, pineal parenchymal tumors with intermediate differentiation, and pineoblastomas [3, 6–8]. Pineocytomas are well-circumscribed masses that compress surrounding structures. They are composed of mature-appearing cells arranged in sheets or illdeﬁned irregular lobules. Fibrillary processes with club-like endings project from these cells, which take up silver stains. Pineocytomatous rosettes are formed around collections of these ﬁbrillary processes. Pineoblastomas are grossly invasive and microscopically highly cellular with small, mitotically active, poorly differentiated cells with scant cytoplasm arranged in pattern-less sheets similar to other primitive neuroectodermal tumors (PNETs) and small blue cell tumors. They resemble retinoblastomas and medulloblastomas. Flexner-Wintersteiner rosette formation is a prominent feature of pineoblastomas. These are rosettes of cells forming around small lumens.

371

372

s.e. schild

Pineoblastomas occurring in patients with bilateral retinoblastomas are called trilateral retinoblastomas. Symptoms from pineal masses are related to the anatomy of the region. They often obstruct the sylvian aqueduct causing hydrocephalus and resulting in increased intracranial pressure, which leads to headache, nausea, vomiting, cognitive dysfunction, and incontinence. Additionally, they can cause pressure on the superior colliculus (midbrain) causing Parinaud syndrome or dorsal midbrain syndrome with a triad of signs consisting of vertical gaze palsy, light-near dissociation of the pupils, and convergence retraction nystagmus. A major scientiﬁc advance took place in 1973: computed tomography (CT) was ﬁrst used in major U.S. medical centers. This tool made the diagnosis and localization of the pineal tumors easier, leading to resurgence of surgical intervention with greater safety. Further improvements took place with the introduction of magnetic resonance imaging (MRI), which aided in localization and the staging of the entire central nervous system (CNS). Prior to 1973, the operative risks of biopsy or resection were substantial with mortality rates as high as 50% [9–11]. At that time, many experts recommended operative intervention be reserved for CSF shunting in patients with hydrocephalus and for the treatment of tumors that progressed after radiotherapy. Conservative therapy, including CSF shunting and radiotherapy, resulted in a generally favorable 5-year survival rate of approximately 70% [9–11]. However, therapy can only be customized to the speciﬁc tumor type if the exact tumor histology is known. Additionally, obtaining histologic conﬁrmation, prior to therapy, allows one to avoid irradiating benign lesions such as cysts [12, 13]. Computers are used to integrate data obtained from CT and MRI for surgical and radiotherapy treatment planning. Stereotactic techniques and operating microscopes have increased surgical and radiotherapy precision, reducing the risks of operative intervention. Cerebrospinal ﬂuid diversion procedures are used to manage hydrocephalus. Corticosteroids are used to decrease edema caused by the tumor or intervention. Collectively, these technological advances have dramatically decreased

the morbidity and mortality previously associated with biopsy or resection of pineal region tumors [14]. We evaluated a multi-institutional cohort of 135 patients with histologically veriﬁed pineal region tumors [6]. Thirty-ﬁve females and 100 males with ages ranging from 3 days to 77 years and a median age of 17 years were included in this study. These tumors are more common in young patients and especially in males. This series was not pure in that primary sites included the pineal gland in 81 patients, the suprasellar region in 31 patients, and the cerebrum in 23 patients. Gross total resections were performed in 26 patients, subtotal resections were performed in 49 patients, and biopsies were performed in 60 patients. There were two groups of neoplasms included in this series: pineal parenchymal tumors and germ cell tumors. The pineal parenchymal tumors included 15 pineoblastomas (PB), 2 mixed pineal parenchymal tumors (mixed PPT), 4 pineal parenchymal tumors with intermediate differentiation (PPTID), and 9 pineocytomas. The germ cell tumors included 48 germinomas, 26 mixed germ cell tumors, 11 mature teratomas, 9 immature teratomas, 6 malignant teratomas (with carcinomatous or sarcomatous components), 2 yolk sac tumors, and 3 choriocarcinomas. At the time of diagnosis, 7 (5%) of the 135 patients were found to have evidence of spinal seeding (Table 35-1). Spinal seeding most commonly occurred with pineoblastomas. One hundred patients underwent radiotherapy (Table 35-2). Regions treated included the craniospinal axis in 35 patients, the whole brain in 26 patients, and the partial brain in the remaining 39 patients. Doses delivered to the primary tumor bed ranged from 450 to 6480 cGy (median dose, 4132 cGy) in 150- to 200-cGy fractions. Eleven patients treated during the earlier years of this study received total doses of less than 3000 cGy. When treated, the whole brain received 2000 to 5040 cGy and the spine received 2000 to 3800 cGy. Chemotherapy was administered to 35 patients as a component of initial therapy (Table 35-2). These patients received various combinations of the following agents: BCNU, CCNU, cyclophosphamide, prednisone, vincristine, vinblastine, cisplatin, etoposide,

TABLE 35-1. Probability of spinal seeding at diagnosis according to tumor type. Histologic types 1. 2. 3. 4. 5. 6.

Mature and immature teratoma Mixed NGGCT Other NGGCT Germinoma Pineocytoma Pineoblastoma, PPTID, mixed PPT

Total

No. with spine seeding/total no. of patients (%) 0/20 (0%) 1/26 (4%)* 0/11 (0%) 2/48 (4%)† 0/9 (0%) 4/21 (19%)‡ 7/135 (5%)

*This patient had a mixed germ cell tumor, composed of immature teratoma and germinoma, and had positive CSF cytology. †Of these two patients with evidence of seeding at diagnosis, one had a positive CSF cytology and the other had radiographic evidence of spinal metastases. ‡Of the four patients with evidence of seeding at diagnosis, two had pineoblastomas (one with clinical evidence of spinal cord compression and one with radiographic evidence of spinal seeding) and two had pineal parenchymal tumors with intermediate differentiation (one with positive CSF cytology and one with radiographic evidence of spinal seeding). Note: All seven patients with evidence of spinal seeding at diagnosis received craniospinal irradiation and of these only one (with a pineoblastoma) had a subsequent treatment failure at the primary site and the spine.

35.

pineal tumors: fractionated radiation therapy perspective

373

TABLE 35-2. Summary of therapy.* Therapy Histologic type

RT

No RT

Chemotherapy

No chemotherapy

1. 2. 3. 4. 5. 6. 7.

2 2 19 5 48 6 18

5 7 6 4 0 1 0

1 0 17 3 8 0 6

6 9 8 6 40 7 12

Mature teratoma Immature teratoma Mixed NGGCT Other pure NGGCT Germinoma Pineocytoma PB, PPTID, mixed PPT

Surgical procedure Histologic type

1. 2. 3. 4. 5. 6. 7.

Mature teratoma Immature teratoma Mixed NGGCT Other pure NGGCT Germinoma Pineocytoma PB, PPTID, mixed PPT

Biopsy

Subtotal resection

Gross total resection

0 3 3 4 29 3 9

3 2 12 3 16 4 6

4 4 10 2 3 0 3

RT, radiotherapy; NGGCT, nongerminomatous germ cell tumor; PB, pineoblastoma; PPTID, pineal parenchymal tumor with intermediate differentiation; mixed PPT, mixed pineal parenchymal tumor. *Excludes 12 of the 135 patients who died postoperatively during the early years of the study.

dactinomycin, bleomycin, chlorambucil, thiotepa, procarbazine, and melphalan. Patients were followed for 0.25 to 37.3 years or until death (median follow-up, 5.3 years). The survival rate for the entire group of patients was 62% at 5 years. Survival was worse for patients treated during the early years of this study. Patients diagnosed before 1973 had a 5-year survival rate of 34% compared with 66% for those patients diagnosed more recently (p = 0.0002). Twelve of the patients who were operated on between 1936 and 1950 died in the immediate postoperative period. These patients were excluded from the remainder of the analysis because their deaths were more likely a result of operative complications than of progressive disease. Tumor histology was evaluated for its relationship to survival. The 5-year patient survival rate was 86% for those with mature teratomas, 86% with pineocytomas, 80% with germinomas, 67% with immature teratomas, 49% with pineal parenchymal tumors other than pineocytomas (PB, PPTIDs, and mixed PPTs), 38% with mixed germ cell tumors, and 17% with the other pure NGGCTs (p = 0.0001). Age, sex, and tumor location were not associated with survival. The extent of resection was evaluated for its effect on patient survival. NGGCTs were the only tumors for which survival was associated with extent of tumor resection. The 3-year survival rate was 0 for patients having a biopsy, 36% for patients having a subtotal resection, and 73% for patients having a gross total resection (p = 0.0002). The relationship between the administration of chemotherapy (as a component of initial therapy) and survival was evaluated. Patients who received chemotherapy had a 5-year survival rate of 45% compared with 65% for those patients who did not (p = 0.37). However, the administration of chemotherapy was associated with improved survival rates in patients with NGGCTs other than mature and immature teratomas. Patients

with NGGCTs who received chemotherapy had a 3-year survival rate of 56% compared with 8% for patients who did not receive chemotherapy (p = 0.0001). Most of the patients (100 of 123) received radiotherapy. However, only four patients with either mature or immature teratomas received radiotherapy. Therefore, the effect of radiotherapy in patients with mature and immature teratomas could not be assessed. Patients with NGGCTs, other than mature and immature teratomas, who received radiotherapy had a 3-year survival rate of 46% compared with 11% for patients who received no radiotherapy (p = 0.0089). There was an association between the dose of radiation administered and survival in patients with NGGCTs other than mature and immature teratomas. Patients who received ≤5000 cGy had a 3-year survival rate of 21% compared with 55% for patients receiving higher doses (p = 0.02). The dose of radiation administered correlated with survival rates in patients with PPTs. Patients with PPTs receiving doses of ≤5000 cGy had a 3-year survival rate of 56%, and patients who received higher doses had a 3-year survival rate of 94% (p = 0.03). There was also an association between the dose of radiation administered and survival in patients with germinomas. Those patients with germinomas receiving doses of ≤4400 cGy had a 5-year survival rate of 70% compared with 92% for patients who received higher doses (p = 0.04). There was no association between patient survival and radiation ﬁeld arrangement (partial-brain irradiation, whole-brain irradiation, or craniospinal axis irradiation). Spinal seeding was the predominant pattern of distant failure. The outcome for patients with evidence of spinal seeding at the time of diagnosis is summarized in the footnotes of Table 35-1. The risk of spinal failure relative to both tumor type and radiotherapy ﬁeld arrangement for patients without evidence of spinal seeding at diagnosis is shown in Table 35-3. Patients with pineal parenchymal tumors (other than pineocytomas) and

374

s.e. schild TABLE 35-3. Probability of spinal failure in patients without evidence of spinal seeding at diagnosis. Histologic types

1. 2. 3. 4. 5. 6.

No. with spine failure/total no. of patients (%)*

Mature and immature teratoma Mixed NGGCT† Other NGGCT† Germinoma Pineocytoma Pineoblastoma, PPTID, mixed PPT

0/16 (0%) 1/24 (4%) 3/9 (33%) 8/46 (17%) 0/7 (0%) 8/14 (57%)

Total

20/116 (17%)

Probability of spinal failure according to the radiotherapy ﬁelds in patients without evidence of spinal seeding at diagnosis‡ (including only seeding malignancies) (%) Histologic types

1. 2. 3.

NGGCT§ Germinoma PPT¶

Total

PB

WB

CSPRT

2/4 (50%) 7/25 (28%) 2/3 (66%)

0/8 (0%) 1/11 (9%) 2/5 (40%)

1/11 (9%) 0/10 (0%) 4/6 (66%)

3/23 (13%) 8/46 (17%) 8/14 (57%)

Total

11/32 (34%)

3/24 (13%)

5/27 (19%)

19/83 (23%)

PB, partial-brain ﬁelds, WB, whole-brain ﬁelds; CSPRT, craniospinal irradiation; NGGCT, nongerminomatous germ cell tumor; PPT, pineal parenchymal tumor; RT, radiotherapy. *Some of these patients received no RT. †Of the NGGCT patients with spinal failure, three of these patients had teratomas with malignant transformation and one had a yolk sac tumor. Not all of the patients with NGGCTs received RT. ‡This included only patients with potentially seeding tumors (NGGCTs excluding mature and immature teratomas, pineal parenchymal tumors excluding pineocytomas, and germinomas) who received RT. One patient with a NGGCT received no RT and later had a spinal failure. §NGGCTs excluding mature and immature teratomas. ¶Excluding pineocytomas.

germ cell tumors (other than mature and immature teratomas) had the greatest risk of spinal failure. Only one patient developed distant metastases outside the central nervous system. This individual had a mixed germ cell tumor with a component of malignant teratoma. After placement of a ventriculo-peritoneal shunt for the treatment of tumor-induced hydrocephalus, he developed peritoneal metastases and died.

Conclusion Pineal region tumors are rare and include a large variety of lesions. Many studies included patients whose tumor types were not histologically veriﬁed because of the risks associated with operative intervention [9, 10, 15–17]. As a result, these series may have included patients with benign processes such as cysts and vascular lesions. Prognosis is dependent upon tumor histology. This ﬁnding has been reported by many investigators [18–31]. Patients with pineocytomas, germinomas, and mature teratomas have the most favorable survival rates. Jennings et al. found that survival was dependent upon the extent of tumor, which emphasizes the importance of careful staging [24]. The staging workup should include a careful history and physical examination, CT or MRI of the brain, MRI of the spine, spinal ﬂuid cytology, complete blood cell count, chemistry panel, and tumor marker studies (AFP and β-HCG from both the CSF and serum). Caution should be exercised when obtaining CSF from patients with increased intracranial pressure, because herniation of the brain may occur if pressure is rapidly reduced.

The treatment of pineal and other primary CNS germ cell tumors requires a precise knowledge of tumor histology. Obtaining histologic diagnosis is important because optimal therapy for the various tumor types differs signiﬁcantly. Improvements in surgical technique and postoperative care have decreased the morbidity once associated with resection or biopsy of these tumors. Popovic and Kelly reported on 34 patients with pineal lesions who underwent 66 stereotactic procedures at the Mayo Clinic (including 37 biopsies and 10 resections) [14]. Diagnostic tissue was obtained in 33 of the 34 patients. No mortality or permanent morbidity was observed. Because the survival of patients with NGGCTs appears dependent upon the extent of resection, total resection should be attempted in all cases where it can be safely performed. Obtaining tissue for diagnosis is recommended in all patients with pineal region tumors whenever possible. Platinum-based multiagent chemotherapy has dramatically improved the outcome in patients with NGGCTs [9]. Our series revealed that a signiﬁcant increase in survival was associated with the administration of chemotherapy to patients with NGGCTs other than mature and immature teratomas. Many other authors have recommended the use of platinum-based multiagent chemotherapy for the treatment of NGGCTs of the brain [19, 20, 24, 25, 31–34]. Radiotherapy should be administered to all patients with pineal region tumors other than possibly the mature and immature teratomas for which little data exist regarding radiotherapy. Careful consideration should be given to the dose, because survival in patients with NGGCTs (other than mature and immature teratomas), PPTs, and germinomas is dependent upon the dose of radiation administered to the primary tumor.

35.

375

pineal tumors: fractionated radiation therapy perspective

The extent and dose of radiation should be tailored to the particular tumor type as outlined below. Patients with pineal parenchymal tumors (other than pineocytomas) have a high risk of spinal failure, and craniospinal axis radiotherapy would appear reasonable. However, patients with no spinal metastasis can be given whole-brain irradiation because the risk of spinal seeding appears no greater after whole-brain irradiation than after craniospinal irradiation (Table 35-3). The primary tumor should receive 5040 to 5400 cGy, and spinal metastases should receive 5040 cGy in 180cGy fractions. Prophylactic therapy can include the delivery of 3000 to 4500 cGy to uninvolved high-risk areas. Patients with pineocytomas can be irradiated to the local tumor alone. None of the pure pineocytomas we studied metastasized. It appears that the radiation dose (>5000 cGy versus ≤5000 cGy) delivered to the primary tumor was the most signiﬁcant factor affecting survival in patients with pineal parenchymal tumors. Patients with germinomas but no spinal seeding have a moderately high risk of spinal failure. However, these patients can be treated with either craniospinal or whole-brain irradiation. No patient who received craniospinal irradiation had a spine failure and only 1 of 11 (9%) patients treated with wholebrain irradiation failed (Table 35-3). There appears to be a substantially greater risk (28%) of spinal failure in those patients who received only partial-brain irradiation. The recommended dose for areas of gross disease is 4500 to 5040 cGy in 180-cGy fractions. Prophylactic therapy can include the delivery of 2000 to 3600 cGy to high-risk areas. If spinal seeding is present, craniospinal irradiation is recommended. Patients with NGGCTs (other than immature and mature teratomas) but no spinal seeding can be treated with either whole-brain or craniospinal irradiation. Spinal failure occurred in two of the four patients treated with partial brain ﬁelds. There was a low risk of spinal failure with either whole-brain or craniospinal irradiation (Table 35-3). Preferably, wholebrain irradiation should be prescribed. These patients also experienced improved survival after the administration of chemotherapy. Newer chemotherapy regimens, including cisplatin, etoposide, and bleomycin, are myelosuppressive. The potential beneﬁts of prophylactic spinal irradiation must be carefully weighed against potential toxicity in patients requiring plati-

num-based multiagent chemotherapy. Craniospinal irradiation and systemic chemotherapy should be considered for patients who present with spinal seeding. The primary tumor should receive 5040 to 5400 cGy and spinal metastases should receive 5040 cGy in 180-cGy fractions. Prophylactic therapy can include the delivery of 2400 to 3600 cGy to uninvolved high-risk areas. Immature and mature teratomas should be resected, if this can be safely achieved. We can not make deﬁnitive recommendations regarding the efﬁcacy of radiotherapy or chemotherapy for patients with immature and mature teratomas, because of the small number of patients in our series treated with these modalities. If progression occurs, further surgical intervention, radiotherapy, and/or chemotherapy should be considered. Table 35-4 lists the relative advantages of stereotactic radiosurgery (SRS) versus fractionated external beam irradiation. Stereotactic radiotherapy and radiosurgery have potential roles in the treatment of pineal tumors. These techniques would have the greatest potential as a single treatment modality in nonseeding tumors. The advantages for radiosurgery in these cases would be as a shortened course compared with fractionated external beam irradiation. Additionally, the dose to surrounding structures is small, which is especially advantageous in pediatric patients. Pineocytomas and teratomas (both mature and immature) represent tumors in which this approach would be most applicable as they do not generally seed the CSF. There are examples in the literature where pineocytomas have been reported to seed, but this may represent confusion regarding the classiﬁcation of PPTs, which should be subdivided into four groups [6–8]. The risks of radiosurgery are likely substantially less than resection. Regarding PPTs, the University of Pittsburgh experience with Gamma Knife radiosurgery would support these contentions. Hasegawa et al. reported the 5-year survival of 10 patients with pineocytomas was 90% utilizing Gamma Knife radiosurgery [35]. In contrast, the results of Gamma Knife treatment with pineoblastomas was poor with death occurring in 4 of 6 patients. Complications including upward gaze paralysis occurred in 2 of 16 (12.5%) and was likely due to the dose delivered to the superior colliculus, which lies in direct contact with pineal tumors. Pineocytomas are round, circumscribed tumors that are generally not large at

TABLE 35-4. Advantages of stereotactic radiosurgery (SRS) versus fractionated external beam radiotherapy (FRT).

Effectively address those at ↑ risk of CSF seeding Most effectively minimizes RT dose to surrounding normal brain More radiobiologically effective in the treatment of small, nonseeding tumors Marrow suppression in patients that in some cases receive chemotherapy May be an effective alternative to resection in histologies in which resection is believed to improve survival Requires less patient time Proven long-term efﬁcacy: Germinomas Nongerminomas Pineocytomas Other PPTs

Advantage

Disadvantage

FRT

SRS

SRS SRS SRS SRS

FRT FRT FRT FRT

SRS FRT FRT FRT SRS FRT

FRT SRS SRS SRS FRT SRS

Note: Ultimately, one might consider the following paradigm for future investigation in seeding pineal tumors (all those other than pineocytomas, mature teratomas, and immature teratomas): Biopsy ﬁrst (and shunting if needed), followed by FRT followed by a specialized boost (SRS, IMRT, heavy ions . . .), and chemotherapy when appropriate. In nonseeding tumors, one might consider: Biopsy ﬁrst (and shunting if needed) followed by a specialized boost technique. The major rationale for this approach is the opinion that complete resection of a pineal tumor incurs a substantially greater overall risk to the patient than does SRS. Additionally, the specialized boost would provide greater local control than FRT alone in unresected tumors.

376

s.e. schild

diagnosis because they cause symptoms as soon as pressure is exerted on the sylvian aqueduct or superior colliculus. However, pineoblastomas, PPTs with intermediate differentiation, and mixed PPTs are often diffuse, invasive, and seeding. These should be treated with either whole-brain radiation therapy followed by boost to the tumor in cases without seeding or craniospinal radiation therapy in cases with seeding followed by a boost to areas of gross disease. The options for boost include external beam irradiation, stereotactic radiotherapy, or radiosurgery. Which boost technique option offers the best outcome is unknown. For all other pineal tumors with seeding potential, radiosurgery or stereotactic radiotherapy could be used as a boost to gross disease after other therapy. Hasegawa et al. also reported on a small series of four patients treated for nongerminomatous germ cell tumors of the brain [36]. These patients received a Gamma Knife boost after other therapies, and three of four were alive without disease at last follow-up. Although this is a small series, it does show potential and should lead to further investigation. The specialized radiation techniques might also be considered as salvage therapeutic maneuvers when other treatments have failed.

References 1. Jemal A, Tiwari R, Murray T. et al. Cancer statistics, 2004. CA Cancer J Clin 2004; 54:8–29. 2. Zulch KJ. Biologie und pathologie der hirgeschwulste. In: Oliverona H, Tonnis J, eds. Handbuch der neurochirurgie. Berlin: Springer-Verlag, 1965:348. 3. Schild SE, Buskirk SJ, Frick LM, et al. Pineal parenchymal tumors: clinical, pathologic, and therapeutic aspects. Cancer 1993; 72: 870–880. 4. Schild SE, Haddock M, Scheithauer B, et al. Non-germinomatous germ cell tumors of the brain. Int J Radiat Oncol Biol Phys 1996; 36:557–563. 5. Haddock M, Schild SE, Scheithauer B, Schomberg P. Radiotherapeutic management of primary CNS germinomomas. Int J Radiat Oncol Biol Phys 1997; 38:915–923. 6. Schild SE, Scheithauer B, Haddock M, et al. Histologically conﬁrmed pineal tumors and other germ cell tumors of the brain: treatment and outcome. Cancer 1996; 78:2564–2571. 7. Burger P, Scheithauer B. Pineal tumors. In: Tumors of the Central Nervous System: Atlas of Tumor Pathology. Bethesda, MD: Armed Forces Institute of Pathology, 1994:227–237. 8. Mena H, Nakazato Y, Jouvet A, Scheithauer BW. Pineal prenchymal tmours. In: Kleihues P, Cavanee WK, eds. World Health Organization Classiﬁcation of Tumours. Pathology and Genetics. Tumors of the Nervous System. Lyon, France: IARC Press, 2000: 115–121. 9. Donat JF, Okazaki H, Gomez MR, et al. Pineal tumors: a 53 year experience. Arch Neurol 1978; 35:736–740. 10. Abay EO, Laws ER, Grado GL, et al. Pineal tumors in children and adolescents: treatment by CSF shunting and radiotherapy. J Neurosurg 1981; 55:889–985. 11. Marsh WR, Laws ER. Shunting and irradiation of pineal tumors. Clin Neurosurg 1985; 32:384–396. 12. Fain JS, Tomlinson FH, Scheithauer BW, et al. Symptomatic glial cysts of the pineal gland. J Neurosurg 1994; 80:454–460. 13. Fleege MA, Miller GM, Fletcher GP, et al. Benign glial cysts of the pineal gland: unusual imaging characteristics with histologic correlation Am Neuroradiol 1994; 15:161–166. 14. Popovic EA, Kelly PJ. Stereotactic procedures for lesions of the pineal region. Mayo Clin Proc 1993; 68:965–970.

15. Linstadt D, Wara WM, Edwards MS, et al. Radiotherapy of primary intracranial germinomas: the case against routine craniospinal irradiation. Int J Radiat Oncol Biol Phys 1988; 15:291–297. 16. Rich TA, Cassady JR, Strand RD, Winston KR. Radiation therapy for pineal and suprasellar germ cell tumors. Cancer 1985; 55:932–940. 17. Wara WM, Jenkin RD, Evans A, et al. Tumors of the pineal and suprasellar region: Childrens Cancer Study Group treatment results 1960–1975: a report from Children’s Cancer Study Group. Cancer 1979; 43:698–701. 18. Bjornsson J, Scheithauer BW, Okazaki H, Leech RW. Intracranial germ cell tumors: pathobiological and immunohistochemical aspects of 70 cases. J Neuropathol Exp Neurol 1985; 44:32–46. 19. Bruce JN, Fetell MR, Stein BM. Incidence of spinal metastases in patients with malignant pineal region tumors: avoidance of prophylactic spinal irradiation (meeting abstract). J Neurosurg 1990; 72:354A. 20. Bruce J, Stein B, Balmaceda C, et al. Management of pineal region germ cell tumors: results and long-term follow-up (meeting abstract). Proc Ann Meet Am Soc Clin Oncol 1994; 13:A505. 21. Dearnaley DP, A’Hern RP, Whittaker S, Bloom HJ. Pineal and CNS germ cell tumors: Royal Marsden Hospital experience 1962– 1987. Int J Radiat Oncol Biol Phy 1990; 18:773–781. 22. Fuller BG, Kapp DS, Cox R. Radiation therapy of pineal region tumors: 25 new cases and a review of 208 previously reported cases. Int J Radiat Oncol Biol Phy 1994; 28:229–245. 23. Hoffman HJ, Otsubo H, Hendrick EB, et al. Intracranial germ-cell tumors in children. J Neurosurg 1991; 74:545–551. 24. Jennings MT, Gelman R, Hochberg F. Intracranial germ-cell tumors: natural history and pathogenesis. J Neurosurg 1985; 63:155–167. 25. Linggood RM, Chapman PH. Pineal tumors. J Neurooncol 1992; 12:85–91. 26. Mineura K, Sasajima T, Sakamoto T, Kowada M. Results of the treatment of intracranial germ cell tumors. Gan No Rinsho - Jpn J Can Clin 1990; 36:2399–2403. 27. Sano K, Matsutani M, Seto T. So-called intracranial germ cell tumors: personal experiences and a theory of their pathogenesis. Neurol Res 1989; 11:118–126. 28. Sano K. Pineal region tumors: problems in pathology and treatment. Clin Neurosurg 1983; 30:59–91. 29. Takakura K. Intracranial germ cell tumors. Clin Neurosurg 1985; 32:429–444. 30. Wolden SL, Wara WM, Larson DA, et al. Radiation therapy for primary intracranial germ-cell tumors. Int J Radiat Oncol Biol Phys 1995; 32:943–949. 31. Yoshida J, Sugita K, Kobayashi T, et al. Prognosis of intracranial germ cell tumors: effectiveness of chemotherapy with cis-platin and etoposide (CDDP and VP-16). Acta Neurochir 1993; 120: 111–117. 32. Gobel U, Bamberg M, Calaminus G, et al. Improved prognosis of intracranial germ cell tumors by intensiﬁed therapy: results of the MAKEI 89 therapy protocol. Klin Padiatr 1993; 205:217–224. 33. Gobel U, Bamberg M, Engert J, et al. Treatment of non-testicular germ cell tumors in children and adolescents with bep and vip: initial results of the Makei 89 therapy study. Klin Padiatr 1991; 203:236–245. 34. Herrmann HD, Westphal M, Winkler K, et al. Treatment of nongerminomatous germ-cell tumors of the pineal region. Neurosurgery 1994; 34:524–529; discussion 529. 35. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. The role of radiosurgery for the treatment of pineal parenchymal tumors. Neurosurgery 2002; 51:880–889. 36. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. Stereotactic radiosurgery for CNS nongerminomatous germ cell tumors. Report of four cases. Pediatr Neurosurg 2003; 38:329–333.

3 6

Pineal Region Tumors: Chemotherapy Perspective Barry Meisenberg and Lavanya Yarlagadda

Introduction Tumors involving the pineal gland are uncommon, accounting for 0.4% to 11% of intracranial tumors in children, and 0.4% to 1% of such tumors in adults [1]. Tumors arising within the pineal gland are commonly of three principal histologic types, each with its own subtypes. It is most appropriate to consider treatment approaches based on the histologic type, not anatomic origin. The three main types of pineal tumors that will be considered are germ cell tumors (GCTs), pineal parenchymal tumors (PPTs), and astrocytomas. Other tumors that are occasionally seen in this area are meningiomas, gangliogliomas, ependymomas, lipomas, and metastatic cancer. Pineal cysts, vascular malformations and aneurysms are benign lesions that can occur in the pineal region, occasionally posing a diagnostic challenge. A biopsy of pineal tumors is strongly recommended prior to the initiation of therapy, because optimal treatment is dependent upon the histology. In addition, staging evaluation is also dependent upon the histology, with cerebrospinal ﬂuid (CSF) sampling and magnetic resonance imaging (MRI) of the spine required in some diagnoses, but not others. The different approaches to obtain biopsy specimens include stereotactic computer-guided biopsy, transventricular endoscopic biopsy, third ventriculostomy, and direct open surgery. The speciﬁc biopsy approach and extent of surgical resection required for each histologic type of tumor is controversial and beyond the scope of this review.

Germ Cell Tumors Primary GCTs account for approximately half of the pineal tumors. Central nervous system GCTs mostly arise in the midline: 80% or more arise in structures near the third ventricle, with the region of the pineal gland being the most common site of origin. Multifocal disease is common. Central nervous system (CNS) GCTs are seen most commonly in childhood up through young adulthood. Histologically, CNS GCTs are analogous to systemic GCTs. These tumors can be divided into germinomas (also called seminomas) and nongerminomatous

GCTs, which include teratoma, embryonal cell carcinoma, endodermal sinus tumor, and choriocarcinoma. Pure germinomas are not associated with signiﬁcant tumor-marker elevation in serum but may have lower levels of CSF β-human chorionic gonadotropin (β-HCG). Nongerminoma GCTs usually have elevated serum and/or CSF α-fetoprotein (AFP) and/or β-HCG levels, which are useful to follow the course of the disease. However, the absence of elevated tumor markers does not rule out either type of GCT.

Pineal Parenchymal Tumors Pineal parenchymal tumors originate from the pineocyte, a cell with photosensory and neuroendocrine functions. Pineal parenchymal tumors range from tumors composed of mature elements to ones consisting of primitive, immature cells that behave aggressively. They are classiﬁed as pineocytoma pineal parenchymal tumor of intermediate differentiation and pineoblastoma. Pineocytomas constitute approximately 45% of all pineal parenchymal tumors. They occur throughout life but are most frequently seen in adults 25 to 35 years of age without any sex predilection. Pineoblastomas also constitute approximately 45% of all pineal parenchymal tumors. They usually occur in the ﬁrst two decades of life. The cells are arranged in a diffuse pattern with occasional formation of Homer-Wright or FlexnerWintersteiner rosettes. The latter indicates retinoblastic differentiation. Pineal parenchymal tumors of intermediate differentiation constitute approximately 10% of all pineal parenchymal tumors. They are monomorphous tumors characterized by moderately high cellularity, mild nuclear atypia, occasional mitosis, and the absence of large pineocytomatous rosettes. They occur at all ages with peak incidence in adulthood.

Astrocytic Tumors Astrocytomas are the most common form of malignant tumor in adults. They can occur anywhere in the CNS, including the

377

378

b. meisenberg and l. yarlagadda

spinal cord and brain stem, but only rarely are found in the pineal region. Astrocytomas are graded based on pathologic criteria, including cellularity, pleomorphism, atypia, and necrosis. The prognosis and treatment approach to astrocytomas varies by histologic grade.

Treatment Because of the rarity of the various types of pineal area tumors, recommendations for treatment are often based on limited experience with few prospective trials.

Germ Cell Tumors GCTs, both germinomas and nongerminomas, are sensitive to radiation and chemotherapy. The role of surgical resection beyond diagnostic biopsy is therefore controversial, as no prospective studies address this issue. A retrospective review, however, suggested no survival advantage to complete resection if subsequent radiation or chemotherapy was administered [2]. In a single-institution, retrospective study of 29 patients with pineal region GCTs, partial resection and complete resection were obtained in ﬁve and eight patients, respectively, and outcomes were compared with 16 who underwent biopsy only. Postoperative neurologic improvement was seen in only two patients, but transient complications occurred in six. One patient had slight hemiparesis as a long-term surgical complication. In this small sample, no difference in long-term outcomes was seen with different types of resection. Germinomas are particularly sensitive to radiation, and 5year survival rate with radiation to the tumor bed and craniospinal axis is greater than 90%. Craniospinal irradiation is usually included because of a 10% to 20% risk of meningeal relapse, though the need for this has been disputed. Because the diagnosis of GCT is usually made in children who may suffer from late sequelae of radiation, attempts have been made to use lower than traditional doses of radiation without sacriﬁcing durable remissions. This is consistent with the approach in peripheral GCT, where seminomas can be cured with much lower doses than traditionally utilized in CNS germinomas. In one retrospective study of patients treated over three decades, the dose of radiation was lowered from the traditional dose of 50 to 60 Gy to a dose based on the estimated tumor volume according to the schedule shown in Table 36-1. Some patients also received craniospinal irradiation. The 10-year relapse-free survival was 91% for 31 patients treated according to the schedule in Table 36-1 [3]. Such a treatment regimen, if adapted, would suggest a potential role for debulking surgery, which would then reduce the radiation dose required.

TABLE 36-1. Radiation dose by tumor volume. >4 cm 2.5 cm to 4.0 cm <2.5 cm Resected masses

50 Gy 45 Gy 40 Gy 36 Gy

Bamberg et al. [4] conducted a multicenter prospective trial within the German Society of Pediatric Oncology and Hematology. Two different doses of radiation were used in two different phases of the study. A pilot study of 11 patients was performed in which patients received 36 Gy (1.8 to 2.0 Gy fractions) to the craniospinal axis with a boost of 14 Gy to the tumor region. A subsequent cohort of 49 patients received 30 Gy to the craniospinal axis on a boost of 15 Gy. The fractionated dose was reduced to 1.5 Gy to reduce the risk for subsequent morbidities. Complete resection had been accomplished in only 6.6% of patients. The results were highly favorable. All patients entered a complete remission after therapy documented by computed tomography (CT) or MRI. Relapse-free survival was 88 ± 4.7% for the 49 patients treated with 30 Gy/15 Gy. Of the 60 total patients, only 5 patients relapsed (all treated with 30 Gy/15 Gy), but 3 of these patients were salvaged with chemotherapy, enjoying subsequent long-term disease-free survival. Overall 5-year survival for the patients treated with 30 Gy/15 Gy was 92 ± 4.6%. The intra-CNS relapse rate was only 1.6% consistent with other studies using higher-dose radiation. Long-term sequelae of treatment were rare with few overt deﬁcits in intellectual functioning, based on scholastic and occupational achievement. However, detailed neuropsychological testing was not done. The most serious neurologic problems, hemiparesis and blindness, were consequences of the tumor mass or surgery. Because of the potential for growth retardation, endocrine dysfunction, and other postradiation syndromes in young patients, attempts to eliminate the radiation have involved the use of chemotherapy as frontline therapy to take the place of craniospinal irradiation. Several studies utilize chemotherapy with radiation therapy limited to the tumor bed. A French cooperative group, The Société Française d’Oncologie Pediatrique (SFOP) [5], enrolled 57 patients in a prospective protocol in which carboplatin 600 mg/m2 day 1, etoposide 150 mg/m2 days 1 to 3 was alternated with ifosphamide 1.8 g/m2 days 1 to 5 plus etoposide 150 mg/m2 dose days 1 to 3. A total of four cycles of chemotherapy was given 21 days apart. Radiation therapy was given to the initial tumor volume in 1.8-Gy fractions up to 40 Gy. Signiﬁcant hematologic toxicity developed with the chemotherapy, but only one patient was forced to stop chemotherapy early. Complete remission occurred in 18 of 38 assessable patients. Twenty patients had partial responses but had residual radiographic abnormalities, which did not necessarily represent active cancer. The 3-year overall survival was 98%, and 3-year event-free survival (no relapse, no progression, and no death from toxicity) was 96.4% (conﬁdence interval, 86.2% to 99.1%). These results, albeit with short follow-up, indicate that chemotherapy may be able to substitute for craniospinal radiation, a particular concern in very young children most at risk for neuropsychological dysfunction or growth retardation. The four relapses were all intracerebral (not within the spinal axis) and raised the question of optimum dose and treatment volume for the brain. In an attempt to eliminate radiation therapy altogether from the treatment regimen, an international CNS GCT study group initiated a study of carboplatin, etoposide, and bleomycin

36.

pineal region tumors: chemotherapy perspective

chemotherapy with radiation therapy only initiated if patients failed to obtain a complete remission. In this study, Balmaceda et al. [6] prospectively studied 71 patients (45 with germinoma and 26 with nongerminoma) CNS GCTs. These patients received four cycles of carboplatin, etoposide, and bleomycin chemotherapy, a regimen adapted from systemic germ cell tumor experience, though one that is considered inferior to the same regimen utilizing cisplatin instead of carboplatin. In this study, patients who obtained a complete response went on to receive two additional cycles of the same chemotherapy. Patients who failed to obtain a complete response received irradiation in addition to two cycles of salvage chemotherapy, usually high-dose cyclophosphamide 65 mg/kg. In some cases where residual masses were seen after chemotherapy, a surgical resection was performed. Thirty-two percent of patients achieved a partial remission and 13% of patients were treatment failures with no response or progression during the chemotherapy. Fifty-seven percent of the patients achieved a complete remission radiographically after the fourth cycle of chemotherapy. Sixteen of 29 patients who achieved less than complete remission had intensiﬁed chemotherapy without irradiation, and 10 of them subsequently achieved a complete remission. Overall, the use of chemotherapy with salvage cyclophosphamide intensiﬁcation and the resection of the residual masses induced 78% of patients into a complete remission without the use of radiation. Patients with nongerminoma and germinoma GCTs were equally likely to respond to chemotherapy. With a median follow-up of 35 months, the overall survival among patients with germinoma was 84% (95% CI, 73% to 95%) and for nongerminoma GCTs it was 62% (95% CI, 43% to 80%). However, the relapse-free survival was disappointing, with approximately 50% of patients with each pathologic subtype experiencing relapse or progression. Twenty-eight of 55 patients in complete remission without irradiation subsequently developed tumor recurrence at a median time of 18 months from diagnosis. Eleven recurrences were local, two were local plus ventricular, three were spinal leptomeningeal, and 12 were ventricular alone. Fortunately, many of those patients could be salvaged with radiation therapy or additional chemotherapy with or without surgical resection. From this experience, it appears that aggressive chemotherapy alone can induce durable complete remissions only in approximately 50% of patients, sparing those patients from the toxicity of radiation therapy, although detailed studies on neurocognitive outcomes are yet to be done. Given the superior results with radiation for germinomas, the use of chemotherapy without radiation cannot be considered routine for these patients though the risks must be individualized. There may be circumstances in which an attempt at chemotherapy-induced complete remission is warranted in these patients at high risk for radiation complications. Even those entering complete remission should be followed carefully. It is worth remembering that chemotherapy is not without chronic side effects as well, including ototoxicity, infertility, and even secondary leukemias related to etoposide use. The role of chemotherapy is much less controversial in patients with nongerminoma GCTs. This is principally because nongerminoma GCTs are only partially radiosensitive. Overall survival rates are 10% to 50% for nongerminoma GCTs when

379

treated with radiation therapy alone [7]. The widespread disparity in reported outcomes has to do with the inclusion in some studies of patients with metastatic disease to the craniospinal axis and the mixture of different histologies that appear to have a different prognosis when treated with radiation therapy. Radiation doses have also differed in various trials. Despite this variation, it is apparent that radiation alone is not adequate therapy for non-GCTs. The inclusion of chemotherapy was initiated as soon as responses began to be seen in the treatment of systemic GCTs, but for advanced germinomas and nongerminomas. In the international CNS GCT study discussed previously [6], Balmaceda used chemotherapy to spare patients the toxicity of radiation therapy. Chemotherapy alone was disappointing, although there was a signiﬁcant response rate of approximately 50% indicating activity of the assets. Combining radiation therapy with chemotherapy became a more appropriate strategy. Robertson et al. [8] reported on a study (combined chemotherapy and radiation therapy) of 18 patients conducted over 8 years in multiple institutions. Fourteen of the 18 patients have had histologic documentation of nongerminoma GCTs, and four patients had this presumptive diagnosis based on elevated serum and CSF α-fetoprotein and/or β-HCG. These patients were treated with three or four cycles of upfront chemotherapy consisting of cisplatin and etoposide. Nine of the 12 evaluable patients had a response, including 5 complete remissions and 4 partial remissions. One patient progressed during chemotherapy. Seventeen of the 18 patients then went on to receive radiation therapy per protocol, consisting of involved-ﬁeld with either craniospinal or whole brain. Twelve of the patients received postradiation chemotherapy consisting of vinblastine, bleomycin, etoposide, and carboplatin. The 4-year event-free survival was 67%, and the overall survival was 74% at 4 years, a substantial improvement from radiation-alone outcomes. The frustratingly high relapse rate and the initial refractoriness of nongerminoma GCTs to chemotherapy has led to the use of high-dose chemotherapy and autologous stem cell transplant for either relapse disease or to consolidate partially responsive disease. This strategy is based on experience with systemic GCTs, as well as the increasing experience with other high-risk primary brain tumors, such as medulloblastoma. One of the largest such experiences was published by Modak et al. [9], a group based in New York and one of the leaders in the ﬁeld of high-dose therapy and stem cell rescue for brain tumors. This was a study of 21 patients with CNS GCTs who had either relapsed or progressed, despite receiving chemotherapy and/or radiation therapy. These patients all received high-dose thiotepa combined with other chemotherapy agents, chieﬂy etoposide, often with carboplatin. Prior to high-dose chemotherapy, patients were treated with aggressive chemotherapy to shrink the tumors and also resection after chemotherapy when feasible. Posttransplant radiation therapy was added focally or to the entire neural axis whenever feasible to consolidate the highdose chemotherapy response. Four of the 12 patients with nongerminoma GCTs had disease-free survival. At the time of publication, 4 patients were in remission 24 to 55 months out from the initial therapy; 7 of the 12 patients relapsed and died of progressive disease. One patient was alive with relapse disease at the time of the publication, 51 months out from the time of treatment.

380

b. meisenberg and l. yarlagadda

Nine patients with recurrent germinomas were also treated. Seven of the 9 survived disease-free, with a median of 48 months (6 to 87 months after high-dose chemotherapy). These results, albeit based on very limited clinical experience, suggest that long-term remissions can be obtained even in a group of patients highly refractory to chemotherapy. It remains for clinical trials to then take these results and move them into the high-risk primary setting, which would include almost all patients with nongerminoma GCTs of the CNS. It remains to be demonstrated that superior outcomes could be obtained with acceptable toxicity, but this is a very important avenue for further investigation.

Pineal Parenchymal Tumors Experience with chemotherapy is scarce among the class of pineal parenchymal tumors called pineocytomas. These are the most common of the pineal parenchymal tumors, and most patients can expect to do well for prolonged periods of time with either surgery with radiation given up front or at time of relapse. Chemotherapy experience is anecdotal only, with many different chemotherapeutic agents utilized.

Pineoblastomas and Pineal Parenchymal Tumors of Intermediate Differentiation Pineoblastomas (PBs) and pineal parenchymal tumors of intermediate differentiation (PPTIDs) are very rare in adults. To date, no prospective studies analyzing different chemotherapy regimens are available. A recent multicenter review [10] indicated the poorer outcomes of these entities compared with pineocytomas. Primary or secondary failure (relapse) rats were about 50% at 10 years, with the worse outcomes seen among patients who present with disseminated disease or who have large-volume residual disease. PB histology compared with PPTID also had worse outcomes. Use of chemotherapy in these tumors is also anecdotal. Chemotherapy regimens similar to those used for medulloblastoma (which pineoblastoma resembles histologically) have been adopted. These agents include cisplatin, carboplatin, etoposide, cyclophosphamide, and vincristine [11]. A study of six patients receiving high-dose cyclophosphamide (4 g/kg) reported no complete responses when given once a month for 4 months [12]. A children’s cancer group prospective study [13] demonstrated the role of combined radiation and chemotherapy for children with primitive neuroectodermal tumors of the pineal region, a histologic category that is often used interchangeably with pineoblastoma. In this study, eight infants less than 18 months old were selected for chemotherapy with an aggressive 8-drugs-in-1-day cocktail. The eight drugs included methylprednisone, vincristine, CCNU, procarbazine, hydroxyurea, cisplatin, cytarabine, and cyclophosphamide. No radiation was given to avoid the toxicities of radiation in these very young patients. Seventeen older patients were treated with craniospinal radiation therapy and randomized to receive either the 8-drugs-in1-day regimen as above or a simpler regimen consisting of vincristine, CCNU, and prednisone.

Unfortunately, all the infants with disease suffered relapse at a median of 4 months from the initiation of therapy. Of the 17 patients receiving combined radiation therapy with either of the chemotherapy regimens, the progression-free survival at 3 years was 61 ± 13%. The authors further noted that persistent residual radiographic mass in the pineal region did not necessarily indicate eventual progression as these masses did persist for many years but only a minority developed progressive disease. The 3-year progression-free survival of 61% indicates the need for radiation therapy in addition to chemotherapy. However, other experience indicates that there continues to be a risk of relapse that persists beyond 3 years, thus more effective therapies will be required. As for GCTs, attempts to intensify the chemotherapy have utilized high-dose chemotherapy and autologous stem cell rescue. Clinically signiﬁcant dose escalation requires autologous bone marrow or autologous peripheral blood hematologic progenitor cells to rescue the bone marrow after the delivery of high-dose chemotherapy. Treatment with high-dose chemotherapy (HDC) and autologous stem-cell rescue (ASCR) in patients with newly diagnosed PB was reported by Gururangan et al. [14]. Twelve patients (six children and six adults) underwent biopsy or surgical resection of primary tumor at the time of diagnosis. Four patients had evidence of metastatic disease limited to the brain and spine, and no patients had extraneural spread. All patients received induction chemotherapy with one of three regimens. Regimen A consisted of cyclophosphamide 2 g m−2 day−1 intravenously (IV) for 2 days with mesna rescue and hydration, given every 4 weeks for four cycles. Regimen B consisted of vincristine 0.05 mg/kg IV on day 1 and then weekly for 2 weeks, cisplatin 2.5 mg/kg IV on day 1, cyclophosphamide 65 mg/kg IV on day 2, etoposide 4 mg/kg IV on days 2 and 3, every 4 weeks for four to six cycles. Regimen C included carboplatin AUC 5 on day 1, cyclophosphamide 1.2 mg/m2 IV on day 2, and etoposide 100 mg/m2 IV on days 2 and 3, every 4 weeks for a total of four cycles, followed by oral etoposide 50 mg m−2 day−1 for 21 days every 4 weeks for four cycles. Bone marrow stem cell harvest and/or peripheral blood stem cells were harvested after the ﬁrst or second cycle of induction chemotherapy after using granulocyte colony-stimulating factor (G-CSF). All patients, except two infants, received craniospinal irradiation at a median dose of 36 Gy given at 1.5 to 1.8 Gy/fraction and pineal region boost at a median dose of 59.4 Gy at 1.8 to 2 Gy/fraction. Eleven patients received HDC with cyclophosphamide 50 mg/kg daily for 4 days followed by melphalan 60 mg m−2 day−1 for 3 days or busulfan 1 mg/kg every 6 hours for 16 total doses over 4 days followed by melphalan 60 mg m−2 day−1 for 3 days. No treatment related deaths were reported. Five of 12 patients underwent complete resection of the primary tumor at diagnosis. After induction chemotherapy, seven patients were assessable for response: one patient achieved complete response, ﬁve patients achieved partial responses, and one patient had stable disease. After HDC and ASCR, nine patients are alive and disease-free at a median follow-up of 62 months (range, 28 to 125 months) from diagnosis including three of four patients with metastatic disease at

36.

pineal region tumors: chemotherapy perspective

presentation and two infants who did not receive radiotherapy. Two of eight patients with localized disease and one of four patients with metastatic disease developed progression and died. This outcome in this small sample of patients using highdose chemotherapy after induction chemotherapy and radiation are superior to less-aggressive regimens but historic controls are inappropriate for major clinical decisions. These results therefore need to be conﬁrmed by broader experience and perhaps randomized trials.

Gliomas of the Pineal Region Glial tumors are the most common form of brain tumor in adults and account for approximately one quarter of all pineal region tumors. Gliomas however are also seen in children and young adults. Various grading systems have been developed for gliomas, most relying on degree of cellularity, nuclear atypia, mitotic activity, microvascular proliferation, and presence of necrosis to grade a tumor. Among the gliomas, histology appears to be the most important prognostic factor with glioblastoma multiforme having substantially worse outcomes than anaplastic astrocytoma. According to an Radiation Therapy Oncology Group (RTOG) recursive partition analysis, among patients over 50, histology was the most signiﬁcant prognostic factor. Other signiﬁcant prognostic factors include the Karnofsky performance status and the results of mental status examination. Size of the original tumor and location do not appear to inﬂuence survival substantially [15]. At times, no biopsy specimen is available, and the radiographic criteria are often very useful for predicting the dominant grade and therefore underlying behavior. Tumors that are diagnosed with needle biopsies only may show behavior that is different than predicted by the small sample obtained by the needle, which represents sampling error from scarce pathologic material. A comprehensive review of the treatment of each of the grades of glioma tumors as well as variants such as anaplastic oligodendroglioma is beyond the scope of this chapter. There are signiﬁcant controversies involved in the treatment of low-grade gliomas including the extent and even necessity of surgery, the utility of upfront radiation therapy at the time of diagnosis, and the role of chemotherapy in patients with either standard-risk or high-risk low-grade brain tumor. Clinical trials have been testing whether temozolomide chemotherapy plus radiation can improve survival compared with radiation alone for high-risk brain tumors. Results of recently completed studies are awaited. Confounding clinical trial design in low-grade brain tumors is the fact that survival is often prolonged even when no initial therapy is rendered after diagnosis. A handful of trials have evaluated the use of adjuvant chemotherapy, either PCV (procarbazine, CCNU, vincristine) or BCNU (carmustine), for grade 3 gliomas. Trial results are somewhat conﬂicting, which may be a result of the fact that patients with other grade of histologies may have crept into the study population, particularly in studies where biopsy-only patients were entered. The Medical Research Council of the United Kingdom published its randomized trial of 113 patients with anaplastic astrocytoma who received either radiation therapy alone or six cycles of PCV chemotherapy after radia-

381

tion therapy [16]. In this study, a small survival beneﬁt was seen among patients receiving the combined therapy, but it did not achieve statistical signiﬁcance. Thus, adjuvant PCV or BCNU is frequently used for patients with anaplastic astrocytoma although the precise survival beneﬁt is hard to quantitate. In general, patients tolerate this chemotherapy well although cytopenias require careful monitoring of patients, and there is a small risk of pulmonary ﬁbrosis associated with nitrosoureas. Glioblastoma multiforme is the most common brain tumor in adults, and the prognosis remains dismal with median survival approximately 9 to 12 months. Recently, chemotherapy has been shown to have a beneﬁt when combined with radiation therapy in two randomized prospective trials. Two randomized clinical trials of the oral alkylating agent temozolomide showed substantial improvement in progressionfree and overall survival compared with radiation therapy alone in patients with glioblastoma multiforme. Stupp et al. reporting for the European Organisation for Research and Treatment of Cancer (EORTC) [17] conducted a trial involving 573 patients from 85 different brain tumor centers. The median age was 56 years and all patients had received at least a biopsy to qualify for the study. All patients received radiotherapy consisting of fractionated focal irradiation at a dose of 2 Gy per fraction given 5 days per week for 6 weeks for a total dose of 60 Gy. Patients were randomized prior to the initiation of radiation therapy to either receive or not receive temozolomide at a dose of 75 mg/m2 daily during radiation therapy (including Saturday and Sunday when radiation therapy was not given). The results demonstrated a signiﬁcant improvement in overall survival associated with receiving temozolomide in addition to radiation therapy. With a median follow-up of 28 months, the hazard ratio for death in the combined treatment arm was 0.63 (95% conﬁdence interval, 0.52 to 0.75; p < 0.001 by the log rank test). The median survival for patients on the combined modality treatment was 14.6 months compared with 12.1 months with radiotherapy alone. The 2-year survival rate was 26.5% in the combined modality therapy group compared with just 10.4% in the radiotherapy-alone group. Very similar data was obtained in a randomized phase II study of temozolomide and radiotherapy in a study performed in Greece. One hundred thirty patients were randomly assigned to receive either temozolomide 75 mg/m2 per day concomitantly with radiation therapy and followed by six cycles of temozolomide or radiation therapy alone. Once again, radiation was delivered in 2-Gy fractions 5 days a week for a total of 60 Gy [18]. In this study, median time to progression in the combined modality arm was 10.8 months compared with 5.2 months for radiation therapy alone (p < 0.0001). The median overall survival was 13.4 months in the combined modality arm versus 7.7 months for the radiation therapy–alone group (p < 0.0001). In both studies other than for hematologic toxicity, the therapy was otherwise generally well tolerated. Taken together, the two studies demonstrate a signiﬁcant role for oral temozolomide in improving time to progression and overall survival in this disease. The chemotherapy is, in general, well tolerated and associated with few signiﬁcant side effects if monitored carefully with frequent patient evaluations and hematologic assessments. The EORTC trial also allowed collaborative correlative investigation into the role of the MGMT (methylguanine

382

b. meisenberg and l. yarlagadda

methyltransferase) promoter, an enzyme that repairs DNA damage caused by alkylating agents and that can be induced by radiation therapy. This study found that in those patients who had samples taken, methylation of the MGMT promoter, which results in silencing of the MGMT gene, was associated with a survival beneﬁt in patients who received combined modality therapy [19]. Indeed, investigators found that almost all of the survival beneﬁt from adding temozolomide to the radiotherapy accrued to the group of patients whose promoters were silenced by methylation. In other words, patients whose MGMT gene promoter was still active did not beneﬁt from the combined modality therapy to nearly the same extent. Unfortunately, the MGMT promotion technique is not a clinically available test currently. Therefore, current practice is to prescribe oral temozolomide to all patients with a new diagnosis of glioblastoma multiforme. The work however does give hope that eventually this test or others like it could be useful in predicting response to treatment or even selecting agents for treatment as new agents become available.

Conclusion

5.

6.

7. 8.

9.

10.

11. 12.

The pineal region tumors are heterogeneous with several different tumors and grades of tumor occurring clinically. Ideal treatment for each of these depends upon exact histologic type so biopsy is strongly recommended when feasible. When technically not feasible, additional information may be obtained from a parentive tumor on MRI or from studies of the cerebral spinal ﬂuid. Chemotherapy may play an important role in several of the common tumor types including germ cell tumors, pineoblastomas, and high-grade gliomas most notably. Brain tumors are ideally optimally managed by multimodality groups with interest and expertise in brain tumors including radiation therapists, neurosurgeons, medical oncologists, neuroradiologists, among other relevant specialists.

14.

References

17.

1. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. The role of radiosurgery for the treatment of pineal parenchymal tumors. Neurosurgery 2002; 51:880–889. 2. Sawamura Y, De Tribolet N, Ishi N, Abe H. Management of primary intracranial germinomas: diagnostic surgery or radical resection? J Neurosurg 1997; 87:262–265. 3. Shibamoto Y, Takahashi M, Abe M. Reduction of the radiation dose for intracranial germinoma: a prospective study. Br J Cancer 1994; 70:984–989. 4. Bamberg M, Kortmann R-D, Calaminus G, et al. Radiation therapy for intracranial germinoma: results of the German Coop-

13.

15. 16.

18.

19.

erative Prospective Trials MAKEI 83/86/89. J Clin Oncol 1999; 17:2585–2592. Bouffet E, Baranzelli MC, Patte C, et al. on behalf of the Société Française d’Oncologie Pediatrique. Combined treatment modality for intracranial germinomas results of a multicentre SFOP experience. Br J Cancer 1999; 79:1199–1204. Balmaceda C, Heller G, Rosenblum M, et al. for the First International Central Nervous System Germ Cell Tumor Study. Chemotherapy without irradiation—a novel approach for newly diagnosed CNS germ cell tumors: results of an international cooperative trial. J Clin Oncol 1996; 14:2908–2915. Packer RJ, Cohen BH, Coney K. Intracranial germ cell tumors. Oncologist 2005; 5:312–320. Robertson PL, DaRosso RC, Allen JC. Improved prognosis of intracranial non-germinoma germ cell tumors with multimodality therapy. J Neurooncol 1997; 32:71–80. Modak S, Gardner S, Dunkel IJ, et al. Thiotepa-based high-dose chemotherapy with autologous stem-cell rescue in patients with recurrent or progressive CNS germ cell tumors. J Clin Oncol 2004; 22:1934–1943. Lutterbach J, Fauchon F, Schild SF, et al. Malignant pineal parenchyma tumors in adult patients: patterns of care and prognostic factors. Neurosurgery 2002; 51:44–56. Mazzola CA, Pollack IF. Medulloblastoma. Curr Treat Options Neurol 2003; 5:189–198. Ashley DM, Longee D, Tien R, et al. Treatment of patients with pineoblastoma with high dose cyclophosphamide. Med Pediatr Oncol 1996; 26:387–392. Jakacki FI, Zeltzer PM, Boyett JM, et al. Survival and prognostic factors following radiation and/or chemotherapy for primitive neuroectodermal tumors of the pineal region in infants and children: a report of the Childrens Cancer Group. J Clin Oncol 1995; 13:1377–1383. Gururangan S, McLaughlin C, Quinn J, et al. High-dose chemotherapy with autologous stem-cell rescue in children and adults with newly diagnosed pineoblastomas. J Clin Oncol 2003; 27: 2187–2191. Buckner JA. Factors inﬂuencing survival in high-grade gliomas. Semin Oncol 2003; 30(Suppl 19):10–14. Medical Research Council. Randomized trial of procarbazine, lomustine and vincristine in the adjuvant treatment of high grade astrocytoma: a Medical Research Council Trial. J Clin Oncol 2001; 19:509–518. Stupp R, Mason WP, van den Bent M, et al. for the European Organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups and the National Cancer Institute of Canada Clinical Trials Group. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005; 352: 987–996. Athanassiou H, Synodinou M, Maragoudakis E, et al. Randomized phase II study of temozolomide and radiotherapy compared with radiotherapy alone in newly diagnosed glioblastoma multiforme. J Clin Oncol 2005; 23:2372–2377. Hegi ME, Diserens A, Gorlia T, et al. MGMT gene silencing and beneﬁt from temozolomide in glioblastoma. N Engl J Med 2005; 352:997–1003.

3 7

Skull Base Tumors Stefanie Milker-Zabel, Young Kwok, and Jürgen Debus

Introduction The greatest challenge to neurosurgeons is tumors located at the skull base. In many patients, complete resection of skull base tumors is often not possible without a high risk of neurologic deﬁcits and morbidity. Therefore, highly precise radiation techniques are required. For small tumor volumes, radiosurgery is often used. In case of large tumor volumes, fractionated stereotactic radiation therapy combines the advantage of spatial precision of radiosurgery and the radiobiological advantage of fractionation with better sparing of normal tissue [1]. Over the past three decades, the use of radiosurgery in the treatment of skull base tumors has signiﬁcantly increased. Radiosurgery is performed for benign tumors of the skull base as well as for patients with malignant skull base tumors as a palliative treatment. Symptom relief is common; especially, facial pain related to the tumors can be substantially improved. The goals of radiosurgery are long-term prevention of tumor growth, maintenance of patient function, and prevention of new neurologic deﬁcits or adverse radiation side effects. Stereotactic radiosurgery is a very precise form of radiation therapy used in localized and small-volume lesions of less than 3 cm to apply a high single dose to the tumor. It is a noninvasive treatment modality and can be offered to inoperable patients due to several comorbidities, after subtotal resection, or to patients with poor prognosis as an alternative treatment to surgical intervention. Stereotactic radiosurgery can be done by using external stereotactic techniques like Gamma Knife, charged particles such as protons, or linac-based stereotactic radiosurgery such as CyberKnife. A stereotactic minimally invasive head ring that is attached to the patient’s skull by four pins is used for stereotactic linac-based radiosurgery. Gamma Knife and linac-based radiosurgery use multiple arcs to treat the tumor volume. Very few complications occur after stereotactic radiosurgery, such as perifocal edema, delayed intratumoral hemorrhage, or radionecrosis requiring neurosurgical intervention [2, 3]. Skull base tumors are relatively rare. Approximately 0.1% of all intracranial tumors are chondrosarcomas and also approximately 0.1% of all intracranial tumors are chordomas. For benign tumors of the skull base such as glomus tumors, local control rates of 90% to 100% have been reported. Local control

rates for chordomas and chondrosarcomas range from 50% to 70%. Skull base tumors are divided into those that occur throughout the skull base and those that are unique to a particular area. They can also be divided into benign and malignant lesions. Tumors may derive from the bone, paranasal sinuses, nasopharynx, inner ear, dura, cranial nerves, and brain. These lesions may be primary tumors, which may invade local structures, or may be metastatic disease. Unfortunately, even benign lesions may cause progressive and unrelenting problems if located in an area that does not allow complete resection and growth cannot be controlled with any therapy.

Anterior Cranial Fossa (Angioﬁbroma and Esthesioneuroblastoma) Angioﬁbroma Disease Pathophysiology of Angioﬁbromas Juvenile angioﬁbroma is one of the most common benign nasal tumors in adolescent males with an average age of 14 to 18 years. The etiology has not been determined. The tumor originates from the broad area of the posterolateral wall of the nasal cavity and usually arises from cells in the sphenopalatine foramen [4]. Fetal histology conﬁrms large areas of endothelial tissues in this region. They often act in a malignant manner by eroding into the surrounding sinuses developing an aggressive growth pattern. Intracranial extension is noted in 10% to 20% of cases. Staging classiﬁcation for nasopharyngeal angioﬁbroma is shown in Table 37-1 [5]. Typical clinical symptoms are frequent epistaxis, nasal obstructions, and rhinorrhea. Because of obstructions of the eustachian tube, a conductive hearing loss may result. By eroding into the cranial fossa, diplopia may occur as well as symptomatic pressure of the chiasm and optic nerves.

Treatment Options Treatment of choice in patients with primary and recurrent juvenile nasopharyngeal angioﬁbroma is surgical resection as

383

384

s. milker-zabel et al.

TABLE 37-1. Staging of nasopharyngeal angioﬁbroma. Stage

Extension

I II III

Conﬁned to the nasopharynx Extension to the nasal cavity and/or sphenoid sinus Extension into the antrum, ethmoid sinus, pterygomaxillary and infratemporal fossae, orbit, and/or cheek Intracranial extension

IV

Source: Chandler JR, Goulding R, Moskowitz L, Quencer RM. Nasopharyngeal angioﬁbromas: staging and management. Ann Otol Rhinol Laryngol 1984; 93:322–329.

exclusive treatment of early-stage tumor when gross total resection can be achieved. In case of incomplete resection, or in advanced-stage lesions, a combination of surgery followed by radiotherapy is indicated due to the high recurrence rate in these patients. Advanced-stage disease with cranial base involvement and intracranial extension often allows only a subtotal resection of the tumor. Complications in advanced-stage angioﬁbromas after surgery include intraoperative blood loss requiring transfusions, neuralgia, hearing loss, and ophthalmoplegia [6]. Preoperative embolization reduces operative bleeding, morbidity, and recurrence rate. Salvage with embolizations of polyvinyl alcohol has been described [7]. Surgical contraindications include involvement of unresectable intracranial contents. For treatment decision, planning computed tomography (CT) scans, magnetic resonance imaging (MRI), and angiography are required. External beam irradiation has been shown to be a useful adjunct to therapy in patients with unresectable recurrent disease. Gamma Knife and linac-based radiosurgery with a dose of 20 Gy to the tumor margin (55% isodose line) is an effective way to deliver high-dose radiation to incompletely resected angioﬁbromas. Equally effective is the use of fractionated conformal radiation therapy with total doses of 30 Gy to 46 Gy to decrease the risk of late effects such as cranial nerve deﬁcits, bone and soft tissue necrosis, and second malignancy, especially in children. Another treatment technique described in the literature is the use of intensity-modulated radiation therapy (IMRT) in three cases [8]. The applied tumor dose varied from 34 Gy to 45 Gy. In all three cases, a reduction of tumor size occurred without signiﬁcant toxicity. Possible side effects after radiation therapy techniques also included an inhibition of facial bone growth.

Patient Selection and Treatment Planning Details for Radiosurgery Gamma Knife radiosurgery with a single dose of 20 Gy to the tumor margin (e.g., 55% isodose line) is an effective way to deliver high-dose radiation to incompletely resected, small-size angioﬁbromas while minimizing exposure to surrounding normal structures [9].

Experiences and Review of the Literature After primary resection alone of extracranial angioﬁbromas, cure rates of nearly 100% can be achieved compared with results in patients with intracranial lesions where the cure rates are approximately 70%. Angioﬁbromas are slow-growing and late-

responding tissues. Therefore, a radiobiological advantage to radiosurgery may be given. Radiosurgery is regarded as a reasonable strategy in small-volume and localized angioﬁbromas. Dare et al. [9] reported on two patients treated with surgical resection followed by Gamma Knife radiosurgery for juvenile nasopharyngeal angioﬁbromas. The residual tumor volume was 3.0 cm3 and 4.7 cm3, respectively. The dose delivered to the tumor margin was 20 Gy. Both patients tolerated the therapy without complications. There was stable tumor size on MRI during the 2-year follow-up period. McGahan and co-workers [10] reported on 10 patients with intracranial extension of nasopharyngeal angioﬁbroma treated with surgical resection and radiotherapy. One patient underwent postradiation surgery for recurrence of disease. In one of the largest series of patients treated with primary radiation therapy, local control rates of 78% in 45 patients are reported. A moderate dose of 30 Gy to 35 Gy was applied [11]. An update of these reports was done by Cummings and coworkers in 1984 [12]. Fifty-ﬁve patients were treated with radiation therapy, 42 treated primary to and 13 after surgical intervention in case of recurrence of disease. Local control rate was 80% after initial therapy. Reddy et al. [13] reported on 15 patients with advanced-stage juvenile nasopharyngeal angioﬁbroma treated with radiation therapy. Six patients were treated with primary radiation alone and nine patients after surgical resection, with a total radiation dose of 20 Gy to 30 Gy. They reported a complete response rate of 85% after a minimum follow-up of 2.5 years. In a series of three patients with intracranial extension, Wiatrak et al. [14] reported on symptomatic control of disease after radiation doses between 36.6 Gy and 50.4 Gy and a follow-up period of 1.7 to 5 years. Because of the potential of long-term side effects, especially the induction of secondary malignancies, the use of conventional radiotherapy in the management of juvenile nasopharyngeal angioﬁbromas has been criticized. Cases of malignant transformation, induction of basal cell carcinomas, or thyroid carcinomas are described [12, 13]. In the literature, only a few reports exist on the use of IMRT in patients with juvenile nasopharyngeal angioﬁbromas. The applied tumor dose varied from 34 Gy to 45 Gy. In all three cases described, a reduction of tumor size was seen without signiﬁcant toxicity [8]. Possible side effects after radiation therapy techniques may include secondary malignancies as mentioned as well as hypopituitarism, glaucoma, optic atrophy, or an inhibition of facial bone growth in children [12–14].

Esthesioneuroblastoma Disease Pathophysiology of Esthesioneuroblastoma Esthesioneuroblastomas are rare tumors originating from the olfactory epithelium of the upper nasal cavity [15]. They account for about 3% of all malignant neoplasms of the nasal cavity and were ﬁrst described by Berger and Luc in 1924 [16]. The sex distribution of esthesioneuroblastomas is uniform. The olfactory nerves perforate the groove in the ethmoid bone in the cribriform plate and continue into the subarachnoid spaces. Therefore, a high incidence of intracranial extension results. These tumors of the frontal skull base are still associated with

37.

TABLE 37-2. Staging of esthesioneuroblastoma according to Kadish et al. Stage

Characteristic

A

Conﬁned to the nasal cavity

B

Conﬁned to the nasal cavity and one or more paranasal sinuses Extending beyond the nasal cavity or paranasal sinuses, including involvement of the orbit, base of skull or intracranial cavity, cervical lymph nodes or distant metastatic sides

C

385

skull base tumors

Source: Kadish S, Goodman M, Wang CC. Olfactory neuroblastoma: a clinical analysis of 17 cases. Cancer 1976; 37:1572–1576.

high rates of tumor recurrence and mortality. This highly dedifferentiated tumor occurs in all periods of life with a bimodal peak at the second and sixth decades. The two most common clinical signs of esthesioneuroblastoma constitute unilateral nasal obstruction and epistaxis. Other clinical symptoms include headache, swelling of the cheek, burred vision, and dental pain. Esthesioneuroblastoma can metastasize to regional lymph nodes, lung, or bones. According to the WHO classiﬁcation system, the terms olfactory neuroblastoma and olfactory neurogenic tumors are used. The Kadish staging classiﬁcation is shown in Table 37-2 [17].

Rationale for Treatment and Alternatives Because of the rarity of esthesioneuroblastoma and its wide variety of clinical behavior, there is no deﬁnitive consensus regarding the optimal treatment. For small, low-grade tumors conﬁned to the ethmoids, surgery alone appears to be an adequate method. Patients with locally advanced disease or highgrade tumors should receive aggressive treatment with combined modalities such as surgery, radiation therapy, and chemotherapy [18]. Most authors recommend en bloc resection, combined with radiation therapy [19–21]. Local failure rates of 44% in low-grade and 60% in high-grade tumors and metastatic rates of 25% in low-grade and 47% in high-grade tumors are described [22]. A signiﬁcantly lower rate with overall 5- and 10-year survival rates of 81% and 54.5% in patients with response to neoadjuvant radiotherapy combined with chemotherapy are reported [23].

Patient Selection and Treatment Planning Details for Radiosurgery The Austrian group described the combined treatment of endoscopic surgery and radiosurgery for olfactory neuroblastoma

[24]. Median marginal doses range from 15 Gy to 34 Gy by a marginal isodose between 45% and 85%. The maximum tumor volume treated with radiosurgery was approximately 20 cm3. Regarding the radiosensitivity of cranial nerves, the optic tract and the chiasm are most sensitive structures and seem to tolerate not more than 9 Gy; similar to the trigeminal nerve and the facial nerve. The nerves to the ocular muscles seem to tolerate a higher dose [25, 26].

Experiences and Review of the Literature Chao and co-workers [27] reported on 25 patients with esthesioneuroblastoma treated at the Mallinckrodt Institute of Radiology. The 5-year actuarial overall survival, disease-free survival, and local control rates were 66.3%, 56.3%, and 73%, respectively. The local control rate for the combination of surgery and radiotherapy was 87.4%, and 51.2% for radiation therapy alone. Because of the close proximity of the tumor to radiosensitive normal structures such as the optic system and brain stem, radiotherapy is a challenging method. Postoperative radiation doses of 50 Gy to 60 Gy are indicated, depending on the status of the surgical margins. Doses of 65 Gy to 70 Gy are delivered with radiation alone in case of inoperability. Stereotactic radiation therapy offers the possibility of improving dose coverage of the target volume, therefore increasing tumor doses while decreasing the risk of normal tissue complications. Walch et al. [28] reported on three patients with olfactory neuroblastoma treated with a combination of endoscopic surgery and Gamma Knife. Stereotactic radiosurgery was performed within the ﬁrst 3 months of surgery. The maximum diameter of the tumors was approximately 24.3 mm and the marginal dose to the tumor varies from 16 Gy to 34 Gy; 1 to 5 isocenters were used. Radiation-induced side effects were nasal discharge and crusts. One patient developed bilateral frontal chronic sinusitis, and in a second endoscopic operation was necessary. Unger et al. [24] described the combined treatment of endoscopic surgery and radiosurgery for esthesioneuroblastoma. The maximum tumor volume was approximately 20 cm3 with a median dose to the tumor margin of 15 Gy to 34 Gy. Median follow-up period was 58 months. Observed radiosurgical side effects were mild and transient, such as cephalea and dizziness. No changes in mental status were observed. No new pathology of the optic pathway was described during follow-up. In large and complex-shaped tumors in critical proximity to radiosensitive normal structures, IMRT is recommended [21] (Table 37-3).

TABLE 37-3. Summary table: Review of the literature for radiosurgery of esthesioneuroblastoma. Author

N

Median follow-up (range)

Median peripheral dose (range)

Local control*

Walch et al. [28] Unger et al. [24]

3 14

50 mo (39–71) 58 mo (13–128)

20 Gy (16–34) 18 Gy (15–34)

100% 78.6%†

*Crude local control rates. †Two other patients failed distantly while achieving local control. ‡All were minor and transient, all resolved within 48 hours.

Morbidity

Nasal discharge/crust, chronic sinusitis Cephalea, dizziness‡

386

s. milker-zabel et al.

Middle Cranial Fossa Tumor Tumors unique to the middle cranial fossa are often benign. These tumors include pituitary adenomas, craniopharyngioma, temporal bone tumors, cholesteatoma, and enchondroma.

Craniopharyngioma Disease Pathophysiology Craniopharyngiomas are benign tumors located at the base of the skull next to the pituitary gland. A differentiation between craniopharyngioma and pituitary can therefore sometimes be difﬁcult on CT or magnetic resonance (MR) scans. Aproximately 5% to 10% of primary brain tumors are craniopharyngiomas. They typically occur in childhood as well as in the sixth to eighth decades [29]. Histopathologically, craniopharyngiomas are benign tumors arising from squamous cell remnants of the Rathke pouch during embryogenesis at the junction of the pituitary stalk and pituitary. Craniopharyngiomas present as a suprasellar lesion, frequently partially calciﬁed and usually including an intrasellar component. These tumors are often composed of solid and cholesterol-rich cystic components. Cystic or solid components of this tumor extension may occur laterally into the middle or into the posterior cranial fossa. Symptoms relate to compression effects of the tumor due to its vicinity to pituitary gland, chiasm, optic nerves, and hypothalamic region. Locally, these tumors can produce signs and symptoms of increased pressure like headache, drowsiness, or vomiting at the time of diagnosis and are due to hydrocephalus by obstruction of the foramen Monro by tumor parts within the third ventricle in 55% to 85% of the patients [30]. Compression of the pituitary and hypothalamic region can produce antidiuretic hormone and growth hormone deﬁciency or obesity in children. Diabetes insipidus is present in approximately 10% of the patients. Patients commonly present with visual ﬁelds defects and decreased vision due to compression of the optic chiasm and optic pathways from prechiasmatic tumors. Approximately 40% to 60% of the patients show visual symptoms at presentation [31, 32].

local control rates of 70% to 83% after 10 years are reported [39, 40] and assumed to be similar to complete surgical resection of the tumor. Treatment-related toxicities after subtotal resection followed by radiotherapy include impairment of hormone function. Impairment of vision is reported for less than 10% of all patients treated with the combination of subtotal resection and irradiation compared with up to 20% after complete tumor resection [41]. Other side effects such as radionecrosis, radiation-induced malignancies, vascular morbidity, and cognitive decline occur less frequently [41, 42]. The major goal of radiotherapy treatment strategies is sparing of critical normal structures. Radiosurgery as well as intracavitary irradiation with stereotactically applied β-emitting radioisotopes are radiation therapy modalities that enable sparing of normal tissue. The cystic nature of craniopharyngioma has led to trials of intracystic applications of β-emitting radioisotopes such as yttrium-90 or phosphorus-32. The use of radiosurgery has been reported in patients with minimal residual or recurrent disease. In patients with greater target volumes and multiple cystic conﬁgured lesions, fractionated stereotactic radiotherapy may be preferred.

Patient Selection and Treatment Planning Details for Radiosurgery The target volume for craniopharyngiomas is narrowly deﬁned to the tumor volume, including solid and cystic components. In cases with cyst aspiration or subtotal resection, it is important to cover the complete cyst wall. Stereotactic radiosurgery is used to irradiate the solid, not the cystic components of the craniopharyngioma. This technique may be used for highly selected patients due to the proximity of the tumor to the optic chiasm and the brain stem. Median doses to the margin of the tumor range from 9 Gy to 16 Gy. Chung and co-workers [43] recommend a margin dose of 12 Gy to induce satisfactory tumor response. The main restriction with radiosurgery treatment is the tolerance dose of the neighboring visual pathway. The dose to the optic nerves and the chiasm has to be kept below 8 Gy with single-dose techniques to avoid severe damage to these structures. Stereotactic radiosurgery has been used to treat small residual or recurrent tumors after surgical intervention.

Rationale for Treatment and Alternatives

Experiences and Review of the Literature

The main treatment modality for craniopharyngiomas is surgery. The ﬁrst report of surgical intervention was published by Lewis in 1910 [33]. As reported in the past two decades, microsurgery allows complete tumor removal in 49% to 100% of the patients with low morbidity and operative mortality [34–36]. Ten-year progression-free survival rates after radical resection between 60% and 93% are reported [37, 38]. Treatment modalities include complete resection of the tumor with radiation therapy at the time of recurrence or subtotal resection followed by radiotherapy. The probability of complete tumor resection decreases with increasing tumor volume. Because of the proximity to critical normal structures and the relatively high association of radical surgery with a relatively high rate of visual loss and impaired hormone function requiring replacement therapy, many authors recommend a less radical surgery (partial resection, biopsy and aspiration of cystic contents) followed by radiation therapy or stereotactic radiosurgery. In the literature,

Surgery alone in the treatment of craniopharyngioma shows local tumor progression rates between 70% and 90%, occurring at 2 and 3 years. Subtotal resection followed by radiotherapy is associated with long-term control rates of approximately 80% to 95% at 5 to 20 years with low risk of long-term side effects [44, 45]. Mokry et al. [46] treated 23 patients with Gamma Knife radiosurgery for craniopharyngioma and found no relevant morbidity. Ten patients had additional therapy with intracystic bleomycin before radiosurgery. Tumor progression was observed in 5 of 23 patients. They conclude that the best results might be obtained in monocystic tumors amenable to stereotactic drainage and intracystic bleomycin treatment. Voges et al. [47] report on 62 patients with cystic craniopharyngiomas treated with stereotactically applied colloidal β-emitting radioactive sources. Chung et al. [43] reported after an average follow-up of 36 months a tumor control rate of 87% for 31 patients treated with Gamma Knife radiosurgery and a

37.

387

skull base tumors

TABLE 37-4. Summary table: Review of the literature for radiosurgery of craniopharyngioma. Author

N

Median follow-up (range)

Mean peripheral dose (range)

Local control*

Chung et al. [43] Mokry [46] Ulfarsson et al. [48] Kobayashi et al. [50]

31 23 21 98

33 mo (5–69)† 23 mo (6–57) 42 mo (6–348)† 66 mo (6–148)

12.2 Gy (9.5–16) 10.8 Gy (8–15) 30 Gy (20–50)‡ 11.5 Gy (NA)

87.2% 78.2% 36.4% 79.5%

Morbidity

Visual ﬁeld deﬁcit (1 patient) None Visual ﬁeld deﬁcit (8 patients) Visual/endocrine (6%)

*Crude local control rates. †Median values. ‡Given in maximum dose.

prescribed dose to the tumor margin from 9.5 Gy to 16 Gy. One patient developed a mildly restricted visual ﬁeld. None of the patients showed additional endocrinologic impairment or neurologic deterioration related to radiosurgery. In the study of the Swedish group, 21 patients were treated with Gamma Knife radiosurgery. They found a statistically signiﬁcant difference between tumor progression and applied dose. A higher progression rate was found in patients treated with less than 6 Gy to the margin than in patients treated with a dose higher than 6 Gy. Four of these patients developed pituitary dysfunction [48]. After fractionated stereotactically guided radiotherapy of 26 patients, an actuarial local control rate and actuarial overall survival rate of 100% after 5 years are reported. Late toxicity after fractionated stereotactic radiotherapy included impairment of hormone function in 3 of 18 patients at risk. No vision impairments or radionecrosis were reported [49]. In the literature, parenchymal injuries of the brain or second malignancies caused by radiotherapy are estimated to be less than 1% to 2% [50] (Table 37-4).

Posterior Cranial Fossa Tumor Chordomas, Chondromas, and Chondrosarcomas

low-grade hyaline chondrosarcoma. The second variant is dedifferentiated chordoma, which contains areas of typical chordoma admixed with components that resemble high-grade or poorly differentiated spindle cell sarcoma. Basisphenoidal chordoma may be difﬁcult to differentiate histologically from chondroma and chondrosarcomas and radiologically from craniopharyngioma, pineal tumor, and hypophyseal and pontine glioma. Chordomas rarely metastasize, but they often invade local structures. Prognostic factors that most inﬂuence choice of treatment are location, local tumor extension, and surgical resectability. Approximately 0.1% of all intracranial tumors are chondrosarcomas. This locally invasive tumor is a malignant variant of a benign chondroma arising from bone and is composed of cartilage. Chondrosarcomas are slowly growing tumors and rarely metastasize. Its most common site is the sphenoid bone or clivus, at the base of the skull. Chondrosarcomas are more common in males than in females. They can be classiﬁed into three grades (I to III). Lower-grade tumors are less aggressive and act clinically similar to chordomas. Historically, skull base chondroma and chondrosarcoma were often pooled together in reported series due to the rarity of these tumors; however, recent published studies have shown important differences with respect to diagnosis, treatment, and prognosis.

Disease Pathophysiology of Chordoma, Chondroma, and Chondrosarcoma

Rationale for Treatment and Alternatives

Chondromas are rare benign tumors arising at the base of the skull, especially in the area close to the pituitary gland. It is a very-slow-growing tumor and might be present for a long time before causing any symptoms. Chondromas are composed of cartilage formed by the meninges and is usually attached to the dura mater. Surgical intervention might be the treatment of primary choice because of their usually well-deﬁned margins. Chordomas are relatively rare, slow-growing, primary bone tumors arising from embryonic remnants of the notochord (chorda dorsalis) at the two extreme ends of the vertebral axis. They are most often diagnosed in the second or third decade of life and comprise less than 1% of intracranial neoplasm [51, 52]. Twenty-ﬁve percent to 40% of the chordomas occur in the spheno-occipital or skull base region. The clivus is the most common site. Chordomas are locally more aggressive with a poorer outcome compared with chondrosarcomas. Chondrosarcomas are malignant tumors composed of cartilage-producing cells encountered in the skull base. Two histologic variants of chordoma have been described. The ﬁrst is chondroid chordoma, a typical chordoma that also contains areas resembling

Treatment of choice is total surgical resection, if feasible, followed by radiation therapy. The best results in the treatment of chordomas have been obtained by complete surgical resection followed by high doses of proton irradiation [53]. Recurrences are the rule, and metastases are extremely uncommon. Complete resection of these tumors at the skull base is extremely difﬁcult and links to higher rates of mortality and morbidity. Furthermore, it is difﬁcult to treat skull base chordomas and chondrosarcomas with radiation therapy alone because of the large tumor size, the extent of inﬁltrated tissues, and because of dose limitations imposed by the sensitivity of adjacent critical normal tissues like brain stem and often the visual pathway [54]. Frequently used doses are 55 Gy to 66 Gy, median 60 Gy. A clear dose-response relationship with an improved outcome after tumor doses of greater than 60 Gy in chordomas are shown [55]. Because of the slow proliferative nature of chordomas, high linear energy transfer may be useful. To yield good results with respect to local tumor control, high-dose radiation therapy with photons alone or combined with photons and particle beam boost are reported [56, 57]. Heavily charged particles

388

s. milker-zabel et al.

such as protons or carbon ions are more suitable for residual tumors because of their ﬁnite range in tissues. This radiation technique offers an excellent chance of cure with acceptable radiation-induced toxicity [58–60]. Another conformal radiation therapy technique avoiding nearby critical structures is the use of IMRT. Brachytherapy can be used for recurrent tumors of the skull base when more aggressive surgical exposure is offered.

Patient Selection and Treatment Planning Details In the literature, lesions with a diameter of less than 30 mm were treated with 17 Gy to 20 Gy to the tumor margin [54, 61, 62]. Better results concerning local control rates are reported using fractionated carbon-ion radiotherapy after surgical intervention [59, 60].

Experiences and Review of the Literature A sequelae of patients treated with high irradiation doses or charged particles include bone or brain necrosis as well as vision impairment or radiation injury to the brain stem [63]. Kondziolka et al. [54] reported on four patients with chordoma and two patients with chondrosarcoma treated with radiosurgery. All tumors were less than 30 mm in diameter and were treated with 20 Gy to the tumor margin. During a mean follow-up of 22 months (range, 8 to 36 months), no progression of the treated tumor was found. Three patients showed improvement of preexisting neurologic deﬁcits. The other three patients remained in stable neurologic condition. Serial follow-up imaging studies showed tumor volume reduction in two patients, whereas the other four patients showed stable tumor size. One patient showed tumor progression outside the irradiated tumor volume. In the study of Feigl et al. [61], 13 patients with chordoma and chondrosarcoma were treated with Gamma Knife radiosurgery. Before starting radiosurgery, all patients had maximal tumor resection. The mean treated tumor volume was 9.7 cm3 with a range from 1.4 cm3 to 20.3 cm3. The mean treatment dose was 17 Gy and the mean isodose was 52%. After a mean follow-up of 17 months, only one recurrence of disease was seen at the margin of the radiation ﬁeld. Pamir et al. [62] reported on 26 skull base chordomas with a mean follow-up period of 48.5 months receiving multimodality treatment with various combinations of conventional surgery, skull base surgical techniques, and Gamma Knife surgery. The mean follow-up after Gamma Knife treatment was 23.3 months. They recommend Gamma Knife radiosurgery immediately after initial surgical intervention if the tumor volume is less than 30 cm3. This suggested treatment algorithm of Pamir and colleagues is shown in Figure 37-1. The results of postoperative proton radiation therapy for skull base chordomas are superior to those achieved with conventional fractionated photon irradiation techniques. Proton radiotherapy allows signiﬁcant improvement in conformal treatment (i.e., radiosurgery and fractionated stereotactic radiotherapy). Initial clinical results of carbon-ion irradiation in patients with chordomas (17 patients) and chondrosarcomas (10 patients) were published by Debus et al. in 2000 [59]. They underwent fractionated carbon-ion irradiation with a median total dose of 60 GyE (GyE = Gy × relative biologic effectiveness). Local tumor control was achieved in 16 of 17 patients

Chordomas

Skull base approaches

Surgical intervention

No residual tumors

Follow-up

Residual tumors

<30 CC

>30 CC

Radiosurgery Consider Radiosurgery as Re-operation soon as recurrence detected FIGURE 37-1. Management algorithm for skull base chordomas according to Pamir et al. [62].

with chordomas. Partial tumor response was seen in 3 of 17 patients with chordomas. The 1-year local control rate was 94%. No severe toxicity was observed. Schulz-Ertner et al. [60] reported on 24 patients with base of skull chordomas and 13 patients with chondrosarcomas treated with carbon-ion radiotherapy. Tumor conformal application of beams was achieved by intensity-controlled raster scanning with pulse-to-pulse energy variation. Median applied tumor dose was 60 GyE. Mean follow-up period of the patients was 13 months. Local tumor control after 1 and 2 years was 96% and 90%, respectively. Progression-free survival was 100% for chondrosarcomas and 83% for chordomas at 2 years (Table 37-5) [64, 65].

Glomus Jugulare Tumors Disease Pathophysiology of Glomus Jugulare Tumors Glomus jugulare tumors are rare neoplasms of the skull base, slow-growing, hypervascular, and histologically benign lesions that usually occur in the temporal bone. They comprise 0.6% of all tumors. Glomus tumors are closely associated with the sympathetic system. They represent chemoreceptors in a diffuse neuroendocrine system. Because of their wide invasion of the temporal bone and compression of adjacent tissues, many local problems can be caused. Glomus jugulare tumors arise from the paraganglia of the chemoreceptor system. Therefore, frequently used synonyms are chemodectoma and paraganglioma. Typical clinical symptoms are gradual hearing loss, unilateral pulsatile tinnitus, or imbalance. Only 2% to 5% malignant transformation rates are reported in the literature with metastatic spread to lung, liver, and bone [66].

Rationale for Treatment and Alternatives The treatment options for glomus jugulare tumors include surgery, radiotherapy, and endovascular occlusion of feeding vessels either in combination or alone [67–69]. Embolization

37.

389

skull base tumors

TABLE 37-5. Summary table: Review of the literature for radiosurgery of chordoma, chondroma, and chondrosarcoma. Author

Krishnan et al. [64] Chordoma Chondrosarcoma Feigl et al. [61] Chordoma Chondrosarcoma Pamir et al. [62] Chordoma Chang et al. [65]§ Chordoma

N

Median follow-up (range)

Median peripheral dose (range)

4.8 yr (0.8–11.4)

15 Gy (10–20)

Localcontrol*

25 4 17 mo (6–36)†

Morbidity

32%‡ 100%

34% (all with combined EBRT)

33% 100% 29%

Cranial nerve deﬁcits, headaches, diplopia

17 Gy (14–18)+

3 10 23.3 mo (NA)†

NA

4 yr (1–9)†

19.4 Gy (18–24)†

NA

7 10

80%

None

*Crude local control rates. †Mean values. ‡Actuarial local control at 5 years. §Only 4 patients received single fractions (5 patients = linac radiosurgery; 5 patients = CyberKnife).

alone does not prevent further tumor progression [70]. Surgical resection is the only treatment option that can offer immediate and complete tumor elimination. Contraindications to surgical intervention of skull base tumors are based on the patients` comorbidities. Surgical resection of glomus jugulare tumors carries a high complication rate, due to their high vascularity and the involvement of critical vascular and neuronal structures, which included stroke, cranial nerve injury with 8% to 40%, and an overall mortality rate of 5% to 13% [71]. Fractionated stereotactic radiotherapy compared with stereotactic radiosurgery may reduce the risk of radiation-induced side effects providing additional radiobiologic sparing.

Patient Selection for Radiosurgery and Treatment Planning Details For radiosurgical techniques, the prescribed dose to the tumor margin ranges from 12 Gy to 25 Gy with typical doses greater than 20 Gy for small- to medium-sized tumors but lower doses to larger tumors due to an increased risk of radiation-induced side effects [72–74]. Median tumor volumes of glomus tumors should be less than 10 cm3 because of the possible increased risk of radiation-induced cranial nerve deﬁcits due to the radiosensitivity of cranial nerves.

Experiences and Review of the Literature For well-deﬁned and noninﬁltrating glomus jugulare tumors, stereotactic radiotherapy should be particularly beneﬁcial.

They usually present in a small size due to their proximity to cranial nerves whose dysfunctions often herald the presence of the tumor. The steep dose gradient achievable with radiosurgery minimizes the irradiation dose to surrounding normal tissue. Whereas diagnostic imaging techniques have been much improved within recent years, they can not reliably separate tumor from adjacent cranial nerve when targeting radiosurgery treatment. Because of the close proximity of glomus tumors to cranial nerves, permanent cranial nerve deﬁcits are possible side effects of radiosurgery. Most published studies reported only transient dizziness or occasional hearing loss (Table 37-6) [75–79]. In the study of Mukherji et al. [80], 70% of the patients treated with external beam radiation therapy showed an improvement of clinical symptoms and approximately 60% tumor volume reduction. Conventional radiotherapy in the upper neck and skull base often releases secondary complications of bone radionecrosis, xerostomia, and possible induction of secondary malignancies [72]. The use of IMRT may reduce side effects, especially xerostomia, but the target conformality is not as high as by using stereotactic radiosurgery [81]. Radiosurgery of glomus jugulare tumors has been reported by few centers. Overall, the number of treated patients is small with only short-term follow-up. Jordan et al. [72] reported on eight patients treated with Gamma Knife radiosurgery between 1990 and 1998. The mean tumor volume of these patients was 9.81 cm3 (range, 17.25 to 4.29 cm3). The mean applied marginal dose was 16.3 Gy (range, 12 Gy to 20 Gy). None of these patients

TABLE 37-6. Summary table: Review of the literature for radiosurgery of glomus jugulare tumors. Author

Jordan et al. [72] Foote et al. [75] Eustacchio et al. [76] Maarouf et al. [77] Liscak et al. [78] Gottfried et al. [79]‡ *Crude local control rates. †Mean values. ‡Literature review.

N

Median follow-up (range)

Median peripheral dose (range)

Local control*

8 25 19 12 52 142

27 mo (9.7–102)† 37 mo (11–118) 7 yr (1.5–10) 4 yr (0.8–9) 24 mo (4–70) 39.4 mo†

16.3 Gy (12–20)† 15 Gy (12–18) 14 Gy (12–20) 15 Gy (11–20) 16.5 Gy (10–30)

100% 100% 95% 100% 100% 98%

Morbidity

Acute vertigo (1 patient) Late vertigo (1 patient) None Moderate facial palsy (1 patient) Tinnitus (2 patients) 8.5% morbidity

390

s. milker-zabel et al.

developed delayed cranial neuropathy or tumor progression during a mean follow-up of 27 months. Lim et al. [73] reported on 10 patients treated with radiosurgery, whereas 4 patients were treated with primary radiosurgery due to several comorbidities. The other ﬁve had prior surgeries for their tumor. Tumor size ranged from 1.2 cm to 3.6 cm at the largest diameter with an average of 2.4 cm. Six patients were treated with a frame-based linac system and four with the CyberKnife. Prescribed dose to the 80% isodose ranged from 16 Gy to 25 Gy to the tumor margin. After a median follow-up of 21.5 months, nine patients had no change of tumor size, whereas one patient showed tumor regression. Nine patients had stable neurologic symptoms and only one patient experienced transient ipsilateral tongue weakness and hearing loss. After an observation period of up to 6.7 years, Saringer and co-workers [68] reported on 13 patients with glomus tumors treated with Gamma Knife radiosurgery. Three of these patients showed tumors size reduction, whereas 10 had stable tumor volume. Clinical symptoms remained unchanged in six, and also six patients showed an improvement on clinical symptoms. One patient failed the follow-up. Two patients developed transient cranial nerve complication like worsening of preexisting swallowing disorders and temporary facial nerve palsy 1 and 12 months after radiosurgery, respectively. The Mayo Clinic experiences were published by Foote et al. in 2001 [75]. They described treatment outcome for a total of 25 patients and long-term results for a cohort of 9 patients with a median largest diameter of 3.3 cm. No acute neurologic toxicity was identiﬁed in all 25 patients. Only one patient experienced clinically signiﬁcant vertigo 8.5 months after treatment. They observed no new or progressive neuropathy of cranial nerves V to XII. After fractionated stereotactic radiation therapy of large chemodectomas of the skull base, Zabel et al. [82] showed a local control rate of 90.4% after 5 and 10 years. Median target volume was 71.8 cm3 (range, 10.5 cm3 to 212.2 cm3); median delivered total dose was 57.6 Gy with 1.8 Gy per fraction. A partial response was seen in 32%, whereas 59% showed stable tumor volume during follow-up. Only two patients developed recurrence after 19 and 32 months, respectively. An improvement of neurologic dysfunction improved or resolved completely in 59% and stabilized in 32%. Impairment on preexisting neurologic dysfunction was seen in 9% of the patients. No patient developed new neurologic deﬁcits after fractionated stereotactic radiation therapy.

Conclusion When radiosurgery is limited to lesions that are 3 cm or less, dose fall-off is sharp, and only a small amount of normal brain tissue receives a high radiation dose. At larger volumes, the radiation fall-off into the surrounding normal tissue is not as steep as in smaller-volume tumors and the risk of delayed radiation-induced side effects increases. Still, many patients with skull base tumors larger than 30 mm can undergo radiosurgery safely because these tumors are often in contact with the brain for only a portion of their surface, and therefore, radiation falloff occurs into bone, the sinuses, or the infratemporal or cervi-

cal regions. For larger lesions, for tumors next to the optic pathway, for patients at greater risk for microsurgical morbidity, or in those with residual or recurrent disease, fractionated radiation therapy is recommended [83]. Currently, the treatment of choice for patients with chordomas and chondrosarcoma is the use of heavily charged particles such as protons, or carbon ions, because of their ﬁnite range in tissues.

References 1. Milker-Zabel S, Debus J, Thilmann C, et al. Fractionated stereotactically-guided radiotherapy and radiosurgery in the treatment of functional and nonfunctional adenomas of the pituitary gland. Int J Radiat Oncol Biol Phys 2001; 50(5):1279–1286. 2. Debus J, Pirzkall A, Schlegel W, et al. Stereotaktische Einzeitbestrahlung (Radiochirurgie). Methodik, Indikationen, Ergebnisse. Strahlenther Onkol 1999; 175:47–56. 3. Pirzkall A, Debus J, Lohr F, et al. Radiosurgery alone or in combination with whole brain radiotherapy for brain metastases. J Clin Oncol 1998; 16:3563–3569. 4. Neel HB, Whicker JH, Devine KD, Weiland LH. Juvenile angioﬁbroma: review of 120 cases. Am J Surg 1973; 126:547–556. 5. Chandler JR, Goulding R, Moskowitz L, Quencer RM. Nasopharyngeal angioﬁbromas: staging and management. Ann Otol Rhinol Laryngol 1984; 93:322–329. 6. Jafek BW, Krekorian EA, Kirsch WM, Wood RP. Juvenil nasopharyngeal angioﬁbroma: management of intracranial extension. Head and Neck Surg 1979; 2(2):119–128. 7. Jacobsson M, Petruson B, Svendsen P, et al. Juvenil nasopharyngeal angioﬁbroma: a report of eighteen cases. Acta Otolaryngol (Stockh) 1988; 105:132–139. 8. Kuppersmith RB, The BS, Donovan DT, et al. The use of intensity-modulated radiotherapy for the treatment of extensive and recurrent juvenile angioﬁbroma. Int J Pediatr Othorhinolaryngol 2000; 52(3):261–268. 9. Dare AO, Gibbons KJ, Proulx GM, Fenstermaker RA. Resection followed by radiosurgery for advanced juvenile nasopharyngeal angioﬁbroma: report of two cases. Neurosurgery 2003; 52:1207– 1211. 10. McGahan RA, Durrance FY, Parke RB Jr, et al. The treatment of advanced juvenile nasopharyngeal angioﬁbroma. Int J Radiat Oncol Biol Phys. 1989; 17(5):1067–1072. 11. Briant TD, Fitzpatrick PJ, Berman J. Nasopharyngeal angioﬁbroma: a twenty year study. Laryngoscope 1978; 88:1247– 1251. 12. Cummings BJ, Blend R, Keane T, et al. Primary radiation therapy for juvenile nasopharyngeal angioﬁbroma. Laryngoscope 1984; 94:1599–1605. 13. Reddy KA, Mendenhall WM, Amdur RJ, et al. Long-term results of radiation therapy for juvenile nasopharyngeal angioﬁbroma. Am J Otolaryngol 2001; 22:172–175. 14. Wiatrak BJ, Koopmann CF, Turrisi AT. Radiation therapy as an alternative to surgery in the management of intracranial juvenile nasopharyngeal angioﬁbroma. Int J Pediatr Othorhinolaryngol 1993; 28:51–61. 15. Takeda A, Shigematsu N, Susuki S, et al. Late retinal complications of radiation therapy for nasal and paranasal malignancies: relationship between irradiated-dose and severity. Int J Radiat Oncol Biol Phys 1999; 4:599–605. 16. Berger L, Luc R. L´esthésioneuroépithéliome olfactif. Bull Assoc Fr Etude Cancer 1924; 13:410–421. 17. Kadish S, Goodman M, Wang CC. Olfactory neuroblastoma: a clinical analysis of 17 cases. Cancer 1976; 37:1572–1576.

37.

skull base tumors

18. Wieden PL, Yarington CT Jr, Richardson RG. Olfactory neuroblastoma: chemotherapy and radiotherapy for extensive disease. Arch Otolaryngol 1984; 110:759–760. 19. Dulguerov P, Calcaterre T. Esthesioneuroblastoma: the UCLA experience 1970–1990. Laryngoscope 1992; 102:843–849. 20. Kempf H-G, Becker G, Weber BP, Ruck P. Diagnostik und Therapie des Olfaktoriusneuroblastoms. HNO 1994; 42:422–428. 21. Zabel A, Thilmann C, Milker-Zabel S, et al. The role of stereotactically guided conformal radiotherapy for local tumor control of esthesioneuroblastoma. Strahlenther Onkol 2002;178(4):187– 191. 22. Morita A, Ebersold MJ, Olsen KD, Foote RL, Lewis JE, Quast LM. Esthesioneuroblastoma: prognosis and management. Neurosurgery 1993; 32:706–715. 23. Polin RS, Sheehan JP, Chenelle AG, et al. The role of preoperative adjuvant treatment in the management of esthesioneuroblastoma: the University of Virginia experience. Neurosurgery 1998; 42:1029–1037. 24. Unger F, Haselsberger K, Walch C, Stammberger H, Papaefthymiou G. Combined endoscopic surgery and radiosurgery as treatment modality for olfactory neuroblastoma (esthesioneuroblastoma). Acta Neurochir (Wien) 2005; 147(6):595– 601. 25. Leber KA, Berglöff J, Pendl G. Dose-response tolerance of the visual pathwaya and cranial nerves of the cavernous sinus to stereotactic radiosurgery. J Neurosurg 1998; 88:43–50. 26. Ling SM, Linzer D, Lu J, et al. Local control of large primary brain tumours or metastasis treated by Leksell Gamma Knife stereotactic radiosurgery: an analysis of dose-volume considerations, local control and complications. J Radiosurg 1999; 2:141–152. 27. Chao KSC, Kaplan C, Simpson JR, et al. Esthesioneuroblastoma: the impact of treatment modality. Head Neck 2001; 23:749–757. 28. Walch C, Stammberger H, Anderhuber W, Unger F, Köle W, Feichtinger K. The minimally invasive approach to olfactory neuroblastoma: combined endoscopic and stereotactic treatment. Laryngoscope 2000; 110:635–640. 29. Sanford RA, Muhlbauer MS. Craniopharyngioma in children. Neurol Clin 1991; 9:453–465. 30. Rutka JT, Hoffman HR, Drake JM, et al. Suprasellar and sellar tumors in childhood and adolescence. Neurosurg Clin North Am 1992; 3:803–820. 31. Fisher PG, Jenab J, Gopldthwaite PT, et al. Outcomes and failure patterns in childhood craniopharyngeomas. Childs Nerv Syst 1998; 14:558–563. 32. Abrams LS, Repka MX. Visual outcome of craniopharyngeoma in children. J Pediatr Ophthalmol Strabis 1997; 34:223–228. 33. Lewis D. A contribution to the subject of tumors of the hypophysis. JAMA 1910; 55:1002–1008. 34. Fahlbusch R, Honegger J, Paulus W, et al. Surgical treatment of craniopharyngiomas: experience with 168 patients. J Neurosurg 1999; 90:237–250. 35. Samii M, Bini W. Surgical treatment of craniopharyngiomas. Zentralblatt Neurochir. 1991; 52:17–23. 36. Yasargil MG, Curcic M, Kis M, et al. Total removal of craniopharyngiomas. Approaches and long-term results in 144 patients. J Neurosurg 1990 ; 73 :3–11. 37. Kalapurakal JA, Goldman S, Hsieh YC, et al. Clinical outcome in children with recurrent craniopharyngioma after primary surgery. Cancer J 2000; 6(6):388–393. 38. Hoffmann HJ, DeSilva M, Humphreys RP, et al. Aggressive surgical management of craniopharyngioma in children. J Neurosurg 1992; 76:47–52. 39. Becker G, Kortmann RD, Skalej M, Bamberg M. The role of radiotherapy in the treatment of craniopharyngiomas— indications, results, side effects. Front Radiat Ther Oncol 1999; 33:100–113.

391

40. Jose CC, Rajan B, Ashley S, et al. Radiotherapy for the treatment of recurrent craniopharyngioma. Clin Oncol 1992; 4:287–289. 41. Sanford RA. Craniopharyngioma: results of survey of the American Society of pediatric Neurosurgery. Pediatr Neurosurg 1994; 21(Suppl 1):39–43. 42. Rajan B, Ashley S, Gorman C, et al. Craniopharyngioma— long-term results following limited surgery and radiotherapy. Radiother Oncol 1993; 26:1–10. 43. Chung WY, Pan DH, Shiau CY, et al. Gamma knife radiosurgery for craniopharyngiomas. J Neurosurg 2000; 93(Suppl 3):47–56. 44. Sung DI, Chang CH, Harisiadis L, et al. Treatment results of craniopharyngiomas. Cancer 1981; 47:847–852. 45. Fischer EG, Welch K, Shillito J Jr, et al. Craniopharyngiomas in children: Long-term effects of conservative surgical procedures combined with radiation therapy. J Neurosurg 1990; 73:534–540. 46. Mokry M. Craniopharyngiomas: a six year experience with Gamma Knife radiosurgery. Stereotact Funct Neurosurg 1999; 72(Suppl 1):140–149. 47. Voges J, Sturm V, Lehrke R, et al. Cystic craniopharyngiomas: long-term results after intracavitary irradiation with stereotactically applied colloidal beta-emitting radioactive sources. Neurosurgery 1997; 35:1001–1010. 48. Ülfarsson E, Lindquist C, Roberts M, et al. Gamma knife radiosurgery for craniopharyngiomas: long-term results in the ﬁrst Swedish patients. J Neurosurg 2002; 97(Suppl 5):613–622. 49. Schulz-Ertner D, Frank C, Herfarth KK, et al. Fractionated stereotactic radiotherapy for craniopharyngiomas. Int J Radiat Oncol Biol Phys 2002; 54:1114–1120. 50. Kobayashi T, Kida Y, Mori Y, et al. Long-term results of gamma knife surgery for the treatment of craniopharyngioma in 98 consecutive cases. J Neurosurg 2005; 103(6 Suppl):482–488. 51. Al Mefty O, Borba LA. Skull base chordomas: a management challenge. J Neurosurg 1997; 86:182–189. 52. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the cranial base: results and follow-up of 60 patients. Neurosurgery 1995; 36:887–896. 53. Hug EB, Loredo LN, Slater JD, et al. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg 1999; 91:432–439. 54. Kondziolka D, Lunsford LD, Flickinger JC. The role of radiosurgery in the management of chordoma and chondrosarcoma of the cranial base. Neurosurgery 1991; 29:38–45. 55. Pearlman AW, Friedman M. Radical radiation therapy of chordoma. AJR 1970; 108:332–341. 56. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlenther Onkol 1999; 175 (Suppl II):57–63. 57. Schulz-Ertner D, Nikoghosyan A, Thilman C, et al. Results of carbon ion radiotherapy in 152 patients. Int J Radiat Oncol Biol Phys 2004; 58(2):631–640. 58. Noel G, Habrand JL, Mammar H, et al. Combination of photon and proton radiation therapy for chordomas and chondrosarcomas of the skull base: the Centre of Protontherapie D’Orsay experience. Int J Radiat Oncol Biol Phys 2001; 51:392–398. 59. Debus J, Haberer T, Schulz-Ertner D, et al. Fractionated carbon ion irradiation of skull base tumors at GSI. First clinical results and future perspectives. Strahlenther Oncol 2000; 176(5):211– 216. 60. Schulz-Ertner D, Nikoghosyan A, Thilmann C, et al. Carbon ion radiotherapy for chordomas and low grade chondrosarcomas of the skull base. Results in 67 patients. Strahlenher Onkol 2003; 179(9):598–605. 61. Feigl GC, Bundschuh O, Gharabaghi A, et al. Evaluation of a new concept for the management of skull base chordomas and chondrosarcomas. J Neurosurg 2005; 102(Suppl):154–170. 62. Pamir MN, Kilic T, Türe U, Özek MM. Multimodality management of 26 skull base chordomas with 4-year mean follow-up:

392

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

s. milker-zabel et al. experience at a single institution. Acta Neurochir (Wien) 2004; 146:343–354. Berson AM, Castro JR, Petti P, et al. Charged particle irradiation of chordoma and chondrosarcoma of the base of skull and cervical spine: The Lawrence Berkely Laboratory experience. Int J Radiat Oncol Biol Phys 1988; 15:559–565. Krishnan S, Foote R, Brown P, et al. Radiosurgery for cranial base chordomas and chondrosarcomas. Neurosurgery 2005;56(4):777– 784; discussion 777–784. Chang SD, Martin DP, Lee E, et al. Stereotactic radiosurgery and hypofractionated stereotactic radiotherapy for residual or recurrent cranial base and cervical chordomas. Neurosurg Focus 2001; 10(3):E5. Konefal JB, Pilepich MV, Spector GJH, et al. Radiation therapy in the treatment of chemodectomas. Laryngoscope 1987; 97:1331– 1335. Foote RL, Coffey RJ, Gorman DA, et al. Stereotactic radiosurgery for glomus jugulare tumors: a preliminary report. Int J Radiat Oncol Biol Phys 1997; 38:491–495. Saringer W, Khayal H, Ertl A, et al. Efﬁciency of gamma knife radiosurgery in the treatment of glomus jugulare tumors. Minim Invas Neurosurg 2001; 44:141–145. Woods CI, Strasnick B, Jackson CG. Surgery for glomus tumors: the otology group experience. Laryngoscope 1993; 60(Suppl): 65–70. Pauw BK, Malek MS, Fisch U, Valvanis A. Preoperative embolization of paragangliomas (glomus tumors) of the head and neck; histopathologic and clinical features. Skull Base Surgery 1993; 3:37–44. Chretien PB, Engelman K, Hoye RC, Geelhoed GW. Surgical management of intravascular glomus jugulare tumors. Am J Surg 1971; 122740–122743. Jordan JA, Roland PS, McManus C, et al. Stereotactic radiosurgery for glomus jugulare tumors. Laryngoscope 2000; 110:35– 38.

73. Lim M, Gibbs IC, Adler JR, et al. The efﬁcacy of linear accelerator stereotactic radiosugery in treating glomus jugulare tumors. Technol Cancer Res Treatm 2003; 2(3):261–265. 74. Maarouf M, Voges J, Landwehr P, et al. Stereotactic linear accelerator-based radiosurgery for the treatment of patients with glomus jugulare tumors. Cancer 2003; 97:1093–1098. 75. Foote RL, Pollock BE, Gorman DA, et al. Glomus jugulare tumors: tumor control and complications after stereotactic radiosurgery. Head Neck 2002; 24:332–339. 76. Eustacchio S, Trummer M, Unger F, et al. The role of Gamma Knife radiosurgery in the management of glomus jugular tumours. Acta Neurochir Suppl 2002; 84:91–97. 77. Maarouf M, Voges J, Landwehr P, et al. Stereotactic linear accelerater-based radiosurgery for the treatment of patients with glomus jugulare tumors. Cancer 2003; 97(4):1093–1098. 78. Liscak R, Vladyka V, Wowra B, et al. Gamma Knife radiosurgery of the glomus jugulare tumour - early multicentre experience. Acta Neurochir (Wien) 1999; 141(11):1141–1146. 79. Gottfried ON, Liu JK, Couldwell WT. Comparison of radiosurgery and conventional surgery for the treatment of glomus jugulare tumors. Neurosurg Focus 2004; 17(2):E4. 80. Mukherji SK, Kasper ME, Tart RP, Mancuso AA. Irradiated paragangliomas of the head and neck: CT and MR appearance. AJNR Am J Neuroradiol 1994; 15(2):357–3363. 81. Zabel A, Thilmann C, Zuna I, et al. Comparison of forward planned conformal radiation therapy and inverse planed intensity modulated radiation therapy for esthesioneuroblastoma. Br J Radiol 2002; 75:356–361. 82. Zabel A, Milker-Zabel S, Huber P, et al. Fractionated stereotactic conformal radiotherapy of chemodectomas of the skull base. Int J Radiat Oncol Biol Phys 2004; 58:1445–1450. 83. Milker-Zabel S, Zabel A, Schulz-Ertner D, et al. Fractionated stereotactic radiotherapy in patients with benign or atypical intracranial meningiomas—long-term experience and prognostic factors. Int J Radiat Oncol Biol Phys 2005; 61(3):810–817.

3 8

Skull Base Tumors: Surgery Perspective James K. Liu, Oren N. Gottfried, and William T. Couldwell

Introduction Tumors of the skull base are frequently intimately involved with surrounding critical neurovascular structures, which makes complete resection difﬁcult. Because tumors of the skull base can extend into adjacent compartments, both intracranial and extracranial structures can be involved. Thus, a multidisciplinary approach is required for optimal surgical management. Because of advancements in microsurgical techniques, modern skull base approaches, electrophysiologic monitoring, and neuroimaging in the past two decades, many of these tumors that were once thought to be inaccessible and unresectable can be safely removed with preservation of neural structures [1]. The microsurgical application of skull base techniques maximizes bone removal and therefore improves the operative ﬁeld of vision, minimizes brain retraction, and provides multiple surgical angles for dissection [2]. Recent applications of frameless stereotaxy and endoscopic techniques have further contributed to modern skull base surgery. The primary goals of surgical removal of skull base tumors are to preserve neurologic function while treating the patient’s pathology. The ultimate goal is cure by gross total tumor resection with minimal morbidity and mortality. Surgical removal of skull base tumors offers the beneﬁts of immediate neural decompression and tumor removal. For benign tumors, such as meningiomas, schwannomas, and glomus jugulare tumors, curative resection can be achieved. For aggressive lesions, such as chordomas, and malignant lesions, such as chondrosarcomas and esthesioneuroblastomas, resection can also offer maximal cytoreduction as a component of multimodal therapy. In the past several years, stereotactic radiosurgery (SRS) has emerged as a treatment option for patients harboring skull base tumors. In properly selected patients, tumor control rates have been reported between 90% and 100% for benign tumors and between 50% and 70% in some chordomas and chondrosarcomas [3]. Whereas the goal of surgery is neural decompression and complete curative resection, the main goal of SRS is tumor control. The beneﬁt of single-stage SRS is limited when treating tumors greater than 30 mm in average diameter, because the radiation fall-off into the surrounding normal tissues is not as steep and therefore places the patient at risk of

radiation-induced complications. This is an important consideration particularly if the lesion is near the optic nerve and brain stem. Larger tumors also tend to have symptoms related to mass effect and require surgery for immediate decompression. Although the efﬁcacy of radiosurgical treatment of skull base tumors versus resection is the subject of much debate, it is important to recognize that SRS is part of the armamentarium of the skull base surgeon and can complement resection in the overall multimodal treatment of some complex tumors. For example, a patient with a large skull base tumor with cavernous sinus invasion can undergo subtotal resection of the noncavernous portion followed by postoperative SRS for the cavernous sinus portion of the tumor. This chapter focuses on the role of surgery in the management of complex skull base tumors. It is beyond the scope of this chapter to discuss at length the clinicopathologic and radiologic aspects and treatment of all the different skull base tumor pathologies; rather, we discuss the surgical results of more commonly encountered skull base pathologies. The role of surgery in the treatment of skull base meningiomas, acoustic neuromas, and head and neck malignancies are discussed in other chapters in this book.

Surgical Considerations The goals of surgery should be tailored to the individual patient, considering the histopathology of the tumor, the location of the tumor, the presenting symptomatology, and the age of the patient. For asymptomatic benign lesions in elderly patients, a period of observation may be warranted unless symptomatic brain-stem compression or progressive cranial neuropathy is evident, as most benign tumors are slow growing. The goal of surgery for benign lesions should be total removal of tumor while preserving or improving neurologic function, because it provides the best chance for a surgical cure or long-term tumor control. The ﬁrst attempt at resection offers the best chance at complete removal when the arachnoidal membranes are intact, as these facilitate dissection of neurovascular structures. Tumor involvement of the cranial nerves or tumor adherent to the brain stem and its vasculature may preclude total removal without increasing signiﬁcant morbidity and incurring new

393

394

j.k. liu et al.

neurologic deﬁcits. In these instances, a subtotal removal with resection of the symptomatic portion of the mass may be chosen. If, for example, a patient has a compressive petroclival tumor that extends into the cavernous sinus but has no signiﬁcant cranial neuropathies (diplopia), a subtotal resection (removing the compressive posterior fossa mass) and subsequent stereotactic radiosurgery to the residual cavernous sinus tumor may minimize postoperative morbidity while maximizing tumor control. This strategy may be considered in the elderly patient with associated medical problems. If, on the other hand, symptomatic cranial neuropathies are present because of tumor in the cavernous sinus, especially in a young patient, a more aggressive radical resection of the tumor may be considered. If tumor is encasing the cavernous carotid artery and an oncologic resection is planned with sacriﬁce of the carotid artery, a cerebrovascular bypass may be necessary. In instances where the tumor encases the basilar artery, adheres to the brain stem or vasculature, or parasitizes the brain-stem perforators, a more conservative approach that leaves a small remnant of tumor may be considered if radical removal might result in potential devastating neurologic morbidity. Many of these decisions regarding the justiﬁcation of radical removal must be made intraoperatively based on the surgeon’s judgment of the risk involved with resection.

Trigeminal Schwannoma Schwannomas that arise from the trigeminal nerve are the second most frequent intracranial schwannomas after vestibular schwannomas and represent 0.07% to 0.36% of all intracranial tumors and 0.8% to 8% of all intracranial schwannomas [4–7]. Clinically, patients can present with facial numbness, trigeminal neuralgia, headaches, gait disturbance, hearing loss, facial weakness, or even hemifacial spasm, depending on the tumor location, size, and the involved cranial nerves. Approximately 50% arise predominately within the middle fossa, 30% arise in the posterior fossa, and 20% are dumbbell-shaped with signiﬁcant extension into both cranial fossae [8]. When these benign tumors become large, they tend to displace the surrounding neurovascular structures rather than engulf them. The majority of these tumors involve the cavernous sinus and can be safely removed through an extradural approach [5]. Total removal is believed to offer the best chance of cure, and subtotal removal is associated with higher recurrence rates (Fig. 38-1). Early attempts at complete resection of these tumors were associated with high rates of mortality [8]. Skull base microsurgical approaches have dramatically improved surgical morbidity and extent of tumor resection. In a comparison study between conventional and skull base approaches, Taha et al. [9] found that skull base approaches allowed better exposure of trigeminal schwannomas, multiple working angles with minimal brain retraction, and more complete removal without increased morbidity. The rate of residual or recurrent tumors was also lower when patients underwent a skull base approach (10% vs. 65%). The tendency to recur is low if these tumors are removed completely. The appropriate skull base approach is determined based on the location of the tumor. For tumors that primarily occupy the middle fossa, a frontotemporal craniotomy is preferred.

FIGURE 38-1. (A, B) Preoperative MRI with gadolinium (A, axial view; B, coronal view) demonstrating a giant trigeminal schwannoma with brain-stem compression in a 54-year-old man who presented with progressive gait ataxia. A right combined petrosal approach was performed to achieve a complete resection of the tumor. (C, D) Postoperative MRI with gadolinium (C, axial view; D, coronal view) shows complete resection.

For posterior fossa tumors, the retrosigmoid approach is usually sufﬁcient for exposure. The combined petrosal approach is useful when approaching large dumbbell-shaped schwannomas that occupy both middle and posterior fossae. For tumors that are more peripheral that extend into the infratemporal fossa, the frontotemporal extradural infratemporal fossa approach is useful. Total resection has been achieved in 79% to 83% of patients with preservation or improvement of cranial nerve function [8]. In a report of 25 patients by AlMefty et al. [5], preoperative trigeminal sensory deﬁcit improved in 44%, facial pain decreased in 73%, and trigeminal motor deﬁcit improved in 80%. Preoperative abducens nerve palsy improved in 63%. Three (12%) patients experienced a persistent new or worse cranial nerve function postoperatively. All cerebellar and brain-stem manifestations resolved postoperatively. SRS has been used recently as primary treatment for patients with small and moderate-sized trigeminal schwannomas [10–12]. In a recent study by Pan et al. [11], tumor growth control was 93% and decrease in tumor volume was seen in 73% of patients. Four patients, however, had tumor progression resulting in worsening of neurologic symptoms. Thirteen (56%) patients had worsening or no improvement of trigeminal nerve symptoms. Because the mean follow-up of SRS has ranged from

38.

skull base tumors: surgery perspective

40 to 68 months, the efﬁcacy and role of SRS has yet to be determined from long-term studies. Complete microsurgical resection as the primary treatment offers the best chance at curing the tumor. SRS should be considered for those patients harboring residual or recurrent tumors that cannot be removed surgically, for patients who cannot tolerate surgery because of medical comorbidities, or for patients who do not wish to have surgery.

Glomus Jugulare Tumors Glomus jugulare tumors are rare lesions that arise from the chief cells of the paraganglia in the adventitia of the dome of the jugular bulb. Although 1% to 5% of these tumors metastasize [13–16], most are benign tumors that can be locally aggressive and inﬁltrate and compress adjacent bone, cranial nerves, or blood vessels [17]. Poor outcomes and signiﬁcant complication rates were once associated with most treatment options for glomus jugular tumors, but advances in imaging, microsurgical skull base techniques, and the delivery of radiation have improved treatment efﬁcacy and safety. Today, microsurgical resection, radiosurgery, vascular embolization, conventional fractionated external beam radiotherapy, or a combination of these are all treatment options. Surgery offers immediate and complete tumor elimination and is the primary treatment modality for many patients. Although previously associated with morbidity and occasional mortality, many glomus jugulare lesions that were once deemed inoperable because of the vascularity and involvement of critical vascular and neural anatomy are now being safely and totally resected [18–24]. We recently reviewed the results of 7 series of glomus jugulare tumors treated with surgery [25]. The authors of these studies used a variety of approaches, including both single [18–21, 23, 26] and staged [20, 22] operations. Total resection on the ﬁrst attempt was achieved in 254 of 288 (88.2%) surgeries, and the surgical tumor control rate ranged from 88.5% to 94.5% [18, 19, 21, 22]. Recurrence occurred in 3.3% of cases (range, 0 to 5.5%) [18–23, 26]. Cranial nerve (CN) preservation was most easily accomplished in smaller tumors and was as high as 80% to 95% [27, 28], although new cranial nerve deﬁcits occurred in 22% to 59% of patients after surgery [19, 29, 30]. The overall mortality rate was 1.3% (5/374 patients; range, 0 to 4.1%) [18–23, 26]. The use of Gamma Knife and linear accelerator (linac) stereotactic radiosurgery as a primary and secondary treatment modality for glomus jugulare tumors has been increasing [17, 31–41]. These tumors are ideal candidates for treatment with radiosurgery because they are well demarcated on magnetic resonance imaging (MRI) and rarely invade the brain, allowing a steep dose decrease at the margin [31]. Radiosurgery involves a shorter treatment time (1 to 2 days vs. 4 to 6 weeks for conventionally fractionated external beam radiation or postoperative recovery after resection) [32], precise stereotactic localization, and a smaller volume of irradiated normal tissue than conventional radiotherapy so the incidence of complications is lower [31, 32]. Proponents argue that radiosurgery has the potential to avoid the hospital costs and potential postoperative morbidity associated with surgery, providing a more

395

cost-effective treatment [31, 32]. In eight recent series [17, 31– 33, 35–40] that used Gamma Knife stereotactic radiosurgery or a frame-based linac system, tumor size decreased in 50 of 137 (36.5%) tumors or was unchanged in 84 of 137 (61.3%) [25]. No tumors were eliminated entirely. Tumor volume decreased as early as 6 months after radiosurgery (median, 20 months) [33]. Three (2.2%) patients had new tumor growth after radiosurgery [38, 40]. Ninety-seven percent of patients had stabilization or improvement of their neurologic exam. Neurologic complications occurred in 12 of 141 (8.5%) patients, although 9 (6.4%) of these were transient. Direct comparison of these modalities is difﬁcult because the end points of treatment are different (total elimination of tumor for surgery or growth inhibition for radiosurgery). Nonetheless, recurrence after either treatment is very low and occurs at similar rates. No mortality was associated with radiosurgery, and mortality secondary to surgery was just above 1%. The biggest difference between surgery and radiosurgery is the morbidity. Although morbidity associated with radiosurgery (8.5%) is easily quantiﬁed, it is difﬁcult to establish the short- and longterm morbidity in the surgical patient. Surgery remains the treatment of choice for a glomus jugulare tumor in an otherwise healthy patient who desires an immediate cure. Invalidating cranial nerve palsies, tumors that are too large for radiosurgery, tumors causing vascular insufﬁciency because of major arterial encasement, or lifethreatening tumors that have signiﬁcant intracranial extension are good candidates for surgery [42]. If a complete resection cannot be guaranteed or the risk of neurologic deﬁcits or morbidity is high with a radical resection, then a subtotal resection can be performed to debulk the tumor and prevent further neurologic compromise. After a subtotal resection, stereotactic radiosurgery is a safe and effective option to treat residual tumor. Patients with advanced age or signiﬁcant comorbidities may instead elect to undergo radiosurgery as a primary treatment modality [17]. Long-term studies evaluating the efﬁcacy of radiosurgery over a 10- to 20-year period are needed to compare these two treatment modalities further.

Chordoma Chordomas are rare tumors of the skull base that develop from remnants of the notochord. They represent approximately 0.1% of all intracranial tumors [43–47]. These slow-growing tumors are usually located in the midline at the clivus and can extend in various directions compressing neighboring neurovascular structures. Although they are histologically benign, they are considered aggressive in behavior because of their tendency to inﬁltrate surrounding structures. This makes complete resection difﬁcult and results in high recurrence rates. Most of the tumors that occur in the cranial region originate in the clivus bone and grow extradurally. At advanced stages, they can eventually invade intradurally and engulf the cranial nerves, arteries, and brain stem [46]. The goal of treatment of skull base chordomas is to maximize the chance of recurrence-free survival. An extensive skull base resection plays a signiﬁcant role in the treatment of skull base chordomas. Two or more skull base procedures

396

j.k. liu et al.

The introduction of adjuvant proton beam radiotherapy for the treatment of skull base chordomas has improved local tumor control and overall survival rates [43, 48, 49]. This is attributed to the Bragg peak effect, where high radiation doses are delivered to the target with a subsequent sharp fall-off of energy beyond the target volume. For proton beam radiotherapy to be effective, the tumor volume must be small and the lesion or resection cavity should be detected on neuroimaging studies. One limitation of proton beam radiotherapy is that only a few centers in the world can offer this treatment. Recently, reports of the use of high-dose SRS for the treatment of residual tumor after skull base surgery of chordomas have emerged [3, 50–52]. Early reports show reasonable local control rates; however, the efﬁcacy of SRS as an adjuvant treatment compared with proton beam radiotherapy awaits longer followup studies. Tumor progression appears to be related to marginal failure where the tumor grows adjacent to the treated area.

Chondrosarcoma

FIGURE 38-2. (A, B) Preoperative MRI with gadolinium (A, axial view; B, coronal view) showing a chordoma involving the clivus and left cavernous sinus in a 35-year-old woman. The tumor has grown despite previous surgical resection and radiosurgery at an outside institution. A two-stage operation was performed (extended transsphenoidal approach followed by a frontotemporal extradural transcavernous approach). (C, D) Postoperative MRI with gadolinium (C, axial view; D, coronal view) demonstrates gross total resection.

may be necessary to achieve a radical removal (Fig. 38-2) [43]. Most patients require combined multimodal treatment including surgery and radiation therapy [43, 47]. The advent of modern skull base techniques has allowed radical resection of these tumors, but some dispute the value of radical resection because of postoperative cranial nerve morbidity and the persistent high rate of tumor recurrence. Lanzino et al. [46] advocate the strategy of cytoreductive resection in which the goal is to remove as much tumor as possible while minimizing complications. It appears that more extensive resection followed by postoperative proton beam radiotherapy (to deliver high-dose radiation to the tumor bed) improves the prognosis of these patients. In a study by Colli and Al-Mefty [43], patients who had radical or subtotal resections were associated with better recurrencefree survival rates than those who underwent partial resection. The 5-year survival estimate was 85.9% for patients with typical chordomas. Adjuvant proton-beam radiotherapy was shown to increase the recurrence-free survival rate when compared with conventional radiotherapy (90.9% vs. 19.4%, respectively, at 4 years after treatment). The recurrence-free survival rates for all patients were 80.1% and 50.7% at 1 year and 5 years, respectively. Twenty-eight percent sustained additional permanent postoperative neurologic deﬁcits and 22% had transient deﬁcits. Of the 107 preoperative cranial nerve palsies, 25.2% were improved after surgery.

Chondrosarcomas are rare malignant tumors of the skull base that are thought to arise from primitive mesenchymal cells or from the embryonal rest of the cartilaginous matrix of the cranium. They are less common than chordomas and represent 0.15% of all intracranial tumors [44, 53]. Chondrosarcomas more commonly arise in a paramedian location rather than in the midline, with the petroclival junction as the most common site of origin. Most classic chondrosarcomas are low grade and grow slowly. Mesenchymal chondrosarcomas have islands of cartilage and undifferentiated mesenchymal cells, and dedifferentiated chondrosarcomas have features of anaplastic sarcomas. These latter two variants are more aggressive and have a higher potential to metastasize than classic chondrosarcomas [54]. Their involvement with neurovascular structures at the base of the skull may pose limitations on radical resection. Most authors agree that maximal tumor resection with minimal morbidity to the cranial nerves and vascular structures is the optimal way to achieve radical tumor cytoreduction (Fig. 38-3) [43, 53–56]. Chondrosarcomas generally have a better prognosis than chordomas. In a study by Colli and Al-Mefty [43], the 5-year survival estimates for chondrosarcomas and chordomas was 100% and 85.9%, respectively. Chondrosarcomas also demonstrated a higher recurrence-free survival than chordomas at 1 year (100% vs. 80.1%) and 5 years (100% vs. 50.7). Proton beam radiotherapy remains a good option for small residual or recurrent tumors. Hug et al. [48, 49] demonstrated a local control of 92% in 25 patients with chondrosarcomas. Tumor volumes greater than 25 cm3 and the presence of brainstem involvement had poorer tumor control rates. Surgical debulking enhanced delivery of full tumoricidal doses. As with the data on chordomas, the data on SRS for the treatment of chondrosarcoma include very few patients with short follow-up periods [50–52, 57]. Whether SRS will play a signiﬁcant role in tumor control for residual or recurrent disease awaits further studies with long-term follow-up.

38.

skull base tumors: surgery perspective

397

Reports of SRS for esthesioneuroblastomas are limited [70–72]. It has been used mostly as adjuvant therapy to treat small residual or recurrent disease, but no long-term follow-up studies have evaluated the efﬁcacy of this treatment.

References

FIGURE 38-3. (A, B) Preoperative MRI (A, unenhanced sagittal view; B, enhanced coronal view) showing a large chondrosarcoma occupying the posterior nasopharynx with involvement of the anterior skull base and clivus. A bifrontal transbasal approach was performed for gross total resection of the tumor. (C, D) Postoperative MRI (C, unenhanced sagittal view; D, enhanced coronal view) shows complete resection of the tumor.

Esthesioneuroblastoma Esthesioneuroblastomas are rare malignant tumors of the upper nasal cavity and anterior skull base. These tumors are thought to arise from olfactory epithelium that has spread locally to involve the paranasal sinuses, nasal cavity, and surrounding structures. They can also cross the cribriform plate and invade the brain or seed the cerebrospinal ﬂuid. Metastasis by lymphatic and hematogenous routes can also occur in 17% to 48% of patients [58–61]. Because of the limited number of patients and the variability in treatment protocols between institutions, a standardized management strategy has not been developed for esthesioneuroblastomas [61]. Multimodal treatment has been advocated with a combination of craniofacial resection, chemotherapy, and radiation therapy. Craniofacial resection, in which the cribriform plate is removed to prevent perineural extension, has been the mainstay of therapy and appears to improve long-term results when combined with adjuvant radiation therapy and chemotherapy [58, 61–66]. Some have shown preoperative radiotherapy to reduce the tumor burden, thus allowing an increased chance of gross total resection [61, 67, 68]. At the University of Virginia, the 5-year disease-free survival rate improved from 37% to 82% with craniofacial resection [59]. The M.D. Anderson Group reported 89% and 81% 5-year and 10-year survival rates in 30 patients, with a mean follow-up of 7.32 years after undergoing craniofacial resection followed by radiation therapy [69]. This approach appeared to be curative for early-stage disease (Kadish A or B).

1. Nanda A, Jawahar A, Sathyanarayana S. Microsurgery for potential radiosurgical skull base lesions: a retrospective analysis and comparison of results. Skull Base 2003; 13(3):131–138. 2. Bulsara KR, Al-Mefty O. Skull base surgery for benign skull base tumors. J Neurooncol 2004; 69(1–3):181–189. 3. Pollock BE, Foote RL. The evolving role of stereotactic radiosurgery for patients with skull base tumors. J Neurooncol 2004; 69(1–3):199–207. 4. Friedman RA, Pensak ML, Osterhaus D, et al. Trigeminal schwannomas: the role of the neurotologist in multidisciplinary management. Otolaryngol Head Neck Surg 1999; 120(3):355–360. 5. Al-Mefty O, Ayoubi S, Gaber E. Trigeminal schwannomas: removal of dumbbell-shaped tumors through the expanded Meckel cave and outcomes of cranial nerve function. J Neurosurg 2002; 96(3):453–463. 6. McCormick PC, Bello JA, Post KD. Trigeminal schwannoma. Surgical series of 14 cases with review of the literature. J Neurosurg 1988; 69(6):850–860. 7. Pollack IF, Sekhar LN, Jannetta PJ, et al. Neurilemomas of the trigeminal nerve. J Neurosurg 1989; 70(5):737–745. 8. Shrivastava RK, Sen C, Post KD. Trigeminal schwannomas. In: Winn HR, ed. Youmans Neurological Surgery, 5th ed. Philadelphia: Saunders, 2004:1343–1350. 9. Taha JM, Tew JM Jr, van Loveren HR, et al. Comparison of conventional and skull base surgical approaches for the excision of trigeminal neurinomas. J Neurosurg 1995; 82(5):719–725. 10. Huang CF, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for trigeminal schwannomas. Neurosurgery 1999; 45(1):11– 16; discussion 16. 11. Pan L, Wang EM, Zhang N, et al. Long-term results of Leksell gamma knife surgery for trigeminal schwannomas. J Neurosurg 2005; 102(Suppl):220–224. 12. Nettel B, Niranjan A, Martin JJ, et al. Gamma knife radiosurgery for trigeminal schwannomas. Surg Neurol 2004; 62(5):435–444; discussion 44–46. 13. Brown JS. Glomus jugulare tumors revisited: a ten-year statistical follow-up of 231 cases. Laryngoscope 1985; 95(3):284–288. 14. Brewis C, Bottrill ID, Wharton SB, et al. Metastases from glomus jugulare tumours. J Laryngol Otol 2000; 114(1):17–23. 15. Bojrab DI, Bhansali SA, Glasscock ME 3rd. Metastatic glomus jugulare: long-term followup. Otolaryngol Head Neck Surg 1991; 104(2):261–264. 16. Manolidis S, Shohet JA, Jackson CG, et al. Malignant glomus tumors. Laryngoscope 1999; 109(1):30–34. 17. Eustacchio S, Leber K, Trummer M, et al. Gamma knife radiosurgery for glomus jugulare tumours. Acta Neurochir (Wien) 1999; 141(8):811–818. 18. Green JD Jr, Brackmann DE, Nguyen CD, et al. Surgical management of previously untreated glomus jugulare tumors. Laryngoscope 1994; 104(8 Pt 1):917–921. 19. Jackson CG, McGrew BM, Forest JA, et al. Lateral skull base surgery for glomus tumors: long-term control. Otol Neurotol 2001; 22(3):377–382. 20. Pareschi RR, Destito S, Raucci AF, Colombo S. Surgery of glomus jugulare tumors. Skull Base 2003; 13:149–157. 21. Whitﬁeld PC, Grey P, Hardy DG, et al. The surgical management of patients with glomus tumours of the skull base. Br J Neurosurg 1996; 10(4):343–350.

398

j.k. liu et al.

22. Patel SJ, Sekhar LN, Cass SP, et al. Combined approaches for resection of extensive glomus jugulare tumors. A review of 12 cases. J Neurosurg 1994; 80(6):1026–1038. 23. Al-Mefty O, Teixeira A. Complex tumors of the glomus jugulare: criteria, treatment, and outcome. J Neurosurg 2002; 97(6):1356– 1366. 24. Al-Mefty O, Fox JL, Rifai A, et al. A combined infratemporal and posterior fossa approach for the removal of giant glomus tumors and chondrosarcomas. Surg Neurol 1987; 28(6):423– 431. 25. Gottfried ON, Liu JK, Couldwell WT. Comparison of radiosurgery and conventional surgery for the treatment of glomus jugulare tumors. Neurosurg Focus 2004; 17(2):E4. 26. Watkins LD, Mendoza N, Cheesman AD, et al. Glomus jugulare tumours: a review of 61 cases. Acta Neurochir (Wien) 1994; 130(1–4):66–70. 27. Jackson CG, Cueva RA, Thedinger BA, et al. Conservation surgery for glomus jugulare tumors: the value of early diagnosis. Laryngoscope 1990; 100(10 Pt 1):1031–1036. 28. Jackson CG, Cueva RA, Thedinger BA, et al. Cranial nerve preservation in lesions of the jugular fossa. Otolaryngol Head Neck Surg 1991; 105(5):687–693. 29. Cece JA, Lawson W, Biller HF, et al. Complications in the management of large glomus jugulare tumors. Laryngoscope 1987; 97(2):152–157. 30. Gjuric M, Rudiger Wolf S, Wigand ME, et al. Cranial nerve and hearing function after combined-approach surgery for glomus jugulare tumors. Ann Otol Rhinol Laryngol 1996; 105(12):949– 954. 31. Saringer W, Khayal H, Ertl A, et al. Efﬁciency of gamma knife radiosurgery in the treatment of glomus jugulare tumors. Minim Invasive Neurosurg 2001; 44(3):141–146. 32. Foote RL, Coffey RJ, Gorman DA, et al. Stereotactic radiosurgery for glomus jugulare tumors: a preliminary report. Int J Radiat Oncol Biol Phys 1997; 38(3):491–495. 33. Liscak R, Vladyka V, Wowra B, et al. Gamma Knife radiosurgery of the glomus jugulare tumour—early multicentre experience. Acta Neurochir (Wien) 1999; 141(11):1141–1146. 34. Pendl G, Schrottner O, Friehs GM, et al. Radiosurgery with the ﬁrst Austrian cobalt-60 Gamma-unit. A one year experience. Acta Neurochir (Wien) 1994; 127(3–4):170–179. 35. Jordan JA, Roland PS, McManus C, et al. Stereotastic radiosurgery for glomus jugulare tumors. Laryngoscope 2000; 110(1):35– 38. 36. Maarouf M, Voges J, Landwehr P, et al. Stereotactic linear accelerator-based radiosurgery for the treatment of patients with glomus jugulare tumors. Cancer 2003; 97(4):1093–1098. 37. Lim M, Gibbs IC, Adler JR Jr, et al. The efﬁcacy of linear accelerator stereotactic radiosurgery in treating glomus jugulare tumors. Technol Cancer Res Treat 2003; 2(3):261–265. 38. Feigenberg SJ, Mendenhall WM, Hinerman RW, et al. Radiosurgery for paraganglioma of the temporal bone. Head Neck 2002; 24(4):384–389. 39. Foote RL, Pollock BE, Gorman DA, et al. Glomus jugulare tumor: tumor control and complications after stereotactic radiosurgery. Head Neck 2002; 24(4):332–338; discussion 8–9. 40. Eustacchio S, Trummer M, Unger F, et al. The role of Gamma Knife radiosurgery in the management of glomus jugular tumours. Acta Neurochir Suppl 2002; 84:91–97. 41. Liscak R, Vladyka V, Simonova G, et al. Leksell gamma knife radiosurgery of the tumor glomus jugulare and tympanicum. Stereotact Funct Neurosurg 1998; 70(Suppl 1):152–160. 42. van der Mey AG, Frijns JH, Cornelisse CJ, et al. Does intervention improve the natural course of glomus tumors? A series of 108 patients seen in a 32-year period. Ann Otol Rhinol Laryngol 1992; 101(8):635–642.

43. Colli B, Al-Mefty O. Chordomas of the craniocervical junction: follow-up review and prognostic factors. J Neurosurg 2001; 95(6):933–943. 44. Crockard A. Chordomas and chondrosarcomas of the cranial base: results and follow-up of 60 patients. Neurosurgery 1996; 38(2):420. 45. Watkins L, Khudados ES, Kaleoglu M, et al. Skull base chordomas: a review of 38 patients, 1958–88. Br J Neurosurg 1993; 7(3): 241–248. 46. Lanzino G, Dumont AS, Lopes BS, et al. Skull base chordomas: overview of disease, management, options, and outcome. Neurosurg Focus 2001; 10(3):E12. 47. Crockard HA, Steel T, Plowman N, et al. A multidisciplinary team approach to skull base chordomas. J Neurosurg 2001; 95(2):175– 183. 48. Hug EB, Slater JD. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. Neurosurg Clin N Am 2000; 11(4):627–638. 49. Hug EB, Loredo LN, Slater JD, et al. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg 1999; 91(3):432–439. 50. Kondziolka D, Lunsford LD, Flickinger JC. The role of radiosurgery in the management of chordoma and chondrosarcoma of the cranial base. Neurosurgery 1991; 29(1):38–45; discussion 46. 51. Krishnan S, Foote RL, Brown PD, et al. Radiosurgery for cranial base chordomas and chondrosarcomas. Neurosurgery 2005; 56(4): 777–784; discussion 784. 52. Muthukumar N, Kondziolka D, Lunsford LD, et al. Stereotactic radiosurgery for chordoma and chondrosarcoma: further experiences. Int J Radiat Oncol Biol Phys 1998; 41(2):387–392. 53. Crockard HA, Cheeseman A, Steel T, et al. A multidisciplinary team approach to skull base chondrosarcomas. J Neurosurg 2001; 95(2):184–189. 54. Sekhar LN, Chanda A, Chandrasekar K, et al. Chordoma and chondrosarcoma. In: Winn HR, ed. Youmans Neurological Surgery, 5th ed. Philadelphia: Saunders, 2004:1283–1294. 55. Sen CN, Sekhar LN, Schramm VL, et al. Chordoma and chondrosarcoma of the cranial base: an 8-year experience. Neurosurgery 1989; 25(6):931–940; discussion 940–941. 56. Lanzino G, Sekhar LN, Hirsch WL, et al. Chordomas and chondrosarcomas involving the cavernous sinus: review of surgical treatment and outcome in 31 patients. Surg Neurol 1993; 40(5): 359–371. 57. Debus J, Schulz-Ertner D, Schad L, et al. Stereotactic fractionated radiotherapy for chordomas and chondrosarcomas of the skull base. Int J Radiat Oncol Biol Phys 2000; 47(3):591–596. 58. Cantrell RW, Ghorayeb BY, Fitz-Hugh GS. Esthesioneuroblastoma: diagnosis and treatment. Ann Otol Rhinol Laryngol 1977; 86(6 Pt 1):760–765. 59. Levine PA, Gallagher R, Cantrell RW. Esthesioneuroblastoma: reﬂections of a 21-year experience. Laryngoscope 1999; 109(10): 1539–1543. 60. Levine PA, McLean WC, Cantrell RW. Esthesioneuroblastoma: the University of Virginia experience 1960–1985. Laryngoscope 1986; 96(7):742–746. 61. Oskouian RJ Jr, Jane JA Sr, Dumont AS, et al. Esthesioneuroblastoma: clinical presentation, radiological, and pathological features, treatment, review of the literature, and the University of Virginia experience. Neurosurg Focus 2002; 12(5):e4. 62. Bilsky MH, Kraus DH, Strong EW, et al. Extended anterior craniofacial resection for intracranial extension of malignant tumors. Am J Surg 1997; 174(5):565–568. 63. Eden BV, Debo RF, Larner JM, et al. Esthesioneuroblastoma. Long-term outcome and patterns of failure—the University of Virginia experience. Cancer 1994; 73(10):2556–2562.

38.

skull base tumors: surgery perspective

64. Levine PA, Debo RF, Meredith SD, et al. Craniofacial resection at the University of Virginia (1976–1992): survival analysis. Head Neck 1994; 16(6):574–577. 65. Shah JP, Galicich JH. Esthesioneuroblastoma. Treatment by combined craniofacial resection. N Y State J Med 1979; 79(1):84–86. 66. Shah JP, Kraus DH, Bilsky MH, et al. Craniofacial resection for malignant tumors involving the anterior skull base. Arch Otolaryngol Head Neck Surg 1997; 123(12):1312–1317. 67. Polin RS, Sheehan JP, Chenelle AG, et al. The role of preoperative adjuvant treatment in the management of esthesioneuroblastoma: the University of Virginia experience. Neurosurgery 1998; 42(5):1029–1037. 68. Sheehan JM, Sheehan JP, Jane JA Sr, et al. Chemotherapy for esthesioneuroblastomas. Neurosurg Clin N Am 2000; 11(4):693– 701.

399

69. Diaz EM Jr, Johnigan RH 3rd, Pero C, et al. Olfactory neuroblastoma: the 22-year experience at one comprehensive cancer center. Head Neck 2005; 27(2):138–149. 70. Unger F, Haselsberger K, Walch C, et al. Combined endoscopic surgery and radiosurgery as treatment modality for olfactory neuroblastoma (esthesioneuroblastoma). Acta Neurochir (Wien) 2005; 147(6):595–602. 71. Unger F, Walch C, Stammberger H, et al. Olfactory neuroblastoma (esthesioneuroblastoma): report of six cases treated by a novel combination of endoscopic surgery and radiosurgery. Minim Invasive Neurosurg 2001; 44(2):79–84. 72. Walch C, Stammberger H, Anderhuber W, et al. The minimally invasive approach to olfactory neuroblastoma: combined endoscopic and stereotactic treatment. Laryngoscope 2000; 110(4):635– 640.

3 9

Skull Base Tumors: Fractionated Stereotactic Radiotherapy Perspective René-Olivier Mirimanoff and Alessia Pica

Introduction Skull base tumors encompass a broad spectrum of different intracranial benign or malignant neoplasms with a great variety of histologies and locations. Taken together, they account for up to a third of all intracranial tumors. The most common types are represented by meningioma, pituitary adenoma, schwannoma of the acoustic nerve, craniopharyngioma, optic chiasm and optic nerve glioma, chordoma and chondrosarcoma, chemodectoma or paraganglioma, giant cell tumors, and carcinoma of the skull base. Some, like the meningioma, can involve almost any part of the anterior, middle, or posterior fossa, whereas others, like the chemodectoma, arise from welldeﬁned anatomic structures such as, for example, the foramen jugulare. Surgery still remains the cornerstone in the management of most skull base tumors [1]. However, in spite of a better understanding of the natural behavior of these tumors, the great progress in anesthesiology, and the reﬁnements in surgical techniques, a complete surgical removal is impossible in many instances or made at a price of considerable neurologic deﬁcits. This is why noninvasive treatments, such as radiotherapy and in particular high-precision radiotherapy, that is, radiosurgery (RS), stereotactic fractionated radiotherapy (STRT), intensity-modulated radiation therapy (IMRT), and particle therapy, play an increasing role in the multidisciplinary management of most skull base tumors. A number of the most common tumors, including meningioma, pituitary adenoma, and schwannoma of the acoustic nerve, will be discussed in separate chapters of this book. We will focus here on chordoma and chondrosarcoma on the one hand, and on the chemodectoma/paraganglioma/glomus tumor category on the other, with particular emphasis on their treatment by RS and STRT.

Chordoma and Chondrosarcoma Chordoma: General Features Chordomas are rare tumors and represent only 0.1% to 0.2% of all primary intracranial neoplasms [2]. Chordomas are of

neuroectodermal origin and are presumed to develop from notochordal remnants [3] (Fig. 39-1). There are several histopathologic subcategories, with or without predominance of particular cell types. Those include the physaliferous type, the epithelioid and the chondroid cell types [4]. Immunohistochemistry shows that almost all tumors stain positively for the epithelial membrane antigen, cytokeratins, S-100 proteins, and are negative for HMB 45 and desmin [4]. One third of all chordomas arise from the skull base, the remainder from the sacrum and the mobile spine. Although they do grow slowly, and rarely metastasize, they are quite often lethal due to their local progression. They commonly involve vital neural or bony structures (Fig. 39-2), thus compromising the effectiveness of surgical or radiation therapy [5]. Intracranial chordomas typically arise from the clivus and can invade the dura, extend in any direction, for example toward the foramen magnum, compress the brain stem, or inﬁltrate anteriorly the cavernous sinus [6].

Chondrosarcoma: General Features Chondrosarcomas are also rare tumors that can arise from the skull base. Like chordoma, they represent 0.15% of all intracranial tumors [7] although certain series suggest that they are even less common than chordoma [8–10]. It is thought that they originate from primitive mesenchymal cells or embryonal remnants of the cartilaginous matrix of the cranium [11] (Fig. 39-3). They are sometimes mistakenly diagnosed as chordoma, but it is quite important to distinguish them from the latter as they have a much better prognosis [8, 12–16]. Several subtypes have been proposed that include the hyaline, the myxoid, and mixed hyaline-myxoid; the mesenchymal and dedifferentiated subtypes are less common and disclose a more aggressive behavior, whereas the clear cell subtype is extremely rare [17, 18]. On immunohistochemistry, almost all chondrosarcomas stain for S-100 protein, but none for keratin and less than 10% for epithelial membrane antigens [18]. Thus, tumor markers are helpful adjuncts to differentiate between chordoma and chondrosarcoma: chordomas express cytokeratin and epithelial membrane antigens, whereas chondrosarcomas lack the former

401

402

r-o. mirimanoff and a. pica

FIGURE 39-1. Pathology slide of a chordoma.

FIGURE 39-3. Pathology slide of a skull base chondrosarcoma.

and rarely stain for the latter. Like chordomas, chondrosarcomas tend to invade local structures. Distant metastases are uncommon; Hassounah et al. in their literature review found 5 cases out of 50 reported patients [19].

cal behavior, and outcome, they have a good number of features in common and their overall management is relatively similar. There is no question that surgery remains the cornerstone of treatment. The surgeon’s goal should be, whenever possible, to carry out an en bloc resection with a gross total removal [20]. Historically, this has been difﬁcult because the areas involved are not easily accessible and contain vital structures [20]. However, major advances in surgical techniques and microsurgery in particular and a more generally accepted multidisciplinary approach have allowed neurosurgeons to perform more macroscopically complete resections. Maximal surgical cytoreduction is feasible both for chordoma and chondrosarcoma, resulting in encouraging survivals. With extensive surgery, Crockard et al. have reported 5-year survival rates of 77% and 93%, respectively, for chordoma and chondrosarcoma [15, 16]. Gay et al. obtained a 67% total or near total resection rate, and for these an 84% recurrencefree rate was achieved [8]. However, these results represent probably a favorable patient selection. In most cases, a genuine curative resection with clear margins in not realistic, and ultimately a majority of patients will recur and die from their disease. Even in chondrosarcoma, with a better prognosis than chordoma, surgery alone is insufﬁcient. In a series of 13 operated patients with chondrosarcoma from The Netherlands, only 3 received postoperative photon (2) or proton (1) radiotherapy, and the recurrence-free survival was only 43% at 5 years [11]. Adjuvant, postoperative radiotherapy has been considered and used for many years both for chordoma and chondrosarcoma. In spite of old claims that chondrosarcoma and chordoma are radioresistant tumors [21], conventional external beam radiotherapy was shown to provide useful and prolonged palliation in overt residual disease [5]. It was however suggested for a long time that these tumors, in order to be con-

Standard Treatment of Chordoma and Chondrosarcoma Although chordoma and chondrosarcoma of the skull base present with important differences in histopathology, biologi-

FIGURE 39-2. MRI of a skull base chordoma.

39.

skull base tumors: fractionated stereotactic radiotherapy perspective

trolled, need to receive relatively high-dose radiotherapy and that the total dose should be increased to 70 to 80 Gy [22, 23]. Because of the presence of neighboring critical structures like the brain stem or the temporal lobe, the total dose was limited to 50 to 60 Gy with conventional techniques, which is clearly insufﬁcient to control even microscopic residual disease in the long-term. Charged particles such as protons, helium, or neons are well suited for extremely precise localization of radiation and permit an increase of the total dose from 15% to 35% compared with conventional X-rays [10]. In their experience with charged particle irradiation of tumors of the skull base from 1997 to 1992, Castro et al. have treated 53 chordomas and 27 chondrosarcomas, with a mean dose of 65 Grayequivalent (GyE) (range, 60 to 85 GyE) [10]. Five-year local control was 63% for chordoma and 78% for chondrosarcoma [10]. Since then, a number of patients were treated with particle therapy, generally proton beam therapy in the United States, France, and Japan, with protocols delivering high doses of irradiation. At the Massachusetts General Hospital (MGH), 68 patients with either chordoma or chondrosarcoma treated with 160-MV proton beam with a median tumor dose of 69 cobalt-Grayequivalent (CGE) were shown to have an overall 5-year actuarial local control of their tumors of 82% [24]. In a later update from MGH, it was found that chordoma had a 31% rate of failure of which 95% were local [25]. It was also shown that the 10-year local control was the highest with chondrosarcoma, followed by male chordoma and female chordoma, with 94%, 65%, and 42% local control rates, respectively [14]. At Loma Linda University, 58 patients with chordoma and chondrosarcoma received proton beam radiotherapy with a mean dose of 70 CGE, with a local control rate of 92% for chondrosarcoma and 78% for chordoma [13]. Experience at Orsay with 45 patients treated with a combination of photon and proton beam radiotherapy (201 MV) to a median dose of 67 CGE disclosed a 3-year local control for chordoma and chondrosarcoma of 83% and 90%, respectively [26]. An update of the French series on 49 patients conﬁrmed their results, with 4-year overall survivals of 87% and 75%, respectively, age being the only pretherapeutic prognostic-related factor in univariate analysis [9]. Thirteen patients with chordoma were treated with proton beam therapy at Tsukaba University; they received a median dose of 72 Gy with 5-year local control and local-speciﬁc survivals of 42% and 72%, respectively [27]. In the experience with proton beam radiotherapy, late grades 3 to 4 toxicities were generally low, given the high radiotherapy dose delivered, and were reported to be between 4.5% and 7% [13, 26]. In conclusion, high-precision, high-dose proton beam radiotherapy is followed by an excellent local control for chondrosarcoma and a reasonably good local control for chordoma, with a low toxicity. Postoperative proton beam radiotherapy is often considered as the standard treatment for these tumors. We will examine next if high-precision and high-dose RS or STRT can be considered as alternative treatments to proton beam therapy.

403

Radiosurgery and Fractionated Stereotactic Radiotherapy of Chordoma and Chondrosarcoma Patient Selection As previously emphasized in the standard management of chordoma and chondrosarcoma, prior to any sort of postoperative treatment, including RS, a maximal surgical resection should be attempted if one wants to maximize the chances of local control [15, 16, 20, 28]. Radiosurgery should be well suited for postoperative residual disease or for inoperable recurrent disease, but one limiting factor is the tumor size. Lesions of more than 30 mm in diameter are generally not considered to be suitable for RS as is the case for other tumor types and locations [29–31]. However, in some situations, skull base tumors larger than 30 mm can undergo RS if they are in contact with the brain or other sensitive structures with only a small portion of their surface. Thus, much of the radiation fall-off occurs into bone, sinuses, and other regions [29]. In order to maximize the dose to the target volume and minimize the dose to the brain stem, the surgeon can sometimes use a dermal fat graft as a “spacer” to maintain a posterior displacement of the brain stem and thus facilitate high-dose RS [28]. Previous external beam radiotherapy is not necessarily a contraindication for RS; in this setting and in selected cases, RS was successfully performed with acceptable toxicity [28, 31, 32]. Obviously, a particularly careful assessment of the previous dose given to the critical structures added to that delivered by RS is mandatory before taking the decision of a retreatment. The indications for STRT are essentially the same as those for RS, except that STRT combines the precision of stereotactic positioning with the radiobiological advantage of fractionation, especially in large tumors [12]. Thus, in tumors larger than 30 mm in diameter, there is an advantage of using STRT rather than RS, at least in theory. In the series from Heidelberg, where STRT was used for chordoma and chondrosarcoma of the skull base, the median target volume was 56 cm3 (range, 17 to 215 cm3) for chordoma and 102 cm3 (range, 24 to 237 cm3) for chondrosarcoma [12]. This is in sharp contrast with the mean or median volumes described in the RS series, which ranged from 4.6 cm3 (range, 0.98 to 10.8 cm3) to 14.6 cm3 (range, 2.9 to 52 cm3) [31–33].

Treatment Techniques Planning techniques are similar to those used for any RS or STRT of other intracranial lesions. Immobilization systems including various stereotactic head frames (RS) or relocatable head masks (STRT) are used for imaging studies under stereotactic guidance and for treatment. All patients should undergo high-resolution computed tomography (CT) and if not contraindicated magnetic resonance imaging (MRI), with image fusion for treatment planning. For radiosurgery, Miller et al. have determined their volume by MRI or CT deﬁned at the 50% isodose volume encompassing radiographically evident tumor, without margin [32]. Kondziolka et al. have also generally used the 50% isodose line to best achieve a sharp fall-off

404

r-o. mirimanoff and a. pica

FIGURE 39-4. (A) BrainLAB plan of a case of chordoma (sagittal). (B) BrainLAB plan of a case of chordoma (axial).

of the speciﬁed dose [30]. In most situations, multiple isocenters were used, from 2 to 17 [31–35]. The average prescribed dose ranged from 12.9 to 18 Gy and generally depended on tumor volume [31–35]. For example, at the Mayo Clinic, for patients without prior or planned external beam radiotherapy, the dose to the 50% isodose line was 20 Gy, 18 Gy, or 16 Gy for tumor volumes of ≤4.2 cm3, 4.2 to 14.1 cm3, and ≥14.1 cm3, respectively [32]. Radiosurgery has been delivered either by a Gamma Knife unit [31–33, 35, 36] or by linac-based radiosurgery [28, 34]. With regard to fractionated STRT, the Heidelberg group have deﬁned the clinical target volume (CTV) as the visible tumor on CT and MRI and the potential residual tumor, taking preoperative imaging into account [12]. They added a 2-mm safety margin for the planning target volume (PTV). The median prescribed dose at isocenter was 66.6 Gy for chordoma and 64.3 Gy for chondrosarcoma, with a median daily dose of 1.8 Gy [12]. Figure 39-4A, B shows an example of an STRT treatment plan at our institution (CHUV) for a 59year-old patient who was not operated due to important comorbidities. The patient, with a skull base chordoma, received a total dose of 63 Gy at 1.8 Gy per fraction, using a micro-multileaf collimator with 6 ﬁxed beams from a 6-MV linear accelerator.

Treatment Outcomes Because of the rare occurrence of chordoma and chondrosarcoma, the published experience with RS and STRT is still very limited. Table 39-1 summarizes the few available series with RS and STRT for chordoma and Table 39-2 the experience

with chondrosarcoma. As can be seen, these small series of Gamma Knife radiosurgery (GKRS) have gathered altogether only 20 cases of chordoma and 16 cases of chondrosarcoma [31–33], with 2 reports on one [28] or two [34] additional patients treated with linear accelerator–based RS. The results suggest encouraging early response and local control, but for chordoma and chondrosarcoma, however, a much longer follow-up period is needed to assess the efﬁcacy of RS in this setting. The unique experience at the University of Heidelberg with STRT also discloses very encouraging results, with a more adequate observation period than that of the RS series [12]. It should also be remembered that patients treated by STRT had on average a much larger residual tumor volume compared with those who received GKRS (vide supra). Altogether, the only STRT series from Germany is more reliable than the RS series in terms of number of patients, length of follow-up, and adequate actuarial reporting of local control and survival. As far as complications of RS are concerned, they appear altogether to be quite acceptable. In the Mayo Clinic series, out of nine patients with chordoma or chondrosarcoma, one developed secondary amenorrhea after RS and external beam radiotherapy, and another patient presented with a gait disorder and short-term memory difﬁculties, but none of the patients suffered from cranial nerve injuries [37]. In the Pittsburgh experience, no single case of neurologic or endocrine toxicity was observed [31]. From the 45 patients treated with STRT, only one presented with symptomatic infarction of the pons [12].

Conclusion In conclusion, RS, either with GKRS or linac RS, and STRT appear to be safe treatments as postoperative complement to surgery or for recurrent chordoma and chondrosarcoma of the base of skull. These treatments appear to be followed by good short-term results, especially in chondrosarcoma. Because of the yet insufﬁcient follow-up, especially after RS, it is too early at this time to consider that either RS or STRT represents a valid alternative to the standard proton beam radiotherapy. These therapies deserve further investigation, and long-term evaluation will be necessary before considering them for standard practice, either for chordoma or chondrosarcoma.

TABLE 39-1. Treatment outcomes, chordoma. Series

Year

No. of patients

Local control

Survival

Median follow-up

Miller et al. [32]

1997

8

GKRS

15

100% (2 yr)

100% (2 yr)

NS

Muthukumar et al. [31]

1998

9

GKRS

18

66%*

NA

40 mo

Feigl et al. [33] Debus et al. [12]

2005 2000

3 37

GKRS Linac STRT

17 66.6

2/3 50% (5 yr)

NA 82% (5 yr)

17 mo 19 mo

Technique

*No separate analysis between chordoma and chondrosarcoma.

Median prescribed dose (Gy)

Comments

Symptomatic improvement in 5% of patients Clinical improvement or stabilization of symptoms in 73% of patients*

39.

skull base tumors: fractionated stereotactic radiotherapy perspective

405

TABLE 39-2. Treatment outcomes, chondrosarcoma. Series

Year

No. of patients

Muthukumar et al. [31]

1998

6

GKRS

18

66%*

NA

40 mo

Feigl et al. [33] Debus et al. [12]

2005 2000

10 8

GKRS Linac STRT

17 64.9

100% 100% (4 yr)

NA 100% (4 yr)

17 mo 19 mo

Technique

Median prescribed dose (Gy)

Local control

Survival

Median follow-up

Comments

Clinical improvement or stabilization of symptoms in 73% of patients*

*No separate analysis between chordoma and chondrosarcoma.

Chemodectoma/Paraganglioma Chemodectoma/Paraganglioma: General Features Chemodectoma are also called paraganglioma and include also glomus tumors. They are rare, with an estimated incidence of 1 per million people [37]. They generally affect patients in their sixth or seventh decade with a female predominance [38]. They arise from paraganglionic tissues that can be found in the bulbus jugularis (glomus jugulare) along the Jacobsen nerve (glomus tympanicum), along the vagus nerve (glomus vagale), and the carotid body [39, 40]. They are generally benign, highly vascular, and slow growing [41–43] (Fig. 39-5). In rare cases, they can present with a malignant behavior [44] with metastatic spread to bone, lung, and liver [45] or have an aggressive local-regional pattern [44]. They can also be multicentric, and this is more common in the familial cases [43]. In some instances, they can reach very large size and are referred to as giant tumors [43]. In general, however, their clinical behavior is indolent with an interval between the ﬁrst symptom and the diagnosis that can be as long as several years [39, 43]. Depending on their location, they can invade or compress adjacent structures, such as the temporal bone, the middle ear, the clivus, the jugular vein, the internal carotid artery, the cavernous sinus, the hypoglossal canal, and a series of cranial nerves from V to XII [41, 42, 46– 48] (Fig. 39-6). There are several classiﬁcations that take into

FIGURE 39-5. Pathology slide of a case of chemodectoma/ paraganglioma.

account tumor location, extent, and size. The Fisch classiﬁcation is one of the most commonly used [49].

Standard Treatment In spite of a long experience with surgery, the treatment of choice of chemodectoma is still a matter of controversy [47]. Surgery is an important but by far not the only option. Thanks to major progress in surgical techniques, neuroradiologic imaging, and cranial nerve monitoring, results in terms of complete removal and decreased immediate complications have been improved. Surgical techniques include conservative jugulopetrosectomy and infratemporal approach [47–49]. However, a careful study of 11 glomus tumors involving the temporal bone compared with 10 normal jugular foramen blocks has demonstrated that within the foramen jugulare, the cranial nerves have a multifascicular histoarchitecture (in particular the tenth cranial nerve) and glomus tumors can invade these nerves despite normal function [50]. Therefore, a total resection will often be impossible without sacriﬁce of these nerves. Not taking into account cranial nerve deﬁcits prior to the operation, postoperative cranial nerve damage can be quite substantial, with potential impairment to cranial nerves VII and IX to XII [47, 48]. In one series, preoperative deﬁcit of cranial nerves X to XII was

FIGURE 39-6. MRI of a case of chemodectoma.

406

r-o. mirimanoff and a. pica

between about 20% and 30% and postoperative deﬁcits increased to 25% to 50% [48]. In the same report, however, facial function recovery was observed in 95% of patients [48]. The rate of complete surgical removal is quite variable and claimed to be between 40% to 96% [47, 48]. Even in so-called complex tumors, gross total resection can be performed in a proportion of patients, but at a cost of a rather high rate of postoperative cranial nerve deﬁcits compared with the preoperative status [43]. In general, data regarding resection rates and the actual occurrence of complications are difﬁcult to interpret due to the heterogeneity and limited size of most series. Data on local control and long-term survival after surgery alone is scant. Nora et al. from the Mayo Clinic have reported on 59 carotid body tumors, of which only 3 (6%) presented with a recurrence after surgery, and only one developed metastatic disease [51]. Batsounis et al. evaluated 17 tumors; with a followup of between 7 months and 20 years, they noticed 2 deaths from other causes, 10 patients were reported to have no recurrence, but as many as 5 were lost to follow-up [52]. Pareschi et al. have operated on 37 glomus jugulare tumors, 96% of which were said to undergo a complete resection; no relapse was recorded, but the mean follow-up was only 4.9 years [47]. In their series of 28 patients with so-called complex glomus jugulare tumors, Al-Mefty and Teixeira were able to perform a complete resection in 24 patients and observed altogether 2 recurrences [43]. In other studies, however, local control was much lower, with 70% to 100% recurrence rates after surgery alone [53, 54]. It is also likely that because these tumors tend to recur at a late date, the actual failure rate is underestimated, especially if the follow-up is short, and/or the proportion of patients lost to follow-up is high, as in three of the quoted series [43, 47, 52]. Also, long-term mortality due to recurrence after surgery is not trivial and was reported to be between 5% and 13% [38, 55, 56]. Claims that paraganglioma are radioresistant tumors and that external beam radiotherapy is associated with long-term complications are still made by some [43]. They are based on older studies in which ancient orthovoltage techniques were used [57]. These claims are in fact unjustiﬁed as modern megavoltage and conformal-beam radiotherapy represent an excellent alternative or sometimes complementary treatment to surgery. Cole et al. have treated 32 tumors of the glomus jugulare or glomus vagale with megavoltage units at doses of 45 Gy in 5 weeks [57]. Their very long-term results have yielded a local control rate of 94% for patients at risk more than 10 years [57]. Konefal et al. have published their 26 cases of chemodectoma, irradiated with 45 Gy to 50 Gy. Of these cases, 15 of 16 glomus tympanicum, and 4 of 6 glomus jugulare lesions achieved longterm control, with a mean observation time of 10.5 years [45]. The largest series is that of the University of Florida, with 80 chemodectoma of the temporal bone, carotid body, or glomus vagale, treated by radiotherapy alone (72) or postoperatively (8) [40]. Crude local control was 94%; 5 local recurrences occurred between 2.6 and 18.8 years, of which 2 were salvaged by surgery [40]. Other series also indicated that a good local control can be obtained with radiotherapy alone or with postoperative radiotherapy [53, 54]. In their experience at the Royal Marsden Hospital, Powell et al. treated 46 patients with glomus jugulare or glomus tympanicum tumors with radiotherapy doses between 45 and 50 Gy and obtained a 75% actuarial control at 25 years [54]. It should be noted that in all previous radiotherapy

series, a proportion of patients (between 18% and 30%) had been previously—and sometimes heavily—pretreated by surgery. On the contrary to older reports using old orthovoltage techniques, complication rates were reported to be low in the most recent experience with megavoltage units [40, 45, 57]. The claim that these tumors are “radioresistant” was also probably based on the fact that after radiotherapy, they shrink but rarely disappear completely on follow-up studies [57]. Thus, long-term local control after radiotherapy can be deﬁned in this tumor group as the absence of both clinical and radiologic progression. External beam radiotherapy can also be used successfully in the rare instance of malignant chemodectoma, probably best in the postoperative setting [44]. In their literature survey, Springate et al. have reviewed the outcome of chemodectoma after various treatments [58]. They noted that local control after surgery, combined surgery and radiotherapy, or radiotherapy alone was 86%, 90%, and 92%, respectively. They also noticed that the treatment-related morbidity after surgical excision was frequent, whereas late complications were rare after radiotherapy [58].

Radiosurgery and Fractionated Stereotactic Radiotherapy of Chemodectoma Patient Selection The main reason to consider a nonsurgical procedure for chemodectoma, as we have seen, is that in many instances it may be very difﬁcult for a surgeon to avoid severe neurologic deﬁcits if a complete surgical removal is anticipated. Because external beam radiotherapy is an efﬁcient alternative, the main argument of the proponents of RS is that the incidence of complications might be expected to be lower with the latter treatment, because of a smaller volume of normal tissue exposed to the effect of radiation [46]. The criteria for patient selection for RS for chemodectoma are comparable with those used for external beam radiotherapy. RS and STRT have been used either as primary treatment, at progression after one or multiple surgeries, after embolization, postoperatively, or even after previous external beam radiotherapy [39, 41, 42, 46, 59]. The proportion of patients treated primarily with RS and STRT varied between 45% and 53% [39, 41, 42, 46]. In one large retrospective and multicenter study, 5 of 66 (7.6%) patients had been previously treated with fractionated external beam radiotherapy [39]. In the latter study, the tumor was limited to the foramen jugulare (Fisch type B or C) in 54%, whereas in 41% there was tumor spread to the skull base (Fisch type C) and in 26% there was intracranial extension (Fisch type D) [39]. As for any RS, one limiting factor is the tumor size. In the Mayo Clinic experience, the mean largest diameter was 3.3 cm (range, 1.7 to 4.7 cm), however 3 narrow tumors with a maximum diameter of more than 4 cm could be included because of a relatively lesser volume [46]. The indications for STRT are essentially the same as those for RS, STRT combining the precision of stereotactic positioning and the potential advantages of fractionation [12, 42]. Certainly, large tumors will be best treated by STRT. In the Heidelberg experience with STRT for chemodectoma, the median target volume was 71.8 cm3 (range, 10.5 to 212 cm3) [42], which is in sharp contrast with the median volumes described in the RS series, which ranged from 5.7 cm3 (range, 0.5 to 27 cm3) to 10.8 cm3 (range, 4.9 to 18 cm3) [39, 41, 46, 59, 60].

39.

skull base tumors: fractionated stereotactic radiotherapy perspective

407

Treatment Techniques Pretreatment imaging includes high-resolution CT, MRI, and preferably both and wherever needed angiography [41, 46]. For treatment planning, CT and MRI fusion are required, especially with linac-based RS/STRT. Head frames (RS) or relocatable head masks (STRT) are used for imaging under stereotactic guidance and treatment. For RS, the tumor volume is generally covered by median isodoses of 50% [41, 46]. For Gamma Knife RS, the median number of isocenters varied from 5 (range, 1 to 7) [41] to 7 (range, 3 to 15) [46]. RS single dose to the margin varied from 10 to 30 Gy. At the Mayo Clinic, the median dose was 15 Gy (range, 12 to 18 Gy) [46], at the University of Graz it was 13.5 Gy (range, 12 to 20 Gy) [41], at the University of Florida 15 Gy (range, 12.5 to 15 Gy) [59], at the University of Texas 16.3 Gy (range, 12 to 20 Gy) [60], and in the large international review 16.5 Gy (range, 10 to 30 Gy) [39]. Radiosurgery was delivered either with a Gamma Knife [39, 41, 46, 60] or with a linac-based SR [59]. With regard to STRT, the Heidelberg group deﬁned the clinical target volume as the macroscopic tumor on MRI plus a 10-mm margin along the involved vessels [42]. They added a 2-mm safety margin for the planning target volume (PTV). The median total dose was 57.6 Gy at a median daily dose of 1.8 Gy [42]. Figure 39-7 shows an example of an STRT treatment plan at our institution (CHUV) for a 55-year-old patient diagnosed with a jugular tympanic chemodectoma. This patient received radiosurgery with a single dose of 12 Gy using a micromultileaf collimator with 6 ﬁxed beams from a 6-MV linear accelerator.

Treatment Outcome At the present time, the experience with RS or STRT for chemodectoma is limited to a few series; all showed encouraging results. Table 39-3 summarizes the available data with RS and STRT. Altogether, the RS studies have gathered 104 patients (the 13 patients of Eustacchio et al. are included in the Liscak multicenter study), of which 5 were treated by linac-based RS and the rest by GKRS. The results disclosed a good local control, with on average 41% of partial responses, 57% of stable lesions, and only 3% of progressions. Similarly, the neurologic status was improved in 40%, stable in 55%, and worsened in 5% of cases. Information on long-term survivals is quite limited, but apparently no patient was reported to have died of tumor; median follow-up in all these series was quite short, on average only 32 months (Table 39-3). The unique Heidelberg series of STRT for chemodectoma also discloses encouraging results, with a 32% rate of partial response, 59% of stable disease, and 9% of progression [42]. The apparently slightly worse rate of progression is likely to be due on one hand to the much longer follow-up than that of the RS series (68.4 months vs. 32 months) and on the other hand to tumors with much larger volumes (vide supra). The neurologic improvement was also quite good, with 59% of patients who improved, 32% who remained stable, and 9% who deteriorated [42]. Altogether, the Heidelberg data are the most reliable in terms of length of follow-up and adequate actuarial reporting of local control and survival.

FIGURE 39-7. BrainLAB plan of a case of chemodectoma/ paraganglioma.

As far as complications of RS are concerned, it is difﬁcult to distinguish between neurologic worsening due to treatment or due to recurrence. However, the rate of reported complications appears to be low. Eustacchio et al. had no single case of complications in their 13 cases [41]. In the multi-institutional report of Liscak et al., 3 cases of neurologic deﬁcits out of 52 patients are described, 2 of which were permanent: facial nerve impairment and deafness in one case, and facial nerve impairment and vertigo in another [39]. Two further cases of inner ear inﬂammatory complications are also noted [39]. From the Mayo Clinic experience, only one case of temporary vertigo is described [46]. Again, this low rate of complications should be interpreted with caution as the average median follow-up of the RS studies is only 24 months in the largest one [39], and development of neuropathies are known to occur as late as 3 years after RS [29, 30]. Out of the 22 patients who beneﬁted from STRT, 3 experienced temporary xerostomia and 2 temporary taste impairment, 4 had middle-ear effusion, but none of these had more than a grade 2 CTC reaction, and all recuperated; no late adverse reactions from STRT were observed at followup [42].

Conclusion In conclusion, RS or STRT appear to be safe and efﬁcient in the primary treatment or at progression for chemodectoma. Only short follow-ups are available for most RS series, and updated results will be needed to assess long-term local control and complication rate. With regard to STRT, the results also appear to be quite good, with longer time of evaluation, but the data are so far based on the unique Heidelberg series and should be conﬁrmed by others. Both RS and STRT deserve further experience, and more data and longer follow-ups are necessary before considering them for standard practice for chemodectoma/paraganglioma.

408

r-o. mirimanoff and a. pica

TABLE 39-3. Treatment outcomes, chemodectoma.

Series

Year

No. of patients

Technique

Eustacchio et al. [41]

1999

13

GKRS

Liscak et al. [39]

1999

Jordan et al. [60]

2000

Foote et al. [46]

2002

66 (follow-up in 52, radiologic follow-up in 47) 8 (radiologic follow-up in 7) 25

Feigenberg et al. [59]

2002

Zabel et al. [42]

2004

Median marginal dose (Gy)

Median follow-up

Local control

Survival (%)

13.5

PR 4/10* SD 6/10

11/13†

46 mo

GKRS

16.6

PR 15/47 SD 28/47

NS

24 mo

GKRS

16.3

PR 4/7 NC 3/7

NS

27 mo

GKRS

15

80% at 5 yr‡

37 mo

5

Linac RS

15

100%

27 mo

22

Linac STRT

57.6

PR 8/25 NC 17/25 PR or SD 3/5 PRO 2/5 PR 7/22 SD 13/22 PRO 2/22 Actuarial 5- & 10-yr LC: 90.4%

Actuarial overall survival at 5 & 10 yr 89.5§

5.7 yr

Neurologic status

Improved in 5/13 Unchanged in 5/13 Stroke in 1/13 Improved in 15/52 Unchanged in 34/52 Worsened in 3/52

Improved in 4/8 Unchanged in 3/8 Worsened in 1/8 Improved in 15/25 Unchanged in 10/25 NS Improved in 13/22 Unchanged in 7/22 Worsened in 2/22

*Two lost to follow-up, one follow-up <1 year. †Two deaths from unrelated causes. ‡Three deaths from unrelated causes. §One patient died of PRO, one patient died of colon cancer.

References 1. Prabhu S, Demonte F. Treatment of skull base tumors. Curr Opin Oncol 2003; 15:209–212. 2. Watkins L, Khudados ES, Kaleoglu M, et al. Skull base chordoma: a review of 38 patients, 1955–88. Br J Neurosurg 1993; 7:241–248. 3. Bouropoulou V, Bosse A, Roessner A, et al. Immunohistochemical investigation of chordomas: histogenic and differential diagnostic apsects. Curr Top Pathol 1989; 80:183–203. 4. Sell M, Sampaolo S, Di Lorio G, et al. Chordomas: a histological and immunohistochemical study of cases with or without recurrent tumors. Clin Neuropathol 2004; 23:277–285. 5. Catton C, O’Sullivan B, Bell R, et al. Chordoma: long-term followup after radical photon irradiation. Radiother Oncol 1995; 41: 67–70. 6. Maire JP, Liguoro D, San Gelli F. Volume tumoral macroscopique (GTV) et volume-cible anatomoclinique (CTV) dans la radiothérapie des tumeurs bénignes de la base du crâne. Cancer Radiother 2001; 5:581–596. 7. Cianfriglia F, Pompili A, Occhipiati E. Intercranial malignant cartilaginous tumors. Report of 2 cases and review of literature. Act Neurochir 1978; 45:163–175. 8. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the cranial base: results and follow-up of 60 patients. Neurosurgery 1995; 36:887–896. 9. Noel G, Jaffret E, de Crevoisier R, et al. Irradiation par association de photons et de protons des chordomes et des chondrosarcomes de la base de crâne et du rachis cervical. Bull Cancer 2002; 89:713–723. 10. Castro JR, Linstadt DE, Bahary JP, et al. Experience in charged particle irradiation of tumors of the skull base: 1977–1992. Int J Radiat Oncol Biol Phys 1994; 29:647–655. 11. Korten AGC, ter Berg HJW, Spincemaille GH, et al. Intracranial chondrosarcoma: review of the literature and report of 15 cases. J Neurol Neurosurg Psychiatry 1998; 65:88–92.

12. Debus J, Schulz-Ertner D, Schad L, et al. Stereotactic fractionated radiotherapy for chordomas and chondrosarcomas of the skull base. Int J Radiat Oncol Biol Phys 2000; 47:591–596. 13. Hug EB, Loredo LN, Slater JD, et al. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg 1999; 91:432–439. 14. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlenther Onkol 1999; S2(Suppl):57–63. 15. Crockard HA, Steel T, Plowman N, et al. A multidisciplinary team approach to skull base chordomas. J Neurosurg 2001; 95:175– 183. 16. Crockard HA, Cheeseman A, Steel T, et al. A multidisciplinary team approach to skull base chondrosarcomas. J Neurosurg 2001; 95:184–189. 17. Koch BB, Karnell LH, Hoffman HT, et al. National Cancer database report on chondrosarcoma of the head and neck. Head Neck 2000; 22:408–425. 18. Roseberg AE, Nielsen GP, Keel SB, et al. Chondrosarcoma of the base of the skull: a clinicopathologic study of 200 cases with emphasis on its distinction from chordoma. Am J Surg Pathol 1999; 23:1370–1378. 19. Hassounah M. Al-Mefty O, Akhtar M, et al. Primary cranial and intracranial chondrosarcoma. A survey. Acta Neurochir 1985; 78:123–132. 20. Neff B, Sataloff RT, Storey L, et al. Chondrosarcoma of the skull base. Laryngoscope 2002; 112:134–139. 21. Krayenbuhl H, Yasargil MG. Cranial chordomas. Prog Neurol Surg 1975; 6:380–434. 22. Rich TA, Schiller A, Suit HD, et al. Clinical and pathologic review of 48 cases of chordoma. Cancer 1985; 56:182–187. 23. Pearlman AW, Friedman M. Radical radiation therapy of chordoma. Am J Radiol 1970; 108:333–341. 24. Austin-Seymour M. Munzenrider J, Goitein M, et al. Fractionated proton radiation therapy of chordoma and low grade chondrosarcoma of the base of the skull. J Neurosurg 1985; 70:13–17.

39.

skull base tumors: fractionated stereotactic radiotherapy perspective

25. Fagundes MA, Hug EB, Liebsch NJ, et al. Radiation 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; 15:579–584. 26. Noel G, Habrand JL, Mammar H. Combination of photon and proton radiation therapy for chordomas and chondrosarcomas of the skull base: the Centre de Protonthérapie d’Orsay experience. Int J Radiat Oncol Biol Phys 2001; 51:392–398. 27. Igaki H, Tokuuye K, Okumura T, et al. Clinical results of proton beam therapy for skull base chordoma. Int J Radiat Oncol Biol Phys 2004; 60:1120–1126. 28. Crockard A, Macaulay E, Plowman PN. Stereotactic radiosurgery VI. Posterior displacement of the brainstem facilitates safer high dose radiosurgery for clival chordoma. Br J Neurosurg 1999; 13:65–70. 29. Pollock BE, Foote RL. The evolving role of stereotactic radiosurgery for patients with skull base tumors. J Neurooncol 2004; 69:199–207. 30. Kondziolka D, Lunsford D, Flickinger JC. The role of radiosurgery in the management of chordoma and chondrosarcoma of the cranial base. Neurosurgery 1991; 29:38–45. 31. Muthukumar N, Kondziolka D, Lunsford LD, et al. Stereotactic radiosurgery for chordoma and chondrosarcoma: further experiences. Int J Radiat Oncol Biol Phys 1998; 41:387–392. 32. Miller RC, Foote RL, Coffey RJ, et al. The role of stereotactic radiosurgery in the treatment of malignant skull base tumors. Int J Radiat Oncol Biol Phys 1997; 39:977–981. 33. Feigl GC, Bundschuh O, Gharabaghi A, et al. Evaluation of a new concept for the management of skull base chordomas and chondrosarcomas. J Neurosurg 2005; S102(Suppl):165–170. 34. Kocher M, Voges J, Staar S, et al. Linear accelerator radiosurgery for recurrent malignant tumors of the skull base. Am J Clin Oncol 1998; 21:18–22. 35. Tanaka T, Kobayashi T, Kidu Y, et al. The results of gamma knife radiosurgery for malignant skull base tumors. No Shinkei Geka 1996; 24:235–239. 36. Pamir MN, Kilic T, Türe U, et al. Multimodality management of 26 skull base chordomas with 4-year mean follow-up: experience at a single institution. Acta Neurochir 2004; 146:343–354. 37. Thedinger BA, Glasscock ME, Cueva RA, et al. Postoperative radiographic evaluation after acoustic neuroma and glomus jugulare tumor removal. Laryngoscope 1992; 102:261–266. 38. Brown JS. Glomus jugulare tumors revisited; a ten-year statistical follow-up of 231 cases. Laryngoscope 1985; 95:284–288. 39. Liscak R, Vladyka V, Wowra B, et al. Gamma-knife radiosurgery of the glomus jugulare tumors—early multicentre experience. Acta Neurochir 1999; 141:1141–1146. 40. Hinerman RW, Mendenhall WM, Amdar RJ, et al. Deﬁnitive radiotherapy in the management of chemodectomas arising in the temporal bone, carotid body and glomus vagale. Head Neck 2001; 23:363–371. 41. Eustacchio S, Leber K. Trummer M, et al. Gamma-knife radiosurgery for glomus jugulare tumors. Acta Neurochir 1999; 141: 811–818.

409

42. Zabel A, Milker-Zabel S, Huber P, et al. Fractionated stereotactic conformal radiotherapy in the management of large chemodectomas of the skull base. Int J Radiat Oncol Biol Phys 2004; 58: 1445–1450. 43. Al-Mefty O, Teixeira A. Complex tumors of the glomus jugulare: criteria, treatment and outcome. J Neurosurg 2002; 97:1256– 1265. 44. Mayer R, Fruhwirth J, Beham A, et al. Radiotherapy as adjunct to surgery for malignant carotid body paragangliomas presenting with lymph node metastases. Strahlenther Onkol 2000; 76:356– 360. 45. Konefal JB, Pilepich MV, Spector GJ, et al. Radiation therapy in the treatment of chemodectomas. Laryngoscope 1987; 97:1331– 1335. 46. Foote RL, Pollok BE, Gorman DA, et al. Glomus jugulare tumors: tumor control and complications after stereotactic radiosurgery. Head Neck 2002; 24:332–339. 47. Pareschi R, Righini S, Destiﬁ D, et al. Surgery of glomus tumors. Skull Base 2003; 13:149–157. 48. Green JD, Brackmann DE, Nguyen CD, et al. Surgical management of previously untreated glomus jugulare tumors. Laryngoscope 1994; 104:917–921. 49. Fisch U, Fagan P, Valavanis A. The infratemporal fossa approach for the lateral skull base. Otolaryngol Clin North Am 1984; 17:513–522. 50. Sen C, Hagne K, Kachara R., et al. Jugular foramen: microscopic anatomic features and implications for neural preservation with reference to glomus tumors involving the temporal bone. Neurosurgery 2001; 48:838–847. 51. Nora JD, Hallet JW, O’Brien PC, et al. Surgical resection of carotid body tumors: long-term survival, recurrence and metastasis. Mayo Clin Proc 1988; 63:348–352. 52. Bastounis E, Maltezos C, Pikolis E, et al. Surgical treatment of carotid body tumors. Eur J Surg 1999; 165:198–202. 53. Reddy EK, Mansﬁeld CM, Hartman GV. Chemodectoma of glomus jugulare. Cancer 1983; 52:337–340. 54. Powell S, Peters N, Harmer C. Chemodectoma of the head and neck: results of treatment in 84 patients. Int J Radiat Oncol Biol Phys 1992; 22:919–924. 55. Spector GJ, Sobol S. Surgery for glomus tumors at the skull base. Otolaryngol Head Neck Surg 1980; 88:524–530. 56. Rosenwasser H. Long-term results of tehray of glomus jugulare tumors. Arch Otolaryngol 1973; 97:49–54. 57. Cole JM, Beiler D. Long-term results of treatment for glomus jugulare and glomus vagale tumors with radiotherapy. Laryngoscope 1994; 104:1461–1465. 58. Springate SC, Haraf D, Weichselbaum RR. Temporal bone chemodectomas—comparing surgery and radiation therapy. Oncology 1991; 5:131–137. 59. Feigenberg SJ, Mendenhall WM, Hinerman RW, et al. Radiosurgery for paraganglioma of the temporal bone. Head Neck 2002; 24:384–389. 60. Jordan J, Roland PS, McManus C, et al. Stereotactic radiosurgery for glomus jugulare tumors. Laryngoscope 2000; 110:35–38.

4 0

Head and Neck Tumors Daniel T.T. Chua, Jonathan Sham, Kwan-Ngai Hung, and Lucullus Leung

Introduction The technique of stereotactic radiosurgery has proved to be effective in the treatment of a variety of intracranial tumors, both benign and malignant, and abnormal conditions such as vascular malformations. Because of the proximity and the frequent involvement of the base of the skull, head and neck tumors as a group have also been explored as targets for extracranial radiosurgery. Because malignant head and neck tumors are usually highly aggressive and inﬁltrative, radiosurgery alone is seldom appropriate as the primary treatment modality for newly diagnosed disease. Radiosurgery, however, may offer signiﬁcant beneﬁts when used as a salvage treatment for recurrent disease or as a boost treatment after conventional radiotherapy. Among all the head and neck cancers, nasopharyngeal carcinoma (NPC) has attracted more attention with regard to the application of radiosurgery due to its unique location near the skull base, pattern of growth and invasion, relative radiosensitivity, and high incidence observed in some countries. In this chapter, the clinical results of using radiosurgery in the management of head and neck tumors are reviewed and discussed, with emphasis on the management of NPC.

Nasopharyngeal Carcinoma

treated in the modern era. In that report, based on the results of 2687 patients treated at all public oncology centers in Hong Kong during the period 1996–2000, the observed 5-year local control and survival rates were 85% and 75%, respectively [2]. The 5-year local control rates for T1-2 and T3-4 disease were 88% and 79%, respectively. These patients were mostly treated by radiotherapy alone, and only 23% had additional chemotherapy. The results showed signiﬁcant improvement when compared with those reported in old series. Despite the improved results, about 11.7% of patients still developed local failures, and 8.2% had isolated local failures without regional or distant failures. In order to further improve local tumor control, combined chemo-radiotherapy has been used in advanced-stage disease, which showed improvement in local control compared with radiotherapy alone [3–6]. Another way to improve local control is by escalating the dose delivered to primary tumor based on the established dose-volumecontrol relationships in NPC [7, 8]. For early-stage disease, dose escalation can be achieved by brachytherapy using intracavitary intubation. For advanced-stage disease, dose escalation is best delivered by either conformal or stereotactic radiotherapy.

Conventional Salvage Treatment Options for Local Failures and Selection Criteria Conventional Salvage Treatment Options

Local Control After Radiotherapy Radiotherapy is the mainstay of treatment for all stages of NPC without distant metastases. The largest series on treatment outcome of NPC was reported by Lee et al. based on 5037 patients treated at a single institution during the period 1976–1985 [1]. The 10-year local control and survival rates were 61% and 43%, respectively. In the past decade, advances in both imaging and radiotherapy techniques have had signiﬁcant impact on the management of NPC. As a result, the outlook of patients with NPC treated in recent years, even for those presenting with a more advanced stage, is likely to have improved signiﬁcantly. This is supported by a recent report on the outcome of patients with newly diagnosed NPC

Aggressive salvage treatment should be considered and offered to patients with local failures of NPC whenever possible because a signiﬁcant proportion of them can still achieve long-term survival after successful treatment. Many salvage treatments are available, and the choice of method depends on several factors including extent of disease, site of involvement of disease, any synchronous nodal relapse, cumulative radiation dose already received by the patient, the patient’s general condition and preference, and expertise available. In general, salvage treatments for NPC can be classiﬁed into three types: surgery using different approaches such as maxillary swing or transmandibular resection; brachytherapy using intracavitary intubation, mold application, or gold grain implantation; and

411

412

d.t.t. chua et al.

external reirradiation using various techniques including stereotactic radiosurgery and radiation.

Patient Selection Criteria for Conventional Salvage Treatment A team approach involving both radiation oncologist and head and neck surgeon is recommended in formulating a treatment plan for local failures of NPC. All patients should be restaged by physical examination, nasopharyngoscopy and biopsy, and computed tomography (CT) and/or magnetic resonance imaging (MRI) of nasopharynx and neck. Patients with disease conﬁned to nasopharynx (rT1) are suitable candidates for either surgery or brachytherapy. For brachytherapy, patients with unilateral small-volume disease are best treated by gold grain implantation, whereas those with more diffuse or bilateral disease should be treated by intracavitary intubation. Surgery should be considered in patients with bulky disease, especially those already receiving a high cumulative radiation dose to nasopharynx. Patients with limited extension to nasal fossa, parapharyngeal space, and oropharynx (rT2) may also be surgical candidates although most would require external reirradiation. For patients with skull base involvement (rT3) or intracranial extension of disease (rT4), external reirradiation is usually the only option available, although reirradiation with curative intent is often difﬁcult due to the large numbers of critical structures in the vicinity of the target that were already irradiated to a high dose during primary radiotherapy.

Results of Conventional Salvage Treatments In salvaging local failures of NPC, a better outcome was usually observed in patients treated for persistent disease after primary radiotherapy then recurrent disease. Brachytherapy in the form of intracavitary intubation or gold grain implantation can be employed in salvaging early-stage local failure. Our institution has been using gold grain implantation as salvage treatment for NPC since 1986, and the reported 5-year local control rates were 87%, 63%, and 23% for persistent disease, ﬁrst recurrence, and second recurrence, respectively [9]. Using highdose-rate intracavitary intubation, Leung et al. reported an excellent 5-year local relapse-free survival rate of 85% in 87 NPC patients with persistent local disease after radiotherapy [10]. The same group also reported a lower control rate when intracavitary brachytherapy alone was used in salvaging local recurrence in a small group of patients (n = 8), with a 3-year local relapse-free survival rate of 42%. Surgical resection of persistent or recurrent NPC has also been attempted using various approaches, with reported control rates of 31% to 52% [11–14], but is not commonly practiced because of limited experience in most centers. For patients requiring external reirradiation usually due to extensive disease not amenable to brachytherapy or surgery, treatment results remained poor especially in advanced T stage. The reported 5-year survival rates after external reirradiation ranged from 7.6% to 36% with the use of conventional two-dimensional treatment planning and radiotherapy [15–17] and 12.4% in a mixed cohort of patients treated either with conventional two-dimensional radiotherapy or three-dimensional conformal radiotherapy [18]. A high incidence of late complication was commonly

observed after external reirradiation, the majority being neurologic damage and soft tissue ﬁbrosis. In our experience, patients with rT1-2 NPC treated by conventional two-dimensional external reirradiation alone had a 5-year survival rate of about 57%, but late neuroendocrine complications were common [17]. The combination of intracavitary brachytherapy and external radiotherapy has also been employed for recurrent NPC with the advantage of sparing more normal tissues while achieving a high total dose to the target. In one series by Lee et al., combined intracavitary brachytherapy and external radiotherapy yielded a superior 5-year local control rate of 45% compared with 32% by external radiotherapy and 29% by brachytherapy [19]. The outcome of patients with local failures of NPC may improve with the use of intensity-modulated radiation treatment, which can achieve better dose distribution while sparing dose-limiting critical structures [20]. Using intensity-modulated radiation treatment for high-dose (68 to 70 Gy) reirradiation of NPC, Lu et al. reported 100% locoregional control rate in 49 patients, although the longest follow-up was only about 1 year [21].

Radiosurgery for Local Failures of NPC: The Queen Mary Hospital Experience (Case Study 40-1 and Case Study 40-2) Patient Population Between March 1996 and February 2005, 48 patients received radiosurgery as a salvage treatment for local failures of NPC at Queen Mary Hospital, Hong Kong. All patients had previously received radical radiotherapy for undifferentiated type of NPC

Case Study 40-1 A 75-year-old man was diagnosed with T1N0 NPC in 1985. He underwent external radiotherapy with a dose of 61 Gy delivered in 26 fractions using a fraction dose of 2.5 to 3 Gy. The biological equivalent dose to 2 Gy (using α/β of 10) was 65 Gy. He was incidentally found to have local recurrence in 2000, after a long latency period of 15 years. Assessment showed a small tumor involving the left lateral wall of the nasopharynx, and biopsy showed undifferentiated carcinoma, hence rT1 disease (Fig. 40-1A). Biopsies from posterior and right lateral walls were negative. Surgery was offered but refused by the patient, and the site of lesion was considered not suitable for brachytherapy because of involvement of the cushion. Stereotactic radiosurgery using a single isocenter with 27.5-mm collimator was performed, with a dose of 12.5 Gy delivered to 80% isodose line (Fig. 40-1B). In this patient, both imaging and endoscopy showed that tumor was lateralized and conﬁned to the lateral wall, hence only part of the nasopharynx was treated. Follow-up CT scan performed at 3 and 12 months after radiosurgery showed complete resolution of tumor (Fig. 40-1C). Endoscopy and biopsy after radiosurgery were also negative. The patient has now been followed-up for more than 4 years with no evidence of relapse or major complications.

40.

head and neck tumors

FIGURE 40-1. (A) Baseline CT of a patient with recurrent T1 nasopharyngeal carcinoma showing mass at left lateral wall of nasopharynx (arrow). (B) Planning CT showing isodose distribution of target covered by single isocenter. The prescribed dose to target periphery was 12.5 Gy. (C) Follow-up CT taken 12 months after radiosurgery showing complete resolution of tumor.

413

ynx or with limited extension to nasal fossa, parapharyngeal space, and oropharynx were ﬁrst evaluated jointly by a surgeon and a radiation oncologist for brachytherapy or surgery, and radiosurgery was offered only to those patients whose disease was deemed not amenable to these salvage treatments. At our institution, brachytherapy for persistent or recurrent nasopharyngeal carcinoma was usually performed using gold grain implantation. This technique was applicable only to patients with a well-deﬁned and relatively small-volume tumor conﬁned to the nasopharynx. Patients with extensive mucosal recurrence involving a large area were not suitable for gold grain implantation. In addition, patients with relapse at the cushion near the opening of eustachian tube were also not suitable for the procedure because gold grain could not be directly implanted into the relatively thin mucosa overlying the cartilaginous cushion. In general, patients with small-volume mucosal tumor were usually treated by gold grain implantation, with surgery reserved for those with more bulky tumor or those with relapse at the cushion. Patients with disease amenable to surgery or brachytherapy who refused these treatments or were considered to be medically contraindicated for these procedures were also offered radiosurgery.

Target Localization with or without concurrent/adjuvant chemotherapy. Although the disease in some patients that relapsed after a long latency period may actually represent a second primary in the nasopharynx, most patients with local failures had residual tumor after prior treatment. We arbitrarily classiﬁed these residual tumors into persistent and recurrent disease based on the time interval from primary radiotherapy: persistent disease was deﬁned as local failure diagnosed within 6 months of completion of primary radiotherapy, and recurrent disease was deﬁned as that occurring beyond 6 months. Using these deﬁnitions, half of the patients were treated for persistent disease and the other half for recurrent disease. None of these patients received additional treatment after radiosurgery unless progression or further failure was documented. Table 40-1 summarizes patient characteristics of the whole group treated by radiosurgery; those with persistent disease and those with recurrent disease. Prior to radiosurgery, all patients were restaged by physical examination, nasopharyngoscopy and biopsy, and CT of nasopharynx and neck. Patients with disease conﬁned to nasophar-

Case Study 40-2 A 45-year-old man was diagnosed with T3N0 NPC in 1997. He received a course of external radiotherapy with a dose of 68 Gy in 34 fractions followed by a booster dose of 10 Gy in 4 fractions to the right parapharyngeal space. He was found to have local recurrence involving the right side of the nasopharynx extending to the nasal fossa about 1 year after treatment (Fig. 40-2A). Biopsies from the nasal fossa and the right posterior wall both showed undifferentiated carcinoma, but biopsies from the left side of the nasopharynx were negative. The recurrent stage was rT2N0, which was not suitable for brachytherapy or surgery. The patient was treated with stereotactic

Tumor extent was assessed by imaging and endoscopic examination before radiosurgery. For imaging study, axial contrast CT with a slice thickness of 2.5 to 3 mm was performed in all patients, supplemented by axial contrast MRI with a slice thickness of 3 mm in 73% of patients. PET-CT was not performed in this patient group for target localization, although we are currently also evaluating its role in salvaging local failures of NPC. Target volume was deﬁned as any abnormal soft tissue mass and/or contrast-enhancing areas as shown in axial imaging plus a margin of about 2 to 3 mm. Endoscopic examination was always performed irrespective of imaging ﬁndings to assess the mucosal extent of tumor, including any spread to nasal cavity or oropharynx. In our experience, endoscopic examination is the most reliable means to access the mucosal extent of disease and should not be omitted even when imaging showed that only part of the nasopharynx was involved. During endoscopy, multiple biopsies were taken from both sides of the nasopharynx to assess the macroscopic as well as microscopic extent of tumor. In patients just completing primary radiotherapy,

radiosurgery using two isocenters with collimator sizes of 25 mm and 27.5 mm. Because of the larger bulk of tumor, a higher dose of 14 Gy was delivered to 80% isodose line (Fig. 40-2B). Endoscopy and biopsies after radiosurgery showed no evidence of residual disease. Follow-up CT was performed at 6 months after treatment, which showed complete resolution of tumor (Fig. 40-2C). He was found to have asymptomatic right temporal lobe necrosis 15 months after radiosurgery. The complication was thought to be related more to the ﬁrst course of radiotherapy, which delivered a high dose to the right temporal lobe to ensure adequate coverage of tumor extent. He was later found to have lung metastases and died of uncontrolled systemic disease 24 months after radiosurgery.

414

d.t.t. chua et al.

FIGURE 40-2. (A) Baseline CT of a patient with recurrent T2 nasopharyngeal carcinoma showing mass lesion at right side of nasopharynx extending to nasal fossa. (B) Planning CT showing isodose distribution

of target covered by two isocenters. The prescribed dose to target periphery was 14 Gy. (C) Follow-up CT taken 6 months after radiosurgery showing complete resolution of tumor.

biopsies were routinely taken from bilateral roofs, lateral walls, and posterior walls. In patients with tumor apparently localized to one part of the nasopharyngeal mucosa, mapping of tumor extent was performed by taking multiple biopsies from grossly uninvolved mucosa surrounding the tumor. The results of mapping were then used to guide the extent of target coverage for radiosurgery.

Radiosurgery Planning and Treatment

TABLE 40-1. Characteristics of patients treated with radiosurgery for local failures of nasopharyngeal carcinoma at Queen Mary Hospital, Hong Kong.

Gender Male (n) Female (n) Age <50 (n) ≥50 (n) Median (years) Range (years) rT classiﬁcation rT1 (n) rT2 (n) rT3 (n) rT4 (n) Prior local failures Yes (n) No (n) Previous salvage Rx Surgery (n) ERT (n) Radiosurgery (n) Cumulative ERT dose 65–68 Gy (n) 70–76 Gy (n) 125–130 Gy (n) Interval from end of ﬁrst ERT ≤6 months (n) >6 months to 1 year (n) >1 to 2 years (n) >2 to 3 years (n) >3 years (n) Median (months) Range (months) ERT, external radiotherapy.

Persistent disease (n = 24)

Recurrent disease (n = 24)

All (n = 48)

17 7

18 6

35 13

14 10 45 32–86

14 10 47 35–84

28 20 46 32–86

18 2 4 0

9 4 7 4

27 6 11 4

0 24

9 15

9 39

0 0 0

4 7 1

4 7 1

22 2 0

16 1 7

38 3 7

24 0 0 0 0 4 3–6

0 4 3 5 12 36 9–197

24 4 3 5 12 8 3–197

Radiosurgery was performed using the commercial XKnife system (Radionics, Burlington, MA) to deliver multiple noncoplanar arcs of photon to the target with a modiﬁed 6-MV linear accelerator (Clinac 600C; Varian, Milpitas, CA). Head immobilization and target localization were performed with the Brown-Roberts-Wells head frame and stereotaxic system (Radionics). The head frame was anchored by placing two anterior head pins to the forehead and two to the occiput. The head frame was not placed in a strict horizontal position but with the posterior side tilted downward to ensure that the localizer ring would include the whole nasopharynx. Using this approach, it was possible to treat the entire nasopharynx and adjacent soft tissues, although special attention should be given to those with disease extending down to the junction of the oropharynx. Although some investigators adopted the approach of anchoring the head ring anteriorly to the zygomas to ensure complete coverage of the nasopharynx, we did not ﬁnd this to be necessary for most patients. The target volume was localized as described above. For rT1 tumor, target conﬁned to one side of nasopharynx was usually covered using one isocenter, whereas target involving both sides was covered using one or two isocenters. For rT2-4 tumors, target was usually covered by one or two isocenters. We found it seldom necessary to use more than two isocenters. The target volume was treated using three to ﬁve arcs of beams with a degree of 90° to 210°. The following dose limits were set to the critical structures: 5 Gy to brain stem, 4 Gy to optic apparatus, and 5 Gy to temporal lobes. An additional dose limit of 8 Gy to internal carotid artery was later set to reduce the risk of hemorrhage. Most patients received a dose of 12.5 Gy delivered to the 80% isodose line. Table 40-2 summarizes the radiosurgery treatment.

Tumor Control After Radiosurgery Thirty-seven (77%) patients achieved complete regression of tumor after radiosurgery. All patients (100%) treated for persistent disease achieved complete regression of disease compared with 13 (54%) patients for recurrent disease. Local failure after radiosurgery occurred in 22 (46%) patients, regional failure in 5 (10%) patients, and distant metastasis in 7 (15%) patients. For those patients who failed locally after radiosurgery, two received brachytherapy with intubation and both were salvaged, four had nasopharyngectomy and three were

40.

415

head and neck tumors

TABLE 40-2. Summary of radiosurgery treatment parameters of patients treated at Queen Mary Hospital, Hong Kong.

1.0 0.9

Persistent disease

Number of isocenters One (n) Two (n) Three (n) Size of collimator ≤20 mm (n) >20 to 30 mm (n) >30 mm (n) Median (mm) Range (mm)

20 4 0

19 4 1

4 19 1 25 12.5–35

Target volume <10 cm3 (n) ≥10 cm3 (n) Median (cm3) Range (cm3)

23 1 4.0 1.3–14.5

Prescribed dose to 80% isodose line 3 <12 Gy (n) 12 to 13 Gy (n) 17 4 >13 Gy (n) Median (Gy) 12.5 Range (Gy) 11.1–18 Dose to isocenter <15 Gy (n) 15 to 16 Gy (n) >16 Gy (n) Median (Gy) Range (Gy)

Recurrent disease

4 16 4 15.6 13.9–22.5

All

39 8 1

1 16 7 30 15–37.5

5 35 8 25 12.5–37.5

15 9 5.7 2.7–30.7

38 10 4.6 1.3–30.7

5 13 6 12.5 8–14

8 30 10 12.5 8–18

5 13 6 15.6 10–17.5

9 29 10 15.6 10–22.5

salvaged, and ﬁve had external radiotherapy and two were salvaged. One patient with local failure outside the radiosurgerytreated volume was salvaged by second radiosurgery. Three-year local relapse-free and overall survival rates for all patients were 52% and 66%, respectively. Patients treated for persistent disease had better outcome than those treated for recurrent disease: 3-year local relapse-free survival rate was 76% in the former and 29% in the latter (Fig. 40-3), and the corresponding

Overall Survival Probability

Parameters

0.8

Persistent disease

0.7 0.6 0.5 0.4 Recurrent disease

0.3 0.2 0.1

p = 0.035

0.0 0

12

24

36

48

60

72

84

96

108 120

Months after Radiosurgery FIGURE 40-4. Comparison of overall survival curves in patients with persistent and recurrent nasopharyngeal carcinoma treated by stereotactic radiosurgery at Queen Mary Hospital.

3-year overall survival rates were 81% and 54% (Fig. 40-4). Patients with advanced T classiﬁcation at the time of relapse had a poorer control rate after radiosurgery: 3-year local relapse-free survival rates were 69% for rT1 disease, 40% for rT2-3 disease, and 0% for rT4 disease (Fig. 40-5). Patients with bulky tumor also had a poorer control rate after radiosurgery: 3-year local relapse-free survival rate was 61% for target volume <10 cm3 compared with 20% for target volume ≥10 cm3 (Fig. 40-6).

Complications Radiosurgery was well tolerated, and we did not observe any acute complications. In assessing late effects after treatment, it was not possible to attribute the cause of complication to radio-

Local Relapse-free Survival Probability

0.9 Persistent disease

0.8 0.7 0.6 0.5 0.4

Recurrent disease

0.3 0.2

p = 0.0004

0.1

Local Relapse-free Survival Probability

1.0 1.0

0.9 0.8 0.7 0.6

rT1

0.5 0.4 rT2-3

0.3 0.2 rT4

0.1

p = 0.0042

0.0 0

0.0 0

12

24

36 48 60 72 84 96 108 120 Months after Radiosurgery FIGURE 40-3. Comparison of local relapse-free survival curves in patients with persistent and recurrent nasopharyngeal carcinoma treated by stereotactic radiosurgery at Queen Mary Hospital.

12

24

36

48

60

72

84

96

108 120

Months after Radiosurgery FIGURE 40-5. Comparison of local relapse-free survival curves according to recurrent T classiﬁcation after stereotactic radiosurgery for local failures of nasopharyngeal carcinoma at Queen Mary Hospital.

416

d.t.t. chua et al.

surgery or external radiotherapy. Some patients also received another course of external radiotherapy either before or after radiosurgery and they would expect to have a relatively high incidence of late complications even without radiosurgery. Table 40-3 summarizes the late complications observed in our patients, which included all observed complications (other than xerostomia) irrespective of whether they were thought to be mainly caused by radiosurgery or not. Overall, 31% of patients developed one or more late complications. The percentage decreased to 24% if patients with two courses of radiotherapy before radiosurgery were excluded, and it further decreased to 19% if those with a second course of radiotherapy before or after radiosurgery were excluded. Patients with more advanced tumor tend to have a higher risk of developing late complications. The percentages of patients who developed late complications were 19% in rT1 disease, 33% in rT2 disease, 46% in rT3 disease, and 75% in rT4 disease. Common late complications observed were cranial neuropathy and temporal lobe necrosis, both occurring in about 15% of patients. Two patients developed severe epistaxis, and both were found to have internal carotid artery aneurysm located at cavernous sinus and retropharyngeal space. Stenting was performed in both patients, which successfully controlled the bleeding, and there were no treatment-related deaths.

Radiosurgery Treatment Results for Local Failures of NPC in Other Series Radiosurgery Alone Firlik et al. ﬁrst reported the use of Gamma Knife radiosurgery in a patient with recurrent NPC and noted complete regression of tumor after delivering a dose of 20 Gy to 50% isodose line [22]. Miller et al. reported another three patients with NPC that were also treated by Gamma Knife: one patient was noted to have symptomatic improvement before disease progression, and one patient had not responded to the treatment [23]. Buatti et al. treated three patients with recurrent NPC by linac radio-

Local relapse-free survival probability

1.0 0.9 0.8 0.7 0.6 Tumor volume < 10 cc

0.5 0.4 0.3 0.2

Tumor volume > or = 10 cc

0.1

p = 0.0041

0.0

TABLE 40-3. Late complications after radiosurgery for locally recurrent nasopharyngeal carcinoma at Queen Mary Hospital, Hong Kong. Complication

Cranial neuropathy Temporal lobe necrosis Hypopituitarism Carotid artery aneurysm Trismus None

One course of ERT (n = 36)

Two courses of ERT (n = 12)

One to two courses of ERT (n = 48)

2 4

5 3

7 7

1 1

1 1

2 2

0 29

1 4

1 33

ERT, external radiotherapy; n, number of patients.

surgery using a dose of 12.5 Gy to 80% isodose line [24]. Two patients also received reirradiation, one before and the other after radiosurgery. Of these two patients, one remained diseasefree 1 year after treatment, whereas the other had local recurrence 6 months after treatment. The third patient received radiosurgery alone for recurrent disease and had neurologic deterioration of uncertain etiology 6 months after the treatment. Kocher et al. treated ﬁve patients with recurrent NPC at the skull base using radiosurgery with a dose of 15 to 24 Gy [25]. Only one patient’s tumor was controlled after treatment. Three patients developed late complications, including two with fatal internal carotid artery hemorrhage. Cmelak et al. reported using linac radiosurgery in treating 12 recurrent NPC lesions in 9 patients [26]. The dose delivered ranged from 15 to 20 Gy with a median of 18 Gy. With a median follow-up of 17 months, the crude local control was 53% (7/12), and only one patient developed radiation-induced cranial neuropathy.

Radiosurgery as a Boost After External Reirradiation Chen et al. reported the outcome of 11 patients with rT3-4 NPC after conformal radiotherapy and linac radiosurgery [27]. The radiosurgery dose ranged from 10 to 19 Gy with a median of 14 Gy. Signiﬁcant regression of tumor was noted in ﬁve patients and limited regression in another three. Chang et al. reported 15 patients with locally recurrent NPC who received external reirradiation followed by radiosurgery using a dose ranging from 8 to 15 Gy [18] and noted a 3-year survival rate of 52%. Pai et al. reported 36 patients with recurrent NPC also treated with external reirradiation followed by radiosurgery boost [28]. The radiosurgery dose to target periphery ranged from 8 to 20 Gy with a median of 12 Gy. A 3-year local control rate of 58% was achieved. In these reports, it is uncertain whether radiosurgery was used to boost the entire or partial tumor volume, with the intention to further escalate the total dose or to selectively boost the volume that was underdosed during external reirradiation.

Fractionated Stereotactic Radiation 0

12

24

36 48 60 72 84 96 108 120 Months after radiosurgery FIGURE 40-6. Comparison of local relapse-free survival curves according to tumor volume after stereotactic radiosurgery for local failures of nasopharyngeal carcinoma at Queen Mary Hospital.

Fractionated stereotactic radiation instead of radiosurgery has also been employed as a salvage treatment for locally recurrent NPC. Mitsuhashi et al. treated three patients with rT1 NPC using stereotactic radiation at a dose of 50 to 64 Gy, and all

40.

417

head and neck tumors

three patients achieved complete response and remained free of local disease at 4 to 61 months [29]. The report by Mitsuhashi et al. also included another patient with mucoepidermoid carcinoma of the nasopharynx treated by stereotactic radiation after previous two courses of external radiotherapy, but the treatment was complicated by rupture of the internal carotid artery resulting in patient death. Using hypofractionated stereotactic radiation at a dose of 24 Gy in two to four fractions, Orecchia et al. reported a less satisfactory outcome in 13 patients with locally recurrent NPC, with a 3-year survival rate of 31% [30]. Ahn et al. treated 12 patients with recurrent NPC by stereotactic radiation using a median dose of 54 Gy and reported a 2-year local control rate of 92% [31]. The largest series of stereotactic radiation for treatment of local failures of NPC was reported by Xiao et al.: 50 patients with persistent or recurrent NPC were treated with a dose ranging from 14 to 35 Gy using a fraction dose of 5 to 15 Gy [32]. Of the 31 evaluable patients with persistent disease, 94% had complete response with a 1-year disease-free survival rate of 47%. Eighteen patients, most of them with rT3-4 tumor, were treated for recurrent disease. The complete response rate was 56% and 1year disease-free survival rate was 47%. In Xiao’s series, however, 16% of patients treated by fractionated radiation developed fatal hemorrhage, probably due to the relatively high cumulative dose delivered.

Radiosurgery as a Boost Treatment After Radiotherapy for Newly Diagnosed NPC Radiosurgery has also been employed as a planned boost treatment after radiotherapy for newly diagnosed NPC. The treatment is similar in principle to the use of intracavitary intubation or parapharyngeal boost commonly given after conventional two-dimensional radiotherapy. Cmelak et al. treated 11 patients using radiosurgery boost with a median dose of 12 Gy (7 to 16 Gy) after a dose of 64.8 to 70 Gy was delivered by external

radiotherapy [26]. The reported control rate was 91% without any major late complications. Tate et al. treated 23 patients with a median radiosurgery boost dose of 12 Gy (7 to 15 Gy) after external radiotherapy; the latter delivered a dose of 64.8 to 70 Gy to the primary [33]. All patients had their local tumors controlled after the treatment although 35% subsequently developed regional or distant failures. Again, no major late complications were reported. In these two series, radiosurgery was delivered after treatment using a radical dose of external radiotherapy with the aim of improving local control by dose escalation. Ahn et al. used stereotactic radiotherapy instead as a boost treatment, with a dose ranging from 8 to 40 Gy [31]. The dose delivered by external radiotherapy ranged from 36 to 61.2 Gy. The reported 4-year local control and survival rates were 89% and 75%, respectively. One patient developed mucosal necrosis after the treatment. The approach by Ahn et al. was to reduce the dose delivered by conventional radiotherapy and substitute it by stereotactic radiotherapy thereby reducing the incidence of late complications.

Summary and Recommendations Table 40-4 summarizes the outcome after radiosurgery for local failures of NPC as reported in different series. Based on these results, there is strong evidence indicating that radiosurgery is an effective salvage treatment for local failures of NPC. There is, however, no data comparing the relative efﬁcacy and complication risks of radiosurgery with other salvage options for NPC. In practice, selection of treatment modalities depends mainly on extent of disease and expertise available. For rT1-2 disease, treatment outcome after radiosurgery appears to be comparable with that of brachytherapy and surgery. For rT3-4 disease, radiosurgery also seems to yield similar results with conventional two-dimensional reirradiation. The advent of three-dimensional conformal radiotherapy and intensitymodulated radiation treatment appears to improve the outcome

TABLE 40-4. Summary of reported treatment outcomes after stereotactic radiosurgery/radiotherapy for local failures of NPC. Authors

No. of patients

rT classiﬁcation

Treatment

SRS/SRT dose

rT1-2: 69% rT3-4: 31% Not reported rT3-4: 100% rT1-2: 64% rT3-4: 36% rT1-2: 67% rT3-4: 33% rT3-4: 100%

SRS

8–18 Gy

SRS SRS ER + SRS

15–20 Gy 15–24 Gy 8–20 Gy

ER + SRS

8–15 Gy

ER + SRS

10–19 Gy

Chua et al. (current report) Cmelak et al [26] Kocher et al. [25] Pai et al. [28]

48

Chang et al. [18]

15

Chen et al. [27]

11

Xiao et al. [32]

50

rT1-2: 38% rT3-4: 62%

SRT

14–35 Gy

Orecchia et al. [30]

13

SRT

24 Gy

Ahn et al. [31]

12

rT1-2: 23% rT3-4: 77% Not reported

SRT

45–65 Gy

9 5 36

Local control and survival rates

Late complications crude rate: all (fatal)

LC: 52% (3-year) OS: 66% (3-year) LC: 58% (crude) LC: 20% (crude) LC: 58% (3-year) OS: 54% (3-year) 52% (3-year survival)

31% (0%)

LC: 45% (crude) OS: 37% (2-year) CR: 76% OS: Persistent disease 78% (2-year) Recurrent disease 42% (2-year) LC: 31% (crude) OS: 31% (3-year) LC: 92% (2-year) OS: 60% (2-year)

Not reported

11% (0%) 60% (40%) 22% (0%) Not separately reported

34% (16%)

0% (0%) Not reported (8%)

SRS, stereotactic radiosurgery; SRT, stereotactic radiotherapy; ER, external reirradiation; LC, local control; OS, overall survival; CR, complete response.

418

d.t.t. chua et al.

Tumor confined to nasopharynx?

Yes

Yes

Unilateral involvement?

No

Brachytherapy (gold grain implantation or intracavitary intubation) or SRS or nasopharyngectomy

No SRS

rT3/4 disease? (skull base or intracranial invasion)

Yes

Persistent disease?

Yes

No

Encasement of carotid artery or tumor volume? 10 cc?

Yes SRT or IMRT SRS

No

No Encasement of carotid artery or tumor volume? 10 cc?

Yes

SRT or IMRT SRS

No

FIGURE 40-7. A proposed decision-tree for salvaging local failure of nasopharyngeal carcinoma. SRS, stereotactic radiosurgery; SRT, stereotactic radiotherapy; IMRT, intensity-modulated radiation therapy.

of recurrent NPC, particularly in those with more advanced tumor. Hence, patients with rT3-4 disease should be treated by external reirradiation using these new techniques with radiosurgery reserved as a boost treatment or for further recurrence. In assessing late complications after radiosurgery, it was always difﬁcult if not impossible to attribute them to either external radiotherapy or radiosurgery, which may lead to underestimation of complication risks of radiosurgery. Although most series reported a relatively low risk of late complications, hemorrhage remains the most severe form of complication after radiosurgery with a possible fatal outcome. Although hemorrhage could also occur as a complication of tumor progression, most cases that developed after radiosurgery were probably due to radiation damage to the carotid artery as a result of high cumulative dose. The complication is probably not related to rapid shrinkage of tumor as suggested by Kocher et al. [25], because hemorrhage is rare after chemotherapy and brachytherapy, both of which can induce rapid tumor shrinkage. To minimize the risk of hemorrhage, it is advised to use radiosurgery only in the absence of direct tumor encasement of carotid artery, otherwise the patient should be treated by stereotactic or fractionated radiotherapy using a small fraction dose. A proposed decision tree for salvaging local failures of NPC is outlined in Figure 40-7. The role of radiosurgery as a boost treatment after radical radiotherapy in newly diagnosed NPC is more difﬁcult to deﬁne, although its impact is likely to be small with the advent of modern conformal radiotherapy, chemotherapy, and intensitymodulated radiation therapy.

Other Head and Neck Tumors Radiosurgery has been employed in the treatment of head and neck tumors other than NPC. A wide range of tumors, both benign and malignant, has been treated using either stereotactic radiosurgery or radiotherapy. Most treated cases were recurrent tumors after prior surgery and/or radiotherapy. The more commonly treated histologic type includes squamous cell carcinoma, adenoid cystic carcinoma, and chordoma. Most lesions were located at the skull base, making them ideal targets for radiosurgery. For lesions located below the skull base, ﬁxation

and immobilization are not as rigid as for intracranial target, which should be taken into consideration during localization of target and adjacent critical structures. In one series reported by Cmelak et al. [26], 36 recurrent head and neck tumors in 29 patients were treated by radiosurgery using a dose ranging from 7 to 35 Gy (median, 20 Gy). Fifteen patients were treated for squamous cell carcinoma, six for adenocarcinoma, four for malignant meningioma, three for adenoid cystic carcinoma, and three for melanoma. Local control rate was 72% but late complications including CSF leakage, trismus, and cranial neuropathy occurred in four patients after treatment. Miller et al. [23] treated 32 new or recurrent head and neck tumors in 29 patients using radiosurgery with or without external radiotherapy. Twelve patients were treated for adenoid cystic carcinoma, eight for squamous cell carcinoma, and eight for chordoma. The dose delivered by radiosurgery was in the range of 12 to 20 Gy (median, 15 Gy). Local control rates were 83% for adenoid cystic carcinoma and 100% for chordoma. Late complications occurred in four patients and included memory loss, hypopituitarism, CSF leakage, brain edema, and optic neuropathy. Kocher et al. [25] treated four patients with recurrent sarcoma using radiosurgery and achieved good tumor control in three. They also treated three patients with recurrent squamous cell or adenoid cystic carcinoma but achieved local control in only one patient. The radiosurgery dose used was in the range of 9 to 20 Gy. One patient developed brain necrosis and cranial neuropathy after radiosurgery. Ryu et al. [34] treated ﬁve patients with recurrent squamous cell or mucoepidermoid carcinoma by radiosurgery with a dose of 12 to 18 Gy and achieved complete response in two patients. They also treated ﬁve recurrent tumors (squamous cell, adenoid cystic, and basal cell carcinoma) and three metastatic cervical nodes by stereotactic radiotherapy using a dose of 30 to 36 Gy in six fractions and noted three complete responses in the former group and one in the latter group. Elshaikh et al. [35] treated 12 patients with recurrent chemodectoma (seven with ﬁrst recurrence after surgery and ﬁve with second recurrence after surgery) using Gamma Knife radiosurgery at a dose of 13 to 16 Gy and achieved 83% control rate without any major late complications. Using stereotactic radiotherapy, Ahn et al. [31] treated three patients with steroid refractory orbital pseudotumor using a dose of

40.

head and neck tumors

20 Gy in 10 fractions and noted good response in all patients. Because of the heterogeneous nature of the patient population treated in these series, it is not possible to conclude the efﬁcacy and indication of radiosurgery. It seems, however, reasonable to recommend stereotactic radiosurgery or radiotherapy in patients with recurrent head and neck tumors not amenable to further surgery if the site and extent of the lesions are a suitable target of radiosurgery.

References 1. Lee AW, Poon YF, Foo W, et al. Retrospective analysis of 5037 patients with nasopharyngeal carcinoma treated during 1976–1985: overall survival and patterns of failure. Int J Radiat Oncol Biol Phys 1992; 23:261–270. 2. Lee AW, Sze WM, Au JS, et al. Treatment results for nasopharyngeal carcinoma in the modern era: the Hong Kong experience. Int J Radiat Oncol Biol Phys 2005; 61:1107–1116. 3. Al-Sarraf M, LeBlanc M, Shanker Giri PG, et al. Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: phase III randomized intergroup study 0099. J Clin Oncol 1998; 16:1310–1317. 4. Chan AT, Teo PM, Ngan RK, et al. Concurrent chemotherapyradiotherapy compared with radiotherapy alone in loco-regionally advanced nasopharyngeal carcinoma: progression-free survival analysis of a phase III randomized trial. J Clin Oncol 2002; 20: 2038–2044. 5. Lin JC, Jan JS, Hsu CY, et al. Phase III study of concurrent chemoradiotherapy versus radiotherapy alone for advanced nasopharyngeal carcinoma: positive effect on overall and progressionfree survival. J Clin Oncol 2003; 21:631–637. 6. Chua DT, Ma J, Sham JST, et al. Long-term survival after cisplatin-based induction chemotherapy and radiotherapy for nasopharyngeal carcinoma: a pooled data analysis of two phase III trials. J Clin Oncol 2005; 23:1118–1124. 7. Chua DT, Sham JS, Kwong DL, et al. Volumetric analysis of tumor extent in nasopharyngeal carcinoma and correlation with treatment outcome. Int J Radiat Oncol Biol Phys 1997; 39:711– 719. 8. Wu PM, Chua DT, Sham JS, et al. Tumor control probability of nasopharyngeal carcinoma: a comparison of different mathematical models. Int J Radiat Oncol Biol Phys 1997; 37:913–920. 9. Kwong DL, Wei WI, Cheng AC, et al. Long term results of radioactive gold grain implantation for the treatment of persistent and recurrent nasopharyngeal carcinoma. Cancer 2001; 91:1105–1113 10. Leung TW, Tung SY, Sze WK, et al. Salvage brachytherapy for patients with locally persistent nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 2000; 47:405–412. 11. Fee WE Jr, Roberson JB Jr, Gofﬁnet DR. Long-term survival after surgical resection for recurrent nasopharyngeal cancer after radiotherapy failure. Arch Otolaryngol Head Neck Surg 1991; 117: 1233–1236. 12. Morton RP, Liavaag PG, McLean M, et al. Transcervicomandibulo-palatal approach for surgical salvage of recurrent nasopharyngeal cancer. Head Neck 1996; 18:352–358. 13. King WW, Ku PK, Mok CO, et al. Nasopharyngectomy in the treatment of recurrent nasopharyngeal carcinoma: a twelve-year experience. Head Neck 2000; 22:215–222. 14. Wei WI. Salvage surgery for recurrent primary nasopharyngeal carcinoma. Crit Rev Oncol Hemat 2000; 33:91–98. 15. Teo PM, Kwan WH, Chan AT, et al. How successful is high dose (≥ 60 Gy) reirradiation using mainly external beams in salvaging local failures of nasopharyngeal carcinoma? Int J Radiat Oncol Biol Phys 1998; 40:897–913.

419

16. Öksüz DÇ, Meral G, Uzel Ö, Çagˇatay P, Turkan S. Reirradiation for locally recurrent nasopharyngeal carcinoma: treatment results and prognostic factors. Int J Radiat Oncol Biol Phys 2004; 60: 388–394. 17. Chua DT, Sham JS, Kwong DL, et al. Locally recurrent nasopharyngeal carcinoma: treatment results for patients with computed tomography assessment. Int J Radiat Oncol Biol Phys 1998; 41: 379–386. 18. Chang JT, See LC, Liao CT, et al. Locally recurrent nasopharyngeal carcinoma. Radiother Oncol 2000; 54:135–142. 19. Lee AW, Foo W, Law SC, et al. Reirradiation for recurrent nasopharyngeal carcinoma: factors affecting the therapeutic ratio and ways for improvement. Int J Radiat Oncol Biol Phys 1997; 38: 43–52. 20. Hsiung CY, Yorke ED, Chui CS, et al. Intensity-modulated radiotherapy versus conventional three-dimensional conformal radiotherapy for boost or salvage treatment of nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 2002; 53:638–647. 21. Lu TX, Mai WY, The BS, et al. Initial experience using intensitymodulated radiotherapy for recurrent nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 2004; 58:682–687. 22. 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; 18:160–166. 23. Miller RC, Foote RL, Coffey RJ, et al. The role of stereotactic radiosurgery in the treatment of malignant skull base tumors. Int J Radiat Oncol Biol Phys 1997; 39:977–981. 24. Buatti JM, Friedman WA, Bova FJ, Mendenhall WM. Linac radiosurgery for locally recurrent nasopharyngeal carcinoma: rationale and technique. Head Neck 1995; 17:14–19. 25. Kocher M, Juergen V, Susanne S, et al. Linear accelerator radiosurgery for recurrent malignant tumors of the skull base. Am J Clin Oncol 1998; 21:18–22. 26. Cmelak AJ, Cox RS, Adler JR, Fee WE Jr, Gofﬁnet DR. Radiosurgery for skull base malignancies and nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 1997; 37:997–1003. 27. Chen HJ, Leung SW, Su CY. Linear accelerator based radiosurgery as a salvage treatment for skull base and intracranial invasion of recurrent nasopharyngeal carcinoma. Am J Clin Oncol (CCT) 2001; 24:255–258. 28. Pai P, Chuang C, Wei K, et al. Stereotactic radiosurgery for locally recurrent nasopharyngeal carcinoma. Head Neck 2002; 24:748– 753. 29. Mitsuhashi N, Sakurai H, Katano S, et al. Stereotactic radiotherapy for locally recurrent nasopharyngeal carcinoma. Laryngoscope 1999; 109:805–809. 30. Orecchia R, Redda MGR, Regona R, et al. Results of hypofractionated stereotactic re-irradiation on 13 locally recurrent nasopharyngeal carcinoma. Radiother Oncol 1999; 53:23–28. 31. Ahn YC, Lee KC, Kim DY, et al. Fractionated stereotactic radiation therapy for extracranial head and neck tumors. Int J Radiat Oncol Biol Phys 2000; 48:501–505. 32. Xiao JP, Xu GZ, Miao YJ. Fractionated stereotactic radiosurgery for 50 patients with recurrent or residual nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 2001; 51:164–170. 33. Tate DJ, Adler JR, Chang SD, et al. Stereotactic radiosurgical boost following radiotherapy in primary nasopharyngeal carcinoma: impact on local control. Int J Radiat Oncol Biol Phys 1999; 45:915–921. 34. Ryu S, Khan M, Yin F, et al. Image-guided radiosurgery of head and neck cancers. Otolaryngol Head Neck Surg 2004; 130: 690–697. 35. Elshaikh MA, Mahmoud-Ahmed AS, Kinney SE, et al. Recurrent head-and-neck chemodectomas: a comparison of surgical and radiotherapeutic results. Int J Radiat Oncol Biol Phys 2002; 52:953–956.

4 1

Head and Neck Tumors: Surgery Perspective Gregory Y. Chin and Uttam K. Sinha

Introduction The extracranial head and neck region is complex and varied with respect to anatomy, physiology, and pathology. Squamous cell carcinoma represents a large proportion of upper aerodigestive tract (sinonasal cavity, oral cavity, nasopharynx, oropharynx, hypopharynx, and larynx) malignancies. However, the salivary glands have pathology ranging from mucoepidermoid to adenoid cystic and adenocarcinoma. Likewise, common malignancies of the thyroid gland include papillary, follicular, and medullary carcinoma. Neuroendocrine lesions occur throughout the region such as carotid body tumors and glomus tumors. Smoking and tobacco products represent a major risk factor in most head and neck malignancies. Surgeons and oncologists face several treatment challenges when dealing with cancer of the head and neck. Many structures in this region have an intimate relationship with form and function of the body. Thus, it is often difﬁcult to preserve quality of life while giving the best chance for cure. In an area such as the larynx, a minute difference in tumor location can be the difference between preservation and loss of phonation or breathing without a tracheotomy. In the pharynx, small changes in tumor size can completely alter a person’s ability to swallow. Likewise, these minor disparities can alter a surgeon’s ability to perform a simple resection with closure versus reconstruction with a regional pedicled or microvascular free ﬂap.

Conventional Treatment The mainstay of treatment for head and neck tumors has been surgery, conventional radiation, or a combination of the two. In recent years, intensity-modulated radiation therapy has become increasingly popular. This provides more targeted therapy to the tumor and decreases morbidity to areas such as the salivary glands. The applications for chemotherapy have increased as well. In the past, it was primarily limited to palliative roles or laryngeal cancer. However, it is gaining more acceptance in the deﬁnitive treatment of selected head and neck cancers. Current protocols for combined chemotherapy and radiation therapy often show success in areas such as the base of the tongue. The functional consequences associated with major head and neck resections are pervasive. They range from minor dys-

phagia to complete loss of speech and swallowing. Many patients end up with a permanent tracheotomy and percutaneous gastrostomy tube. Some lesions require disﬁguring extirpations that are difﬁcult to disguise even with the best reconstructive techniques. The side effects of high-dose conventional radiation are also well-known. The effects on the salivary glands are particularly troubling to patients. Despite the high morbidity of these procedures, the results are often disappointing. In the case of advanced cancers, massive surgical resections usually have a low cure rate and may not prolong life. They often worsen the patient’s already tenuous swallowing, speech, or breathing.

Conventional Treatment Results Squamous cell cancer of the neck and upper aerodigestive tract is the most common malignancy faced by head and neck surgical oncologists. Stage I lesions are generally well treated with surgery or radiation, and 5-year survival rates exceed 80% in many anatomic locations. Stage II lesions often respond to a combination of surgery and radiation and are cured in a majority of patients. More problematic are stage III and IV patients where the primary lesion is massive or there is involvement of lymph nodes. In these situations, combination surgery and radiation or sometimes chemotherapy and radiation are the standard of care. In most anatomic locations, 5-year survival for stage III patients is less than 50% and for stage IV patients is less than 30% [1]. Patients with recurrent squamous cell cancer have an even worse prognosis. If left untreated, median survival is as little as a few months. Even with surgical salvage, average survival is seldom greater than 1 year. Revision surgeries have increased morbidity because scar tissue and loss of tissue planes caused by prior treatments make dissection of important structures much more difﬁcult. Also, healing is impaired by radiation. When radiation is used alone, the recurrence rate is greater than 80%. Chemotherapy in selected cases has added some beneﬁt and is being used with increasing frequency [1]. Patients with unresectable lesions or those who cannot tolerate surgery for medical reasons also face a grim prognosis in the case of advanced disease. The relative contraindications for surgery vary from center to center. However, many surgeons would deem carotid artery encasement or deep neck

421

422

g.y. chin and u.k. sinha

musculature involvement to be reasons for not performing surgery or stopping it if discovered during the operation. Distant metastases, most commonly in the lung, would also be a contraindication.

Stereotactic Radiosurgery Stereotactic radiosurgery (SRS) is one of the newest technologies to be applied to the head and neck region. This modality in its various forms, such as Gamma Knife and CyberKnife, is not unknown to the head and neck surgeon. Indeed, it is often used on lesions in the anterior skull base where surgical resection is extremely difﬁcult and the lateral skull base where benign tumors lie in close proximity to critical structures. These areas are discussed thoroughly elsewhere in this book. In this chapter, we will focus on the use of SRS in non–skull base regions traditionally treated with surgery. This includes areas such as the sinonasal cavity, oral cavity, oropharynx, hypopharynx, larynx, and neck. The application of SRS is of great interest because of its ability to target tumors with minimal damage to surrounding normal tissues. Surgery and conventional radiation can affect large amounts of normal tissue and lead to unnecessary morbidity. SRS research on cancer of the neck and upper aerodigestive tract is still in its infant stage. Little data exists for the application of this technology to primary lesions as either a solo modality or adjuvant treatment. More data exists on patients with recurrent or unresectable cancers. In most of these studies, some combination of surgery, radiation, and chemotherapy has been used prior to SRS. Quality of life is of paramount importance when pondering treatment of malignancy. Loss of ability to swallow, phonate, and breathe can result from advancing local disease. Disﬁguring cosmesis and pain are also of utmost concern to a patient. SRS offers the opportunity to improve local control without signiﬁcant morbidity. Because of the complex three-dimensional nature of head and neck tumors and concerns regarding movement of structures during treatment, the use of this modality was extremely limited in the past. However, with improving technology and immobilization techniques, it is now possible to accurately target lesions in the head and neck region [2]. It is also possible to deal with the irregular shapes of many of these lesions.

and hypopharynx cancer. The treatments included ﬁve fractionated does of 2 to 3 Gy over a 1-week period. Good responses were seen with all lesions. In addition, the treatments were well tolerated, and no major side effects developed [3]. Elshaikh used Gamma Knife radiosurgery for salvage of head and neck chemodectomas initially treated with surgery. The tumors included glomus jugulare, carotid body, and glomus tympanicum lesions. Patients were followed for 4.5 years on average. Gamma Knife was used in 12 patients and revision surgery was performed in 17 patients. For glomus jugulare and glomus tympanicum tumors, 100% of the stereotactic radiosurgery group was disease-free at 5 years. The salvage surgery group showed a 62% 5-year disease-free rate. In contrast, recurrent carotid body tumors were better controlled with surgery. However, postoperative complications occurred in nearly half the surgical group. These included cranial nerve palsies, meningitis, infection, and cerebrospinal ﬂuid leak. None of these occurred in the radiotherapy salvage group [4]. Foote also treated patients with glomus jugulare tumors using the Leksell Gamma Knife. Twenty-ﬁve patients were prospectively followed for an average of 3 years. Seventeen patients had stable lesions and eight had regression of the tumor. No cranial neuropathies developed. They concluded that the lower cranial nerves could safely tolerate a dose of 12 to 18 Gy [5]. Ryu treated 13 patients with single-dose or fractionated radiosurgery. Twelve to 18 Gy was used for single-dose radiosurgery, and 30 Gy was used for fractionated radiosurgery given twice a week for 3 weeks. All patients except one had prior treatments including surgery, conventional radiation, and chemotherapy. The lesions included squamous cell cancer in various regions: ﬂoor of mouth, pharyngeal wall, base of tongue, parotid, lip, and cervical metastases. Mucoepidermoid and adenoid cystic carcinoma were also evaluated. Six patients had a complete response, three had >50% tumor reduction, and three had no progression of disease. All experienced pain relief and had preservation of function and cosmesis. Patients were followed at least 6 months and developed no signiﬁcant complications [6]. Lee evaluated skull base recurrence of salivary gland tumors. Eight patients had 16 skull base recurrences that were treated with Gamma Knife. All of the patients experienced decreased or resolution of pain. The average local freedom from progression was 15 months. This included 100% freedom from progression in all patients for at least 1 year. Local control was achieved in several patients from 4 to 13 months. One patient suffered radiation necrosis [7].

Stereotactic Radiosurgery Results Ahn studied 48 patients with extracranial head and neck tumors over a period of 4 years. The patients were given fractionated stereotactic radiation therapy via the XKnife-3 system and GillThomas-Cosman frame. Half of the patients were treated with this as the sole modality, and half were treated with this as a boost in addition to conventionally delivered external radiation therapy. The lesions in the combination therapy group included nasopharynx cancer, lacrimal gland adenoid cystic cancer, orbital lymphoma, and skull base recurrence of maxillary sinus adenoid cystic cancer. The stereotactic radiosurgery–only group included nasopharynx cancer, orbital pseudotumor, rhabdomyosarcoma, lymphoma, neuroblastoma, nasal cancer and skull base recurrence of maxillary sinus cancer, salivary gland cancer,

Discussion Thus far, there are relatively few studies regarding the efﬁcacy of SRS in extracranial head and neck lesions. The number of patients in the studies has been small, and the methodologies have not been uniform. Also, because patient numbers were small, patients with various tumors in different locations were often lumped together. Local control seems promising. Ahn showed good local control for primary lesions in a variety of locations [3]. Elshaikh showed that SRS was superior to salvage surgery for glomus jugulare and glomus tympanicum tumors [4]. Foote demonstrated promising results for glomus jugulare tumors [5]. Ryu

41.

head and neck tumors: surgery perspective

TABLE 41-1. Indications for use of stereotactic radiosurgery in head and neck tumors.

Paragangliomas except carotid body tumors Squamous cell cancer with skull base extent Squamous cell cancer without skull base extent Salivary gland cancer with skull base extent Salivary gland cancer without skull base extent

Primary lesion

Recurrent lesion

Yes

Yes

Unknown

Unknown

Unknown

Unknown

Unknown

Yes

Unknown

Yes

found that successful treatment of squamous cell cancer could be accomplished with single-dose or fractionated therapy depending on the situation [6]. Lee showed local freedom from progression greater than a year for all his patients with recurrent salivary gland malignancies of the skull base [7]. All the studies have shown SRS to be well tolerated. Few signiﬁcant complications were reported. The study by Elshaikh is particularly striking in the number of complications suffered by the surgical salvage group despite the fact that they had a lower 5-year disease-free rate [4]. The radiation necrosis suffered by one patient in Lee’s study is certainly a known complication of irradiating bone [7]. Also, the patients in these studies did not suffer a signiﬁcant loss of function [3–6]. Unfortunately, there is no data on stereotactic radiosurgery showing improved survival rate, particularly in the case of squamous cell cancer. Studies have only shown local control and delayed progression. Also, there is no data comparing the effect of surgical salvage versus SRS. Until these questions are answered, it would be difﬁcult to recommend SRS for squamous cell cancer in any situation except for palliation. Its use as a boost in conjunction with conventional therapy is currently being studied. Table 41-1 identiﬁes situations in which SRS may be of beneﬁt. At the University of Southern California, we use CyberKnife as an adjunct to surgery and conventional radiation in selected primary and recurrent cases of squamous cell cancer in the head and neck. Figure 41-1 demonstrates a primary squamous cell

FIGURE 41-1. Left infratemporal fossa squamous cell cancer with extension into the neck.

423

cancer of the infratemporal fossa extending to the neck. This patient was treated by surgical resection, conventional radiation, and CyberKnife. Our rational for treating certain lesions with conventional radiation and a boost from CyberKnife is related to the relative dosing of radiation throughout the head and neck. Due to anatomic reasons, the submental area, infratemporal fossa, and apex of the posterior neck triangle (i.e., area around the mastoid tip) receive less radiation than other parts of the head and neck. Thus, they are ideal areas for incorporating stereotactic radiosurgery when they are involved by tumor. Another situation in which we use CyberKnife is when there is involvement of a cranial nerve near the skull base. The area around the foramen of the cranial nerve is then targeted because it is difﬁcult to treat surgically. Research is ongoing.

Conclusion There is little data on the long-term efﬁcacy of SRS in extracranial head and neck tumors. However, short-term reduction of pain and local control has been documented in several studies. Glomus tumors seem to have a good response to this modality. Patients with recurrent salivary gland cancer might beneﬁt as well. Large long-term studies are needed to address local control and survival beneﬁt in both primary and recurrent tumors. The beneﬁt of SRS in squamous cell cancer is unclear. It may be a useful adjuvant in recurrent or unresectable lesions when all other modalities have been exhausted. Despite its morbidity and often disappointing results, surgical salvage has shown a survival beneﬁt for recurrent lesions. Surgery is also the patient’s only hope for cure when prior surgery, radiation, and chemotherapy have failed. However, if a patient’s primary goal is palliation, SRS can be considered.

References 1. Schaefer U, Micke O, Schueller P, et al. Recurrent head and neck cancer: retreatment of previously irradiated areas with combined chemotherapy and radiation therapy—results of a prospective study. Radiology 2000; 216:317–376. 2. Kassaee A, Das IJ, Tochner Z, et al. Modiﬁcation of Gill-ThomasCosman frame for extracranial head-and-neck stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 2003; 57:1192–1195. 3. Ahn YC, Lee KC, Kim DY, et al. Fractionated stereotactic radiation therapy for extracranial head and neck tumors. Int J Radiat Oncol Biol Phys 2000; 48:501–505. 4. Elshaikh MA, Mahmoud-Ahmed AS, Kinney SE, et al. Recurrent head and neck chemodectomas: a comparison of surgical and radiotherapeutic results. Int J Radiat Oncol Biol Phys 2002; 52: 953–956. 5. Foote RL, Pollock BE, Gorman DA, et al. Glomus jugulare tumor: tumor control and complications after stereotactic radiosurgery. Head Neck 2002; 24:332–338. 6. Ryu S, Khan M, Yin FF, et al. Image-guided radiosurgery of head and neck cancers. Otolaryngol Head Neck Surg 2004; 130:690– 697. 7. Lee N, Millender LE, Larson DA, et al. Gamma knife radiosurgery for recurrent salivary gland malignancies involving the base of skull. Head Neck 2003; 25:210–216.

4 2

Head and Neck Malignancies: Chemotherapy and Radiation Perspective Mohan Suntharalingam, Kathleen Settle, and Kevin J. Cullen

Introduction Squamous cell carcinomas of the head and neck represent approximately 5% of cancers diagnosed in the United States each year. More than 40,000 patients will be diagnosed this year, and worldwide this number will exceed 500,000 [1]. Although patients presenting with early-stage disease have high cure rates with either radiation therapy or surgery alone, those diagnosed with locally advanced disease have historically had poorer prognoses and limited therapeutic options. Although most of these cancers tend to remain clinically limited to the head and neck region, locally advanced tumors can still present a signiﬁcant therapeutic challenge. These tumors are often characterized by invasion into surrounding anatomic structures such as muscle, bone, nerves, or blood vessels and often present with clinically detectable lymph node metastases. Surgical resection with adjuvant radiotherapy does offer a chance of cure for a portion of these patients; however, the opportunity for cure has come at a signiﬁcant cost in terms of function and cosmesis [2, 3]. These patients have historically had poor longterm survival, and a signiﬁcant portion die as a result of locoregional disease progression. The management of patients diagnosed with locally advanced squamous cell carcinomas of the head and neck has undergone a major paradigm shift during the past two decades. The standards of monotherapy of either surgery or radiation alone has been replaced by the multidisciplinary approach of combined modality therapy [4]. For many patients presenting with locally advanced disease, the concurrent application of systemic therapy and radiation has offered the opportunity for cure while simultaneously achieving organ preservation. Most recently, researchers have focused their efforts on optimizing

radiotherapy delivery schedules, deﬁning effective chemotherapy combinations, and the incorporation of novel biologic targeted therapies into treatment regimens [4–9].

Locoregional Control and Overall Survival The link between locoregional control and its effect on overall survival has long been appreciated in head and neck cancer patients. Historically, the overall survival of patients presenting with locally advanced disease has primarily been determined by the ability to achieve local control. Systemic chemotherapy agents with radiosensitizing properties have been administered with deﬁnitive doses of radiotherapy in an attempt to improve the outcome for this patient population. The integration of chemotherapy into the overall treatment scheme has been pursued with the goal of improving locoregional control and thereby affecting overall survival rates while simultaneously achieving organ preservation and maintaining function. A review of the concurrent chemotherapy and radiation literature reveals a consistent improvement in local control associated with combined modality therapy for patients presenting with advanced disease [6–8]. Whereas these ﬁndings have been encouraging, a critical review of these experiences has revealed signiﬁcant room for improvement. This important end point has served as the rationale for the design of current trials focused on the control of local regional disease. Despite the fact that combination chemotherapy and radiation trials have been conducted for the past four decades, the absolute survival beneﬁt associated with combined modality therapy had historically remained poorly deﬁned. In order to better deﬁne the potential advantage of cytotoxic therapy, four

425

426

m. suntharalingam et al.

large meta-analyses have been performed comparing chemotherapy (delivered as neoadjuvant, concurrently, or as adjuvant) plus local therapy versus local therapy alone [10–13]. The primary end point analyzed in the studies was overall survival. These four reports included three studies that were literature based [10–12] and one that updated actual patient data for the meta-analysis [13]. The total number of cases included in each analysis ranged from 4292 (from 28 trials) to as high as 10,741 (63 trials). The median follow-up for all patients ranged from 2 years to as long as 6.8 years [10–13]. It is interesting to note that the four separate meta-analyses reached similar conclusions despite signiﬁcant differences in the databases analyzed. The results from all of the studies conﬁrmed a small, albeit reproducible, survival beneﬁt in favor of the addition of chemotherapy. The magnitude of this beneﬁt ranged from an absolute survival advantage of 2.8% to 6.5% (Munro) [12]. The study by Pignon identiﬁed an overall survival advantage at both 2 and 5 years of 4%. At 5 years, the overall survival rate was 36% for those patients who received combination therapy compared with 32% for radiotherapy alone [11]. All four studies noted that the survival beneﬁt associated with the addition of chemotherapy was a result of the concurrent application of systemic therapy with radiation. This beneﬁt ranged from 8% to as high as 12% throughout the analyses.

Concurrent Chemotherapy and Radiation Multiple randomized trials have now documented the beneﬁt to the concurrent application of chemotherapy with external beam radiation in patients presenting with locally advanced squamous cell carcinomas of the head and neck. Forastiere and colleagues recently reported the results of a phase III randomized trial from the Radiation Therapy Oncology Group (RTOG) that examined the role of sequential versus concurrent chemotherapy with radiation for patients with stage III/IV laryngeal cancer [6]. The induction chemotherapy arm consisted of cisplatin (100 mg/m2) and 5-ﬂuorouracil (1000 mg/m2) followed by 70 Gy. The concurrent arm included cisplatin (100 mg/m2) delivered days 1, 22, and 43 of RT (70 Gy). This trial conclusively deﬁned the advantage of concurrent chemotherapy and radiation when local regional control was improved from 68% to 80% (p = 0.004). These trial results also emphasized the importance of achieving local control. Given the location of the primary tumors eligible for this study, a total laryngectomy was the surgical procedure required for salvage. The successful salvage of patients who did not achieve local tumor control with nonoperative therapy led to the ﬁnding of no signiﬁcant difference in overall survival between the treatment arms (76% vs. 74%, p = NS). This trial offers direct evidence of the connection between local control and the eventual impact on overall survival. Among the various anatomic subsites included in the category of head and neck malignancies, carcinomas involving the nasopharynx continue to stand out in terms of response to combined modality therapy and the associated improvement in overall survival. In 1998, Al-Sarraf and colleagues published the results of the Intergroup Phase III randomized trial involving patients presenting with stage III and IV squamous cell carci-

nomas of the nasopharynx [14]. This trial was designed to compare standard radiation (70 Gy) delivered in 7 weeks to 3 cycles of cisplatin (100 mg/m2) with the same radiation schedule. Patients randomized to the concurrent chemotherapy arm also received 3 cycles of cisplatin (80 mg/m2) and 5-ﬂuorouracil (1000 mg/m2). One hundred forty-seven patients were eligible for primary analysis of survival and toxicity. The 3-year progression-free survival was 24% for RT alone versus 69% (p < 0.001) in the chemoradiation arm. The 3-year overall survival rate was 47% versus 78% (p = 0.005) in favor of the combined modality treatment scheme. This study was the ﬁrst randomized trial to demonstrate an overall survival beneﬁt associated with the use of concurrent chemoradiation followed by adjuvant chemotherapy compared with radiotherapy alone in any head and neck cancer patient population. These results have helped to deﬁne the current standard of care for patients presenting with locally advanced squamous cell carcinomas of the nasopharynx. Once again, a direct correlation can be made between improving local regional control and its ultimate impact on overall survival. Multiple institutional trials have now documented local control rates between 80% and 90% for these patients when combined chemotherapy and modern radiotherapy are utilized [15, 16]. In fact, longer follow-up of these patients revealed that a majority of those who ultimately succumbed to their cancer did so as a result of metastatic disease with no evidence of local recurrence. Given the outstanding local control that has been achieved, this pattern of failure has led investigators to design future trials focused on addressing the risk of metastatic disease spread.

Targeted Therapies as Part of a Combined Modality Strategy The current limitations of most systemic cytotoxic agents are often deﬁned by the nonspeciﬁc toxicities of healthy tissues. To improve the therapeutic ratio, investigators have focused their interests on newer biologic agents targeting cellular protein receptors. Epidermal growth factor receptor, a member of the ErbB family of receptor tyrosine kinases, is one of four receptors critical to cellular proliferation, differentiation, and survival [17]. The EGFR pathway is important in controlling cell-cycle events that directly affect survival. Its expression has been associated with a poor outcome in squamous cell carcinomas of the head and neck (SCCHN) [18]. Given these facts, anti-EGFR therapy has been proposed as a potentially powerful agent in future combined modality treatment strategies. Anti-EGFR agents such as cetuximab (an IgG1 monoclonal antibody against the ligand binding domain of EGFR) have been studied in conjunction with radiotherapy. In vitro studies have suggested a synergistic relationship between RT and cetuximab in human squamous cell carcinomas of the head and neck cell lines [19]. A number of possible mechanisms have been purported, including (a) induction of G1 cell-cycle arrest, (b) inhibition of cellular proliferation, (c) promotion of radiation-induced apoptosis, (d) inhibition of radiation-induced damage repair, and (e) inhibition of tumor angiogenesis [20]. A recent phase III, randomized trial has deﬁned the beneﬁt of adding cetuximab to radiation for patients with squamous

42.

head and neck malignancies: chemotherapy and radiation perspective

cell carcinomas of the head and neck. Bonner et al. reported a signiﬁcant improvement in terms of locoregional control (hazard ratio 0.68, p = 0.005), progression-free survival (3-year PFS 42% vs. 31%, p = 0.04), and overall survival (median survival 49 months vs. 29.3 months, p = 0.03) associated with the addition of cetuximab. The 3-year overall survival rate was 55% for those treated with cetuximab and RT versus 45% for those treated with RT alone [21]. Notably, these improvements in outcome were achieved without a signiﬁcant increase in any grade 3 or greater acute reactions (other than acneform rash and infusion reactions). The ability to improve survival over radiation alone without increasing mucositis opens the door to the possibility of adding these types of agents to chemoradiation strategies.

Intensity-Modulated Radiation Therapy The delivery of deﬁnitive doses of radiation has, at times, been limited by the risk of normal tissue damage. Given the anatomic proximity of vital structures such as the spinal cord, salivary glands, optic nerves, and the mandible, tumors of the head and neck can present a signiﬁcant therapeutic challenge. The longterm consequences of tumoricidal doses of radiation delivered to this anatomic region can be quite signiﬁcant. Standard beam arrangements and treatment planning techniques are able to achieve uniform dose delivery throughout an entire anatomic region. The ability to deliver a homogeneous dose to a tumorbearing region is advantageous, however this proves to be a disadvantage in terms of doses delivered to the surrounding normal tissue. Nasopharyngeal carcinoma has lent itself to the study of technological advances of radiotherapy delivery given its unique anatomic location and the close proximity of multiple critical structures. The complexity of treatment planning is made more difﬁcult given the dose constraints of the base of the skull, associated cranial nerves, retina, and brain stem. The advent of three-dimensional treatment planning allowed treating physicians to gain a better understanding of the potential toxicities associated with the use of external beam radiation. Recent advances such as intensity-modulated radiation therapy (IMRT) have clearly improved the ability to safely deliver effective total doses of radiation to tumors in this anatomic location while simultaneously decreasing the risk of untoward side effects [22]. Multiple institutions have now published their experiences describing the utilization of intensity-modulated radiotherapy techniques in conjunction with concurrent chemotherapy for nasopharyngeal carcinoma patients. The results have uniformly documented the ability to improve the precise delivery of radiation to targets in close proximity to the base of the skull while respecting the surrounding normal tissue dose tolerances. The long-term outcomes for these patients presenting with locally advanced disease have deﬁned local control rates between 85% and 90% [15, 16]. Recently, the RTOG has conducted a multiinstitutional phase II trial utilizing IMRT and concurrent chemotherapy. This study represents the ﬁrst cooperative group trial focused on use of this technology in conjunction with systemic therapy. Accrual to this trial is now complete; however, longer follow-up will be necessary prior to reporting the outcome results.

427

Altered Fractionation Schedules Given the primary role that radiation therapy has historically played in the management of head and neck cancers, it is not difﬁcult to understand the rationale behind the investigation of altered fractionation in this patient population. Numerous strategies have been tested in an attempt to improve tumor control rates. Altered fractionation radiotherapy refers to the delivery of multiple fractions per day in order to increase the total dose without signiﬁcantly increasing the overall treatment time. Hyperfractionated radiotherapy is characterized by an increase in the total dose and the number of fractions, a decrease in the dose per fraction, and maintaining the overall treatment time [23]. In accelerated fractionation schemes, the overall treatment time is signiﬁcantly reduced while the number of fractions, the total dose, and fraction size remain relatively unchanged [24]. The radiobiologic principles of both of these schemes are based on the desire to increase tumor cell kill without increasing the long-term toxicities. These schedules are typically delivered with the understanding that patients often experience increased acute toxicities. Altered fractionation schedules have provided physicians with the opportunity to improve locoregional control for patients presenting with advanced tumors [25]. These improvements in local control have generally not translated into better overall survival. Researchers have therefore evaluated concurrent chemotherapy with altered fractionation in an attempt to improve both the local control and overall survival for this patient population. Brizel and colleagues reported the results of a institutional randomized trial designed to compare hyperfractionated radiotherapy delivered at 1.25 Gy twice a day to a total dose of 75 Gy versus 70 Gy (1.25 Gy twice a day) to 70 Gy with the addition of cisplatin (12 mg m−2 day−1) and 5-ﬂuorouracil (600 mg m−2 day−1) given weeks 1 and 6 of radiation [8]. Patients randomized to receive concurrent therapy were given two additional cycles of chemotherapy at the completion of radiation. One hundred twenty-two patients were entered onto study and the median follow-up was 41 months when it was initially reported. The complete response rate was 73% for the RT-alone group versus 88% for those receiving CT + RT (p = 0.52). The locoregional control was 44% in the radiotherapy arm versus 70% with combined modality treatment (p = 0.01). This improvement in local control did not translate into a statistically signiﬁcant overall survival advantage (RT alone 34 % vs. CT + RT 55%, p = 0.07) when the trial was originally published. With longer follow-up, the local control advantage was maintained. Jeremic et al. published the results of another randomized trial again comparing hyperfractionated radiotherapy alone (77 Gy in 1.1 Gy/twice daily) versus the same radiotherapy schedule delivered in conjunction with daily, low-dose cisplatin (6 mg m−2 day−1) [26]. This trial included patients with tumors located in a variety of anatomic locations. With a median follow-up of 79 months, this trial reported statistically signiﬁcant improvements associated with the delivery of combined modality therapy in all end points evaluated. The complete response rates were signiﬁcantly higher for the combined modality group (75%) versus those receiving RT alone (45%) (p = 0.002). At 5 years, the overall survival was 46% for CT + RT versus 25% for the RT arm (p = 0.041). The locoregional

428

m. suntharalingam et al.

TABLE 42-1. Local control and survival with concurrent chemotherapy and radiation. Author

Site

Brizel et al.

Aldelstein et al.

OC, OP, HYP, LAR, PNS OC, OP, NP, HYP, LAR OC, OP, HYP, NP, MaxSinus, LAR OC, OP, LAR, HYP

Wendt et al. Forastiere et al.

Jeremic et al. Al-Sarraf et al.

Dose per fraction

Chemo

70 Gy

Total dose

1.25 Gy twice daily

77 Gy

1.1 Gy twice daily

Cisplatin, 5-FU weeks 1,6 (2 cycles out back) Daily bolus cisplatin

LC

66–73.8 Gy

1.8–2 Gy daily

66–72 Gy

1.8–2 Gy daily

OC, OP, HYP, LAR

70.2 Gy

1.8 Gy daily

LAR

70 Gy

2 Gy daily

Cisplatin every 3 weeks Cisplatin, 5-FU every 3 weeks Cisplatin, 5-FU every 3 weeks Cisplatin every 3 weeks

DFS/PFS

3-yr 70% 5-yr 50%

OS

3-yr 55% 5-yr 46%

5-yr 46%

4-yr 43%

4-yr 34%

5-yr 77%

5-yr 50%

3-yr 35%

3-yr 48%

2-yr 80%

2-yr 75% 5-yr 55%

5-FU, 5-ﬂuorouracil.

progression-free survival rates were 50% and 36% (p = 0.041), respectively, while the distant metastasis–free survival were 86% and 57% (p = 0.0013). It was interesting to note that the radiation-induced toxicities were similar in both groups (including grade III/IV xerostomia and subcutaneous ﬁbrosis), whereas high-grade hematologic toxicities were more commonly seen in the combined modality group. These two trials clearly demonstrate that systemic therapy can be safely delivered with altered fractionated radiotherapy. This treatment scheme holds tremendous potential as the therapeutic gains may lead to signiﬁcant improvements in local control, distant metastatic rates, and, ultimately, overall survival. At the same time, they serve as an important reminder that these advances often come at the cost of increased toxicities associated with this aggressive treatment schedule. Concurrent chemotherapy and external beam radiotherapy has had a dramatic impact in improving local control and in many circumstances survival for patients with locally advanced squamous cell carcinomas of the head and neck. Table 42-1 depicts the outcomes of multiple phase III trials that have documented these beneﬁts over radiotherapy alone. These results also give insight into the possible role for emerging technologies such as extracranial stereotactic radiosurgery. The local control for patients presenting with advanced disease still leaves signiﬁcant room for improvement. It is certainly conceivable that the safe application of focused high doses of radiotherapy could have a dramatic impact in appropriately selected patients.

Challenges of Stereotactic Therapy The basic tenets of stereotactic therapy must be respected as one considers the use of this therapy as a means of improving local control in this patient population. In order to minimize

normal tissue exposure, patient and organ motion must be minimized. This is especially important given the close proximity of tumor targets to the surrounding critical structures. The concept of delivering a stereotactic boost to tumor-bearing regions at the completion of standard fractionated therapy is attractive based on the local control probabilities associated with conventional fractionated radiotherapy and concurrent chemotherapy (Table 42-2). The desire to improve the primary tumor control must be balanced against the realties of the toxicity proﬁles that may be encountered when adding a high-dose stereotactic boost after conventional therapy [27]. When one considers the volume of tumor typically associated with locally advanced head and neck cancer, these volumes would be required to be included in any “boost,” and the ﬁeld sizes may prove to be difﬁcult to safely deliver. Radiosurgery has been considered as an alternative therapeutic option for salvage therapy for those patients presenting with recurrent disease after having undergone deﬁnitive radiotherapy [28]. This has been especially true in the management of recurrent nasopharyngeal carcinoma. In order to appropriately frame the results that have been reported to date with the use of stereotactic therapy, it is important to understand the data that currently exists regarding the external beam reirradiation.

The Option of Reirradiation Patients found to have recurrent squamous cell carcinoma of the head and neck have historically had a dismal prognosis. These patients have typically been managed with systemic therapy alone. Most combination therapies have documented response rates between 20% and 30% and median survivals of 5 to 6 months. Against this background, investigators began

TABLE 42-2. IMRT with chemotherapy for nasopharyngeal cancer. Author, institution

Site

Total dose

Dose per Fraction

Chemo

LC

DFS/PFS

OS

Wolden et al., MSK

NP

70 Gy

1.8/1.5 Gy CB

3-yr 91%

3-yr 67%

3-yr 83%

Lee, USCF

NP

65–70 Gy

1.8–2 Gy daily

Platinum-based; concurrent and adjuvant Cisplatin concurrently; CP/5-FU adjuvant

4-yr local progression-free survival 97%

4-yr local-regional progression-free survival 98%

4-yr 88%

5-FU, 5-ﬂuorouracil.

42.

429

head and neck malignancies: chemotherapy and radiation perspective

TABLE 42-3. Reirradiation with concurrent chemotherapy. Author

Total dose

Dose per fraction

Chemo

TX delivery

LC

DFS/PFS

OS

Haraf et al. Spencer et al.

Median 50 Gy 40–60 Gy

HU, 5-FU HU, 5-FU

CRT given on alternate weeks CRT given on alternate weeks

5-yr 20%

5-yr 24%

5-yr 15% 2-yr 20%

RTOG 96-10 RTOG 99-10

60 Gy 60 Gy

1.5 Gy twice daily 2 Gy daily 1.2 Gy twice daily 1.5 Gy twice daily 1.5 Gy twice daily

HU, 5-FU Cisplatin, paclitaxel

CRT given on alternate weeks CRT given on alternate weeks

1-yr 62%

1-yr 42% 2-yr 20%

5-FU, 5-ﬂuorouracil.

examining the potential of delivering second courses of deﬁnitive doses of radiotherapy with concurrent chemotherapy (Table 42-3). Spenser and colleagues from the University of Alabama at Birmingham published the results of their phase I/II trial of reirradiation with concurrent chemotherapy [29]. The radiation dose escalation strategy ultimately deﬁned the ability to deliver an additional 60 Gy (delivered at 1.5 Gy twice daily) given with concurrent 5-ﬂuorouracil and hydroxyurea. All therapy was delivered on a week-on, week-off schedule. The response rate of 74% was quite encouraging. The median survival for the entire group was 10.5 months, and the 2-year overall survival was 20%. Most importantly, this was one of the ﬁrst series to accurately deﬁne the risks associated with reirradiation as 4 of the original 35 patients developed severe late toxicities (2 radiation myelitis, 2 esophageal strictures). These ﬁndings were subsequently validated by a multi-institutional phase II trial through the RTOG. RTOG 96-10 conﬁrmed the ﬁndings of Spenser et al. and documented the safety and efﬁcacy of this approach in patients presenting with unresectable recurrent or second primary malignancies involving previously irradiated sites in the head and neck [30]. Investigators at the Fox Chase Cancer Center have also published their experience with reirradiation and concurrent chemotherapy for unresectable, recurrent head and neck cancer. Their treatment scheme included a similar radiation schedule of 1.5 Gy twice daily to a total dose of 60 Gy, however, the systemic therapy was signiﬁcantly different. Daily cisplatin (15 mg/ m2) and paclitaxel (20 mg/m2) were given along with radiation on a week-on, week-off schedule. Again, toxicities were acceptable, and 2-year survivals were approximately 20% [31]. These results were again validated by an RTOG phase II multi-institutional trial. Horowitz et al. presented the results of this trial at the 2005 ASTRO annual meeting [32]. The median survival was 12.5 months, and the 2-year overall survival for the entire cohort of 99 patients was 26%. It is of note that there were eight grade 5 toxicities (ﬁve acute and three late) encountered with the regimen. This ﬁnding is consistent with other reirradiation experiences. It is important to note that although these treatment schemes have achieved impressive response rates and 2-year survivals of approximately 20%, there is a treatment-related mortality ranging from 5% to 15%.

Conclusion Patients presenting with locally advanced head and neck cancer represent one of the most challenging patient groups in oncology. The past few decades have borne witness to a fundamental

paradigm shift toward the routine application of combinedmodality therapy. This approach most commonly employs the use of concurrent chemotherapy and radiation in order to achieve local control, improve survival, and simultaneously provide organ preservation. Investigators continue to explore new agents and technology in the continued pursuit of these worthwhile goals. The consideration of stereotactic radiation therapy may ultimately prove itself to be an effective tool in the effort to improve local control. It is, however, imperative that we proceed with caution as we attempt to add high doses per fraction to our target volumes. The proximity and sensitivity of the surrounding normal structures must be appreciated when tumors are delineated and dose constraints are deﬁned. The concurrent application of chemotherapy and radiation has set the bar by which any new therapies will be judged. Any improvements in local control or overall survival should be evaluated in light of the existing data.

References 1. Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. CA Cancer J Clin 2004; 54:8–29. 2. Cooper JS, Pajak TF, Forastiere AA, et al. Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head and neck. N Engl J Med 2004; 350:1937– 1944. 3. Bernier J, Domenge C, Ozsahin M, et al. Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer. N Engl J Med 2004; 350:1945–1952. 4. Adelstein DJ, Li Y, Adams GL, et al. An intergroup phase III comparison of standard radiation therapy and two schedules of concurrent chemoradiotherapy in patients with unresectable squamous cell head and neck cancer. J Clin Oncol 2003; 21: 92–98. 5. Calais G, Le Floch O. [Concomitant radiotherapy and chemotherapy in the treatment of cancers of the upper respiratory and digestive tracts]. Bull Cancer Radiother 1996; 83:321–329. 6. Forastiere AA, Goepfert H, Maor M, et al. Concurrent chemotherapy and radiotherapy for organ preservation in advanced laryngeal cancer. N Engl J Med 2003; 349:2091–2098. 7. Al-Sarraf M, Pajak TF, Marcial VA, et al. Concurrent radiotherapy and chemotherapy with cisplatin in inoperable squamous cell carcinoma of the head and neck. An RTOG Study. Cancer 1987; 59:259–265. 8. Brizel DM, Albers ME, Fisher SR, et al. Hyperfractionated irradiation with or without concurrent chemotherapy for locally advanced head and neck cancer. N Engl J Med 1998; 338:1798– 1804. 9. Forstiere AA, Maor M, Weber R, et al. Phase III trial to preserve the larynx: induction chemotherapy and radiotherapy versus concomitant chemoradiotherapy versus radiotherapy alone: results of

430

10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

m. suntharalingam et al. the Intergroup trial R91-11. Proceedings of ASCO. J Clin Oncol 2001; 20:2a. Stell PM. Adjuvant xhemotherapy in head and neck cancer. Semin Radiat Oncol 1992; 2:195–205. Pignon JP, Bourhis J, Domenge C, et al. Chemotherapy added to locoregional treatment for head and neck squamous-cell carcinoma: three meta-analyses of updated individual data. MACHNC Collaborative Group. Meta-analysis of chemotherapy on head and neck cancer. Lancet 2000; 355:949–955. Munro AJ. An overview of randomised controlled trials of adjuvant chemotherapy in head and neck cancer. Br J Cancer 1995; 71:83–91. Bourhis J, Pignon JP. Meta-analyses in head and neck squamous cell carcinoma. What is the role of chemotherapy? Hematol Oncol Clin North Am 1999; 13:769–775. Al-Sarraf M, LeBlanc M, Giri PG, et al. Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: phase III randomized Intergroup study 0099. J Clin Oncol 1998; 16:1310–1317. Lee N, Puri DR, Blanco AI, Chao KS. Intensity-modulated radiation therapy in head and neck cancers: an update. Head Neck 2007; 29:387–400. Wolden SL, Chen WC, Pﬁster DG, et al. Intensity-modulated radiation therapy (IMRT) for nasopharynx cancer: update of the Memorial Sloan-Kettering experience. Int J Radiat Oncol Biol Phys 2006; 64:57–62. Salomon DS, Brandt R, Ciardiello F, et al. Epidermal growth factor-related peptides and their receptors in human malignancies. Crit Rev Oncol Hematol 1995; 19:183–232. Ang KK, Andratschke NH, Milas L. Epidermal growth factor receptor and response of head-and-neck carcinoma to therapy. Int J Radiat Oncol Biol Phys 2004; 58:959–965. Herbst RS, Langer CJ. Epidermal growth factor receptors as a target for cancer treatment: the emerging role of IMC-C225 in the treatment of lung and head and neck cancers. Semin Oncol 2002; 29:27–36. Huang SM, Bock JM, Harari PM. Epidermal growth factor receptor blockade with C225 modulates proliferation, apoptosis, and radiosensitivity in squamous cell carcinomas of the head and neck. Cancer Res 1999; 59:1935–1940. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N Engl J Med 2006; 354:567–578.

22. Lee N, Xia P, Fischbein NJ, et al. Intensity-modulated radiation therapy for head-and-neck cancer: the UCSF experience focusing on target volume delineation. Int J Radiat Oncol Biol Phys 2003; 57:49–60. 23. Mendenhall WM, Million RR, Stringer SP, et al. Squamous cell carcinoma of the glottic larynx: a review emphasizing the University of Florida philosophy. South Med J 1999; 92:385– 393. 24. Wang CC, Eﬁrd J, Nakfoor B, et al. Local control of T3 carcinomas after accelerated fractionation: a look at the “gap”. Int J Radiat Oncol Biol Phys 1996; 35:439–441. 25. Fu KK, Pajak TF, Trotti A, et al. A Radiation Therapy Oncology Group (RTOG) phase III randomized study to compare hyperfractionation and two variants of accelerated fractionation to standard fractionation radiotherapy for head and neck squamous cell carcinomas: ﬁrst report of RTOG 9003. Int J Radiat Oncol Biol Phys 2000; 48:7–16. 26. Jeremic B, Shibamoto Y, Milicic B, et al. Hyperfractionated radiation therapy with or without concurrent low-dose daily cisplatin in locally advanced squamous cell carcinoma of the head and neck: a prospective randomized trial. J Clin Oncol 2000; 18: 1458–1464. 27. Ahn YC, Lee KC, Kim DY, et al. Fractionated stereotactic radiation therapy for extracranial head and neck tumors. Int J Radiat Oncol Biol Phys 2000; 48:501–505. 28. Cmelak AJ, Cox RS, Adler JR, et al. Radiosurgery for skull base malignancies and nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 1997; 37:997–1003. 29. Spencer SA, Wheeler RH, Peters GE, et al. Concomitant chemotherapy and reirradiation as management for recurrent cancer of the head and neck. Am J Clin Oncol 1999; 22:1–5. 30. Spencer SA, Harris J, Wheeler RH, et al. RTOG 96-10: reirradiation with concurrent hydroxyurea and 5-ﬂuorouracil in patients with squamous cell cancer of the head and neck. Int J Radiat Oncol Biol Phys 2001; 51:1299–1304. 31. Langer C, Damsker J, Ridge A. Reirradiation: exploring new territory in the therapy of recurrent head and neck cancer. Clin Adv Hematol Oncol 2003; 1:424–429, 440. 32. Horowitz E HJ, Langer C, Nicolaou N, et al. RTOG 99-11: Phase II study of concurrent chemotherapy and re-irradiation for patients with recurrent squamous cell cancer of the head and neck. Int J Biol Phys Radiat Oncol 2005; 63(Suppl):185.

4 3

Spinal Tumors Robert L. Dodd, Iris Gibbs, John R. Adler Jr., and Steven D. Chang

Introduction Although spinal neoplasms constitute only a fraction of central nervous system tumors [1, 2], treatment of these lesions remains an important component of neurosurgical practice. The protective function provided by the spinal column requires a tightly conﬁned relationship between complex skeletal anatomy and delicate neuronal structures. As a result, access for tumor removal is often limited, and treatment strategies other than microsurgery are frequently employed. Radiation therapy has become a mainstay in the treatment of both primary [3] and metastatic disease of the spine [4] and spinal cord [5]. Such therapy frequently provides signiﬁcant palliative beneﬁt and, in rare cases, can even be locally curative. Standard radiation techniques for treating spinal tumors typically rely on treatment ﬁelds extending up to two vertebral segments above and below the target to compensate for set-up errors and patient movement during treatment. Of even greater importance, the close proximity of the spinal cord typically makes it impossible to exclude it from the highest dose region. These technical limitations result in a signiﬁcant amount of normal tissue being included in the radiation ﬁeld, the most important being the spinal cord. In an effort to protect these normal structures, conventional radiation therapy is highly fractionated. Despite such radiobiologic manipulations, the dose tolerance of the spinal cord ultimately precludes the administration of larger and more effective quantities of radiation, thereby limiting the effectiveness of radiation therapy in the deﬁnitive management of most spinal lesions [6]. After standard radiotherapy for spinal lesions, relapse within the treatment ﬁeld is not uncommon [7–9]. Despite the higher rate of failure, it is deemed unacceptable to use larger and more therapeutically beneﬁcial doses, given the risk of spinal cord injuries such as radiation-induced myelopathy. Shaping the irradiation beam to strictly involve the target volume would theoretically maximize the dose to the lesion while minimizing the risk of injury to the spinal cord. The newest radiation technology, intensity-modulated radiation therapy, attempts to achieve this goal [10, 11]. This technology enables the shape of the treatment ﬁeld to more closely match that of the lesion being treated. However, most of the advantages of such dose conformity go unrecognized because the method of targeting radiation remains relatively imprecise. When performing conventional radiotherapy, the radiation

beams are aimed at the lesion by manually positioning the patient by means of surface markers attached to the skin. The inherent nature of this process results in limited spatial accuracy and reproducibility. As a consequence, even this most modern of radiation therapy techniques fails to address the unique demands of optimally irradiating the spinal axis. Ultimately, the dilemma with all current irradiation methods is that greater success can only be achieved at the cost of substantially increasing the risk of devastating complications.

Development of Stereotactic Spinal Radiosurgery Stereotactic radiosurgery combines the principles of stereotactic localization with multiple radiation beams from a highly collimated radiation source to deliver a high-dose irradiation to a target while limiting exposure to normal tissue [12]. Stereotactic radiosurgery has long been used to treat intracranial lesions. Most current radiosurgical methods rely on a rigid frame ﬁxed directly to a patient’s skull and presume a ﬁxed relationship between the target and the cranium. In the past 15 years, the results of stereotactic radiosurgery have redeﬁned the standards of treatment for a range of intracranial and skull-based pathology. Furthermore, the extremely favorable intracranial experience with radiosurgery has inspired the development of new techniques for stereotactically irradiating the spine. Spinal lesions also often have a more or less ﬁxed relationship to the bony spine. Capitalizing on this, stereotactic methods for targeting within the spine were developed that, not surprisingly, bear a strong resemblance to the many techniques for intracranial localization that were developed in parallel. Woroschiloff performed the ﬁrst spinal stereotactic surgery in 1874 when he used a spinal localization device attached directly to the spinous processes to make cord lesions with knife or electrode in conjunction with electrophysiologic studies [13]. Over the years, modiﬁcations of his original design have been employed to perform cervical cordotomies [14], high cervical tractotomies [15], and thermocoagulation of lumbosacral centers of the spinal cord [16]. However, it was not until the mid-1990s that Hamilton and colleagues [17–19] developed a spinal stereotactic radiosurgery system using a modiﬁed linear accelerator (linac) and a skeletal ﬁxation frame.

431

432

r.l. dodd et al.

This system employed a patient-encircling rigid frame for reference and immobilization. This device could be ﬁrmly attached to the spinous processes by means of a series of small incisions made under general anesthesia [20]. Although preliminary results were promising, frame ﬁxation was cumbersome and resulted in long procedures. Moreover multiple treatments (i.e., treatment fractionation) were impractical. Another drawback of frame-based radiosurgical systems is that they typically use overlapping ﬁxed isocenters to achieve a degree of ﬁeld shaping when treating nonspherical targets. This approach, however, results in an inhomogeneous dose of radiation within the treatment area that can cause undertreatment (cold spot) and/or overtreatment (hot spot) of the targeted lesion. Such inhomogeneity is particularly problematic when administering aggressive doses of radiation immediately adjacent to the spinal cord. The CyberKnife, a frameless image-guided radiosurgery system developed by Accuray, Inc (Sunnyvale, CA), overcomes many of the limitations encountered by Hamilton et al. [11, 21, 22]. An external ﬁxation device is eliminated because targeting is based on internal radiographic features such as skeletal anatomy or implanted ﬁducials. Because this design allows real-time images to be acquired during treatment, the system is capable of measuring any change in target location and then precisely retargeting the radiation beam in order to compensate. Finally, in contrast with other radiosurgical techniques, a ﬁxed isocenter is not used by the CyberKnife. As a result, more conformal treatment of irregularly shaped targets is possible, and greater in-ﬁeld homogeneity can be achieved. Since 1994, the CyberKnife has been used at more than 30 sites, including Stanford University Medical Center. More than 8000 patients with benign and malignant intracranial lesions have undergone CyberKnife radiosurgery with outcomes similar to conventional frame-based techniques [23–27]. Because many lesions in the spine are histopathologically similar to those in the brain, it is not unreasonable to assume that treatment outcomes using the same modality would be comparable. Beginning in 1996, the Stanford CyberKnife has been used to test this hypothesis in a highly selected group of patients. Treatment was initially performed under an investigational device exemption (IDE) until the Food and Drug Administration (FDA) approved the CyberKnife for extracranial applications in 2001. Since that

TABLE 43-1. Indications and contraindications of stereotactic spinal radiosurgery. Indications Progressive but minimal neurologic deﬁcit Postresection local irradiation Disease progression despite previous surgery and/or irradiation Patients with severe medical comorbidities that preclude surgery Inoperable lesions Contraindications Severe neurologic deﬁcit with signiﬁcant cord compression Neurologic deﬁcit caused by bony compression Spinal instability Lesion not responsive to irradiation Maximal tolerable doses of irradiation already given to adjacent spinal cord

TABLE 43-2. Treated/treatable neoplasms with CyberKnife radiosurgery. Benign Meningioma, schwannoma, neuroﬁbroma, hemangioblastoma, chordoma, paraganglioma, ependymomas, epidermoid Malignant/metastatic Breast, renal, non–small cell lung, colon, gastric, and prostate metastases, squamous cell (laryngeal, esophageal, and lung) tumors; osteosarcoma; carcinoid; multiple myeloma; clear cell carcinoma; adenoid cystic carcinoma; malignant nerve sheath tumors; endometrial carcinoma; malignant neuroendocrine tumor; Ewing sarcoma

time, the CyberKnife has been successfully used to treat a broad range of lesions throughout the spine, chest, and abdomen [24–27].

Indications for Spinal Radiosurgery The indications for spinal radiosurgery have evolved rapidly as experience with this relatively new technology has accumulated (Table 43-1). The primary indications include minimal but progressive neurologic deﬁcits, the need for postresection local irradiation, radiographic progression despite previous surgery or irradiation, or treatment of patients with inoperable lesions. Similar to cranial radiosurgery, spinal lesions appropriate for CyberKnife radiosurgery need to be reasonably well circumscribed and not so large as to negate the beneﬁts of steep dose gradient. The selected lesions were usually either unresectable or associated with signiﬁcant medical comorbidities, thereby precluding surgery. A small subset of patients undergoing spinal radiosurgery had refused open surgical resection. A wide spectrum of spinal neoplasms has been treated, including primary and metastatic tumors (Table 43-2). Spinal radiosurgery is contraindicated in patients with signiﬁcant spinal cord compression causing a substantial neurologic deﬁcit. This contraindication applies particularly to tumors for which the temporal response to treatment is delayed, such as benign lesions. Furthermore, spinal radiosurgery should be avoided when there is evidence of signiﬁcant spine instability or when a patient’s neurologic deﬁcit is caused mainly by bony compression. If the adjacent spinal cord has already received the maximally tolerated dose of radiation, further irradiation can be dangerous. Although it can be a difﬁcult decision to make, this latter group of patients would usually be better treated with surgery and/or chemotherapy, if possible.

Rationale for Radiosurgical Treatment of Benign Spinal Tumors Several publications report preliminary evidence of the feasibility of spinal radiosurgery, mostly for spine metastases [11, 28– 30]. However, because the life expectancy of most of these patients is measured in months, and because radiation injury can take even years to manifest itself, the longer-term efﬁcacy and safety of spinal radiosurgery is yet to be established.

43.

spinal tumors

Benign intradural extraaxial tumors of the spine consist mainly of spinal meningiomas, schwannomas, and neuroﬁbromas the majority of which are noninﬁltrative and can be completely and safely resected by experienced surgeons. After complete tumor removal, recurrence is low, ranging from 0 to 14% [2, 31–34]. Moreover, advanced operative techniques now enable surgical access to and the means to stabilize all areas of the thoracic [35] and upper cervical spine [36] so that even ventral and lateral tumors in these regions can be removed with modest morbidity [2, 37, 38] and a surgical mortality of less than 3% [2, 33]. Nevertheless, certain patients are less than ideal candidates for standard surgical resection because of age, medical comorbidities, the recurrent nature of a tumor, or because multiple lesions occur in the setting of one of the familial phakomatoses. It is in such clinical circumstances that radiosurgery could be an attractive clinical option. Despite this theoretical attraction, the literature on malignant spinal metastases provides an insufﬁcient basis for judging the possible beneﬁts of radiosurgical ablation of benign spine tumors. Because patients with such lesions have prolonged life expectancies, the potential for delayed and possibly catastrophic radiation myelopathy is a special concern. In addition, benign spine tumors have their own unique presentation, spatial relationship to the spinal cord, and radiobiologic response to radiosurgery, any of which present unique challenges to the safe and effective application of radiosurgical ablation. Throughout the modern era of neurosurgery, surgical resection has been the mainstay for managing benign spinal neoplasms. Nevertheless, there are sporadic accounts of radiation therapy being used successfully, as a surgical adjuvant, in small numbers of patients [33, 39, 40]. In addition, one would expect benign spinal lesions to respond to radiosurgery much the same as their intracranial counterparts, given their biologic similarities. Backed by these assumptions, and equipped with recent image-guided radiosurgical technologies, a few recent investigators have reported, in a preliminary manner, the results of treating a small number of benign spinal lesions with radiosurgery. Ryu et al. published the ﬁrst description of stereotactic radiosurgery being used for benign spinal lesions [11]. Their report of 16 spinal lesions included two schwannomas, one meningioma, and a hemangioblastoma. Subsequently, Gerszten et al. published a series of 15 benign spinal tumors treated with the CyberKnife, nine of which were neuroﬁbroma, schwannoma, or meningioma [41]. No tumor progression was reported in either study, however the minimum follow-up in both publications was 6 months, and the mean follow-up period in the latter study was a mere 12 months. Meanwhile, in the recent series by DeSalles et al. of 14 patients treated with spinal radiosurgery, there were individual cases of neuroﬁbroma and schwannoma, both of which were stable in size after a mean follow-up of 6 months [30]. The small size of all the above published series and the very short follow-up for such a group of relatively slow-growing lesions make it impossible to conclude anything other than that spinal radiosurgery is feasible. In 1997, the ﬁrst patient with a benign spinal lesion was treated using the CyberKnife, and since that time 101 patients with a variety of benign spinal tumors and vascular malformations have undergone radiosurgical ablation at our institution.

433

CyberKnife Stereotactic Radiosurgery System The image-guided CyberKnife Stereotactic Radiosurgery (SRS) System was used to administer spinal radiosurgery in every case in this series. The instrument has been previously described in detail by Adler et al. [21]. In brief, the device consists of a 6-MV linac mounted on a computer-controlled robotic arm, which is coupled to an X-ray tracking system that monitors and adjusts the linac’s alignment based on changes in the target’s position in near real time (Fig. 43-1). The image-guidance system eliminates the need for skeletal frame immobilization. The components of the imaging system are at ﬁxed positions within the treatment room, providing a stationary frame of reference to locate the patient’s spine. It consist of two orthogonally positioned ceiling-mounted diagnostic X-ray tubes, aimed at two amorphous silicon ﬂat-panel digital detectors (dpiX LLC, Palo Alto, CA) capable of generating high-resolution images, which are coupled to a high-speed Silicon Graphics workstation (SGI, Mountain View, CA). Once the spine has been located within the imaging system’s coordinate frame, the position of the lesion is known. Targeting is based on the assumption of a ﬁxed relationship between the lesion and the spine. Real-time radiographs are automatically registered to digitally reconstructed radiographs (DRRs) derived from the treatment planning computed tomography (CT) scan. The most recent CyberKnife software measures both translation and rotation of the anatomy by iteratively changing the position of the anatomy in the DRR until an exact match of the two radiographs and the two DRRs is achieved [27]. This algorithm does not require a database of precomputed DRRs or the array processor used in the previous software, thus eliminating the need to ﬁx the orientation of the patient during the treatment. Once the location of the spine is determined, a control loop between the imaging system and the

FIGURE 43-1. CyberKnife frameless stereotactic radiosurgery suite. (A) A modiﬁed 6-MV X-band linac designed speciﬁcally for radiosurgery is mounted on a highly maneuverable robotic manipulator (KUKA Roboter GmH, Augsburg, Germany). (B) Two high-resolution X-ray cameras are mounted orthogonally to the headrest. (C) One of the two X-ray sources is mounted in the ceiling projecting onto the camera. (D) The treatment couch is mobile, allowing the X-ray sources to image targets at any point along the neuroaxis.

434

r.l. dodd et al.

robotic arm adjusts the pointing of the linac therapeutic beamlet to the observed position of the target. If the patient moves, the change is detected and the beam is precisely realigned with the target in less than 1 second with submillimeter accuracy [10, 22, 42]. The CyberKnife system is designed so that when a patient is positioned within the treatment area, the target is near the center of an 80-cm3 sphere that is deﬁned with respect to the patient’s anatomy. There are 100 equally spaced points on the surface of this sphere, and these points are called nodes. At each node, the robot deﬁnes up to 12 beams of radiation that intersect various portions of the tumor volume. The robotic arm moves the linac to, and stops at, each node, where the beam is precisely aimed and a prescribed dose of irradiation is administered. Treatment plans are formed from a subset of the entire constellation of possible radiation beam trajectories, which maximizes irradiation of the target and minimizes normal tissue irradiation. In practice, not all nodes are available because objects within the treatment room either interfere with the Xray beam or prevent the robotic arm from positioning the linac at a particular node. During system installation, these obscured nodes are disabled. At least 50 nodes, and often more than 200 beams, are utilized during treatment. Because the robot can aim a beam anywhere within the tumor volume (as opposed to only the center of the sphere like most radiosurgical systems), highly conformal treatment plans can be created for complex geometries in a “non-isocentric” manner [27].

Medical Products, Inc., Akron, OH). These devices were used to noninvasively immobilize the spine during CT imaging (used later in treatment planning) and also to restrict patient movement during the radiosurgery itself. To optimize our accuracy, ﬁne-cut 1.25-mm slices are typically obtained. For those tumors that enhanced poorly with iodinated contrast or when severe allergy precluded the acquisition of a contrast CT, a contrast magnetic resonance imaging (MRI) scan of the relevant spine was also acquired and fused to the pretreatment CT scan. Just prior to radiosurgery, a library of DRRs (i.e., computer simulated X-rays) are calculated from the perspective of a pair of X-ray sources and cameras used throughout radiosurgery. This array of images encompasses those vertebral elements, along with embedded ﬁducials, that are in close proximity to the radiosurgical target. Treatment planning begins with the treating neurosurgeon deﬁning the target volume and critical structures within the CT/MR images using software tools provided on the CyberKnife workstation [21, 25, 43] (Figs. 43-2 and 43-3). A proprietary inverse planning computer algorithm uses the above inputs to determine the number, direction, and duration of treatment beamlets so as to optimize dose conformality and minimize irradiation of critical structures. Visual inspection and analysis of dose-volume histograms for the target region and adjacent critical anatomy are performed in each patient as part of the process of ﬁnding the best radiosurgical solution.

Treatment Delivery

Treatment Procedure Targeting Spinal Lesions with Fiducial Placement The process of image-guided radiosurgical targeting of spine lesions begins with an outpatient procedure that inserts small stainless steel markers percutaneously into vertebral segments above and below the benign spinal tumor that is to be treated [11]. Placement of these ﬁducials increases the targeting accuracy above that attainable by using bony landmarks of the spine alone [11]. The ﬁducial markers typically used are 2 × 6-mm stainless steel self-tapping screws placed in a non-coplanar manner on the lamina and/or facets around a spinal target. Under ﬂuoroscopic guidance, three to four ﬁducials are inserted into the posterior spinal elements through stab incisions in the skin so as to be at least 25 mm apart. By virtue of their ﬁxed relationship to the bony spine, any movement in the vertebrae is detected as movement of the ﬁducials. Such movement is automatically detected and compensated for by the CyberKnife system.

Treatment Planning During the next step, each patient was ﬁtted with an individualized, simple, nonrigid immobilization device. In general, patients with tumors of the cervical spine were immobilized with an Aquaplast face mask (WFR/Aquaplast Corp., Wyckoff, NJ) attached to a radiographically transparent timo headrest as previously described [43]. Patients with thoracic, lumbar, or sacral lesions were immobilized in a custom-ﬁtted Alpha cradle mold fabricated with the patient in a supine position (Smithers

During radiosurgery, the patient lies on the operating table in the Alpha cradle mold. Once positioned, sequential paired digital X-rays of the target region are then obtained by ceilingmounted, orthogonally directed, rigidly ﬁxed X-ray tubes. A computer workstation performs rapid image-to-image correlation between the acquired images and the previously calculated DRRs. The robotic arm moves the linac sequentially through the planned beam positions delivering the calculated dose to the target. Before each beamlet of radiation is administered, the X-ray imaging system determines target location and communicates the answer to the robot. At each position, the imaging system reestablishes the location of the target and sends corrective pointing directions to the robotic arm. The robot then adjusts for small patient movements by automatically realigning the treatment beam with submillimeter accuracy [22].

Radiosurgical Treatment Parameters The patients in the current series were treated with a range of doses and hypofractionated regimens. Initial radiosurgical treatment parameters were loosely based on our own and published experience with intracranial ablation. It is important to emphasize that as we embarked on the current experience, an underlying dread of causing potentially catastrophic spinal cord injury was also an important inﬂuence that resulted in radiosurgical parameters erring on the side of caution. Over the past several years as we have gained experience and conﬁdence in the overall safety of spinal radiosurgery, dose has been gradually escalated and the number of fractions decreased. Consequently, the dose-fraction regimens described in this report should not be interpreted as being optimal for the current

43.

spinal tumors

435

FIGURE 43-2. Treatment plan example in a patient with metastatic non–small cell lung cancer to T5 treated with 24 Gy to the 81 isodose line in three stages. (A) Axial, (B) sagittal, and (C) coronal postcontrast CT images used in treatment planning are demonstrated. Red lines (with or without solid squares) demarcate the outline of the lesion, solid

green lines demarcate the 81% isodose curve, and the purple lines demarcate the 50% isodose curve. The green line with the squares demarcates the spinal cord identiﬁed as a critical structure. (D) Threedimensional reconstruction of the actual beam paths used in the treatment.

FIGURE 43-3. Treatment plan example in a patient with T10 schwannoma treated with 22 Gy to the 81 isodose line in two stages. (A) Pretreatment axial T1 postcontrast MRI. (B) Axial, (C) sagittal, and (D) coronal postcontrast CT images used in treatment planning are demonstrated. Red lines (with or without solid squares) demarcate the

outline of the lesion, solid green lines demarcate the 88% isodose curve, and the purple lines demarcate the 50% isodose curve. The green line with the squares demarcates the spinal cord identiﬁed as a critical structure. (E) Three-dimensional reconstruction of the actual beam paths used in the treatment.

436

r.l. dodd et al.

application but merely useful starting points for future investigation.

Results Benign Tumors Since 1999, 51 patients (28 men, 23 women; median age, 46 years; range, 12 to 86 years) with 55 intradural extramedullary benign spinal tumors have been treated with staged radiosurgery using the CyberKnife treatment system. The patient characteristics are detailed in Table 43-3. All tumors were known or presumed to be spinal meningioma, neuroﬁbroma, or schwannoma based on existing pathology, characteristic radiographic appearance on MRI, and/or history of neuroﬁbromatosis. As expected, a female predominance was observed among spinal meningiomas, whereas the male to female ratio in schwannomas and neuroﬁbromas was 1.6 : 1 and 2 : 1, respectively. Twenty-six (51%) patients had previous surgery and were being treated for residual or recurrent tumor. Four patients developed tumors in radiation ﬁelds for other cancers (e.g., Hodgkin lymphoma, breast adenocarcinoma). One patient developed a traumatic schwannoma after surgery for removal of a synovial cyst. Seventeen patients carried clinical diagnosis of either neuroﬁbromatosis type 1 (NF1) or NF2. Tumors were observed throughout the entire spinal axis (Table 43-4) and varied in conﬁguration from entirely intraspinal to dumbbell shaped to predominately foraminal. Presenting symptoms (pain, radiculopathy, and myelopathy) varied depending on spinal location and precise relationship of the tumor to the adjoining nerves/spinal cord (Table 43-5). Few patients had either moderate or severe cord compression prior to radiosurgery; myelopathy upon presentation largely represented a neurologic deﬁcit stemming from previous spinal surgery. Eight asymptomatic patients (15%) underwent preemptive radiosurgical ablation because of the size, location, and/or growth of their tumor on serial MRI.

Radiosurgical Doses and Fractionation Table 43-6 summarizes the radiosurgical dosimetry used for benign tumors in this series. Target volumes ranged from 0.136 to 24.6 cm3 (mean, 4.29 cm3). Treatment plans were designed to

TABLE 43-3. Characteristics of patients with benign intradural extraaxial tumors. Age (years) Mean Range Sex, no. (%) Female Male Previously resected, no. (%) Subtotal Gross total Previously radiated, no. (%) NF1, no. (%) NF2, no. (%)

46.5 12.6–86.5

TABLE 43-4. Characteristics of benign intradural extraaxial tumors. Characteristic

Level Cervical Thoracic Lumbar Sacral Histology Meningioma Schwannoma Neuroﬁbroma

24 (47%) 2 (4%) 4 (8%) 7 (14%) 10 (20%)

38 (69%) 7 (13%) 8 (15%) 2 (4%) 29 (53%) 17 (31%) 9 (16%)

deliver 1600 to 3000 cGy to an average 80%ile isodose line as deﬁned at the margin of the treated lesion. The maximum intratumoral dose ranged from 1975 to 3435 cGy (mean, 2506 cGy). Similar amounts of radiation were delivered to all three tumor histologies, as the α/β was believed similar for these slowgrowing benign lesions. The major limitation to maximizing the tumor dose was the desire to limit the risk of injury to the adjacent spinal cord. Somewhat arbitrarily, but drawing from established optic nerve tolerance to radiosurgery, an attempt was made to construct treatment plans that limited irradiation of the spinal cord to a maximum of 800 cGy per fraction. The speciﬁc fractionation schedule (median, 2; range, 1 to 5) was based on the size and volume of the treated tumor, as well as the length and total dose administered to the spinal cord. In most patients, radiosurgery was delivered in one (37%) or two (42%) stages. However, additional daily stages were administered in three (8 cases), four (2 cases), or ﬁve (1 case) sessions. Acute toxicity was rare and limited to short-lived nausea.

Response of Symptoms to Spinal Radiosurgery Symptomatic patients typically complained of axial pain, signs associated with nerve root compression, and/or early myelopathy. In these cases, neurologic function secondary to spinal cord compression was deemed not severe enough to warrant immediate decompression. Patients with poor functional reserve or signs of frank spinal cord compression were surgically decompressed. After radiosurgery, most clinical symptoms either remained stable or improved. Unfortunately, the small sample size for various symptoms precluded any signiﬁcant ﬁndings, except for pain. Across all tumor histologies, pain (both local and radicular) was the most common clinical complaint. The treated spinal tumors were painful in 78%, 66%, and 53% of patients with neuroﬁbromas, schwannomas, and meningioma, respectively. Approximately 70% and 50% of patients with TABLE 43-5. Presenting symptoms in patients with benign intradural extraaxial tumors. Symptom

23 (45%) 28 (55%)

No. (%)

Local or radicular pain Radicular sensory loss Radicular weakness Myelopathic weakness Axial sensory loss Bladder paresis Asymptomatic

No. (%)

34 (63%) 26 (48%) 22 (41%) 12 (22%) 4 (7%) 3 (6%) 8 (15%)

43.

437

spinal tumors

TABLE 43-6. CyberKnife treatment dosimetry characteristics for benign intradural extraaxial tumors.

Average tumor volume, TV (cm3) Average prescribed dose, TD (cGy) Average dose/fraction, Df (cGy) Average maximum tumor dose, Dmax (cGy)

Meningioma

Neuroﬁbroma

Schwannoma

All

2.55 (0.14–7.57) 2025 (1600–3000) 1184 (500–1800) 2613 (1975–3435)

4.31 (0.62–14.5) 1978 (1800–3000) 1075 (700–2000) 2646 (2168–2985)

5.76 (0.68–24.6) 1917 (1800–2100) 1275 (500–2000) 2402 (2022–3000)

4.29 1933 1243 2506

spinal meningioma or schwannoma, respectively, reported signiﬁcant reduction in pain 12 months after CyberKnife SRS. None of the patients with seven neuroﬁbromas presenting with pain reported improvement after CyberKnife treatment. The time course of pain amelioration varied from a few weeks to months, though one patient reported complete pain abatement after 6 days. Some patients reported a temporary increase in radicular symptoms 2 to 3 weeks after radiosurgery, which improved with corticosteroids and later resolved without additional analgesia. This is hypothesized to be secondary to radiation-induced tumor swelling, which coincidentally may correspond with the slight increase in tumor size occasionally observed at 6 to 8 months. The next most frequent symptoms in this series were radicular weakness and sensory loss, which were present in one third and two thirds of patients, respectively. Trends toward modest improvements in these signs were also observed throughout post-radiosurgical follow-up.

Tumor Growth Control Eighteen of the 55 lesions had greater than 24 months follow-up (six >4 years, two >3 years, and ten >2 years). The mean followup interval in this group was 38.7 months (median, 30.8 months; range, 24 to 66 months). All lesions in this group were either stable (64%) or smaller (36%). No tumor in this group increased in size. Of the patients with greater than 6 months follow-up, 14 of the 17 meningioma patients had radiographic imaging. The size of the treated lesions was either stable (57%) or decreased (43%) at last follow-up. Importantly, no tumor increased in size. Twenty-four of the 29 patients with spinal schwannomas had radiographic follow-up, which demonstrated that the tumor was either stable (70%) or reduced (25%) in size. Among those patients with NF1 who had spinal neuroﬁbromas, six of the seven underwent postoperative MRI, which demonstrated all tumors to be stable in size at last follow-up. Among the entire population of patients treated in this series, there were four lesions that enlarged slightly by less than 10%, three at the 6-month imaging evaluation and a fourth at the 12-month study. Two of these original lesions later regressed on MRI, and in doing so demonstrated typical patterns of central necrosis when contrast was administered. The other two patients underwent surgical removal of their tumors with the primary goal of decompressing the spinal cord and reversing myelopathy.

Neurologic Deterioration After Radiosurgery Three patients (one meningioma, one schwannoma, and one neuroﬁbroma) underwent surgical resection of a CyberKnifeablated lesion at 8, 12, and 13 months, respectively, because of persistent or worsening symptoms. Two of those three tumors

were slightly larger radiographically; the other was unchanged in size. The surgeons removing the tumors reported nothing unusual about these resections. Another treatment-related complication included a 29-year-old woman with a CyberKnifetreated, recurrent C7/T1 spinal meningioma, who developed the new onset of posterior column dysfunction 8 months after radiosurgery. The treatment dose in this case of presumed radiation myelopathy was 2400 cGy in three stages to a tumor volume of 7.56 cm3 and a maximum tumor dose of 3435 cGy. Posttreatment neurologic deterioration also occurred in a 38-year-old patient with NF1 and multiple large cervical neuroﬁbromas who underwent CyberKnife SRS to a large C2 neuroﬁbroma secondary to neck and scapula pain. He had minimal response both clinically and radiographically to the treatment at 6 and 12 months. Fifteen months after radiosurgery, this patient suffered a fall that resulted in a complete C5 quadriparesis. Two patients died from causes unrelated to their tumors. One was an 82-year-old woman who died from respiratory failure complicating a long history of chronic pulmonary obstructive disease 7 months after radiosurgery for a C2 meningioma. A second patient with severe NF2 died 6 months after treatment of a C5 schwannoma. This 49-year-old woman suffered from bilateral vocal cord paralysis and chronic aspiration prior to radiosurgery, a condition that ﬁnally led to her death. Histopathologic examination of this patient’s postmortem tumor revealed central necrosis.

Malignant Tumors Since 1996, 31 patients (20 women, 11 men; mean age, 59 years; range, 28 to 76 years) with 33 spinal metastases were treated using CyberKnife image-guided radiosurgery at Stanford. The patient characteristics are detailed in Table 43-7. All patients TABLE 43-7. Characteristics of patients with malignant/metastatic tumors. Age (years) Mean Range Sex, no. (%) Female Male Karnofsky (KPS) Median Range Previous treatment, no. (%) XRT Chemotherapy Surgery None

59 28–76 20 (65%) 11 (35%) 80 40–90 22 17 1 3

(71%) (55%) (3%) (10%)

438

r.l. dodd et al.

TABLE 43-8. Characteristics of malignant tumors. Characteristic

Level Cervical Thoracic Lumbar Sacral Histology Schwannoma Neuroﬁbroma Meningioma Histology Breast Renal Non–small cell lung Gastrointestinal Head and neck Other

No. (%)

6 20 4 1

(10%) (65%) (13%) (3%)

17 (31%) 9 (16%) 29 (53%) 10 4 4 4 4 6

(32%) (13%) (13%) (13%) (13%) (19%)

had an established primary histologic diagnosis of malignancy, and the median Karnofsky performance score (KPS) was 80 (range, 40 to 90). Twenty patients had lesions involving the thoracic spine, and only a single patient had a sacral lesion (Table 43-8). There were 26 vertebral column lesions, 5 intramedullary lesions, and 1 extramedullary lesion. Breast cancer, lung cancer, and renal cell carcinoma were the most common primary site histologies accounting for 59% of all treated lesions. Four lesions had a gastrointestinal primary histology. Twenty-six patients presented with symptoms including pain and neurologic dysfunction; 5 patients were asymptomatic (Table 43-9). Twenty-three of the 26 symptomatic patients had pain as a component of their symptoms. Twenty-eight patients received some form of treatment prior to CyberKnife therapy. Twenty-two of these patients received radiotherapy as a component of the prior therapy to a radiation ﬁeld adjacent to or encompassing the CyberKnife treatment site.

Radiosurgical Doses and Fractionation The median tumor volume for all patients was 18.3 cm3 with one outlier having a volume of 166.59 cm3 (Table 43-10). A single patient received a dose and fraction schedule outside of the intended range. This lesion received 14 Gy in four stages. The remaining 31 lesions received 15 to 25 Gy in one to three stages with a dose per fraction ranging from 7 to 20 Gy. The range of prescription isodose was 61% to 87% indicating mean dose homogeneity of 77% as normalized to the maximum dose. TABLE 43-9. Presenting symptoms in patients with malignant/metastatic tumors. Symptom

Local or radicular pain Sensory loss Weakness Bladder paresis Asymptomatic

No. (%)

23 18 6 2 5

(74%) (58%) (19%) (6%) (16%)

TABLE 43-10. CyberKnife treatment dosimetry characteristics for malignant/metastatic tumors. Average tumor volume, TV (cm3) Average prescribed dose, TD (cGy) Average dose/fraction, Df (cGy) Average maximum tumor dose, Dmax (cGy)

23.4 (0.10–166.6) 1890 (1100–2500) 1200 (350–2000) 2470 (1500–3690)

Survival The mean follow-up time was 11 months. One patient had no additional follow-up after treatment. At the time of last followup, 19 patients were alive and 12 were dead. One patient died of sepsis within 1 month after the CyberKnife treatment, during a leukopenic nadir period caused by chemotherapy. The other 11 deaths were attributed to metastatic disease progression. One-year actuarial survival was 68.2% for the entire group. Patients with breast cancer had the longest survival with a mean survival of 12 months.

Response of Symptoms to Spinal Radiosurgery Fourteen of the 26 (54%) patients who presented with symptoms reported improvement of symptoms after treatment. The clinical disease status as of the last follow-up or at the time of death is shown in Table 43-11. Over half of the patients experienced further systemic disease progression. Only four patients developed signs of local disease progression referable to the treated lesion.

Complications After CyberKnife treatment, four patients developed clinical signs of myelopathy (Table 43-12). One of these patients was the most severely ill of the entire cohort who had a KPS of 40 and preexisting myelopathic symptoms that continued to progress despite the treatment. This patient died within 1 month of treatment of systemic disease spread and was quadriplegic at the time of death. Three patients developed treatmentrelated chronic myelopathy; one patient was initially asymptomatic. The mean time to onset of signs and symptoms was 7 months (range, 6 to 10 months). In these three patients, classic radiographic signs of spinal cord injury were apparent at the onset of clinical neurologic deterioration and evolved from spinal cord edema at the time of symptom onset to contrast-enhancement within the cord at the level of the treated lesion. Characteristically, the edema resolved over the ensuing 3 to 6 months, even though the contrast enhancement persisted. All three patients are alive with a marked diminishment of mobility. A variety of patient characteristics and treatment parameters including age, gender, primary site, anatomic location, anatomic level, previous treatment, total dose, dose per fraction, maximum dose, maximum spinal cord dose, and tumor volume were analyzed by logistic regression. Logistic regression analysis failed to identify factors that predicted for complications despite the fact that all three patients were female with lesions of the thoracic spine.

58 59 76 61

55

71 59

28

68 69 40 62

20 21 22 23

24

25 26

27

28 29 30 31

F F M M

M

M F

F

F M F M

F F F F F F F F F F F M M M F M F M F

Gender

Thoracic Thoracic Thoracic Cervical

Cervical

Thoracic Thoracic

Thoracic

Lumbar Lumbar Thoracolumbar Thoracic

Lumbar Cervical Cervical Cervical Thoracic Thoracic Thoracic Thoracic Lumbar Cervical Thoracic Thoracic Thoracic Thoracic Thoracic Thoracic Sacral Thoracic Thoracic

Level

Vertebral Vertebral Vertebral Intramedullary

Intramedullary

Vertebral Extramedullary

Vertebral

Vertebral Vertebral Vertebral Vertebral

Vertebral Vertebral Intramedullary Intramedullary Vertebral Vertebral Vertebral Vertebral Vertebral Intramedullary Vertebral Vertebral Vertebral Vertebral Vertebral Vertebral Vertebral Vertebral Vertebral

Anatomic location

Uterine Uterine Ewing sarcoma Melanoma

HN/neuroendocrine

HN/acinic cell HN/acinic cell

HN/adenoid cystic

Cholangio Esophagus/adenoCa Colon/adenoCa Colon/adenoCa

Breast Breast Breast Breast Breast Breast Breast Breast Breast Breast NSCLC NSCLC NSCLC NSCLC Renal cell Renal cell Renal cell Renal cell Melanoma

Primary site/histology

60 70 60 80

90

80 90

90

70 90 80 80

80 80 40 50 90 70 80 80 80 50 90 70 90 90 90 60 70 90 60

KPS

Pain Pain/sphincter dysf Pain/motor/sensory Motor/sensory

Motor

Pain None

None

Pain/motor/sensory None Pain Pain

Pain Pain/sensory Pain/motor/sensory Motor/sensory Pain Pain/sphincter dysf Pain/sensory Pain Pain Pain/motor/sensory Pain Pain Pain None None Pain Pain Pain Pain

Presenting symptoms

Chemo Surg Chemo/RT None

Chemo/RT

RT Chemo/RT

RT

Chemo/RT Chemo/RT Chemo/RT Chemo/RT

RT Chemo None Chemo/RT Chemo/RT RT Chemo/RT Chemo/RT Chemo/RT Chemo Chemo/RT RT RT RT None RT RT Surg Chemo

Previous treatment

Local/systemic progression Unknown Systemic progression Stable NED

Systemic progression Systemic progression Systemic progression Systemic progression Stable Stable Stable Systemic progression Stable Stable Systemic progression Systemic progression Systemic progression Stable Systemic progression Systemic progression Stable Stable Local/systemic progression LTF Stable Stable Local/systemic progression Local/systemic progression Systemic progression Systemic progression

Current status

Dead Dead Dead Alive

Alive

Alive Alive

Alive

Alive Dead Alive Alive

Dead Dead Dead Dead Alive Alive Alive Alive Alive Alive Dead Alive Dead Alive Alive Dead Alive Alive Dead

Mortality status

1 24 1 14

13

11 3

23

LTF 5 4 5

22 22 1 1 24 15 11 12 9 6 3 10 1 11 12 7 9 8 8

Length of follow-up (months)

spinal tumors

KPS, Karnofsky Performance Scale; F, Female; M, Male; RT, radiotherapy; Chemo, chemotherapy; NSCLC, non–small cell lung carcinoma; GI, renal cell carcinoma; Cholangio, cholangiocarcinoma; adenoCa, adenocarcinoma; HN, head and neck; NED, no evidence of disease; dysf, dysfunction; LTF, lost to follow-up.

73 38 29 66 59 74 56 63 48 33 72 73 74 55 59 65 69 53 55

Age (years)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Patient ID

TABLE 43-11. Characteristics of 31 patients treated with CyberKnife for metastatic spinal lesions.

43.

439

6.37 18.90 69.74 10.51 19.50 14.61 1.94 39.51 34.51 45.45 19.43 13.72 18.32 5.25 166.59 22.40 0.88

5.73 3.78 50.18 32.58 11.38

23.70 3.20 7.63

27.28 6.95 35.73 0.10

4 5 6 7 8 9 10 11 12 13 14 15 16 16 17 18 19

20 21 22 23 24

25 26 27

28 29 30 31

20 12 21 20

21 20 20

18 15 18 18 18

14 21 16 20 20 20 22 18 16 18 20 25 16 16 18 24 11

16 16 19

Total dose (Gy)

2 1 3 1

3 1 2

1 2 2 2 1

4 2 2 2 2 2 2 2 1 1 2 2 1 1 2 2 1

2 1 2

Number of fractions

28.6 15.0 25.0 25.7

27.3 28.2 26.3

22.5 20.0 23.6 24.3 22.8

17.0 30.0 19.0 25.0 24.1 24.0 26.0 23.0 21.6 22.2 26.0 31.3 26.2 18.4 23.7 36.9 17.0

20.0 19.0 27.0

Maximum dose (Gy)

70 80 83 78

76 71 76

80 75 76 74 79

83 69 83 80 83 83 84 79 74 81 77 80 61 87 76 65 65

80 83 70

Prescription isodose (%)

1.65 NA 1.59 1.19

1.50 1.67 1.61

1.48 NA 1.14 1.42 1.10

NA 1.14 1.41 1.31 1.49 1.45 1.11 1.35 1.39 1.36 1.11 1.26 1.36 1.02 1.38 1.46 NA

NA NA NA

Modiﬁed conformity index

17.3 7.4 22.1 25.7

22.2 5.0 20.0

12.1 5.0 19.6 18.2 11.2

16.9 13.9 16.7 19.2 21.4 19.7 23.4 19.7 14.4 11.4 15.3 26.2 12.2 15.2 18.5 23.7 7.1

10.0 10.4 21.2

Maximum total spinal cord dose

NA, not available; Ipr→PD, initial partial response followed by progressive disease; PD, progressive disease; LTF, lost to follow-up.

47.89 5.20 0.40

1 2 3

Patient ID

Tumor volume (cm3)

8.6 7.4 7.4 25.7

7.4 5.0 10.0

12.1 2.5 9.8 9.1 11.2

4.2 7.0 8.3 9.6 10.7 9.9 11.7 9.8 14.4 11.4 7.6 13.1 12.2 15.2 9.3 11.9 7.1

5.0 10.4 10.6

Maximum spinal cord dose per fraction

TABLE 43-12. Description of CyberKnife dosimetry, imaging response, and complications in 31 patients with 32 spinal metastases.

IPR→PD NA Stable NA Stable

iPR→PD Stable Stable

iPR→PD NA Stable Stable PD

NA Stable NA Stable Stable Stable PR NA Stable NA Stable Stable Stable Stable Stable Stable

Stable Stable NA

Imaging response

None None None None

None None None

None None None None None

None None Progressive preexisting myelopathy None Severe myelopathy None Severe myelopathy None None None None None None None Severe myelopathy None None None None None

Complication

1 24 1 14

11 3 13

LTF 5 4 5 23

1 24 15 11 12 9 6 3 10 1 11 12 7 — 9 8 8

22 22 1

Length of follow-up (months)

440 r.l. dodd et al.

43.

spinal tumors

Conclusion At the start of this study, we harbored grave concerns that radiosurgical ablative doses could prove injurious to the adjacent spinal cord. Selection of doses and fractions were based on our own and published experience with intracranial ablation, which in hindsight may have been more conservative than needed. As we have gained experience and conﬁdence in the overall safety of spinal radiosurgery, doses have been gradually escalated and the number of fractions decreased. Thus, the dose-fraction regimens described here should not be viewed as optimal but merely as useful starting points for future investigation. Nevertheless, this study demonstrates both the relative safety and early evidence of efﬁcacy for spinal radiosurgery and provides a rough framework for treating patients going forward. It is not possible to say anything deﬁnitive about long-term efﬁcacy given the average length of follow-up to date, and this will remain the subject of ongoing study at our institution. Nonetheless, long-term studies in patients with benign brain tumors who were treated with radiosurgery suggests that tumor control at 3 years is very likely to be durable [44]. The current study does suggest that CyberKnife SRS could someday serve as a useful adjunct to the neurosurgical armamentarium for managing selected benign spinal tumors.

References 1. Sundaresan N, Boriani S, Rothman A, Holtzman R. Tumors of the osseous spine. J Neurooncol 2004; 69:273–290. 2. Parsa AT, Lee J, Parney IF, et al. Spinal cord and intraduralextraparenchymal spinal tumors: current best care practices and strategies. J Neurooncol 2004; 69:291–318. 3. Ozaki T, Flege S, Liljenqvist U, et al. Osteosarcoma of the spine: experience of the Cooperative Osteosarcoma Study Group. Cancer 2002; 94:1069–1077. 4. Bilsky MH, Lis E, Raizer J, et al. The diagnosis and treatment of metastatic spinal tumor. Oncologist 1999; 4:459–469. 5. Bowers DC, Weprin BE. Intramedullary spinal cord tumors. Curr Treat Options Neurol 2003; 5:207–212. 6. Kopelson G, Linggood RM, Kleinman GM, et al. Management of intramedullary spinal cord tumors. Radiology 1980; 135:473– 479. 7. McCunniff AJ, Liang MJ. Radiation tolerance of the cervical spinal cord. Int J Radiat Oncol Biol Phys 1989; 16:675–678. 8. Marcus RB Jr, Million RR. The incidence of myelitis after irradiation of the cervical spinal cord. Int J Radiat Oncol Biol Phys 1990; 19:3–8. 9. Schultheiss TE, Kun LE, Ang KK, Stephens LC. Radiation response of the central nervous system. Int J Radiat Oncol Biol Phys 1995; 31:1093–1112. 10. Ryu S, Fang Yin F, Rock J, et al. Image-guided and intensitymodulated radiosurgery for patients with spinal metastasis. Cancer 2003; 97:2013–2018. 11. Ryu SI, Chang SD, Kim DH, et al. Image-guided hypofractionated stereotactic radiosurgery to spinal lesions. Neurosurgery 2001; 49:838–846. 12. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316–319. 13. Woroschiloff C. Der Verlauf der motorishen und senililen Bahnen durch das Lendenmark des Kaninchens. Ber Vehr Sach Ges Wiss Leipzig 1874; 26:248–304. 14. Rand RW, Bauer RO, Smart CR, Jannetta PJ. Experiences with percutaneous stereotaxic cryocordotomy. Bull Los Angeles Neurol Soc 1965; 30:142–147.

441

15. Hitchcock E. An apparatus for stereotactic spinal surgery. Lancet 1969; 1:705–706. 16. Nadvornik P. Woroschiloff’s locating device for interventions on the spinal cord and its inﬂuence on spinal stereotaxis. Appl Neurophysiol 1985; 48:247–251. 17. Hamilton AJ, Lulu BA, Fosmire H, et al. Preliminary clinical experience with linear accelerator-based spinal stereotactic radiosurgery. Neurosurgery 1995; 36:311–319. 18. Hamilton AJ, Lulu BA. A prototype device for linear acceleratorbased extracranial radiosurgery. Acta Neurochir Suppl 1995; 63: 40–43. 19. Hamilton AJ, Lulu BA, Fosmire H, Gossett L. LINAC-based spinal stereotactic radiosurgery. Stereotact Funct Neurosurg 1996; 66:1–9. 20. Murphy MJ. An automatic six-degree-of-freedom image registration algorithm for image-guided frameless stereotaxic radiosurgery. Med Phys 1997; 24:857–866. 21. Adler JR Jr, Chang SD, Murphy MJ, et al. The CyberKnife: a frameless robotic system for radiosurgery. Stereotact Funct Neurosurg 1997; 69:124–128. 22. Chang SD, Main W, Martin DP, et al. An analysis of the accuracy of the CyberKnife: a robotic frameless stereotactic radiosurgical system. Neurosurgery 2003; 52:140–146; discussion 146–147. 23. Chang SD, Meisel JA, Hancock SL, et al. Treatment of hemangioblastomas in von Hippel-Lindau disease with linear accelerator-based radiosurgery. Neurosurgery 1998; 43:28–34; discussion 34–35. 24. Murphy MJ, Martin D, Whyte R, et al. The effectiveness of breath-holding to stabilize lung and pancreas tumors during radiosurgery. Int J Radiat Oncol Biol Phys 2002; 53:475–482. 25. Murphy MJ, Adler JR Jr, Bodduluri M, et al. Image-guided radiosurgery for the spine and pancreas. Comput Aided Surg 2000; 5:278–288. 26. Gerszten PC, Ozhasoglu C, Burton SA, et al. Evaluation of CyberKnife frameless real-time image-guided stereotactic radiosurgery for spinal lesions. Stereotact Funct Neurosurg 2003; 81:84–89. 27. Chang SD, Le Q, Martin DP, Adler JR. The CyberKnife. In: Dickman CA, ed. Spinal Cord and Spinal Column Tumors: Principles and Practices. New York: Thieme, 2005. 28. Gerszten PC, Welch WC. CyberKnife radiosurgery for metastatic spine tumors. Neurosurg Clin N Am 2004; 15:491–501. 29. Rock JP, Ryu S, Yin FF. Novalis radiosurgery for metastatic spine tumors. Neurosurg Clin N Am 2004; 15:503–509. 30. De Salles AA, Pedroso AG, Medin P, et al. Spinal lesions treated with Novalis shaped beam intensity-modulated radiosurgery and stereotactic radiotherapy. J Neurosurg 2004; 101(Suppl 3):435– 440. 31. Schick U, Marquardt G, Lorenz R. Recurrence of benign spinal neoplasms. Neurosurg Rev 2001; 24:20–25. 32. Cohen-Gadol AA, Zikel OM, Koch CA, et al. Spinal meningiomas in patients younger than 50 years of age: a 21-year experience. J Neurosurg Spine 2003; 98:258–263. 33. Gottfried ON, Gluf W, Quinones-Hinojosa A, et al. Spinal meningiomas: surgical management and outcome. Neurosurg Focus 2003; 14:e2. 34. 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; 61:34–43; discussion 44. 35. McCormick PC. Surgical management of dumbbell and paraspinal tumors of the thoracic and lumbar spine. Neurosurgery 1996; 38:67–74; discussion 74–75. 36. McCormick PC. Surgical management of dumbbell tumors of the cervical spine. Neurosurgery 1996; 38:294–300. 37. Gambardella G, Gervasio O, Zaccone C. [Approaches and surgical results in the treatment of ventral thoracic meningiomas.

442

r.l. dodd et al.

Review of our experience with a postero-lateral combined transpedicular-transarticular approach.] Acta Neurochir (Wien) 2003; 145:385–392; discussion 392. 38. Asazuma T, Toyama Y, Maruiwa H, et al. Surgical strategy for cervical dumbbell tumors based on a three-dimensional classiﬁcation. Spine 2004; 29:E10–14. 39. Gezen F, Kahraman S, Canakci Z, Beduk A. Review of 36 cases of spinal cord meningioma. Spine 2000; 25:727–731. 40. Roux FX, Nataf F, Pinaudeau M, et al. Intraspinal meningiomas: review of 54 cases with discussion of poor prognosis factors and modern therapeutic management. Surg Neurol 1996; 46:458–463; discussion 463–464.

41. Gerszten PC, Ozhasoglu C, Burton SA, et al. CyberKnife frameless single-fraction stereotactic radiosurgery for benign tumors of the spine. Neurosurg Focus 2003; 14:e16. 42. Murphy MJ, Cox RS. The accuracy of dose localization for an image-guided frameless radiosurgery system. Med Phys 1996; 23:2043–2049. 43. Murphy MJ, Chang SD, Gibbs IC, et al. Patterns of patient movement during frameless image-guided radiosurgery. Int J Radiat Oncol Biol Phys 2003; 55:1400–1408. 44. Kondziolka D, Nathoo N, Flickinger JC, et al. Long-term results after radiosurgery for benign intracranial tumors. Neurosurgery 2003; 53:815–821; discussion 821–822.

4 4

Spine Tumors: Surgery Perspective Gabriel Zada and Michael Y. Wang

Introduction Tumors of the spine comprise a heterogeneous group of neoplasms that can be characterized by their pathology, anatomic location, and degree of invasiveness. The majority of spine tumors encountered by clinicians are metastatic, accounting for approximately 70% of all spine tumors [1]. Less frequently encountered neoplasms of the spine and spinal cord include primary lesions such as meningiomas, schwannomas, osseous tumors of the spine, ependymomas, and gliomas. At the time of death, approximately 70% to 90% of all terminal cancer patients have evidence of metastatic disease on postmortem examination [2–4], with the spinal column being the most common site for bony metastases [5]. Each year, an estimated 18,000 patients will be diagnosed with metastatic spine disease [6, 7]. Metastatic lesions most frequently originate from primary tumors in the breast, lung, and prostate [5, 8–10]. Frequent clinical presentations include back pain and symptoms from spinal cord/nerve root compression, resulting in pain and disability. Furthermore, it is likely that the observed incidence of metastatic spine disease will continue to increase over the next several decades, as diagnostic capabilities improve [11, 12]. The treatment of spinal neoplasms requires a team approach including neurosurgeons, medical oncologists, radiation oncologists, and primary care physicians. In the current era, a number of treatment modalities are available for treating spinal tumors, including medical therapy, chemotherapy, surgery, radiation therapy, and radiosurgery. Quite frequently, combinations of these therapies are utilized to maximize clinical beneﬁts and outcomes. Formulating an ideal treatment plan for a patient with spinal metastases is a complicated process and requires an interdisciplinary team of practitioners that must consider each patient’s individual needs and desires. Over the past two decades, radiosurgical modalities have provided minimally invasive and efﬁcacious alternatives to surgery, thereby potentially avoiding the associated morbidity and mortality risks. The role of these newer therapies is extremely promising, but the speciﬁc indications for their use continue to evolve. Despite the beneﬁt of minimal invasiveness inherent to radiation therapy or radiosurgery, there are numerous situations in which surgery remains the optimal treatment for patients with spinal tumors [13–17]. Although surgery has

not been widely accepted as a primary treatment for metastatic spine disease in the past, a shift in this ideology is currently under way, with many physicians advocating more aggressive surgical resection of metastatic spine lesions [7, 14, 18]. Recent series have demonstrated that surgery can increase the overall quality of life as well as prolong survival in particular patients with metastatic spine disease [8, 13, 15, 18, 19]. Careful patient selection is the key to maximizing the potential beneﬁts of surgical intervention. This chapter will describe the current role of surgery in the treatment of spinal tumors and the speciﬁc factors that make surgery the preferred treatment in a given situation. Common surgical procedures utilized by neurosurgeons for metastatic spine disease will be reviewed, focusing on the indications and patient selection considerations for each.

Tumor Location Tumors of the spine are frequently characterized according to their location with respect to the spinal canal, meninges, and spinal cord. The overwhelming majority (95%) of spine metastases are extradural, and the remaining are intraduralextramedullary (4%) or intramedullary (0.5%) [20]. Furthermore, metastases are most commonly found in the posterior half of the vertebral body, usually at the junction of the vertebral body and pedicles [21, 22]. They are infrequently conﬁned to the posterior spinal elements (i.e., lamina and spinous process). The thoracic spine, with a narrower canal diameter relative to the spinal cord [23], is the most common spinal region to present with neural compression secondary to spinal metastases (70% of patients), followed by the lumbar spine (20%) and cervical spine (10%) [9, 13, 24]. The thoracolumbar junction and the T4 vertebral body, in particular, are the most common spinal levels to present with metastases [9, 24]. Spinal metastases are located in multiple, noncontiguous levels in 10% to 38% of cases at the time of presentation [25–27]. Multiple surgical staging systems have been devised in the past to assist in surgical planning based on anatomic criteria. The surgical considerations based on spinal level and relationship to the spinal cord will be discussed later in this chapter.

443

444

g. zada and m.y. wang

Deciding on a Treatment Modality Following a diagnosis of metastatic spine disease, a decision of surgery, radiation, or medical treatment must be made. The key medical therapies used to optimize a patient’s condition and manage pain prior to surgery include steroids, bisphosphonates, and opioids. As with any type of surgery, the age, performance status, and comorbidities of the patient must be taken into account when considering operative management for metastatic spine disease. Comorbidities such as coagulopathy, anemia, malnutrition, liver disease, and renal disease are frequently encountered in cancer patients and may complicate the surgery or postoperative course. The patient’s expected survival time is a crucial factor to consider in cancer patients, as many may be terminal patients in which surgery may not provide an overall quality of life beneﬁt. Additionally, the potential requirement of adjuvant therapies such as chemotherapy or radiation, and the order these should be done in, will guide the choice of treatment. The primary tumor type should be factored into this decision, as some primary tumors (such as renal and thyroid cancer) may require preoperative embolization [11]. Whereas radiation therapy was considered the mainstay of treatment for metastatic spinal disease for several years [28, 29], recent series have demonstrated superior outcomes with surgical decompression [30, 31]. Perhaps the most convincing evidence comes from a recent study by Patchell et al. This prospective, multi-institutional study randomized patients with metastatic spinal cord compression to either surgical decompression plus radiotherapy or radiotherapy alone. In the surgical decompression group, 42 of 50 (84%) patients were able to walk after surgery, compared with 29 of 51 (57%) patients in the radiotherapy group (p = 0.001). The number of nonambulatory patients that speciﬁcally regained the ability to walk after surgery was also signiﬁcantly higher in the surgery group (62% vs. 19%, p = 0.01). Finally, patients treated with surgery retained the ability to walk for much longer than did patients in the radiotherapy group (mean 122 vs. 13 days, p = 0.003). The study was terminated prematurely secondary to a beneﬁcial outcome observed in the surgical group after an interim analysis [32]. Furthermore, a recent meta-analysis by Klimo and colleagues, reviewing 24 surgical series (999 patients) and 4 large irradiation series (543 patients), demonstrated ambulatory success rates in 85% of surgical patients compared with 64% of radiation patients. Surgical patients in this study were 1.3 times as likely to remain ambulatory after surgery and twice as likely to regain ambulation when lost preoperatively (rescue procedure) [31]. In general, results from recent series of patients undergoing surgical intervention for metastatic spinal disease have been promising. Some series have reported that greater than 90% of patients have beneﬁted from surgery in terms of pain relief as well as maintaining ambulatory status [14, 16, 18]. Furthermore, survival beneﬁts have been reported with extensive resections in patients with solitary bone lesions or oligometastatic disease, with some 5-year survival rates being greater than 15% [18, 33]. Despite recent data supporting surgical decompression as a primary treatment for spinal metastases, radiation therapy remains the primary treatment option, par-

ticularly for radiosensitive tumors such as breast, lymphoma, multiple myeloma, small cell lung cancer, seminoma, neuroblastoma, and Ewing sarcoma. In addition, limited anticipated patient survival time, inability to tolerate a surgical procedure, greater than 24 to 48 hours of total neurologic compression, or multilevel/diffuse spinal involvement remain strong motivations to opt for radiation therapy [34]. The role of radiosurgical treatment is currently in evolution both as a primary and an adjunct therapy for spinal metastatic disease. A recent study including 31 patients (35 tumors) treated with the Novalis radiosurgical system showed improvements in pain levels and neurologic status in a majority of patients. Thirtytwo of 34 treatments resulted in rapid pain relief. Furthermore, 22 of 35 (63%) patients experienced sustained improvements in their neurologic status. Regarding the toxicity of this treatment, one patient experienced neurologic deterioration (cauda equine syndrome) despite therapy. Two patients each suffered from transient radiculitis and laryngitis. Furthermore, both patients experiencing radiculitis underwent therapy with biological equivalent doses (BEDs) greater than 60 Gy [35]. A retrospective study involving 61 lesions in 49 patients treated with single-dose radiosurgery (12 to 16 Gy) also showed promising results in pain relief as well as radiologic control of tumors. Eighty-ﬁve percent of patients experienced some degree of pain relief within 1 to 14 days after treatment. Only 7% of patients experienced recurring pain at the site previously treated. Three (4%) patients experienced radiologic progression of tumor to adjacent spinal levels at 6 and 9 month intervals after treatment [36]. Another retrospective study included 22 spinal lesions in 14 patients. All patients included multimodal treatment consisting of some combination of conventional radiotherapy, surgery, and radiosurgery. Radiosurgery was given using IMRS in the majority of cases; dynamic arcs were used in ﬁve cases, and conformal beams in two cases. Seven of 11 patients with preoperative pain showed some degree of improvement in their pain levels. There were varying results regarding post-radiosurgical tumor control. Two patients of nine with postoperative magnetic resonance imaging (MRI) showed decreases in tumor size, whereas two patients developed increased tumor size. The remaining ﬁve patients had stable lesions on follow-up scans. One patient died from progression in tumor size after treatment [37]. Recent studies describing the use of radiosurgery as both primary and adjunct therapies in patients with metastatic spine disease have showed encouraging results in pain relief as well as tumor burden control. The speciﬁc indications for the implementation of varying radiosurgical modalities as well as the optimal timing of this intervention remain to be determined. More prospective analyses would greatly beneﬁt in creating guidelines for the current radiosurgical management of patients with metastatic spine disease.

Guidelines for Surgical Decision Making In the past, various scoring systems have been developed to guide physicians in evaluating the surgical potential of a given patient based on multiple patient and tumor criteria

44.

445

spine tumors: surgery perspective

TABLE 44-1. Classiﬁcation algorithms in surgical decision making for spinal metastases. Study

Criteria

Recommendations

Harrington et al. (1986) [38]

Five categories: (1) No neurologic dysfunction (2) Involvement of bone without collapse or instability (3) Neurologic dysfunction without bony destruction (4) Vertebral collapse with secondary mechanical pain (5) Vertebral collapse or instability with major neurological deﬁcits Zero to 2 points for each of the following categories: (1) Primary tumor type (2) Neurologic grade of the lesion (3) Physical condition of the patient (4) Number of lesions (5) Presence of visceral lesions (6) Spinal sites Based on three categories: (1) Grade of malignancy: Slow growth, 1 point Moderate growth, 2points Rapid growth, 4 points (2) Visceral metastases No metastasis, 0 points Treatable, 2 points Untreatable, 4 points (3) Bone metastases Solitary/isolated, 1 point Multiple, 2 points MAPS Assessment: (1) Method of resection (2) Anatomy of spinal disease (3) Patient ﬁtness level (4) Stabilization

Based on category: (1) and (2): nonoperative management (3): left to judgment of surgeon (4) and (5): surgery indicated

Tokuhashi et al. (1990) [39]

Tomita et al. (2001) [40]

Fourney et al. (2005) [42]

(Table 44-1). The complexity of these algorithms is indicative of how many factors must be taken into consideration for a given scenario. Such frameworks for decision making offer useful guidelines to assist physicians with surgical decision making but cannot encompass the entire spectrum of individual factors that must be considered for each patient situation. In 1986, Harrington developed a staging system for spinal tumors based on neurologic dysfunction and bony destruction. His classiﬁcation included the following ﬁve categories: (1) no neurologic dysfunction, (2) involvement of bone without collapse or instability, (3) neurologic dysfunction without bony destruction, (4) vertebral collapse with secondary mechanical pain, (5) vertebral collapse or instability with major neurologic deﬁcits. Recommendations for categories 1 and 2 were for nonoperative therapy, whereas surgical intervention was advocated in categories 4 or 5. Patients in category 3 may or may not be surgical candidates and represent a gray area left to the judgment of the surgeon [38]. In 1990, Tokuhashi and colleagues developed a system for evaluating patients for surgery based on the (1) primary tumor type, (2) neurologic grade of the lesion, (3) physical condition of the patient, (4) number of lesions, (5) presence of visceral lesions, and (6) spinal sites [39]. A score of 0 to 2 is assigned

Based on total points (0–12): ≤5: nonoperative management 6–8: left to judgment of surgeon ≥9: surgery indicated

Based on total points (2–10): ≤3: wide/marginal excision (long-term local control) 4–5: marginal/intralesional excision (middleterm control) 6–7: surgery for short-term palliation ≥8: nonoperative supportive care

Considerations by category: (1) En bloc spondylectomy, piecemeal excision, palliative decompression (2) Tumor level, location, staging (3) Comorbidities, irradiation, patient age (4) Anterior, posterior, combined

for each category, with the total possible combined score being 12. In general, surgery was recommended for a total score of 9 or greater, and palliation was recommended for a score of 5 or less. This scale has also received criticism for some of its statistical methods, as well as for not providing input as to what type of surgical procedure should be attempted [40]. The Tokuhashi scoring system has been validated as being a useful scoring system in preoperative patient evaluation, except in the value assigned to unknown primary malignancies [41]. In 2001, Tomita and colleagues developed a separate system for evaluating patients based on a retrospective analysis of 67 patients [40]. This scoring system is based on the grade of the malignancy, the extent of visceral disease, and the extent of bony metastases. A score of 0 to 2 is possible for each category, with a possible combined prognostic score of between 2 and 10. They made the following recommendations based on the total score: Patients with a score of ≤3 should undergo a procedure for long-term local control. Those with a score of 4 to 5 should undergo a procedure for middle-term control. A score of 6 to 7 suggested surgery for short-term palliation, and a score of ≥8 supported nonoperative supportive care. Their point system offers guidance not only for consideration of surgery but also for the goal of the procedure, when indicated.

446

g. zada and m.y. wang

FIGURE 44-1. (A) A 54-year-old female developed 1 week of progressive lower extremity weakness and attendant bowel and bladder dysfunction. She was wheelchair-bound at the time of presentation, and a CT scan showed a pathologic fracture at T7. (B) Her past medical history was signiﬁcant for invasive ductal carcinoma treated 10 years prior and in remission. An MRI demonstrated this to be consistent with a metastatic lesion with a soft tissue component com-

pressing upon the spinal cord. (C) SPECT studies showed no evidence of other lesions. (D, E) The patient underwent an urgent T7 corpectomy via a transthoracic approach. The vertebral body was replaced by a PEEK vertebrectomy spacer packed with rib autograft, and supplemental plating was also utilized. The patient went on to have a complete neurologic recovery, ambulating without assistance 5 days after surgery.

44.

spine tumors: surgery perspective

Based on the combined score, Tomita’s scale can assist in deciding whether to perform a wide/marginal excision, intralesional excision, palliative surgery, or no surgery. Furthermore, the Tomita scale takes the degree of growth of the primary lesion into consideration. Slow-growing neoplasms like breast, thyroid, and prostate cancer are given 0 points. Lesions with moderate growth rates, including renal and uterine cancer, are given 1 point. Fast-growing lesions such as lung, liver, stomach, and colon cancer receive 2 points. Additionally, metastatic disease of unknown primary origin receives 2 points. In this series, 28 patients were deemed to have a score of between 2 and 5 and underwent an en bloc excision (vertebrectomy or corpectomy). The mean patient survival time was greater than 3 years [40] (Fig. 44-1). Recently, Fourney and colleagues have introduced the “MAPS” system in order to simplify surgical decision making when deciding on an ideal approach. The MAPS system refers to four categories for consideration, including (1) method of resection, (2) anatomy of spinal disease, (3) patient ﬁtness level, and (4) stabilization [42]. This more comprehensive approach reﬂects the evolution of algorithms to encompass all of these complex factors in surgical decision making.

Goals of Therapy In deciding on the optimal treatment for a patient, the goal of therapy remains a key factor. In patients with metastatic disease, the ultimate rationale for surgical intervention must be considered as disease may be widespread and survival time is often limited, thus favoring palliative treatment. The possible goals for surgery in patients with spinal tumors include (1) diagnosis, (2) neural decompression, (3) stabilization, (4) curative resection, and (5) control of intractable pain.

Diagnosis Since the advent of minimally invasive techniques for tumor biopsy of the spine, such as computed tomography (CT)-guided and ﬂuoroscopic-guided biopsy, diagnosis is usually no longer a primary indication for surgical intervention. One study reported that a tissue diagnosis was possible in 86% of patients using CT-guided biopsy [43]. Another retrospective study of patients undergoing CT-guided biopsy reported that a deﬁnitive diagnosis was obtained with one biopsy in 71% of patients [44].

Neural Decompression The most important indication for surgery remains neural decompression. Ambulatory status is a key prognostic marker for terminal patients with spinal disease [27]. Nonambulatory cancer patients have a mean survival time of 45 days, due to multiple causes [16]. Maintaining neurologic function in patients with spinal metastases may therefore prolong survival time in addition to improving quality of life. Patients with rapidly progressive or far-advanced paraplegia, especially for less than 24 to 48 hours, require emergent surgical intervention [11]. In patients with complete spinal cord dysfunction for

447

greater than 24 hours or those with bowel/bladder dysfunction, however, surgery offers little beneﬁt, and radiotherapy remains the mainstay of treatment [34]. Patients with progressive neurologic deterioration that is refractory to radiation therapy should also be considered for surgical decompression [34]. Cord compression caused by bone or disk fragments in the spinal canal that may result from pathologic fractures should also undergo primary surgical decompression instead of radiation therapy [45]. Whereas posterior laminectomy used to be the primary surgical option used by neurosurgeons to treat cord compression, it has since lost favor for the majority of metastatic spinal tumors because multiple studies failed to demonstrate a functional or survival beneﬁt in patients undergoing surgery and irradiation compared with radiation therapy alone [8, 10, 25, 46]. In the past few decades, however, surgical decompression of the spinal canal has regained favor because of more recent series utilizing one of many extensive circumferential surgical approaches for decompression and subsequent instrumentation. These surgical approaches offer the beneﬁt of direct decompression based on tumor location relative to the spinal canal. Many series have demonstrated greater than 90% maintenance in ambulatory status and pain improvement after decompression [16, 19, 47, 48]. A recent prospective randomized trial of surgical decompression plus radiotherapy versus radiotherapy alone showed a quality of life beneﬁt in the surgical arm (reduction in steroid and morphine requirements) that was statistically signiﬁcant as well as an increase in survival beneﬁt in the surgical arm [30, 32, 49]. This data represents class I evidence favoring surgical decompression over traditional conservative radiotherapy in patients with spinal metastatic disease. More randomized prospective trials are necessary to corroborate this evidence.

Stabilization Maintaining spinal stability is another indication for surgical intervention in patients with metastatic spinal disease. Stability has been deﬁned by Panjabi et al. as allowing the degree of motion that prevents pain, neurologic deﬁcit, and abnormal spinal angulation [50]. Spinal instability in patients with metastases can result from related pathologic fractures, surgical intervention, or diffuse changes in bone density and strength. Instability has been correlated with the following ﬁndings in patients with spine tumors: two-column injury, vertebral body collapse greater than 50%, kyphosis greater than 20° to 30°, or involvement of the same column in greater than 2 levels [41, 51, 52]. After most cases of surgical decompression, especially in cancer patients, instrumentation is required to provide stability. In nonterminal patients, postoperative stability is ideally achieved via a solid bone fusion after resection. In cancer patients, however, a bony fusion may not be a realistic goal due to multiple potentially complicating factors. Solid bony fusion may be prevented by shorter patient survival times, anemia, malnutrition, or frequent requirements for further chemotherapy, steroids, and radiation therapy. Although the ability to study spinal fusion in cancer patients has been compromised by decreased follow-up times, one study studying fusion rates in 25 patients undergoing anterior decompression and interbody fusion with bone struts followed by irradiation reported a pseudarthrosis rate of 16% at 1 year after surgery [53].

448

g. zada and m.y. wang

Several options are currently available for spinal instrumentation in cancer patients. Rigid spinal ﬁxation systems are often used to provide spinal stability [16, 54]. With anterior approaches, constructs are often placed one level above and below the level of resection. Cortical bone integrity must be preserved in order to stabilize anteriorly, however, and loss of integrity may necessitate a posterior fusion or even a combined anterior-posterior approach in cases of signiﬁcant instability [11]. In posterior approaches, two to three levels of construct are generally used above and below the deﬁcit [16, 54]. Polymethylmethacrylate (PMMA) is frequently used for reconstruction in conjunction with mechanical instrumentation if the estimated patient survival time is estimated to be less than 6 months. PMMA may be inserted inside of a titanium cage or chest tube and can provide immediate stability [54, 55]. PMMA has been shown to be a safe option for patients requiring subsequent radiation therapy [56] and is an inexpensive, relatively easy to use, and safe alternative to bony fusion [57]. Other alternatives include minimally invasive procedures such as vertebroplasty and kyphoplasty, which can be utilized to provide relatively instantaneous stability in cases of pathologic fractures (see below).

Curative Resection and Local Control Although a curative surgical resection of spinal metastases was not considered a reasonable goal in patients with metastatic spinal disease for many years, recent series have demonstrated that survival times can be signiﬁcantly prolonged in particular patients that have undergone surgical intervention [18, 32, 58]. In fact, surgery is the only treatment that has demonstrated a potential for cure of spinal metastases, as opposed to local control of tumor available with radiation therapy [33]. In the past, posterior laminectomy has mainly offered the beneﬁts of limited decompression and diagnosis and has caused many practitioners to doubt the utility of surgical intervention in metastatic spine disease. More recently, procedures such as en bloc spondylectomy and vertebrectomy have become key developments in the ability to aggressively treat and even potentially cure metastatic spinal disease. In recent years, more aggressive surgical intervention has been advocated for patients with solitary bone metastases as well as for few, localized lesions [18]. The term oligometastatic disease has been used to describe a situation in which there is one focus of neoplastic disease with or without multiple adjacent or distant small foci in a previous cancer patient [18, 33]. Patients with bony metastases as opposed to visceral metastases, and those who present after a long disease-free interval, are considered to have a low tumor burden and may be optimal candidates for curative resection [18, 33, 40, 58, 59]. Such patients may beneﬁt from more aggressive surgical management in terms of pain control and quality of life and should be treated with the intent to cure [33, 58]. In order to pursue curative therapy for this degree of disease, the patient must be a good surgical candidate, and the surgical center should be experienced enough to provide patients with the lowest morbidity and mortality rates, thus offering the best beneﬁt to risk ratio [18, 60]. The potential for a curative resection also depends on the source of the primary tumor. For instance, tumors of renal origin should be treated with the intent

to cure [61]. Furthermore, local malignancies such as chordomas, sarcomas, chondrosarcomas, Pancoast tumors, giant cell tumors, and osteoblastomas should also be considered for curative en bloc resection [16].

Pain Control Pain control is a challenging issue in patients with metastatic bone disease. Pain from metastatic lesions can be a result of inﬂammation and mechanical forces, as well as direct nerve compression or damage [47]. Radicular symptoms have been reported in up to 90% of patients with lumbar epidural cord compression, 79% of patients with cervical disease, and 55% of patients with thoracic metastases [25]. Corticosteroids and opioids are the primary medications available for pain management secondary to cord or root compression. Radiation therapy remains the primary intervention reserved for patients with terminal disease that are experiencing signiﬁcant pain. However, the beneﬁcial effects of irradiation in relieving pain may take up to 2 weeks and may not provide adequate pain relief [62]. In cases where pain is refractory to radiation therapy, surgery remains a viable option. Many surgical procedures offer a signiﬁcant beneﬁt in pain relief, with recent series reporting pain relief success rates of greater than 90% [13, 18, 63]. Percutaneous vertebroplasty and kyphoplasty are minimally invasive procedures that have been shown to improve pain in patients who may not be candidates for more aggressive decompressive surgery. Several series demonstrate both short- and long-term pain relief rates of 85% to 100% associated with percutaneous vertebroplasty and kyphoplasty [64–67].

Timing of Combined Therapy Because of the lack of substantial class I evidence showing the beneﬁt of surgery as a primary treatment for spinal metastases, radiation therapy remains the primary modality for treatment at many medical institutions. After a radiologic diagnosis of spinal metastases, many patients are directly referred for radiation therapy without a surgeon seeing the patient [68]. The results of this are twofold. First, opting for radiation as a primary measure may not provide the same survival beneﬁt for the patient as a wide marginal or radical excision could potentially provide. Second, patients that fail radiotherapy and ultimately require surgical decompression have signiﬁcantly higher chances of developing wound infections, especially in cases where irradiation was administered within 1 week prior to surgery [32, 68]. One series of 85 patients undergoing posterior decompression with or without instrumentation determined that patients with prior irradiation had a 32% chance of developing wound infections compared with a 12% chance in the de novo surgery group [68]. Furthermore, patients with irradiation in the week prior to surgery had a 46% chance of developing a postoperative wound infection. Multiple other studies have demonstrated similar results [30, 31, 33, 68, 69]. The propective phase III Patchell study also showed that the likelihood of regaining the ability to ambulate among those crossover patients who received

44.

spine tumors: surgery perspective

surgery for salvage after irradiation was less effective than upfront surgery [32]. For these reasons, many surgeons currently advocate considering de novo surgery as the primary intervention for symptomatic metastatic spinal disease and have emphasized the spine surgeon’s involvement in the decision to irradiate patients.

Deciding on a Surgical Approach Once surgery has been decided upon, choosing the correct surgical approach is a key factor in providing the ideal degree of exposure for resection and subsequent stabilization. This decision is based on the anatomic location of the spinal metastases, the degree of surgical intervention a patient is able to tolerate, and the ability of the surgeon to perform a particular procedure safely. Many surgeons have advocated the use of a surgical planning guideline based on radiographic studies to make a surgical plan. The Weinstein, Boriani, and Biagini staging system can be used to help guide surgeons in choosing an approach for surgical decompression based on the anatomic location of the tumor in regard to proximity to the dura as well as anterior to posterior localization [70]. This system maps tumors based on 12 radiating zones in a clockwise fashion, as well as ﬁve concentric layers. It can assist in deciding on a type of excisional procedure, including en bloc spondylectomy, vertebrectomy, sagittal resection, or posterior arch resection.

Decompressive Posterior Laminectomy In the past, the primary surgical procedure utilized by neurosurgeons to relieve compression secondary to spinal metastases was the posterior laminectomy. This procedure was initially favorable because it could be performed by the majority of neurosurgeons within relatively fast time frames. For several reasons, however, this procedure fell out of favor for the primary treatment of spinal metastases. Class I evidence from a casecontrol study comparing decompressive laminectomy plus radiation therapy with radiation therapy alone did not show a signiﬁcant functional beneﬁt in ambulatory status after surgical intervention [46]. Class II and III evidence also failed to demonstrate signiﬁcant functional beneﬁts or improvement in pain [8, 10, 25, 71]. The inability of posterior procedures to provide adequate decompression has been largely attributed to the majority of metastatic lesions being located anterior to the thecal sac. Retraction of the thecal sac would thus be necessary to complete a gross total resection from a posterior approach in the majority of patients with metastases. In addition, the risks of decompressive laminectomies in cancer patients are not to be overlooked. The risk of spinal instability is especially concerning after laminectomy in cancer patients with preexisting disease in the anterior and middle spinal columns [13, 48]. Furthermore, rates of wound infection are higher in patients that have undergone posterior approaches [54]. Decompressive laminectomy is now primarily reserved for metastatic lesions located within the posterior spinal elements or in cases where patients cannot tolerate radiation or an anterior approach. In

449

contrast, wide posterior laminectomies remain the procedure of choice for the majority of intradural-extramedullary and intramedullary primary lesions, often in conjunction with spinal SSEP and MEP monitoring. Some surgeons maintain that even patients with ventrally located neoplasms can be adequately and safely decompressed with posterior laminectomy if internal ﬁxation is used in conjunction. This practice has been reported to realign the spine as well as distract the vertebral bodies out of the spinal canal [72]. Several studies have provided evidence that decompressive laminectomy with internal ﬁxation provides better functional outcomes, pain relief, and spinal stability than laminectomy alone [48, 73–77]. This procedure remains a useful option in patients who are unable to undergo more extensive anterior approaches or for patients with multilevel disease [78]. In many cases, however, a circumferential approach is preferred to achieve adequate spinal cord decompression.

Newer Approaches for Spinal Cord Decompression An increased interest in surgical decompression for metastatic spine disease has grown over the past two decades, with numerous surgical centers reporting their experiences with extensive procedures for spinal cord decompression that are tailored speciﬁcally to the anatomic location of the tumor. To date, multiple studies have been published describing the efﬁcacy of various circumferential approaches for pain relief and preservation or improvement of neurologic function [13–15, 18, 48, 51, 79–81]. Postoperative pain relief has ranged from 71% to 100% in these series [82]. Overall, the proportion of ambulatory patients who retained ambulatory function postoperatively (deﬁned as the “success” of such a procedure) has ranged between 72% and 98% but has been reported as lower when looking at only posterior approaches [82]. The proportions of nonambulatory patients that regained ambulatory status postoperatively (deﬁned as a “rescue” procedure) have been less consistent, ranging between 0 and 94% in these series [82]. Finally, surgical morbidity rates for circumferential decompressions have ranged between 7% and 65%, and mortality rates have ranged between 0 and 31% [82]. The majority of series report mortality rates of less than 10%. More recently, evidence from a randomized prospective series has provided class I evidence for aggressive surgical decompression in metastatic spine disease [30]. In this study, patients were randomized to one of two study arms, one with decompressive surgery plus radiation therapy (50 patients) and the other with radiation therapy alone (51 patients). Both groups received steroids and the same regimen of radiation therapy. Fifty-six percent of patients in the surgical arm regained the ability to walk compared with only 19% in the radiation arm. Patients undergoing surgical resection retained ambulatory and sphincter function for longer than patients in the radiation arm [30]. Approaches to the spine can be broadly classiﬁed as anterior or posterior. Anterior approaches include traditional anterior cervical approaches, transthoracic, and transperitoneal approaches. Posterior approaches include traditional laminectomy and posterolateral approaches such as transpedicular, costotransversectomy, and lateral extracavitary

450

g. zada and m.y. wang

approaches. All of these approaches require subsequent reconstruction and stabilization.

Anterior Approaches In general, an anterior approach is preferable for achieving adequate decompression in patients that are able to tolerate such an approach because spinal metastases tend to localize to the anterior spinal column. In general, series of anterior procedures demonstrate superior rates of neurologic function and survival times than do posterior procedures [13, 16, 72]. In addition to decompression, reconstruction of the vertebral body after resection is also facilitated by an anterior approach. In the cervical spine, reconstruction after an extensive resection is accomplished using bone autograft or allograft, PMMA, or one of many titanium cages or spacers [83]. Stabilization is then usually achieved by way of anterior instrumentation, with plating being the most common technique in the cervical spine. A subsequent posterior approach for stabilization is sometimes done to provide better stabilization (see below) [15, 48, 57]. Posterior stabilization is often accomplished using screw/rod ﬁxation and may be done with a staged anterior than posterior approach to minimize morbidity [15, 48, 57]. Bicortical screw purchase is preferable in order to improve stability [54]. In the thoracic spine, a transthoracic approach can be utilized for anterior decompression and subsequent instrumentation. The side of approach is based on the symmetry of disease, but the left side is often preferable in order to avoid the liver and to more easily identify the aorta. Rib resection, lung deﬂation, and chest tube insertion are hallmarks of this procedure at the thoracic level [45]. An anterior approach at the thoracolumbar junction (T11-L1) can be done via a thoracoabdominal approach. At this level, a portion of the diaphragm requires mobilization in order to achieve adequate exposure. At the lumbar level, dissection via the retroperitoneum is required, usually along the transversalis fascia. The psoas muscles and aorta are mobilized to provide exposure to the lumbar vertebrae [45].

pression and subsequent instrumentation in this region is comparatively high, with two series reporting stabilization failure rates of 36% [84, 85]. Le et al. advocate a posterior/posterolateral approach to the cervicothoracic junction because of lower incidences of hardware failure. The use of a posterolateral approach at the cervicothoracic junction has been reported to be sufﬁcient in achieving adequate exposure to anteriorly positioned lesions as well as the posterior elements for subsequent instrumentation [84]. The transpedicular approach can be used for disease in the dorsal and elements of the vertebrae. The advantages to using this approach over an anterior approach include early identiﬁcation of the spinal canal, ability to resect tumor in the posterior elements, and the ability to provide long segment ﬁxation [80, 86]. A costotransversectomy provides exposure to the posterior and lateral elements of the vertebral column via a midline or paramedian incision from T2 to L3. Currently, costotransversectomy allows resection of the lamina, facet, transverse process, pedicle, rib head, and some of the vertebral body to provide adequate exposure. Advantages of this procedure include avoidance of a thoracotomy in patients that may not be able to tolerate one, access to multilevel or discontinuous spine disease, and the ability to perform both anterior and posterior ﬁxation. Disadvantages with this procedure include a possibly limited or indirect exposure, possible entry into the lung space requiring chest tube placement, and possible injury to the nerve roots or dura resulting in a CSF ﬁstula. The lateral extracavitary approach can be used for spinal decompression and ﬁxation in the thoracolumbar spine (T4-L3) and provides good lateral exposure to the vertebral column without violation of the pleural or abdominal space or mobilization of the diaphragm [87]. Like the costotransversectomy, both anterior and posterior ﬁxation are possible with the lateral extracavitary approach (LECA). It is commonly used in patients requiring complete resection of one to three vertebral bodies. Exposure is achieved via dissection through the thoracolumbar fascia and rib resection and corpectomy. Reconstruction of the vertebral body is then performed followed by posterior with or without anterior ﬁxation [87].

Combined Anterior and Posterior Approaches Posterolateral Approaches As previously mentioned, posterior approaches for decompression have not demonstrated an efﬁcacy comparable with anterior procedures, except in certain cases. Lesions located in the posterior spinal elements should be approached in this fashion. Patients that may not be able to tolerate a more extensive anterior approach or those with multilevel disease may beneﬁt from posterior decompression with stabilization. Finally, particular regional nuances may contribute to the decision to approach posteriorly. One series reported 100% pain relief and greater than 75% recovery of ambulatory status after posterior decompression and stabilization [72]. The disadvantages of a posterior approach include limited anterior exposure with limited tumor resection and higher rates of wound complications [54]. The cervicothoracic junction (C7 to T4) marks a transition zone from the mobile, lordotic cervical spine to a rigid, kyphotic thoracic spine. The hardware failure rates with surgical decom-

While providing the beneﬁt of excellent exposure to the relevant elements of the spinal canal, anterior decompression also carries the potential drawback of decreased spinal stability. Several series report using anterior decompression with subsequent posterior instrumentation to further stabilize the spinal column [15, 48]. Particular indications for this combination approach include involvement of metastatic disease in all three spinal columns, signiﬁcant instability, marked kyphosis, involvement of more than one vertebral body, junctional site involvement, and prior laminectomy [15, 48]. One study including 110 patients undergoing spinal decompression determined that approximately half (48%) of all patients undergoing decompressive surgery required a combined anterior-posterior approach for added stability. This combined approach is associated with a signiﬁcantly higher complication rate, with 35% to 48% of patients having postoperative complications. Wound breakdown and infection were the most commonly reported complications [15].

44.

spine tumors: surgery perspective

En Bloc Spondylectomy In recent years, aggressive surgical resection of spinal metastases has made curative resection of spinal metastatic lesions a possible outcome. As mentioned, patients with solitary bone metastases or oligometastatic disease can be treated with intent to cure if the patient is a surgical candidate for an aggressive surgical procedure. Patient selection is especially crucial when evaluating patients for an en bloc spondylectomy, as this procedure tends to be long and complex with relatively signiﬁcant blood loss and a frequent requirement for an open thoracotomy and more signiﬁcant reconstruction. Surgical planning in terms of approach and degree of resection has been greatly emphasized, with some surgeons favoring the use of a staging system such as the Boriani/Weinstein system in planning this type of procedure. Results from recent series of en bloc resections show promising results. One such study reported a mortality rate of less than 1%, a morbidity rate of less than 10%, and a mean survival time of greater than 3 years [33]. The mean patient survival time in this study was greater than 3 years, with 48% of patients showing neurologic improvement. Almost half of all patients improved one Frankel grade. Pain improved in 95% of patients after surgery, with 76% achieving total pain control. Another recent series of patients undergoing en bloc spondylectomy demonstrated morbidity and mortality rates less than 1% [18]. A smaller series with 28 patients undergoing en bloc spondylectomy reported a mean survival time of 38 months, with 78% of patients demonstrating neurologic improvement after surgery and 93% of patients achieving local control [88]. Once an en bloc spondylectomy or vertebrectomy has been performed, vertebral body replacement or reconstruction is necessary to provide stabilization. There have been many materials and instruments to replace the vertebral body after vertebrectomy that have been described in the literature, including bone cement, ceramic, ceramic/glass, carbon ﬁber, and titanium spacers or cages [89]. Recently, expandable cages have been developed for replacing the vertebral body and disks. The results from a retrospective study with 15 patients revealed an average correction of the kyphotic angle of 20°, immediate stability, and no complications of cage placement or evidence of hardware failure in all patients after more than 1 year of followup [89]. These expandable cages offer the beneﬁts of relatively easy intraoperative placement, restoration of the height and sagittal alignment of the spine, and immediate stability. However, further anterior or posterior instrumentation after cage placement is still necessary. Such cages may offer a beneﬁt in patients with metastatic spine disease because fusion may not be a primary goal of the procedure, but immediate stability is provided.

Percutaneous Vertebroplasty and Kyphoplasty Vertebroplasty is a relatively noninvasive surgical procedure that has been increasingly utilized in recent years to treat pain and instability in vertebral body collapse associated with metastatic disease. Since the procedure’s inception for treating vertebral body angiomas in 1990 [90], indications for treatment have been broadened to include osteoporotic and osteolytic

451

lesions secondary to metastatic disease. The procedure is often used for patients with severe, localized mechanical back pain without epidural involvement that may not be able to tolerate more aggressive procedures [91]. Kyphoplasty is a related procedure in which an inﬂatable balloon is used initially to restore the loss of vertebral height and create a cavity for subsequent injection of cement. Because percutaneous vertebroplasty and kyphoplasty are relatively quick procedures with minimal invasiveness, they can be used in cancer patients with limited survival time, those with poor surgical potential, and in patients that have received prior radiation therapy [91]. It has been shown that vertebroplasty can be used prior to the initiation of radiation therapy, without compromising the efﬁcacy of either [56]. Many series have advocated this combination of vertebroplasty followed by radiation therapy [65, 66, 92]. These procedures should not be used in patients with spinal instability, cord compression, or epidural extension that are able to undergo surgical intervention. Further contraindications for these procedures include vertebral collapse to less than one third of its original height, coagulopathies, inability to lie prone, and inability to perform an emergent decompressive surgery if necessary [65, 66, 92]. Vertebroplasty has been shown to provide its beneﬁt by restoring strength and stiffness to the spine and preventing micromotion [93, 94]. Additionally, PMMA has been shown to destroy nerve endings that cause pain associated with micromotion [65, 92, 95]. It has also been postulated that PMMA may provide an antitumor effect secondary to thermal effects, cytotoxicity, and ischemia, which prevents tumor reinvasion [65, 95]. Percutaneous vertebroplasty can be done with the assistance of CT or ﬂuoroscopy imaging and under local anesthesia. A biopsy can be done during the same procedure prior to the injection of the ﬁlling agent. In a series of 101 patients, Deramond reported an improvement in pain and quality of life in 80% of patients [65, 95]. Patients after vertebroplasty have experienced short-term (within 48 hours) and long-term (at 3 months postoperatively) pain relief rates of 85% to 97% and 89% to 100%, respectively [64, 66]. At 6 months after vertebroplasty, pain relief was still reported by 76% of patients, the majority of which developed either new metastases or epidural involvement of metastases [95]. Despite its many beneﬁts and a low degree of invasiveness, vertebroplasty is not without its complications. The overall complication rate has been reported as approximately 10%, with the majority being short-term complications and only 1.7% resulting in long-term complications [65]. Leakage of PMMA has been reported to occur in approximately 70% to 75% of cases, and in the majority this does not present a clinical problem [65, 91]. Signiﬁcant clinical risks are increased with posterior cortical wall destruction or epidural involvement. The major clinical risk of this procedure is radiculopathy secondary to PMMA leakage into the neuroforamina, or back pain caused by a local inﬂammatory reaction to PMMA [65, 94]. In many cases, radiculopathies result in intercostal neuralgias that may improve with local injection. Although uncommon, pulmonary embolism has been reported as a result of leakage into the inferior vena cava [96]. In summary, vertebroplasty and kyphoplasty are minimally invasive procedures that may be performed for selected patients

452

g. zada and m.y. wang

with metastatic spine disease in order to provide fast pain relief and improved spinal stability. It is often the procedure of choice as a palliating measure for patients with limited survival or for those who are otherwise unable to tolerate surgical intervention.

Conclusion In summary, surgical management remains a viable and increasingly efﬁcacious option for the management of spinal tumors. The primary goals for surgery are tissue diagnosis, neural decompression, spinal stabilization, curative resection, and pain management. Radiation therapy remains the primary treatment for metastatic spine disease in many instances, yet recent prospective, randomized series have demonstrated improved neurologic function and survival outcomes with surgery compared with radiation therapy. Curative resection of solitary or localized metastatic disease is now a potential goal with aggressive surgical methods. Additionally, newer radiosurgical modalities are becoming increasingly utilized because they potentially offer the combination of minimal invasiveness as well as accurate tumor decompression. Careful patient selection is the key factor to maximizing the potential beneﬁts and avoiding the associated risks of any treatment modality, be it surgery or radiotherapy. More clinical studies, especially prospective analyses, are required to deﬁne the ideal roles for each of these useful modalities.

References 1. Posner JB. Back pain and epidural spinal cord compression. Med Clin North Am 1987; 71(2):185–205. 2. Jaffe W. Tumors and Tumorous Conditions of the Bones and Joints. Philadelphia: Lea and Febiger, 1958:589–618. 3. Cobb CA 3rd, Leavens ME, Eckles N. Indications for nonoperative treatment of spinal cord compression due to breast cancer. J Neurosurg 1977; 47(5):653–658. 4. Wong DA, Fornasier VL, MacNab I. Spinal metastases: the obvious, the occult, and the impostors. Spine 1990; 15(1):1–4. 5. Bohm P, Huber J. The surgical treatment of bony metastases of the spine and limbs. J Bone Joint Surg Br 2002; 84(4):521–529. 6. Perrin RG, Laxton AW. Metastatic spine disease: epidemiology, pathophysiology, and evaluation of patients. Neurosurg Clin N Am 2004; 15(4):365–373. 7. Wu AS, Fourney DR. Evolution of treatment for metastatic spine disease. Neurosurg Clin N Am 2004; 15(4):401–411. 8. Sorensen S, Borgesen SE, Rohde K, et al. Metastatic epidural spinal cord compression. Results of treatment and survival. Cancer 1990; 65(7):1502–1508. 9. Nottebaert M, von Hochstetter AR, Exner GU, Schreiber A. Metastatic carcinoma of the spine. A study of 92 cases. Int Orthop 1987; 11(4):345–348. 10. Constans JP, de Divitiis E, Donzelli R, et al. Spinal metastases with neurological manifestations. Review of 600 cases. J Neurosurg 1983; 59(1):111–118. 11. Jacobs WB, Perrin RG. Evaluation and treatment of spinal metastases: an overview. Neurosurg Focus 2001; 11(6):e10. 12. Bailar JC 3rd, Gornik HL. Cancer undefeated. N Engl J Med 1997; 336(22):1569–1574. 13. Gokaslan ZL, York JE, Walsh GL, et al. Transthoracic vertebrectomy for metastatic spinal tumors. J Neurosurg 1998; 89(4): 599–609.

14. Sundaresan N, Digiacinto GV, Hughes JE, et al. Treatment of neoplastic spinal cord compression: results of a prospective study. Neurosurgery 1991; 29(5):645–650. 15. Sundaresan N, Steinberger AA, Moore F, et al. Indications and results of combined anterior-posterior approaches for spine tumor surgery. J Neurosurg 1996; 85(3):438–446. 16. Vrionis FD, Small J. Surgical management of metastatic spinal neoplasms. Neurosurg Focus 2003; 15(5):e12. 17. Fourney DR, Abi-Said D, Lang FF, et al. Use of pedicle screw ﬁxation in the management of malignant spinal disease: experience in 100 consecutive procedures. J Neurosurg 2001; 94(1 Suppl): 25–37. 18. Sundaresan N, Rothman A, Manhart K, Kelliher K. Surgery for solitary metastases of the spine: rationale and results of treatment. Spine 2002; 27(16):1802–1806. 19. Wai EK, Finkelstein JA, Tangente RP, et al. Quality of life in surgical treatment of metastatic spine disease. Spine 2003; 28(5): 508–512. 20. Perrin RG, Livingston KE, Aarabi B. Intradural extramedullary spinal metastasis. A report of 10 cases. J Neurosurg 1982; 56(6): 835–837. 21. Posner JB, Howieson J, Cvitkovic E. “Disappearing” spinal cord compression: oncolytic effect of glucocorticoids (and other chemotherapeutic agents) on epidural metastases. Ann Neurol 1977; 2(5):409–413. 22. Byrne TN. Spinal cord compression from epidural metastases. N Engl J Med 1992; 327(9):614–619. 23. Grant R, Papadopoulos SM, Greenberg HS. Metastatic epidural spinal cord compression. Neurol Clin 1991; 9(4):825–841. 24. Livingston KE, Perrin RG. The neurosurgical management of spinal metastases causing cord and cauda equina compression. J Neurosurg 1978; 49(6):839–843. 25. Gilbert RW, Kim JM, Posner JB. Epidural spinal cord compression from metastatic tumor: diagnosis and treatment. Ann Neurol 1978; 3(1):40–51. 26. O’Rourke T, George CB, Redmond J 3rd, et al. Spinal computed tomography and computed tomographic metrizamide myelography in the early diagnosis of metastatic disease. J Clin Oncol 1986; 4(4):576–583. 27. Ruff RL, Lanska DJ. Epidural metastases in prospectively evaluated veterans with cancer and back pain. Cancer 1989; 63(11): 2234–2241. 28. Ippolito V, Micheletti E, Saccalani M, et al. Radiotherapy and spinal brace: still ﬁrst-choice treatment for vertebral metastases from breast cancer. Chir Organi Mov 1998; 83(1–2):177–183. 29. Katagiri H, Takahashi M, Inagaki J, et al. Clinical results of nonsurgical treatment for spinal metastases. Int J Radiat Oncol Biol Phys 1998; 42(5):1127–1132. 30. Regine WF, Tibbs PA, Young A, et al. Metastatic spinal cord compression: a randomized trial of direct decompressive surgical resection plus radiotherapy vs radiotherapy alone. Int J Radiat Oncol Biol Phys 2003; 57(1 Suppl):S125. 31. Klimo P Jr, Thompson CJ, Kestle JR, Schmidt MH. A metaanalysis of surgery versus conventional radiotherapy for the treatment of metastatic spinal epidural disease. Neurooncology 2005; 7(1):64–76. 32. 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. Lancet 2005; 366:643–648. 33. Yao KC, Boriani S, Gokaslan ZL, Sundaresan N. En bloc spondylectomy for spinal metastases: a review of techniques. Neurosurg Focus 2003; 15(5):e6. 34. Ryken TC, Eichholz KM, Gerszten PC, et al. Evidence-based review of the surgical management of vertebral column metastatic disease. Neurosurg Focus 2003; 15(5):e11.

44.

spine tumors: surgery perspective

35. Benzil DL, Saboori M, Mogilner AY, et al. Safety and efﬁcacy of stereotactic radiosurgery for tumors of the spine. J Neurosurg 2004; 101(Suppl 3):413–418. 36. Ryu S, Rock J, Rosenblum M, Kim JH. Patterns of failure after single-dose radiosurgey for spinal metastasis. J Neurosurg 2004; 101(Suppl 3):402–405. 37. De Salles A, Pedroso AG, Medin P, et al. Spinal lesions treated with Novalis shaped beam intensity-modulated radiosurgery and stereotactic radiotherapy. J Neurosurg 2004; 101(Suppl 3):435– 440. 38. Harrington KD. Metastatic disease of the spine. J Bone Joint Surg Am 1986; 68(7):1110–1115. 39. Tokuhashi Y, Matsuzaki H, Toriyama S, et al. Scoring system for the preoperative evaluation of metastatic spine tumor prognosis. Spine 1990; 15(11):1110–1113. 40. Tomita K, Kawahara N, Kobayashi T, et al. Surgical strategy for spinal metastases. Spine 2001; 26(3):298–306. 41. Enkaoua EA, Doursounian L, Chatellier G, et al. Vertebral metastases: a critical appreciation of the preoperative prognostic Tokuhashi score in a series of 71 cases. Spine 1997; 22(19):2293– 2298. 42. Fourney DR, Gokaslan ZL. Use of “MAPs” for determining the optimal surgical approach to metastatic disease of the thoracolumbar spine: anterior, posterior, or combined. Invited submission from the Joint Section Meeting on Disorders of the Spine and Peripheral Nerves, March 2004. J Neurosurg Spine 2005; 2(1): 40–49. 43. Ghelman B, Lospinuso MF, Levine DB, et al. Percutaneous computed-tomography-guided biopsy of the thoracic and lumbar spine. Spine 1991; 16(7):736–739. 44. Kornblum MB, Wesolowski DP, Fischgrund JS, Herkowitz HN. Computed tomography-guided biopsy of the spine. A review of 103 patients. Spine 1998; 23(1):81–85. 45. Fourney DR, Gokaslan ZL. Anterior approaches for thoracolumbar metastatic spine tumors. Neurosurg Clin N Am 2004; 15(4):443–451. 46. Young RF, Post EM, King GA. Treatment of spinal epidural metastases. Randomized prospective comparison of laminectomy and radiotherapy. J Neurosurg 1980; 53(6):741–748. 47. Weinstein SM, Walton O. Management of pain associated with spinal tumor. Neurosurg Clin N Am 2004; 15(4):511–527. 48. Fourney DR, Abi-Said D, Rhines LD, et al. Simultaneous anterior-posterior approach to the thoracic and lumbar spine for the radical resection of tumors followed by reconstruction and stabilization. J Neurosurg 2001; 94(2 Suppl):232–244. 49. 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. Lancet. 2005; 366(9486):643–648. 50. Panjabi MM, Thibodeau LL, Crisco JJ 3rd, White AA 3rd. What constitutes spinal instability? Clin Neurosurg 1988; 34:313– 339. 51. Kostuik JP, Errico TJ, Gleason TF, Errico CC. Spinal stabilization of vertebral column tumors. Spine 1988; 13(3):250–256. 52. Denis F. The three column spine and its signiﬁcance in the classiﬁcation of acute thoracolumbar spinal injuries. Spine 1983; 8(8):817–831. 53. Emery SE, Hughes SS, Junglas WA, et al. The fate of anterior vertebral bone grafts in patients irradiated for neoplasm. Clin Orthop Relat Res 1994; (300):207–212. 54. Steinmetz MP, Mekhail A, Benzel EC. Management of metastatic tumors of the spine: strategies and operative indications. Neurosurg Focus 2001; 11(6):e2. 55. Errico TJ, Cooper PR. A new method of thoracic and lumbar body replacement for spinal tumors: technical note. Neurosurgery 1993; 32(4):678–680; discussion 680–681.

453

56. Murray JA, Bruels MC, Lindberg RD. Irradiation of polymethylmethacrylate. In vitro gamma radiation effect. J Bone Joint Surg Am 1974; 56(2):311–312. 57. Liu JK, Apfelbaum RI, Chiles BW 3rd, Schmidt MH. Cervical spinal metastasis: anterior reconstruction and stabilization techniques after tumor resection. Neurosurg Focus 2003; 15(5):e2. 58. Singletary SE, Walsh G, Vauthey JN, et al. A role for curative surgery in the treatment of selected patients with metastatic breast cancer. Oncologist 2003; 8(3):241–251. 59. Hortobagyi GN. Can we cure limited metastatic breast cancer? J Clin Oncol 2002; 20(3):620–623. 60. Abdu WA, Provencher M. Primary bone and metastatic tumors of the cervical spine. Spine 1998; 23(24):2767–2777. 61. Boriani S, Biagini R, De Iure F, et al. En bloc resections of bone tumors of the thoracolumbar spine. A preliminary report on 29 patients. Spine 1996; 21(16):1927–1931. 62. Gilbert HA, Kagan AR, Nussbaum H, et al. Evaluation of radiation therapy for bone metastases: pain relief and quality of life. AJR Am J Roentgenol 1977; 129(6):1095–1096. 63. Hosono N, Yonenobu K, Fuji T, et al. Vertebral body replacement with a ceramic prosthesis for metastatic spinal tumors. Spine 1995; 20(22):2454–2462. 64. Cortet B, Cotton A, Boutry N, et al. Percutaneous vertebroplasty in patients with osteolytic metastases or multiple myeloma. Rev Rhum Engl Ed 1997; 64(3):177–183. 65. Deramond H, Depriester C, Galibert P, Le Gars D. Percutaneous vertebroplasty with polymethylmethacrylate. Technique, indications, and results. Radiol Clin North Am 1998; 36(3):533–546. 66. Kaemmerlen P, Thiesse P, Bouvard H, et al. [Percutaneous vertebroplasty in the treatment of metastases. Technic and results]. J Radiol 1989; 70(10):557–562. 67. Fourney DR, Schomer DF, Nader R, et al. Percutaneous vertebroplasty and kyphoplasty for painful vertebral body fractures in cancer patients. J Neurosurg 2003; 98(1 Suppl):21–30. 68. Ghogawala Z, Mansﬁeld FL, Borges LF. Spinal radiation before surgical decompression adversely affects outcomes of surgery for symptomatic metastatic spinal cord compression. Spine 2001; 26(7):818–824. 69. Kim RY, Smith JW, Spencer SA, et al. Malignant epidural spinal cord compression associated with a paravertebral mass: its radiotherapeutic outcome on radiosensitivity. Int J Radiat Oncol Biol Phys 1993; 27(5):1079–1083. 70. Boriani S, Weinstein JN, Biagini R. Primary bone tumors of the spine. Terminology and surgical staging. Spine 1997; 22(9):1036– 1044. 71. Black P. Spinal metastasis: current status and recommended guidelines for management. Neurosurgery 1979; 5(6):726–746. 72. Olerud C, Jonsson B. Surgical palliation of symptomatic spinal metastases. Acta Orthop Scand 1996; 67(5):513–522. 73. Bauer HC. Posterior decompression and stabilization for spinal metastases. Analysis of sixty-seven consecutive patients. J Bone Joint Surg Am 1997; 79(4):514–522. 74. Jonsson B, Sjostrom L, Olerud C, et al. Outcome after limited posterior surgery for thoracic and lumbar spine metastases. Eur Spine J 1996; 5(1):36–44. 75. Kluger P, Korge A, Scharf HP. Strategy for the treatment of patients with spinal neoplasms. Spinal Cord 1997; 35(7):429–436. 76. Rompe JD, Hopf CG, Eysel P. Outcome after palliative posterior surgery for metastatic disease of the spine—evaluation of 106 consecutive patients after decompression and stabilisation with the Cotrel-Dubousset instrumentation. Arch Orthop Trauma Surg 1999; 119(7–8):394–400. 77. Sherman RM, Waddell JP. Laminectomy for metastatic epidural spinal cord tumors. Posterior stabilization, radiotherapy, and preoperative assessment. Clin Orthop Relat Res 1986; (207):55– 63.

454

g. zada and m.y. wang

78. Klimo P Jr, Schmidt MH. Surgical management of spinal metastases. Oncologist 2004; 9(2):188–196. 79. Harrington KD. Anterior decompression and stabilization of the spine as a treatment for vertebral collapse and spinal cord compression from metastatic malignancy. Clin Orthop Relat Res 1988; (233):177–197. 80. Bilsky MH, Boland P, Lis E, et al. Single-stage posterolateral transpedicle approach for spondylectomy, epidural decompression, and circumferential fusion of spinal metastases. Spine 2000; 25(17):2240–2249, discussion 2250. 81. Wise JJ, Fischgrund JS, Herkoitz HN, et al. Complication, survival rates, and risk factors of surgery for metastatic disease of the spine. Spine 1999; 24(18):1943–1951. 82. Klimo P Jr, Kestle JR, Schmidt MH. Treatment of metastatic spinal epidural disease: a review of the literature. Neurosurg Focus 2003; 15(5):e1. 83. Liu JK, Apfelbaum RI, Schmidt MH. Surgical management of cervical spinal metastasis: anterior reconstruction and stabilization techniques. Neurosurg Clin N Am 2004; 15(4):413–424. 84. Le H, Balabhadra R, Park J, Kim D. Surgical treatment of tumors involving the cervicothoracic junction. Neurosurg Focus 2003; 15(5):e3. 85. Kraus DH, Huo J, Burt M. Surgical access to tumors of the cervicothoracic junction. Head Neck 1995; 17(2):131–136. 86. Cahill DW, Kumar R. Palliative subtotal vertebrectomy with anterior and posterior reconstruction via a single posterior approach. J Neurosurg 1999; 90(1 Suppl):42–47. 87. Schmidt MH, Larson SJ, Maiman DJ. The lateral extracavitary approach to the thoracic and lumbar spine. Neurosurg Clin N Am 2004; 15(4):437–441.

88. Murakami H, Kawahara N, Abdel-Wanis ME, Tomita K. Total en bloc spondylectomy. Semin Musculoskelet Radiol 2001; 5(2): 189–194. 89. Thongtrangan I, Balabhadra RS, Le H, et al. Vertebral body replacement with an expandable cage for reconstruction after spinal tumor resection. Neurosurg Focus 2003; 15(5):e8. 90. Deramond H, Galibert P. Percutaneous acrylic vertebroplasty as a treatment of vertebral angioma as well as painful and debilitating diseases. Chirurgie 1990; 116(3):326–334. 91. Pilitsis JG, Rengachary SS. The role of vertebroplasty in metastatic spinal disease. Neurosurg Focus 2001; 11(6):e9. 92. Cotten A, Dewatre F, Cortet B, et al. Percutaneous vertebroplasty for osteolytic metastases and myeloma: effects of the percentage of lesion ﬁlling and the leakage of methyl methacrylate at clinical follow-up. Radiology 1996; 200(2):525–530. 93. Liebschner MA, Rosenberg WS, Keaveny KM. Effects of bone cement volume and distribution on vertebral stiffness after vertebroplasty. Spine 2001; 26(14):1547–1554. 94. Tohmeh AG, Mathis JM, Fenton DC, et al. Biomechanical efﬁcacy of unipedicular versus bipedicular vertebroplasty for the management of osteoporotic compression fractures. Spine 1999; 24(17):1772–1776. 95. Weill A, Chiras J, Simon JM, et al. Spinal metastases: indications for and results of percutaneous injection of acrylic surgical cement. Radiology 1996; 199(1):241–247. 96. Padovani B, Kasriel O, Brunner P, Peretti-Viton P. Pulmonary embolism caused by acrylic cement: a rare complication of percutaneous vertebroplasty. AJNR Am J Neuroradiol 1999; 20(3): 375–377.

4 5

Spinal Metastases: Fractionated Radiation Therapy Perspective Eric L. Chang and Almon S. Shiu

Introduction Approximately 40% of all cancer patients will develop metastatic spinal disease [1]. Some patients may be satisfactorily treated with conventional radiotherapy for palliation of pain and neurologic symptoms arising from spinal metastases. For other patients, a conventional dose of 30 to 45 Gy may be inadequate to achieve durable control. Furthermore, reirradiation of the spine with conventional means is rarely an option. The American Society for Therapeutic Radiology and Oncology (ASTRO) has described a new ﬁeld of stereotactic body radiation therapy (SBRT) in conjunction with the American College of Radiology (ACR). SBRT is a “newly emerging radiotherapy treatment method to deliver a high dose of radiation to the target, utilizing either a single dose or small number of fractions with a high degree of precision within the body. The ability to deliver a single or a few fractions of high dose ionizing radiation with high targeting accuracy and rapid dose falloff gradients encompassing tumors within a patient provides the basis for the development of SBRT [2].” This perspective explores the respective merits of single-dose and hypofractionated SBRT to the spine.

Fractionated Versus Single-Session SBRT SBRT to spinal tumors has the potential to expand on treatment options currently available. Potential indications for SBRT to the spine include primary treatment for a single or oligometastases in the spine, postoperative treatment, or salvage treatment after surgery or previous irradiation. SBRT has advantages, such as the ability to give higher doses to the spinal tumor while minimizing dose to the spinal cord. Before SBRT to the spine can be considered a viable treatment, clinical safety, which is predicated upon accurate and precise treatment, should be demonstrated. Whereas there is some ability to adjust or compensate for positional setup errors after hypofractionated SBRT, no such remedy exists to make corrections after single-session SBRT. This statement has profound implications for new investigators

choosing between fractionated and single-session SBRT. Thus, at the inception of our SBRT program at The University of Texas M. D. Anderson Cancer Center, we conservatively chose hypofractionated SBRT. Our conservative philosophy recognizes the reality that a learning curve exists with any new procedure that must be climbed by the entire treatment team while meeting the simultaneous requirement of safety. Single-fraction SBRT should be deferred until the entire team has sufﬁcient conﬁdence in the entire procedure. The treatment plans representing SBRT to a typical T10 vertebral body target volume prescribed to 18 Gy in 1 fraction, 27 Gy in 3 fractions, and 30 Gy in 5 fractions while limiting the spinal cord to 8 Gy, 9 Gy, and 10 Gy, respectively, are shown in Figure 45-1. On the other hand, single-fraction SBRT has the advantage of convenience to the patient and treatment team and is thus attractive, if and only if it can be delivered safely. The spinal cord tolerance may be better deﬁned for single-fraction SBRT based on extrapolation of data from radiosurgery to the optic chiasm of approximately 8 Gy making it easier to achieve complete target volume coverage. This can be appreciated by examining dose-volume histograms from treatment plans representing 30 Gy in 5 fractions, 27 Gy in 3 fractions, and 18 Gy in 1 fraction (Fig. 45-2). Furthermore, single-fraction SBRT follows the successful paradigm established in the arena of intracranial radiosurgery and may lead to superior local control. Single-fraction SBRT may be preferable for single tumors that are small in volume and anatomically favorable in terms of greater distance from the spinal cord. Fractionated SBRT may be preferable for spinal tumors with paraspinal components spanning multiple vertebral body levels or with large prevertebral components juxtaposed to small bowel, because the biological dose to surrounding normal tissues can be minimized to a greater degree. For tumors surrounding the spinal cord, involving the epidural space, or causing compression of the thecal sac, it may be prudent to request that the neurosurgeon resect the disease in close proximity to the spinal cord so that a more favorable anatomic conﬁguration can be obtained for SBRT. A comparison of modalities including single-fraction and hypofractionated SBRT, conventional radiotherapy, and surgery is shown in Table 45-1.

455

456

e.l. chang and a.s. shiu

FIGURE 45-1. Comparison of treatment plans for 1-, 3-, and 5-fraction SBRT.

Normalized volume (%)

100

T10GTV 1RX

80

3RX

Cord 60

5RX

40

1RX 3RX

20

5RX

0 0

1000

2000

3000

4000

Dose (cGy) FIGURE 45-2. Comparison of dose-volume histograms for 1-, 3-, and 5-fraction SBRT.

45.

457

spinal metastases: fractionated radiation therapy perspective

TABLE 45-1. Comparison of various spinal metastasis treatments. Single-fraction SBRT

Hypofractionated SBRT

Conventional radiotherapy

Surgery

Safety

Unforgiving. No possibility for correction after delivery of treatment; spinal cord tolerance extrapolated from optic chiasm; may be less appropriate in previously irradiated patients

30 Gy in 10 fractions has track record of safety in treating a large number of patients worldwide

Usual potential risks of surgery including wound infection, bleeding, damage to nerves, and spinal cord

Efﬁcacy

Theoretically may have greater ability than hypofractionated SBRT to overcome radioresistance especially for renal cell carcinoma, melanoma, and sarcoma by getting over shoulder of survival curve Ultimate in convenience to patient and treatment team, must be balanced against potential risks

Some compensation possible if initial setup error made; spinal cord tolerance for hypofractionated radiotherapy ill-deﬁned; good approach for setting up new programs during learning curve; may be safer for previously irradiated patients Needs to be prospectively compared with single-fraction SBRT and conventional radiotherapy

Mainstay for palliation of spinal metastases; limited by spinal cord tolerance so dose may be inadequate; reirradiation usually not possible

Rapid resolution of neurologic symptoms due to spinal cord compression or pain caused by mechanical instability

Greater expenditure of resources required for laborintensive procedure; may be required during learning curve or for previously irradiated patients near cord tolerance More costly due to multiple iterations of the same procedure; however, this cost may need to be incurred initially by new programs to ensure patient safety until track record for safety established

Can start immediately, easy to give, widely available

Postoperative recovery time, wound healing, rehabilitation time to achieve ambulation

Cost-effective and appropriate for majority of diffuse spinal metastases involving multiple levels

Costly but deﬁnite indications for spinal cord compression and salvaging previously irradiated metastases

Convenience

Cost-effectiveness

May be less costly than fractionated SBRT from a resource standpoint, and therefore more costeffective if efﬁcacy is proved to be equivalent

Published Clinical Results At selected centers, most SBRT programs are either in the process of being initiated or are still in their infancy. Therefore, the opportunity to collect large experiences or long-term clinical data has not yet occurred, but initial reports are emerging on fractionated SBRT. Bilsky et al. from Memorial SloanKettering reported on their initial clinical experience in treating 16 paraspinal tumors with intensity-modulated stereotactic radiotherapy. Metastatic tumors received a dose of 20 Gy in 4 to 5 fractions while limiting the spinal cord to a maximum dose of 6 Gy. Of the 15 patients who underwent radiographic followup, 13 demonstrated either no interval growth or a reduction in tumor size in a median follow-up period of 12 months (range, 2 to 23 months) [3]. Our group at The University of Texas M. D. Anderson Cancer Center reported on 15 consecutive patients with metastatic spinal disease who underwent 75 treatments involving 90 isocenter setups on a phase I clinical trial involving intensity-modulated, computed tomography (CT) image-guided SBRT. Patients uniformly received 30 Gy in 5 fractions while the spinal cord was constrained to a maximum dose of 10 Gy [4]. The procedure was technically feasible to perform in all patients, and no neurologic toxicity was observed in any patient with a median follow-up time of 9 months (range, 6 to 16 months). Because SBRT for spinal metastases is usually given in the context of palliation, and avoiding spinal cord damage is paramount, we initially chose to prescribe a hypofractionated dose of 30 Gy in 5 fractions on alternating days to allow for sufﬁcient sublethal damage repair while conservatively limiting the spinal cord dose to 10 Gy. It is important to note that treat-

ment planning programs approximate doses to critical structures and are limited by the accuracy of beam data taken by ion chamber measurements, which is affected by the size or volume of the ion chamber, and whether leakage between the leaves used in multileaf collimators and leaf shape are taken into account. Recognizing that treatment plans may over- or underestimate the actual doses delivered, it is prudent to conservatively limit doses to the spinal cord. A dose of 10 Gy in 5 fractions is considered clinically insigniﬁcant. The biologic equivalence (BED2Gy) of 30 Gy in 5 fractions is 40 Gy for earlyresponding tissues assuming an α/β of 10 Gy and 54 to 64 Gy assuming an α/β of 1.5 to 3 Gy for the spinal cord. With the increased conﬁdence of our safety and setup data demonstrating consistent daily patient setups within 1 mm of isocenter [4], we have proceeded to shorten our treatments to a total dose of 27 Gy in 3 fractions on alternating days to allow for sufﬁcient repair while limiting the spinal cord dose to 9 Gy. Ultimately, we plan to evaluate giving single-fraction SBRT to the spine for selected cases that have not been previously irradiated and have sufﬁcient space between disease and spinal cord. Dose escalation may be necessary in the future depending upon outcomes data and analysis of failure patterns. There are also several publications documenting experience with single-fraction SBRT. Ryu et al. reported on patterns of failure for 49 patients with 61 spinal metastases who underwent single-dose radiosurgery (10 to 16 Gy) at Henry Ford Hospital. Follow-up ranged from 6 to 24 months. Complete and partial pain relief was achieved in 85% of the lesions treated, but relapse of pain at the treated site was 7%, and there was progressive metastasis in the immediate adjacent spine in 5% [5].

458

e.l. chang and a.s. shiu

Benzil et al. treated 31 patients who had 35 tumors with stereotactic radiosurgery to the spine using the Novalis unit at New York Medical College. Twenty-six tumors were spinal metastases and 9 were primary tumors of the spine. In general, metastases not previously treated with radiation received 25 Gy in 10 fractions of external-beam radiation followed by a 6- to 8-Gy single-fraction boost. Patients previously irradiated with 30 Gy using conventional radiation were treated with 10 Gy in 2 fractions. Rapid and signiﬁcant pain relief was noted after treatment in 32 of 34 treated tumors within 72 hours and up to 3 months later. Two patients experienced transient radiculitis, and one patient had permanent neurologic deterioration due to pathologically conﬁrmed radiation necrosis. The authors concluded that a biological equivalent dose >60 Gy was associated with an increased risk of radiculitis [6]. Gerszten et al. from the University of Pittsburgh treated 115 consecutive patients with single-fraction radiosurgery using CyberKnife (45 cervical, 30 thoracic, 36 lumbar, and 14 sacral). There were 17 benign and 108 metastatic tumors. Tumor volume ranged from 0.3 to 232 cm3 and tumor dose was 12 to 20 Gy prescribed to the 80% isodose line (mean, 14 Gy). No acute radiation toxicity or new neurologic deﬁcits occurred during the follow-up period (range, 9 to 30 months; median, 18 months), and axial and radicular pain improved in 74 of 79 patients who were symptomatic before treatment. The authors concluded that the CyberKnife system was found to be feasible, safe, and effective [7]. DeSalles et al. from the University of California Los Angeles using the Novalis system treated 14 patients with 11 metastases, 2 neuroﬁbromas, and 1 meningioma. Twelve patients underwent previous conventional irradiation. A mean dose of 12 Gy (range, 8 to 21 Gy) was prescribed. With a mean follow-up of 6.1 months (range, 1 to 16 months), no complications were seen, and 7 patients experienced signiﬁcant pain relief. Seven of 22 lesions that were treated progressed [8]. Although this collective body of preliminary data is encouraging, it should be cautioned that because the ﬁeld is young, long-term complications are unknown. It is currently unknown to what extent high doses of radiation will affect vertebral stability [9]. It also remains to be deﬁned what is the threshold of spinal cord tolerance with hypofractionated or single-dose schedule of irradiation in the setting of primary or reirradiation.

Conclusion Reports of the relative safety of fractionated and single-fraction SBRT to the spine are beginning to be published from several institutions using different delivery systems. Larger experiences with long-term follow-up data are needed to more ﬁrmly establish the safety of high-dose SBRT to spinal tumors. Welldesigned prospective clinical investigations are needed to help address important clinical issues such as optimal dose prescription, integration of SBRT with conventional treatments, determination of spinal cord tolerance, and long-term sequelae of high doses of irradiation to the spine.

References 1. Klimo P, Schmidt MH. Surgical management of spinal metastases. Oncologist 2004; 9:188–196. 2. Potters L, Steinberg M, Rose C, et al. American Society for Therapeutic Radiology and American College of Radiology Practice Guideline for the Performance of Stereotactic Body Radiation Therapy. Int J Radiat Oncol Biol Phys 2004; 60:1026–1032. 3. Bilsky MH, Yamada Y, Yenice KM, et al. Intensity-modulated stereotactic radiotherapy of paraspinal tumors: a preliminary report. Neurosurgery 2004; 54:823–830. 4. 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 Radiat Oncol 2004; 59:1288–1294. 5. Ryu S, Rock J, Rosenblum M, et al. Patterns of failure after single-dose radiosurgery for spinal metastases. J Neurosurg 2004; 101(Suppl 3):402–405. 6. Benzil DL, Saboori M, Mogilner AY, et al. Safety and efﬁcacy of stereotactic radiosurgery for tumors of the spine. J Neurosurg 2004; 101(Suppl 3):413–418. 7. Gerszten PC, Ozhasoglu C, Burton SA, et al. CyberKnife frameless stereotactic radiosurgery for spinal lesions: clinical experience in 125 cases. Neurosurgery 2004; 55:89–98. 8. DeSalles AAF, Pedroso AG, Medin P, et al. Spinal lesions treated with Novalis shaped beam intensity-modulated radiosurgery and stereotactic radiotherapy. J Neurosurg 2004; 101(Suppl 3):435–440. 9. Rock JP, Ryu S, Yin F-F, et al. The evolving role of stereotactic radiosurgery and stereotactic radiation therapy for patients with spine tumors. J Neurooncol 2004; 69:319–334.

4 6

Arteriovenous Malformation Bruce E. Pollock

Introduction Intracranial arteriovenous malformations (AVMs) are congenital lesions arising from abnormal blood vessel formation [1–3]. Whereas normal embryogenesis results in the differentiation of primordial vascular channels into mature arteries, veins, and capillaries, patients with AVMs develop direct arteriovenous shunts without the appropriate intervening vascular beds. Recent large, prospective, population-based studies have determined the incidence of newly diagnosed AVM patients to range from 1.12 to 1.34 per 100,000 person-years [4, 5]. The majority of patients become symptomatic in the second through fourth decades of life with the most common presentation being intracranial hemorrhage (ICH). Patients may also have seizures or headaches, and the number of incidentally discovered intracranial AVMs continues to rise as more patients undergo magnetic resonance imaging (MRI) of the head. The estimated annual risk of ICH from AVMs has ranged from 2% to 4% [5–12]; the combined annual morbidity and mortality from intracranial AVMs is approximately 1% [11]. Because most AVM patients are diagnosed at a point when their life expectancy is long, the cumulative hemorrhage risk is substantial. For example, a 30year-old person carries approximately a 75% lifetime chance of ICH. Factors associated with an increased risk of bleeding include prior hemorrhage [9, 10, 12, 13], increasing age [14, 15], single or deep draining veins [12, 13], associated arterial aneurysm [16], and a diffuse AVM nidus [12]. Pediatric patients [17], patients with AVMs located in the basal ganglia, thalamus, brain stem, or cerebellum [12, 18], and patients with small AVMs [19] are more likely to present after a hemorrhage. It remains uncertain whether these patients actually carry an increased annual hemorrhage risk or whether they are just unlikely to develop other symptoms (seizures or headaches) that would permit the diagnosis.

AVM Management The management options for patients diagnosed with intracranial AVMs include observation, surgical resection, and stereotactic radiosurgery (Fig. 46-1). Embolization of AVMs is frequently performed in conjunction with either surgical resection or radiosurgery but is rarely curative in and of itself.

Conventional fractionated radiation therapy has been utilized in the past but results in a low AVM cure rate [20]. As mentioned before, the estimated risk of hemorrhage from intracranial AVMs has traditionally been considered from 2% to 4% annually, with an elevated risk of rebleeding in the ﬁrst months after a bleed. Patients who present after a large ICH require surgical evacuation to remove the mass effect associated with the hematoma. In some instances, surgery is performed before the patient has undergone cerebral angiography to delineate the AVM morphology. Once the blood clot has been removed and abnormal vessels discovered, most neurosurgeons recommend stopping surgery and performing angiography to determine a deﬁnitive plan for the residual nidus. After a recent hemorrhage, surgical resection is the preferred treatment for patients with accessible AVMs even if they do not require immediate clot evacuation. The beneﬁt of surgical resection compared with radiosurgery is the immediate elimination of future hemorrhage risk. The Spetzler-Martin grading system is the most frequently utilized scale to predict outcomes after the surgical excision of cerebral AVMs [21]. This grading scale is based on AVM size, location, and the presence or absence of deep venous drainage and has been proved over the past two decades to accurately predict patient outcomes after AVM surgery in a number of large AVM series [22–27]. As an alternative to surgical resection, stereotactic radiosurgery has been shown to be a safe and effective method to manage patients with cerebral AVMs. In 1972, Dr. Ladislau Steiner and colleagues from the Karolinska Institute recognized that single-fraction, high-dose irradiation caused the progressive obliteration of AVMs and subsequent cure from the risk of later hemorrhage [28]. In fact, because AVMs could be visualized through angiography before the development of axial imaging, AVMs were a common indication treated in the early radiosurgical series of Leksell. From 1968 until 1982, 204 of the ﬁrst 762 (27%) patients having Gamma Knife radiosurgery had AVMs [29]. Concurrent with the work of Leksell and Steiner, Kjellberg and Fabrikant were using heavy charged particles instead of photons to irradiate AVMs [30, 31]. Similar to the Swedish experience, these innovators noted that focused radiation techniques could obliterate a high percentage of irradiated AVMs. More recent reports have shown that modiﬁed linear accelerators (linacs) could also be used to perform AVM radio-

459

460

b.e. pollock

FIGURE 46-1. Treatment algorithm for intracranial AVMs.

surgery [32–34]. In fact, if one compares the reported number of patients having surgical resection or radiosurgery of their intracranial AVMs from centers of excellence over the past 30 years, it appears that radiosurgery is actually more commonly performed than surgical resection [35]. Observation is also a recognized management strategy for AVM patients especially if the AVM has not previously bled. The decision to treat an AVM is simply a comparison of the estimated risk of intervention to the calculated risk if the malformation is left untreated. If the risk of treatment is perceived to be less than the lifetime risk based on the natural history of untreated AVMs, then treatment is recommended. Prior to the advent of modern imaging, improved microsurgical techniques, and stereotactic radiosurgery, the majority of AVMs were not treated directly, and patients discovered to have AVMs were treated for their seizure disorder or headaches. Yet, once it was recognized that selected AVM patients could undergo treatment directed at the AVM with acceptable morbidity, the trend shifted so that surgical excision or radiosurgery was recommended for the majority of patients [36]. Despite technological advances, it is recognized that the risk of treatment for patients with large AVMs is not insigniﬁcant, and observation of Spetzler-Martin grade IV and V AVMs is generally recommended unless the patient has hemorrhaged or is suffering from a progressive neurologic deﬁcit [37]. However, two factors have led to a reexamination of observation for patients discovered to have smaller, unruptured brain AVMs [38]. First, the morbidity associated with AVM bleeding may be less than previously thought. A review of 119 AVM patients found that 47% of patients suffered no disability related to their ﬁrst ICH, and an additional 37% of patients remained independent in their activities of daily living [39]. Moreover, 20 of 27 (74%) patients having both an initial and follow-up hemorrhage maintained a Modiﬁed

Rankin Score (MRS) of zero or one. Second, neurologic deﬁcits after surgical resection of AVMs may be greater than previously described. Hartmann et al. reported outcomes for 124 AVM patients followed prospectively after surgical resection [40]. At last follow-up, 38% of patients had new postoperative neurologic deﬁcits. Six percent of patients were disabled after surgery. Factors related to a decline in neurologic function were female gender, AVM size, and deep venous drainage. Likewise, Lawton et al. found that patients with unruptured AVMs were 2.3 times more likely to experience a decline in their MRS compared with patients with ruptured AVMs [41]. Consequently, future studies on AVM management must carefully consider the risk-beneﬁt ratio associated with observation of small, unruptured AVMs.

AVM Radiosurgery Patient Selection Proper patient selection is essential for successful AVM radiosurgery. Once it has been determined that intervention is preferred over observation, several factors must be considered when considering radiosurgery for AVM patients. Most importantly, has the patient bled, and if so, how recently. Patients with a recent ICH and a surgically accessible AVM typically are best managed with surgical resection. However, patients with a recent hemorrhage and a surgically inaccessible AVM are generally good candidates for radiosurgery assuming that the AVM is not too large. Also, patients with a hemorrhage months or years earlier can certainly be considered for radiosurgery because they have passed the time when a re-hemorrhage is most likely to occur. In this situation, a comparison of the chance of AVM elimination without new deﬁcits risk between

46.

arteriovenous malformation

surgical resection and radiosurgery should be undertaken. Standardized scales such as the Spetzler-Martin grade [21] and the Pollock-Flickinger score [42] should be used to estimate the efﬁcacy of surgical resection and radiosurgery, respectively, for individual AVM patients. Of course, the success of any intervention must take into account the physician’s experience with a particular group of patients, rather than relying on published reports from experienced centers. Other considerations regarding treatment choice include the presence or absence of associated aneurysms, history of seizures, and need for pretreatment embolization. Aneurysms on feeding vessels of AVMs are likely caused by the stress on the arterial walls due to increased blood ﬂow to the AVM. Patients with associated aneurysms and surgically accessible AVMs can often have both operated in a single procedure, thereby immediately removing the hemorrhage risk. Patients with associated aneurysms and surgically inaccessible AVMs can be considered for pre-radiosurgical surgical clipping or endovascular treatment of their aneurysm, with radiosurgery performed during a separate later procedure. Conversely, if the associated aneurysm is small, radiosurgery alone is generally sufﬁcient because the aneurysm will likely either stabilize or regress as the blood ﬂow to the AVM decreases after radiosurgery. Another consideration when discussing surgical resection or radiosurgery for AVM patients is a history of seizures. Approximately 15% to 20% of AVM patients will have a seizure [5, 41, 43]. However, very few patients will have medically resistant epilepsy, and the use of standardized scales to determine seizure frequency is uncommon in AVM papers. Piepgras et al. studied seizure outcomes after AVM surgery [44]. Preoperative seizure burden was deﬁned as less or more than four seizures prior to treatment. In the “low-seizure group” with a follow-up longer than 2 years, 93% were seizure-free, 2% improved but continued to have seizures, and 5% had worsening of their seizures. In patients with more than four seizures preoperatively, seizure freedom measured at 2 years of follow-up or longer was achieved in 76% of patients, 21% were improved, and 3% remained unchanged. Overall, 83% of patients remained seizure-free at last followup in this study. We retrospectively reviewed 65 AVM patients with one or more seizures having radiosurgery at our center between 1990 and 1998 [45]. Twenty-six (51%) patients were seizure-free (aura-free) after radiosurgery at 3-year follow-up; 40 (78%) patients had an excellent outcome (nondisabling simple partial seizures only) at 3-year follow-up. Factors associated with seizure-free or excellent outcomes include a low seizure frequency score (<4) before radiosurgery and smaller size and diameter AVM. Twenty-three patients had intractable partial epilepsy prior to treatment. Twelve of 23 (52%) and 11 of 18 (61%) patients with medically intractable partial epilepsy had excellent outcomes at years 1 and 3, respectively. Hoh et al. reviewed 110 patients with seizures undergoing AVM treatment [46]. Patients with a short seizure history, seizures related to an ICH, surgical resection, and complete AVM obliteration were more likely to have an Engel class I outcome. However, selection bias with regard to the treatment given prevents any signiﬁcant conclusion as to whether surgical resection or radiosurgery correlates with improved seizure outcomes.

461

Technique The goal of stereotactic radiosurgery is to accurately deliver a high dose of radiation to an imaging-deﬁned target in a single fraction [47]. To accomplish this goal, placement of a stereotactic head frame is needed to ensure rigid ﬁxation and minimize patient movement during imaging and radiation delivery. Head-frame placement for adults is performed under local anesthesia supplemented by a low dose of benzodiazepine. Patients less than 16 years typically require general anesthesia for radiosurgery. Once the stereotactic head frame has been placed, most patients undergo a post-gadolinium MRI and biplanar cerebral angiography. Reliance on angiography alone for radiosurgical dose planning increases the chance of treating too much adjacent normal brain tissue due to the often irregular shape of AVMs [48]. In addition, for AVMs located in the posterior fossa and lateral temporal regions, the chance of not including a portion of the nidus in the prescription isodose volume (PIV) is greater secondary to the difﬁculty in visualizing the AVM clearly on angiography. The addition of MRI minimizes the chance of such marginal or geographic misses. More recently, we have become increasingly conﬁdent not in excluding MRI, but rather eliminating the angiogram for dose planning purposes [49]. Patients with relatively small, compact, hemispheric AVMs with simple venous drainage are ideal candidates for using MRI alone for dose planning. Currently, approximately 20% of our patients do not have cerebral angiography as part of the imaging performed for dose planning. Nonetheless, we continue to insist that patients have a complete diagnostic angiogram, including appropriate external carotid injections, before any decision is made about the feasibility of radiosurgery and to determine if the patient has any associated aneurysms. Dose planning is performed after appropriate imaging has been obtained and imported into the computer workstation. The goal is to create a conformal dose plan that precisely covers the three-dimensional shape of the nidus. Feeding arteries and draining veins are not included in the dose plan if possible. Inclusion of these vessels will increase the volume covered, which may decrease the radiation dose prescribed. Dose prescription must take into account two conﬂicting considerations: the chance of obliteration versus the chance of radiation-related complications. Increasing radiation dose directly correlates with the chance of obliteration. Assuming that the radiation is well targeted, the chance of AVM cure is approximately 70%, 80%, and 90% for radiation doses of 16 Gy, 18 Gy, and 20 Gy, respectively [50, 51]. However, the likelihood of radiationrelated complications after AVM radiosurgery increases at higher radiation doses and larger AVM volumes. Early dose prescription generally followed either Kjellberg’s 3% isodose line [30] or Flickinger’s integrated logistic formula [52] to predict the probability of radiation-related complications. More recent studies have correlated the chance of radiation-related complications after AVM radiosurgery relates to some measure of the radiation dose to the surrounding tissue [53–55]. Patients with AVMs in the thalamus, basal ganglia, and brain stem are more likely to develop neurologic deﬁcits secondary to imaging changes noted on MRI [53]. Also, the prescribed dose is typically reduced for patients having repeat AVM radiosurgery to minimize the chance of radiation-related complications.

462

b.e. pollock

After the dose plan is reviewed by all members of the radiosurgical team and found to be acceptable, the patient undergoes radiation delivery. After radiation delivery, the patient is discharged home either the day of the procedure or after an overnight observation period. Immediate complications are rare. Some patients complain of pin site discomfort, neck pain, or headache, but these are usually temporary and can be managed with over-the-counter medications. Follow-up after radiosurgery typically consists of clinical examination and MRI at 1, 2, and 3 years after radiosurgery. If MRI suggests that the AVM has gone on to complete obliteration, then follow-up angiography is recommended 3 or more years after radiosurgery to deﬁnitively determine the status of the AVM [56]. Patients with residual AVM on follow-up angiography are evaluated for repeat radiosurgery or surgical resection based on their age, clinical condition, and the AVM response from the ﬁrst radiosurgical procedure.

Obliteration After Radiosurgery The primary goal of AVM radiosurgery is complete nidus obliteration to eliminate a patient’s risk of future hemorrhage. Lindquist and Steiner in 1988 deﬁned AVM obliteration as follows: “we have considered a result satisfactory only when the arteriogram has shown a normal circulation time, complete absence of pathological vessels in the former nidus of the malformation, and the disappearance or normalization of draining veins from the area” [57]. Generally, AVM obliteration occurs between 1 and 5 years after radiosurgery. Sequential histopathologic changes after AVM radiosurgery include early damage to the endothelial cells, followed by progressive thickening of the intimal layer secondary to proliferation of smooth muscle cells that produce an extracellular matrix containing type IV collagen, then cellular degeneration and hyaline transformation [58]. The effect of high-dose single-fraction radiation has been studied in cell cultures taken from patients undergoing AVM resection [59]. Within 5 days after irradiation, the proliferation index decreased and remained decreased over the observation period. Other notable ﬁndings were immunohistochemical evidence of apoptosis and transformation of ﬁbroblastic cells into activated myoﬁbroblasts. Histologic and electron microscopic studies of seven AVMs resected after bleeding 10 to 52 months after radiosurgery revealed spindle cell proliferation in the connective tissue stroma and in the subendothelial region of irradiated vessels [60]. It was concluded that the characteristics of the spindle cells were similar to myoﬁbroblasts noted during wound healing, and these cells likely contributed to the occlusive process and ﬁnal obliteration of AVMs after radiosurgery. The most important factor associated with obliteration after AVM radiosurgery is the margin dose delivered to the nidus [50, 51, 61–64]. Several authors have worked on models to predict the chance of AVM cure after radiosurgery. Karlsson et al. reported the K index as a method to predict obliteration after AVM radiosurgery [51]. Based on 945 AVM patients having radiosurgery from 1970 until 1990, they found a logarithmic relationship between minimum dose and AVM obliteration increasing to a maximum of 87%. A higher average

dose also shortened the latency to AVM obliteration. The correlation of obliteration rate to the product minimum dose (AVM volume)1/3 was termed the K index. The obliteration rate increased linearly with the K index up to a value of approximately 27, and for higher K values, the obliteration rate had a constant value of approximately 80%. For the group of patients receiving an AVM margin dose of at least 25 Gy (n = 273), the obliteration rate at 2 years was 80%. If obliterations that occurred beyond 2 years are included, the obliteration rate increased to 85%. Schwartz et al. developed the obliteration prediction index (OPI) as a method to predict success or failure for individual AVM patients [65]. Analyzing a total of 426 patients having either Gamma Knife or linac-based radiosurgery, a relationship was noted between the calculated OPI (AVM margin dose, Gy/lesion diameter, cm) and AVM obliteration. Although both the K index and the OPI correlate with AVM elimination, neither takes into account the likelihood of producing radiation-related complications for a given radiation dose. Perhaps more instructive than studies analyzing factors relating to AVM obliteration has been a number of papers that have reviewed patients with incompletely obliterated AVMs after radiosurgery [66–69]. The most common reasons for incomplete nidus obliteration are targeting errors, recanalization of a portion of the AVM that was previously embolized, re-expansion of nidus after hemorrhage, and low radiation dose. These studies have emphasized the need for complete nidus coverage at the time of radiosurgery. The routine use of computed tomography (CT) or MRI in conjunction with angiography has been the most important change to reduce the chance of “missing” a portion of the AVM with the prescription isodose volume [48, 49]. Certainly, part of the problem in AVM radiosurgery is deﬁning the nidus accurately. Buis et al. had six independent clinicians contour the nidus of AVM patients based off digital subtraction angiography (DSA) [70]. They noted signiﬁcant interobserver variation when outlining the nidus and concluded that this may contribute to failure in some AVM radiosurgical cases. Despite the role of each imaging modality, there has been a general trend to forego angiography for AVM dose planning whenever possible. Yu et al. compared dose plans based on a combination of angiography and MRI to those derived from MRI alone [71]. They concluded that MRI-based dose planning without angiography should be limited to patients with smaller AVMs and compact niduses. Other considerations to minimize the chance of incomplete AVM obliteration are waiting for hematoma absorption for patients with ruptured AVMs and to be sure that all possible feeding arteries are injected at the time of radiosurgery, including the external carotid arteries for peripherally located AVMs. Still, a certain percentage of AVMs will fail to obliterate despite proper target delineation, complete radiation coverage, and adequate dose (radiobiological resistance) [69]. A recent study from the University of Florida has suggested that AVM morphology is also an important factor associated with obliteration [64]. Speciﬁcally, they noted patients with a diffuse nidus structure and associated neovascularity were at a higher risk of incomplete nidus obliteration when compared with patients with compact AVMs.

46.

arteriovenous malformation

Hemorrhage After Radiosurgery The primary drawback of AVM radiosurgery when compared with surgical resection is that the patient remains at risk for hemorrhage until the AVM has gone onto complete obliteration (Fig. 46-2). The issue of AVM bleeding during this latency interval has been extensively studied. Early papers on AVM radiosurgery suggested that the risk of bleeding initially increased before AVM obliteration occurred [33, 72]. Colombo et al. reported 180 patients having linac-based AVM radiosurgery [33]. Twenty-seven patients with large AVMs underwent partial treatment; the mean follow-up in this series was 43 months. Fifteen patients bled after radiosurgery. In totally irradiated cases, the bleeding risk decreased from 4.8% in the ﬁrst 6 months after radiosurgery to 0 starting from 1 year after radiosurgery. Patients with partially irradiated AVMs had bleeding 4% to 10% over the ﬁrst 2 years, then no bleeding thereafter. A review of 65 patients with Spetzler-Martin grade I to II AVMs found that ﬁve (8%) patients sustained an ICH after radiosurgery [72]. Although the annual hemorrhage rate was 3.7% over the latency interval, all the observed hemorrhages occurred within the ﬁrst 8 months after the procedure. Despite these early papers suggesting an increased risk of bleeding after radiosurgery, larger and more detailed analysis of this question has conﬁrmed that the risk of bleeding is either unchanged [73–75] or decreased [76, 77] after AVM radiosurgery. Maruyama et al. performed a retrospective observational study of 500 AVM patients having radiosurgery [77]. Comparing the risk of bleeding before and after radiosurgery, they found

463

a 54% reduction in the bleeding risk during the latency interval. The risk reduction was greatest among patients who presented with a hemorrhage. Likewise, Karlsson et al. analyzed the large AVM experience at the Karolinska Institute and found that some measure of protection occurred as early as 6 months after radiosurgery for patients receiving an AVM margin dose of 25 Gy [76]. Other than patient presentation, the presence of nidal or associated aneurysms has also been associated with an increased risk of ICH after AVM radiosurgery [74, 75]. Despite the general contention that the risk of bleeding after angiographic obliteration is zero, reports of bleeding after presumed AVM have been reported [77–80]. What are the salient points to be learned from these four papers outlining a total of eight AVM patients who bled after angiography showed complete obliteration? First, AVMs in the pediatric population are dynamic lesions that can recur after complete obliteration after either resection or radiosurgery. For this reason, follow-up angiography is indicated for these patients when they reach adulthood. Second, the risk of hemorrhage after angiographic obliteration is extraordinarily small. In the series of Shin et al. [79], the crude risk was approximately 2%. Yet, if one considers that almost 25,000 patients underwent Gamma Knife AVM radiosurgery through 1999 worldwide (not including the patients having linac-based or heavy-particle AVM radiosurgery), it becomes readily apparent that the risk is signiﬁcantly less than 1 in 50. Third, if bleeding does occur, the clinical sequelae are generally minimal. Last, and most important, is what exactly are we discussing? Are we talking about acute blood clots causing mass effect on the adjacent brain, small regions of methemoglobin detected on MRI, or hemosiderin deposition at the site of the obliterated AVM (chronic encapsulated hematoma). In fact, some of the reported cases of delayed hemorrhage are remarkably similar to patients we have described as having cyst formation or other late radiation-related complications after AVM radiosurgery (Fig. 46-3) [81]. The pathology description of one of our patient’s resected AVM was “arteriovenous malformation with hemorrhage, hemosiderinophages, and gliosis consistent with radiation,” yet we have not considered that patient to have bled after angiographic obliteration. Other similarities to our patients are the relative ease of surgical resection, minimal clinical consequences, and the relative incidence (2% for our patients having AVM radiosurgery before 2000). It is possible that we are discussing the same process, but deﬁning it in different ways.

Radiation-Related Complications

FIGURE 46-2. Neuroimaging studies of a 26-year-old woman with an incidentally discovered left basal ganglia AVM. (A) Lateral and (B) anterior-posterior (B) left internal carotid angiograms showing AVM on the day of radiosurgery. The AVM volume was 6.5 cm3; the AVM margin dose was 15 Gy. (C) CT performed 6 months after radiosurgery showing both intraparenchymal and intraventricular hemorrhage.

Advances in imaging and radiosurgical devices have signiﬁcantly improved patient outcomes after stereotactic radiosurgery. Nonetheless, the tissue adjacent to a radiosurgical target does receive a dose of radiation. Assuming that the dose plan conforms to the three-dimensional shape of the target, the major factors that determine the amount of radiation delivered to nearby structures are the PIV and the prescribed radiation dose. As the PIV increases, the fall-off of radiation becomes less steep, resulting in more radiation to the adjacent brain. It is this general principle that typically limits the size of intracranial radiosurgical targets to 3 cm or less in average diameter. Likewise, there is a direct correlation between higher radiosurgery

464

b.e. pollock

doses and the dose received by nearby tissues. For example, assume that a dose plan is created that covers an AVM at the 50% isodose line and the 20% isodose line is in contact with the optic nerve. A maximum radiation dose of 40 Gy would limit the optic nerve exposure to 8 Gy, whereas a maximum dose of 50 Gy exposes the optic nerve to a maximum dose of 10 Gy. Recognition of the relationship between the overall radiation exposure and the chance of radiation-related complications has resulted in studies correlating imaging changes on MRI to measures of radiation such as the 10 Gy volume (V10) [55], the 12 Gy volume (V12) [53], and the mean dose received by 20 cm3 surrounding the maximum radiation point (Dmean20) [54]. Areas of increased signal on long-TR MRI are noted in approximately 30% to 50% of AVM patients after radiosurgery. The chance of developing these imaging changes increases at larger V10, V12, or Dmean20. Levegrun et al. studied the correlation of radiation-induced imaging changes and dose distribution parameters in 73 patients having AVM radiosurgery [82]. The AVM target volume correlated signiﬁcantly with the development of large edema and breakdown of the blood-brain barrier. They concluded that each measure studied (including V10, V12, and Dmean20) yielded similar results, and no parameter was favored over the others. Yet, although such imaging changes are frequently noted within the ﬁrst year after AVM radiosurgery, the majority of patients remain asymptomatic. Comparison of AVM patients with non-AVM patients has demonstrated that the incidence of these imaging changes is greater for AVM patients. Consequently, it is believed that many “radiation-

FIGURE 46-4. (Left) Post-gadolinium and (right) long-TR MRIs performed 3 years after radiosurgery of a left sylvian ﬁssure AVM demonstrates persistent enhancement and edema consistent with radiation necrosis.

related” imaging changes related not to radiation damage per se but rather to alterations in the regional blood ﬂow in the brain adjacent to the AVM. Early closure of draining veins before occlusion of the nidus may be a signiﬁcant factor that contributes to the development of such imaging changes [83, 84]. Patients with AVMs in the thalamus, basal ganglia, and brain stem are more likely to develop neurologic deﬁcits secondary to imaging changes noted on MRI [53, 85]. Although most imaging changes detected on MRI after AVM radiosurgery are temporary and resolve by 2 years, a small percentage of patients will continue to demonstrate imaging changes consistent with radiation necrosis (Fig. 46-4). MRI ﬁndings consistent with radiation necrosis include persistent enhancement at the irradiated site beyond 2 years with associated edema and mass effect. Again, the chance that a patient will have symptoms related to radiation necrosis relates primarily to the location of the AVM. In addition to radiation necrosis, other late complications have been noted after AVM radiosurgery including diffuse white matter changes [86], cyst formation [80, 81, 87], and stenosis of major intracranial vessels [88]. The most devastating complication after any radiationbased procedure is a radiation-induced neoplasm. To date, only one case of a radiation-induced tumor has been reported after AVM radiosurgery [89]. In that patient, a glioblastoma multiforme developed at the site of a previously irradiated AVM. Although the precise incidence of this complication will not be known for years, it is estimated that the risk of a radiationinduced tumor after radiosurgery will be signiﬁcantly less than commonly associated with fractionated radiation therapy [90].

Repeat AVM Radiosurgery

FIGURE 46-3. MRIs of a 35-year-old woman who underwent linacbased radiosurgery 7 years earlier for an incidentally discovered left frontal AVM. The patient developed increasing headaches, seizures, and syncope. (A, B) Post-gadolinium and long-TR images showing cyst formation, edema, and mass effect. Angiography revealed complete obliteration. (C, D) Post-gadolinium and long-TR images 2 months after surgical resection of the obliterated AVM showing resolution of edema and mass effect.

Patients with residual nidus after AVM radiosurgery remain at risk for ICH. A number of studies have analyzed the results of repeat AVM radiosurgery [91–94]. Karlsson et al. reviewed 112 patients and tested whether models developed after one radiosurgery also predicted obliteration and complication rates after repeat AVM radiosurgery [92]. Sixty-two of 101 patients with angiographic follow-up had complete obliteration (predicted number of obliterations, 65). Fourteen (13%) patients had radiation-related complications after a second radiosurgical

46.

arteriovenous malformation

procedure. They concluded that the obliteration rate after repeat radiosurgery is similar to primary procedures, but the complication rate increases with the overall amount of radiation given. Maesawa et al. noted complete obliteration in 21 of 30 (70%) patients with angiographic follow-up after repeat AVM radiosurgery [93]. Permanent adverse radiation effects were noted in two (5%) patients; the annual risk of hemorrhage was 1.6%. It was their opinion that similar or higher radiation doses were required to achieve the same obliteration rates noted in AVMs that had never been radiated before. Foote et al. analyzed 52 patients undergoing repeat AVM radiosurgery at the University of Florida between 1991 and 1998 [91]. The mean volume at ﬁrst radiosurgery was 13.8 cm3 compared with 4.7 cm3 at the time of repeat radiosurgery (66% volume reduction). The median radiation doses on initial and second procedures were 12.5 Gy and 15 Gy, respectively. The cure rate after repeat radiosurgery was 60%. Schlienger et al. noted complete obliteration in 19 of 32 (59%) patients having repeat AVM radiosurgery [94]. Overall, repeat radiosurgery should be considered a safe and effective option for the majority of patients with subtotal AVM obliteration after their initial radiosurgical procedure.

Radiosurgery of Large AVMs Although AVM obliteration is signiﬁcantly related to higher radiation doses, dose prescription must also take into account the likelihood of radiation-related complications. Studies on the dose-volume relationship of AVM post-radiosurgical radiationrelated complications have demonstrated a high morbidity for

Case Study 46-1 A 10-year-old boy complained of headaches that continued to increase despite medical therapy for migraines. In addition, he developed episodes of numbness and tingling involving his left hand and arm that also were increasing in frequency and intensity. Of note, the boy was left-handed. An MRI of his head revealed a large AVM located in the temporal, frontal, and parietal lobes with extension into the lateral portion of the thalamus (Fig. 46-5). The AVM measured 4.5 cm × 4.2 cm × 3.5 cm. There was no evidence of either recent or remote hemorrhage. An electroencephalogram showed diffuse slowing in the region of the AVM but no distinct epileptiform activity. A cerebral angiogram showed the AVM had arterial supply from the right anterior, middle, and posterior cerebral arteries. The AVM had both superﬁcial and deep venous drainage. There were no intranidal or feeding vessel aneurysms and no evidence of venous outﬂow restriction. The case was reviewed at our combined cerebrovascular conference, and the options of observation, surgical resection, and radiosurgery were discussed. It was believed that the AVM was becoming symptomatic, most likely from a “steal phenomenon,” so treatment of the AVM was recommended. The risk of neurologic injury from surgical resection of this Spetzler-Martin grade IV AVM in the patient’s dominant hemisphere was considered high, especially as the AVM

465

large AVMs [53–55, 82]. Miyawaki et al. reported that after linear accelerator–based radiosurgery of AVMs greater than 14 cm3 receiving 16 Gy or more, the incidence of post-radiosurgical abnormalities on long-TR MRI was 72% [95]. The incidence of radiation necrosis for these patients was 22%. Consequently, radiosurgery is generally performed only for patients with AVMs of an average diameter of 3 cm or less (approximately 14 cm3). Alternative strategies that utilize radiation in the management of large AVMs include embolization followed by radiosurgery, fractionated radiation techniques, and staged-volume radiosurgery (Case Study 46-1). Planned embolization and radiosurgery has been used for many years to manage patients with large AVMs as an alternative to surgical resection for these difﬁcult lesions [96–101]. As opposed to embolization prior to resection where ﬂow reduction is the goal, the goal of pre-radiosurgical embolization is permanent volume reduction. Gobin et al. published the results in 96 patients undergoing acrylate embolization followed by radiosurgery [96]. Complete AVM obliteration was documented in 53 of 90 (59%) evaluable patients. Sixteen (13%) patients had complications from the embolization procedures, and two patients died of intracerebral hemorrhages prior to having radiosurgery. Recanalization of the AVM was seen in 14% of patients. A similar rate of recanalization (15%) was noted after particle embolization prior to radiosurgery in an evaluation on the causes of failed AVM radiosurgery [69]. Wikholm et al. reported the complications associated with AVM embolization as a primary treatment or in preparation for either surgical resection or radiosurgery [100, 101]. Eighty-ﬁve percent of patients had grade III or higher AVMs. The incidence of complications was

was unruptured. Because of large size (estimated volume ≈30 to 35 cm3) and location of the AVM, the Pollock-Flickinger AVM score was estimated to exceed 3.5. Therefore, radiosurgery of the entire lesion in a single procedure was also considered a poor option. After discussing the options of planned embolization followed by radiosurgery or staged-volume radiosurgery, the patient’s family consented to staged-volume radiosurgery. The procedures were performed under general anesthesia using stereotactic MRI and angiography for dose planning. The ﬁrst radiosurgery covered the medial portion of the AVM (Fig. 46-6). The dose plan consisted of 11 isocenters of radiation to cover a volume of 15.5 cm. The margin dose was 15 Gy and the maximum dose was 30 Gy. The patient returned 4 months later and underwent a second radiosurgery to cover the lateral portion of the AVM (Fig. 46-7). The dose plan consisted of 13 isocenters of radiation to cover a volume of 17.0 cm. The margin dose was 15 Gy and the maximum dose was 30 Gy. The patient had a reduction in his headaches and remained neurologically intact after the second radiosurgery. MRI performed 6 months after the second procedure showed the AVM to be signiﬁcantly smaller with a small amount of edema surrounding the irradiated nidus (Fig. 46-8). Continued follow-up each year with MRI is planned until the fourth or ﬁfth year, and then angiography will be performed to determine the ﬁnal status of the AVM.

466

b.e. pollock

FIGURE 46-7. (A) Lateral and (B) anterior-posterior right carotid angiograms showing dose plan for the second radiosurgery covering the medial portion of the AVM.

FIGURE 46-5. Axial post-gadolinium MRI of a 10-year-old boy with a large right AVM located in the temporal, frontal, and parietal lobes with extension into the lateral portion of the thalamus.

high (40%), but only 7% were considered severe. The procedural mortality rate was 1.3%. Overall, embolization was able to either completely eliminate or reduce the AVM to a size amenable to radiosurgery in two thirds of the patients. They concluded that the combined death and complication rate of 8% was equivalent to approximately 3.2 years of natural history for untreated AVMs. Fractionated radiation therapy has also been used to treat patients with large AVMs [20, 102–105]. Redekop et al. reported 15 patients with inoperable AVMs ranging from 1.5 cm to 6.5 cm in diameter who underwent conventional radiation therapy between 1955 and 1985 [104]. Fourteen of 15 patients received a total radiation dose of 40 Gy or more in 15 to 28 fractions. At a mean follow-up of 8.1 years, the annual hemorrhage rate was 3.3%, and later angiography conﬁrmed AVM obliteration in 2 of 12 (17%) patients. Karlsson et al. reviewed the outcomes of 28 AVM patients undergoing fractionated radiation therapy

FIGURE 46-6. (A) Lateral and (B) anterior-posterior right carotid angiograms showing dose plan for the ﬁrst radiosurgery covering the medial portion of the AVM. Injection of the vertebral artery ﬁlled an additional posterior component of the AVM not visualized on the carotid injection, but which was included in the dose plan.

between 1980 and 1985 [20]. The median volume treated was 78 cm3. Using a fractionation scheme of 42 Gy in 12 fractions, only two (8%) patients demonstrated angiographic cure. The annual hemorrhage rate after radiation therapy was 6%. Although the information on low-dose fractionation for AVM patients is rather limited, it appears that little protection is provided against future bleeding after this technique. More recently, a number of centers have examined the efﬁcacy of hypofractionated radiation schedules using higher radiation doses per fraction to treat patients with intracranial AVMs [102, 103, 105]. Lindvall et al. used several fractionation schemes (generally 30 to 35 Gy/5 fractions) to treat 36 AVM patients [103]. Two-year follow-up angiography showed that 48% of the patients were cured; angiographic obliteration rose to 76% on 5-year angiography. Although the median AVM volume was only 8.5 cm3 for the entire group, obliteration for patients with AVMs larger than 10 cm3 (n = 10) was 70% at last follow-up. Veznedarglu and colleagues from Jefferson Hospital managed 30 patients with large AVMs from 1995 to 1998 using

FIGURE 46-8. Axial long-TR MRI performed 6 months after the second radiosurgical procedure. The AVM is signiﬁcantly smaller with a small amount of edema surrounding the irradiated nidus.

46.

arteriovenous malformation

a combination of preradiation embolization and stereotactic radiation therapy (SRT) [105]. Patients received either 42 Gy in six fractions (n = 7) or 30 Gy in six fractions (n = 23). Angiographic obliteration was noted on 5-year follow-up angiography in 83% of patients in the 42-Gy group (5 of 6 patients) compared with only 22% of patients in the 30-Gy group (4 of 18 patients). However, the overall morbidity in the higher dose group was 43% (14% permanent deﬁcits). Chang et al. treated 33 AVM patients with SRT using a treatment regimen of 25 to 35 Gy in four daily fractions [102]. Interestingly, they maintained their patients in a stereotactic head frame for the duration of the treatment. Fifteen of the 33 (45%) patients had AVMs larger than 2.5 cm in diameter. The 5- and 6-year actuarial obliteration rates were 61% and 71%, respectively. One (3%) patient developed radiation necrosis after SRT. More information is needed to determine the most effective dosing regimen to achieve AVM obliteration at an acceptable rate of radiation-related morbidity. Recognition of the limitations of the various management options available to patients with large AVMs has encouraged some radiosurgical centers to begin staged-volume radiosurgery for these patients [106–108]. Volume staging of large AVMs into multiple radiosurgical sessions separated by several months allows a higher radiation dose to be delivered to the entire AVM volume while minimizing the radiation exposure to the adjacent brain. We previously performed a dosimetric study comparing staged-volume AVM radiosurgery with hypothetical single-session procedures for our ﬁrst 10 patients [108]. Stagedvolume radiosurgery decreased the V12 by an average of 11.1%, and the non-AVM V12 was reduced by an average of 27.2%. The accuracy of transforming the isocenter coordinates from the initial frame placement to subsequent frame placements has been conﬁrmed using both ﬁducial markers ﬁxed to the skull [109] and less invasive image transformation techniques using mutual information between different MRI data sets [110]. At our center, we utilize ﬁxed anatomic structures (trigeminal nerves, anterior and posterior commissures) as internal ﬁducials for the transformation of dose plans from one radiosurgical procedure to the next. A potential drawback of staged-volume AVM radiosurgery is that partial AVM irradiation may increase the bleeding risk by redistributing blood ﬂow within the AVM to nonirradiated and thus unprotected regions. Columbo et al. found the hemorrhage risk for partially treated AVMs was greater (7 of 27 patients, 26%) in comparison with completely irradiated AVMs (8 of 153, 5%) [33]. For patients with partially treated AVMs, the hemorrhage risk increased from 4% in the ﬁrst 6 months to more than 10% for months 6 to 18. The bleeding rate decreased to 5.5% from 18 to 24 months, then no bleeding was observed after 2 years in this group of patients. Although the authors stated that partial volume irradiation was performed in large AVMs to reduce the risk of radionecrosis, no additional information was provided in regard to history of prior hemorrhage, AVM size, or previous treatments that would allow a complete comparison between the different treatment groups. Mathematical modeling of blood ﬂow within AVMs after radiosurgery has also been performed [111, 112]. Lo calculated changes in ﬂow rates and pressure gradients for both step-function obliteration and with uniform luminal narrowing secondary to thickened vessel walls. It was demonstrated that the biomechanical stress

467

on the vessel walls within an AVM typically decrease after radiosurgery, even though a transient increase is possible under speciﬁc conditions. At his point, very little information is available on patient outcomes after staged-volume AVM radiosurgery, although it is likely that published reports on the safety and efﬁcacy of this technique will be published in the near future. By minimizing the radiation exposure to the adjacent brain, it is hoped that the incidence of delayed white matter changes or cyst formation after staged-volume radiosurgery of large AVMs will be decreased compared with single-session procedures.

Radiosurgery-Based AVM Grading System The Spetzler-Martin grading system has become widely accepted as an accurate method to predict patient outcomes after surgical resection of AVMs [21]. Composed of three components (AVM size, location, and pattern of venous drainage), this system has been validated prospectively [23] and by numerous cerebrovascular centers of excellence [24–27]. Although some authors have noted discrepancies between AVM grade and patient outcomes, especially in regard to grade III AVMs [22, 113, 114], the general consensus supports this grading scale as practical and reliable. Unfortunately, this grading scale does not appear to correlate with successful AVM radiosurgery [115, 116] (Fig. 46-9). This should not be surprising because it is insensitive to important factors such as AVM volume and location. For example, a 1-cm-diameter AVM has an approximate volume of 0.6 cm3, whereas a 3-cm-diameter AVM has an approximate volume of 14 cm3; the expected obliteration rates for these AVMs should be approximately 90% and 50%, respectively. Yet, both would be considered small (<3 cm) in the Spetzler-Martin system. Likewise, deep structures (thalamus, basal ganglia, or brain stem) and critical cortical areas are both considered “eloquent” brain regions. However, radiationrelated complications are more likely to occur in patients with deeply located AVMs compared with hemispheric malformations [53, 85]. Consequently, a valid instrument capable of accurately predicting outcomes after AVM radiosurgery is necessary to adequately compare the expected results of microsurgery and radiosurgery for individual AVM patients. Successful AVM radiosurgery results in complete nidus obliteration without new or worsened neurologic deﬁcits. Karlsson et al. reported the K index as a method to predict

FIGURE 46-9. Lateral carotid angiograms of two patients with Spetzler-Martin grade II AVMs. (A) A 32-year-old man with a 12.8 cm3 frontal AVM (Pollock-Flickinger score = 1.92; chance of obliteration without new deﬁcit after radiosurgery ≈50%). (B) A 35-year-old man with a 2.0 cm3 frontal AVM (Pollock-Flickinger score = 0.90; chance of obliteration without new deﬁcit after radiosurgery ≈95%).

468

b.e. pollock

obliteration after AVM radiosurgery [51]. In a similar fashion, Schwartz et al. proposed the OPI as a means to estimate the chance of AVM obliteration for individual patients [65]. Although both correlate with AVM obliteration after radiosurgery, neither takes into account the chance of producing radiation-related complications for a given radiation dose. Also, the K index and OPI are based on the radiation dosimetry used (AVM margin dose) at the time of treatment and not on patient and AVM characteristics alone. Therefore, our center in collaboration with the University of Pittsburgh developed a radiosurgery-based AVM grading system that accounts for these shortcomings and predicted the chance of successful, singlesession AVM radiosurgery based solely on patient and AVM variables [42]. The grading system was developed for AVM radiosurgery based on the multivariate analysis of 220 patients treated between 1987 and 1991 at the University of Pittsburgh. The dependent variable in all analyses was excellent patient outcomes (complete AVM obliteration without new neurologic deﬁcit). The grading scale was then tested on a separate set of 136 AVM patients treated between 1990 and 1996 at the Mayo Clinic. Overall, 121 of 220 (55%) of the Pittsburgh patients had excellent outcomes. Multivariate analysis found ﬁve variables related to excellent patient outcomes: AVM volume, patient age, AVM location, prior embolization, and number of draining veins. Regression analysis modeling permitted removal of two signiﬁcant variables (prior embolization, number of draining veins) and resulted in an equation to predict patient outcomes after AVM radiosurgery (Table 46-1). Seventy-nine of 136 (58%) Mayo Clinic patients had excellent outcomes. Testing of the radiosurgery-based AVM grading system on the Mayo patients showed the AVM score predicted patient outcomes after radiosurgery. All patients with an AVM score ≤1.0 had an excellent outcome compared with only 39% of patients with an AVM score >2.0. Consequently, despite signiﬁcant differences in preoperative patient characteristics and dose prescription guidelines at the two centers, the AVM grading system developed strongly correlated with patient outcomes after singlesession radiosurgery for both patient groups. Subsequent testing has demonstrated that the radiosurgerybased AVM grading system (Pollock-Flickinger score) appears to be a valid instrument for predicting patient outcomes after AVM radiosurgery. We have performed subset analyses and found signiﬁcant relationships between the AVM score and outcomes for patients with deep AVMs (thalamus, basal ganglia, and brain stem) as well as for pediatric AVM patients. Moreover, although it was created to predict outcomes after one TABLE 46-1. Radiosurgery-based AVM grading system. Factor

Coefﬁcient 3

AVM volume (cm ) Patient age (years) AVM location Frontal, temporal = 0 Parietal, occipital, intraventricular, corpus callosum, cerebellum = 1 Basal ganglia, thalamus, brain stem = 2 AVM score = (0.1)(AVM volume) + (0.02)(Patient age) + (0.3)(AVM location)

0.1 0.02 0.3

radiosurgery, it also correlates with excellent outcomes after overall radiosurgical management (one or more radiosurgical procedures) [117]. Other centers have evaluated this grading system and found it to reliably predict patient outcomes after radiosurgery of corpus callosum AVMs [118] and deep AVMs [119]. Notably, the series of Andrade-Souza et al. from the University of Toronto examining outcomes for patients with deep AVMs was based on linac-based radiosurgery. Higher radiosurgery-based AVM scores have also been found to correlate with an increased risk of neurologic complications after radiosurgery of patients with brain-stem AVMs [120]. We are currently in the process of evaluating whether the radiosurgerybased AVM score predicts neurologic decline after AVM radiosurgery, as well as the prospective testing of this grading system. I believe that this system allows a more accurate prediction of outcomes from radiosurgery to guide choices between surgical and radiosurgical management for individual AVM patients.

AVM Radiosurgery Compared with Microsurgery Much has been made about the debate regarding the best treatment for patients with AVMs less than 3 cm in diameter (Spetzler-Martin grades I to III) [25–27]. When interpreting these results of surgical resection and radiosurgery, it does appear that surgical resection provides cure from the risk of future hemorrhage more frequently than radiosurgery for these patients. Moreover, approximately 5% to 10% of patients suffer an AVM bleed after radiosurgery, and radiation-related complications also occur in approximately 5% of patients. If the chance of late radiation-related complications such as cyst formation or radiation-induced neoplasms are also factored into the equation, radiosurgery looks even less appealing as an option for AVM patients. These considerations are frequently cited by cerebrovascular neurosurgeons in making their argument for resection of cerebral AVMs. Conversely, proponents of radiosurgery state that it is impossible to directly compare the two techniques because the patient characteristics are dissimilar [42]. For example, although the percentage of patients with Spetzler-Martin grade III AVMs ranges from 18% to 39% in microsurgical recent series [23–27], only 4% to 11% of patients have AVMs located in the basal ganglia, thalamus, or brain stem. In contrast, between 20% and 35% of patients having radiosurgery have AVMs in these deep locations. Morgan et al. analyzed the results of surgery for 92 patients with AVMs supplied by the middle cerebral artery [114]. Morbidity for grades I to II (n = 56) and III (n = 26) was 2% and 31%, respectively. Of note, ﬁve of six (83%) patients with grade III AVMs supplied by lenticulostriate arteries suffered a new neurologic deﬁcit. In fact, many patients only undergo radiosurgery after they have been reviewed for consideration of AVM removal and the risk of new neurologic deﬁcit has been deemed unacceptable [121]. Thus, comparisons based solely on grading scales designed to predict patient outcomes after AVM resection without taking into account patient selection are fundamentally ﬂawed and misleading. Advances in endovascular techniques, microsurgical resection, and stereotactic radiosurgery have all contributed to improving outcomes for AVM patients treated over the past

46.

arteriovenous malformation

TABLE 46-2. Patient characteristics. Factor

Mean patient age (range) Location, no. (%) Lobar Cerebellum Thalamus Basal ganglia Brain stem Other Mean AVM volume (range) Mean AVM margin dose (range) Mean maximum AVM dose (range)

38.5 years (3–82) 263 (66%) 26 (7%) 35 (9%) 13 (3%) 12 (3%) 49 (12%) 6.4 cm3 (0.1–45.8) 18 Gy (14–25) 37 Gy (22–50)

several decades. To provide the best care for AVM patients, we should utilize every tool available to individualize their management taking into account the patient’s age, presentation, nidus size and morphology, AVM location, and patient preference.

Mayo Clinic Experience From January 1990 until September 2005, 389 patients underwent radiosurgery for 400 AVMs at the Mayo Clinic, Rochester, Minnesota. Nine patients had two AVMs, whereas one patient had three AVMs. Three hundred twenty-two (83%) patients had a single procedure, 44 (11%) patients had initial and repeat radiosurgery, and one patient underwent initial and repeat radiosurgery twice. Twenty-two (6%) patients underwent planned staged-volume radiosurgery for large AVMs in an attempt to minimize the chance of radiation-related complications. Three of these patients have undergone repeat radiosurgery. Overall, a total of 473 AVM radiosurgical procedures have been performed over this interval. Table 46-2 outlines the clinical and dosimetric characteristics of our AVM patients. Three hundred ﬁfty-ﬁve (91%) patients have undergone imaging 1 or more years after AVM radiosurgery. Obliteration has been conﬁrmed in 145 of 180 (81%) patients having angiography at an average of 40 months after radiosurgery. Obliteration on MRI has been noted in 47 additional patients at a mean of 33 months after radiosurgery. Additional information on these patients and their management has been outlined in a number of publications [42, 45, 81, 83, 85, 108, 117, 122]. Overall, we have found the majority of AVM patients are protected from the risk of future hemorrhage and continue their normal daily activities after radiosurgery. Late complications requiring treatment are rare but can occur many years after a patient is considered cured of his or her AVM.

References 1. McCormick WE. Pathology of vascular malformations of the brain. In: Wilson CB, Stein B, eds. Intracranial Arteriovenous Malformations. Baltimore: Williams and Wilkins, 1984:44–63. 2. Padget DH. The cranial venous system in man in reference to development, adult conﬁguration, and relation to arteries. Am J Anat 1956; 98:307–355. 3. Takashima S, Becker LE. Neuropathology of cerebral arteriovenous malformations in children. J Neurol Neurosurg Psychiatry 1980; 43:380–385.

469

4. Al-Shahi R, Bhattacharya JJ, Currie DG, et al. Prospective, population-based detection of intracranial vascular malformations in adults: the Scottish Intracranial Vascular Malformation Study (SIVMS). Stroke 2003; 34:1163–1169. 5. Stapf C, Mast H, Sciacca RR, et al. The New York Islands AVM Study: design, study progress, and initial results. Stroke 2003; 34: 29–33. 6. Brown RD Jr, Wiebers DO, Forbes G, et al. The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 1988; 68:352–357. 7. Crawford PM, West CR, Park YG, et al. Arteriovenous malformations of the brain: natural history in unoperated patients. J Neurol Neurosurg Psychiatry 1986; 49:1–10. 8. Fults D, Kelly DL Jr. Natural history of arteriovenous of the brain: a clinical study. Neurosurgery 1984; 15:658–662. 9. Graf CJ, Perret GB, Torner JC. Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg 1983; 58:331–337. 10. Itoyama Y, Uemura S, Ushio Y, et al. Natural course of unoperated intracranial arteriovenous malformations: study of 50 cases. J Neurosurg 1989; 71:805–809. 11. Ondra SK, Troupp H, George ED, et al. The natural history of symptomatic arteriovenous malformations of the brain: a 24-year follow-up assessment. J Neurosurg 1990; 73:387–391. 12. Pollock BE, Flickinger JC, Lunsford LD, et al. Factors that predict the bleeding risk of cerebral arteriovenous malformations. Stroke 1996; 27:1–6. 13. Mast H, Young WL, Koennecke HC, et al. Risk of spontaneous hemorrhage after diagnosis of cerebral arteriovenous malformation. Lancet 1997; 350:1065–1068. 14. Karlsson B, Lindquist C, Steiner L. The effect of gamma knife surgery on the risk of rupture prior to AVM obliteration. Minim Invas Neurosurg 1996; 39:21–27. 15. Stapf C, Khaw AV, Sciacca RR, et al. Effect of age on clinical and morphological characteristics in patients with brain arteriovenous malformations. Stroke 2003; 34:2664–2669. 16. Stapf C, Mohr JP, Pile-Spellman J, et al. Concurrent arterial aneurysms in brain arteriovenous malformations with haemorrhagic presentation. J Neurol Neurosurg Psychiatry 2002; 73:294–298. 17. Kondziolka D, Humphreys RP, Hoffman HJ, et al. Arteriovenous malformations of the brain in children: a forty year experience. Can J Neurol Sci 1992; 19:40–45. 18. Khaw AV, Mohr JP, Sciacca RR, et al. Association of infratentorial brain arteriovenous malformations with hemorrhage at initial presentation. Stroke 2004; 35:660–663. 19. Spetzler RF, Hargraves RW, McCormick PW, et al. Relationship of perfusion pressure and size to risk of hemorrhage from arteriovenous malformations. J Neurosurg 1992; 76:918–923. 20. Karlsson B, Lindqvist M, Blomgren H, et al. Long-term results after fractionated radiation therapy for large brain arteriovenous malformations. Neurosurgery 2005; 57:42–49. 21. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg 1986; 65:476–483. 22. de Oliveria E, Tedeschi, H, Raso J. Comprehensive management of arteriovenous malformations. Neurol Res 1998; 20:673–683. 23. Hamilton MG, Spetzler RF. The prospective application of a grading system for arteriovenous malformations. Neurosurgery 1994; 34:2–7. 24. Heros RC, Korosue K, Diebold PM. Surgical excision of cerebral arteriovenous malformations. Neurosurgery 1990; 26:570–578. 25. Pik JHT, Morgan MK. Microsurgery for small arteriovenous malformations of the brain: results in 110 consecutive patients. Neurosurgery 2000; 47:571–577. 26. Pikus HJ, Beach ML, Harbaugh RE. Microsurgical treatment of arteriovenous malformations: analysis and comparison with stereotactic radiosurgery. J Neurosurg 1998; 88:641–646.

470

b.e. pollock

27. Schaller C, Schramm J. Microsurgical results for small arteriovenous malformations accessible for radiosurgical or embolization treatment. Neurosurgery 1997; 40:664–674. 28. Steiner L, Leksell L, Grietz T, et al. Stereotaxic radiosurgery for cerebral arteriovenous malformations: report of a case. Acta Chir Scand 1972; 138:459–464. 29. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry 1983; 46:797–803. 30. Kjellberg RN, Hanamura T, Davis KR, et al. Bragg-peak protonbeam therapy for arteriovenous malformations. N Engl J Med 1983; 309:269–274. 31. Steinberg GK, Fabrikant JI, Marks MP, et al. Stereotactic heavycharged-particle Bragg-peak radiation for intracranial arteriovenous malformations. N Engl J Med 1990; 323:96–101. 32. Betti OO, Munari C, Rosler R. Stereotactic radiosurgery with the linear accelerator: treatment of arteriovenous malformations. Neurosurgery 1989; 24:311–321. 33. Columbo F, Pozza F, Chierego G, et al. Linear accelerator radiosurgery of cerebral arteriovenous malformations: an update. Neurosurgery 1994; 34:14–21. 34. Friedman WA, Bova FJ, Mendenhall WM. Linear accelerator radiosurgery for arteriovenous malformations: the relationship of size to outcome. J Neurosurg 1995; 82:180–189. 35. Pollock BE, Meyer FB. Evaluating arteriovenous malformation radiosurgery. J Neurosurg 2004; 101:390–392. 36. Heros RC, Yu YK. Is surgical therapy needed for unruptured arteriovenous malformations? Neurology 1987; 37:279–286. 37. Han PP, Ponce FA, Spetzler RF. Intention-to-treat analysis of Spetzler-Martin grades IV and V arteriovenous malformations: natural history and treatment paradigm. J Neurosurg 2003; 98: 3–7. 38. Choi JH, Mohr JP. Brain arteriovenous malformations in adults. Lancet Neurol 2005; 4:299–308. 39. Hartmann A, Mast H, Mohr JP, et al. Morbidity of intracranial hemorrhage in patients with cerebral arteriovenous malformation. Stroke 1998; 29:931–934. 40. Hartmann A, Stapf C, Hofmeister C, et al. Determinants of neurological outcome after surgery for brain arteriovenous malformations. Stroke 2000; 31:2361–2364. 41. Lawton MT, Du R, Tran MN, et al. Effect of presenting hemorrhage on outcome after microsurgical resection of brain arteriovenous malformations. Neurosurgery 2005; 56:485–493. 42. Pollock BE, Flickinger JC. A proposed radiosurgery-based grading system for arteriovenous malformations. J Neurosurg 2002; 96: 79–85. 43. Zhao J, Wang S, Li J, et al. Clinical characteristics and surgical results of patients with cerebral arteriovenous malformations. Surg Neurol 2005; 63:156–161. 44. Piepgras DG, Sundt TM Jr, Ragoowansi AT, et al. Seizure outcome in patients with surgically treated cerebral arteriovenous malformations. J Neurosurg 1993; 78:5–11. 45. Schäuble B, Cascino GD, Pollock BE, et al. Seizure outcomes after stereotactic radiosurgery for cerebral arteriovenous malformations. Neurology 2004; 63:683–687. 46. Hoh BL, Chapman PH, Loefﬂer JS, et al. Results of multimodality treatment for 141 patients with brain arteriovenous malformations and seizures: factors associated with seizure incidence and seizure outcome. Neurosurgery 2002; 51:303–309. 47. Pollock BE, Lunsford LD. A call to deﬁne stereotactic radiosurgery. Neurosurgery 2004; 55:1371–1373. 48. Blatt DR, Friedman WA, Bova FJ. Modiﬁcations based on computed tomographic imaging in planning the radiosurgical treatment of arteriovenous malformations. Neurosurgery 1993; 33: 588–595. 49. Kondziolka D, Lunsford LD, Kanal E, et al. Stereotactic magnetic resonance angiography for targeting in arteriovenous malformation radiosurgery. Neurosurgery 1994; 35:585–591.

50. Flickinger JC, Pollock BE, Kondziolka D, et al. A dose-response analysis of arteriovenous malformation obliteration by radiosurgery. Int J Radiat Onc Biol Phys 1996; 36:873–879. 51. Karlsson B, Lindquist C, Steiner L. Prediction of obliteration after gamma knife surgery for cerebral arteriovenous malformations. Neurosurgery 1997; 40:425–431. 52. Flickinger JC. An intregrated logistic formula for prediction of complications from radiosurgery. Int J Radiat Oncol Biol Phys 1989; 17:879–885. 53. Flickinger JC, Kondziolka D, Lunsford LD, et al. Development of a model to predict permanent symptomatic post-radiosurgery injury for arteriovenous malformation patients. Int J Radiat Onc Biol Phys 2000; 46:1143–1148. 54. Lax I, Karlsson B. Prediction of complications in gamma knife radiosurgery of arteriovenous malformations. Acta Oncol 1996; 35:49–56. 55. Voges J, Treuer H, Lehrke R, et al. Risk analysis of linac radiosurgery in patients with arterio-venous malformations (AVM). Acta Neurochir 1997; 68:118–123. 56. Pollock BE, Kondziolka D, Lunsford LD, et al. Magnetic resonance imaging: an accurate method to evaluate arteriovenous malformations after stereotactic radiosurgery. J Neurosurg 1996; 85:1044–1049. 57. Lindquist C, Steiner L. Stereotactic radiosurgical treatment of arteriovenous malformations. In: Lunsford LD, ed. Modern Stereotactic Neurosurgery. Boston: Martinus Nijhoff Publishing, 1988:491–505. 58. Schneider BF, Eberhard DA, Steiner LE. Histopathology of arteriovenous malformations after gamma knife radiosurgery. J Neurosurg 1997; 87:352–357. 59. Major O, Szeifert GT, Fazekas I, et al. Effects of single high-dose gamma irradiation on cultured cells in human cerebral arteriovenous malformation. J Neurosurg 2002; 97(5 Suppl):459–463. 60. Szeifert GT, Major O, Kemeny AA. Ultrastructural changes in arteriovenous malformations after gamma knife surgery: an electron microscopic study. J Neurosurg 2005; 102(Suppl):289– 292. 61. Shin M, Maruyama K, Kurita H, et al. Analysis of nidus obliteration rates after gamma knife surgery for arteriovenous malformations based on long-term follow-up data: the University of Tokyo experience. J Neurosurg 2004; 101:18–24. 62. Smyth MD, Sneed PK, Ciricillo SF, et al. Stereotactic radiosurgery for pediatric intracranial arteriovenous malformations: the University of California at San Francisco experience. J Neurosurg 2002; 97:48–55. 63. Touboul E, Al Halabi A, Buffat L, et al. Single-fraction stereotactic radiotherapy: a dose-response analysis of arteriovenous malformation obliteration. Int J Radiat Oncol Biol Phys 1998; 41: 855–861. 64. Zipfel GJ, Bradshaw P, Bova FJ, et al. Do the morphological characteristics of arteriovenous malformations affect the results of radiosurgery? J Neurosurg 2004; 101:393–401. 65. Schwartz M, Sixel K, Young C, et al. Prediction of obliteration of arteriovenous malformations: the obliteration prediction index. Can J Neurol Sci 1997; 24:106–109. 66. Ellis TL, Friedman WA, Bova FJ, et al. Analysis of treatment failure after radiosurgery for arteriovenous malformations. J Neurosurg 1998; 89:104–110. 67. Gallina P, Merienne L, Meder JF, et al. Failure in radiosurgery treatment of cerebral arteriovenous malformations. Neurosurgery 1998; 42:996–1004. 68. Kwon Y, Jeon SR, Kim JH, et al. Analysis of the causes of treatment failure in gamma knife radiosurgery for intracranial arteriovenous malformations. J Neurosurg 2000; 93(Suppl 3):104–106. 69. Pollock BE, Kondziolka D, Lunsford LD, et al. Repeat stereotactic radiosurgery of arteriovenous malformations: factors associated with incomplete obliteration. Neurosurgery 1996; 38:318–324.

46.

arteriovenous malformation

70. Buis DR, Lagerwaard FJ, Barkhof F, et al. Stereotactic radiosurgery for brain AVMs: role of interobserver variation in target deﬁnition on digital subtraction angiography. Int J Radiat Oncol Biol Phys 2005; 62:246–252. 71. Yu C, Petrovich Z, Apuzzo ML, et al. Study of magnetic resonance imaging—based arteriovenous malformation delineation without conventional angiography. Neurosurgery 2004; 54:1104–1107. 72. Pollock BE, Lunsford LD, Kondziolka D, et al. Patient outcomes after stereotactic radiosurgery for “operable” arteriovenous malformations. Neurosurgery 1994; 35:1–8. 73. Friedman WA, Blatt DL, Bova FJ, et al. The risk of hemorrhage after radiosurgery for arteriovenous malformations. J Neurosurg 1996; 84:912–919. 74. Nataf F, Ghossoub M, Schlienger M, et al. Bleeding after radiosurgery for cerebral arteriovenous malformations. Neurosurgery 2004; 55:298–305. 75. Pollock BE, Flickinger JC, Lunsford LD, et al. Hemorrhage risk after stereotactic radiosurgery of cerebral arteriovenous malformations. Neurosurgery 1996; 38:652–661. 76. Karlsson B, Lax I, Söderman M. Risk of hemorrhage during the 2-year latency period following gamma knife radiosurgery for arteriovenous malformations. Int J Radiat Oncol Biol Phys 2001; 49:1045–1051. 77. Maruyama K, Kawahara N, Shin M, et al. The risk of hemorrhage after radiosurgery for cerebral arteriovenous malformations. N Engl J Med 2005; 352:146–153. 78. Lindqvist M, Karlsson B, Guo W, et al. Angiographic long-term follow-up data for arteriovenous malformations previously proven to be obliterated after gamma knife radiosurgery. Neurosurgery 2000; 46:803–810. 79. Shin M, Kawahara N, Maruyama K, et al. Risk of hemorrhage from an arteriovenous malformation conﬁrmed to have been obliterated on angiography after stereotactic radiosurgery. J Neurosurg 2005; 102:842–846. 80. Yamamoto M, Jimbo M, Hara M, et al. Gamma knife radiosurgery for arteriovenous malformations: long-term follow-up results focusing on complications occurring more than 5 years after irradiation. Neurosurgery 1996; 38:906–914. 81. Pollock BE, Brown RD Jr. Management of cysts arising after radiosurgery of intracranial arteriovenous malformations. Neurosurgery 2001; 49:259–265. 82. Levegrun S, Hof H, Essig M, et al. Radiation-induced changes of brain tissue after radiosurgery in patients with arteriovenous malformations: correlation with dose distribution parameters. Int J Radiat Oncol Biol Phys 2004; 59:796–808. 83. Pollock BE. Occlusive hyperemia. A radiosurgical phenomenon? Neurosurgery 2000; 47:1178–1184. 84. Chapman PH, Ogilvy CS, Loefﬂer JS. The relationship between occlusive hyperemia and complications associated with the radiosurgical treatment of arteriovenous malformations: report of two cases. Neurosurgery 2004; 55:228–234. 85. Pollock BE, Gorman DA, Brown PD. Radiosurgery for arteriovenous malformations of the basal ganglia, thalamus, and brainstem. J Neurosurg 2004; 100:210–214. 86. Yamamoto M, Ban S, Ide M, Jimbo M. A diffuse white matter ischemic lesion appearing 7 years after stereotactic radiosurgery for cerebral arteriovenous malformations: case report. Neurosurgery 1997; 41:1405–1409. 87. Kihlström L, Guo W, Karlsson B, et al. Magnetic resonance imaging of obliterated arteriovenous malformations up to 23 years after radiosurgery. J Neurosurg 1997; 86:589–593. 88. Yamamoto M, Ide M, Jimbo M, et al. Middle cerebral artery stenosis caused by relatively low-dose irradiation with stereotactic radiosurgery for cerebral arteriovenous malformations: case report. Neurosurgery 1997; 41:474–477. 89. Kaido T, Hoshida T, Uranishi R, et al. Radiosurgery-induced brain tumor: case report. J Neurosurg 2001; 95:710–713.

471

90. Loefﬂer JS, Niemierko A, Chapman PH. Second tumors after radiosurgery: tip of the iceberg or a bump in the road? Neurosurgery 2003; 52:1436–1442. 91. Foote KD, Friedman WA, Ellis TL, et al. Salvage retreatment after failure of radiosurgery in patients with arteriovenous malformations. J Neurosurg 2003; 98:337–341. 92. Karlsson B, Kihlstrom L, Lindquist C, et al. Gamma knife surgery for previously irradiated arteriovenous malformations. Neurosurgery 1998; 42:1–5. 93. Maesawa S, Flickinger JC, Kondziolka D, et al. Repeated radiosurgery for incompletely obliterated arteriovenous malformations. J Neurosurg 2000; 92:961–970. 94. Schlienger M, Nataf F, Lefkopoulos D, et al. Repeat linear accelerator radiosurgery for cerebral arteriovenous malformations. Int J Radiat Oncol Biol Phys 2003; 56:529–536. 95. Miyawaki L, Dowd C, Wara W, et al. Five year results of LINAC radiosurgery for arteriovenous malformations: outcome for large AVMs. Int J Radiat Oncol Biol Phys 1999; 44:1089–1096. 96. Gobin YP, Laurent A, Merienne L, et al. Treatment of brain arteriovenous malformations by embolization and radiosurgery. J Neurosurg 1996; 85:19–28. 97. Lawton MT, Hamilton MG, Spetzler RF. Multimodality treatment of deep arteriovenous malformations: thalamus, basal ganglia, and brain stem. Neurosurgery 1995; 37:29–36. 98. Mathis JA, Barr JD, Horton JA, et al. The efﬁcacy of particulate embolization combined with stereotactic radiosurgery for treatment of large arteriovenous malformations of the brain. AJNR 1995; 16:299–306. 99. Paulsen RD, Steinberg GK, Norbash AM, et al. Embolization of basal ganglia and thalamic arteriovenous malformations. Neurosurgery 1999; 44:991–997. 100. Wikholm G, Lundqvist C, Svendsen P. Embolization of cerebral arteriovenous malformations: part I—technique, morphology, and complications. Neurosurgery 1996; 39:448–459. 101. Wikholm G, Lundqvist C, Svendsen P. Embolization of cerebral arteriovenous malformations: part II—aspects of complications and late outcome. Neurosurgery 1996; 39:460–469. 102. Chang T, Shirato H, Aoyama H, et al. Stereotactic irradiation for intracranial arteriovenous malformations using stereotactic radiosurgery or hypofractionated stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 2004; 60:861–870. 103. Lindvall P, Bergström P, Löfroth P, et al. Hypofractionated conformal stereotactic radiotherapy for arteriovenous malformations. Neurosurgery 2003; 53:1036–1042. 104. Redekop GJ, Elisevich KV, Gaspar LE, et al. Conventional radiation therapy of intracranial arteriovenous malformations: longterm results. J Neurosurg 1993; 78:413–422. 105. Veznedaroglu E, Andrews D, Benitez R, et al. Fractionated stereotactic radiotherapy for the treatment of large arteriovenous malformations with or without previous partial embolization. Neurosurgery 2004; 55:519–530. 106. Firlik AD, Levy EI, Kondziolka D, et al. Staged volume radiosurgery followed by microsurgical resection: a novel treatment for giant cerebral arteriovenous malformations: technical case report. J Neurosurg 1998; 43:1223–1228. 107. Pendl G, Unger F, Papaefthymiou G, et al. Staged radiosurgical treatment for large benign cerebral lesions. J Neurosurg 2000; 93(Suppl3):107–112. 108. Pollock BE, Kline RW, Stafford SL, et al. The rationale and technique of staged-volume arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 2000; 48:817–824. 109. Cernica G, de Boer SF, Diaz A, et al. Dosimetric accuracy of a staged radiosurgery treatment. Phys Med Biol 2005; 50:1991– 2002. 110. Bond JE, Smith V, Yue NJ, et al. Comparison of an image registration technique based on normalized mutual information with a standard method utilizing implanted markers in the staged radio-

472

111.

112.

113.

114.

115.

116.

b.e. pollock surgical treatment of large arteriovenous malformations. Int J Radiat Oncol Biol Phys 2003; 57:1150–1158. Lo EH. A theoretical analysis of hemodynamic and biomechanical alterations in intracranial AVMs after radiosurgery. Int J Radiat Oncol Biol Phys 1993; 27:353–361. Lo EH, Fabrikant JI, Levy RP, et al. An experimental compartmental ﬂow model for assessing the hemodynamic response of intracranial arteriovenous malformations to stereotactic radiosurgery. Neurosurgery 1991; 2:251–259. Lawton MT. Spetzler-Martin grade III arteriovenous malformations: surgical results and modiﬁcation of the grading scale. Neurosurgery 2003; 52:740–748. Morgan MK, Drummond KJ, Grinnell V, et al. Surgery for cerebral arteriovenous malformations: risks related to lenticulostriate arterial supply. J Neurosurg 1997; 86:801–805. Meder JF, Oppenhiem C, Blustajn J, et al. Cerebral arteriovenous malformations: the value of radiologic parameters in predicting response to radiosurgery. AJNR Am J Neuroradiol 1997; 18:1473– 1483. Pollock BE, Flickinger JC, Lunsford LD, et al. Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 1998; 42:1239–1247.

117. 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– 1297. 118. Maruyama K, Shin M, Tago M, et al. Gamma Knife surgery for arteriovenous malformations involving the corpus callosum. J Neurosurg 2005; 102(Suppl):49–52. 119. Andrade-Souza Y, Zadeh G, Scora D, et al. Radiosurgery for basal ganglia, internal capsule, and thalamus arteriovenous malformations: clinical outcomes. Neurosurgery 2005; 56:56– 64. 120. Maruyama K, Kondziolka D, Niranjan A, et al. Stereotactic radiosurgery for brainstem arteriovenous malformations: factors affecting outcome. J Neurosurg 2004; 100:407–413. 121. Vates GE, Lawton MT, Wilson CB, et al. Magnetic source imaging demonstrates altered cortical distribution of function in patients with arteriovenous malformations. Neurosurgery 2002; 51:614– 627. 122. Yamamoto Y, Coffey RJ, Nichols DA, et al. Interim report on the radiosurgical treatment of cerebral arteriovenous malformations. The inﬂuence of size, dose, time, and technical factors on obliteration rate. J Neurosurg 1995; 83:832–837.

4 7

Arteriovenous Malformations: Surgery Perspective Ricardo J. Komotar, Elena Vera, J. Mocco, and E. Sander Connolly Jr.

Introduction

Open Surgery

An arteriovenous malformation (AVM) is a congenital, nonhereditary defect of the vascular system, where a normal capillary system does not exist between an artery and a vein, and arterial blood passes directly into the venous system. In this localized arteriovenous shunt exists a tangle of vascular channels and veins with ﬁbrointimal thickening and elastic tissue destruction [1–5]. AVMs begin to develop in utero and grow unpredictably over time. Although AVMs may occur anywhere along the neuraxis, they most commonly occur in the brain and may be detected using computed tomography (CT) or magnetic resonance imaging (MRI). Cerebral angiography conﬁrms the diagnosis. Although relatively uncommon, AVMs may cause hemorrhagic stroke [6, 7]. Symptoms may include severe headache, signs of stroke, or loss of consciousness, a result of the AVM bleeding into adjacent tissue or subarachnoid space. Such hemorrhages are potentially life-threatening, with many surviving patients sustaining permanent disabilities [7]. The risk of hemorrhage from AVM is approximately 4% per year, with the highest bleeding risk occurring from 10 to 55 years of age [8]. With appropriate treatment, the risk of bleeding from most AVMs may be reduced. Therapeutic options include open surgery, stereotactic radiation, and endovascular embolization. Although these interventions have similar aims—diminished rate of hemorrhage or improvement in symptomatology—their indications vary and determining the best treatment is not always straightforward, as the various modalities may complement one another. Because the success of AVM treatment is dependent mainly on lesion size and location, existence and severity of previous hemorrhage, and the experience of treating health care professionals, each case must be assessed individually. In this chapter, we review the relative indications, perceived beneﬁts, and potential risks involved with treating AVMs using open surgery and radiosurgery. In addition, we discuss how these modalities may be used in combination with preoperative embolization for optimal patient outcome, cost-effectiveness, and quality of life preservation.

In open surgery, the malformation is completely excised after ligation of the feeding arteries. Whereas certain malformations are rarely operated on, such as those situated deep within the brain, the indications for open surgical resection of AVMs remain controversial. Appropriate case selection is critical for successful surgical management. The Spetzler-Martin grading scale is the most widely accepted system for risk stratiﬁcation when selecting patients to undergo surgical resection of their AVMs [9]. This scale incorporates AVM size (0 to 3 cm = 1 point, 3 to 6 cm = 2 points, >6 cm = 3 points), location (eloquent = 1 point, noneloquent = 0 points), and venous drainage (deep = 1 point, superﬁcial = 0 points). Dr. Spetzler and Dr. Martin found the following correlation between AVM grade and risk of neurologic deﬁcit after resection: grade I, 0%; grade II, 5%; grade III, 16%; grade IV, 27%; and grade V, 31%. Clinicians may use this system when preoperatively evaluating patients and deciding the most appropriate course of treatment, as AVMs with higher point totals represent a greater surgical risk of permanent morbidity and mortality. Prior studies have supported the Spetzler-Martin grading scale by demonstrating that the presence of permanent deﬁcits after surgical resection of AVMs is associated with increasing size, deep venous drainage, and high eloquence [9–14]. More speciﬁcally, one such series with overall surgical morbidity of 15.3% found AVM location and type of venous drainage to be highly correlated with poor outcome, as patients with lesions located in eloquent regions and deep venous drainage had complications rates of 20.2% and 29.5%, respectively, versus 8.2% and 7.5% for those in noneloquent locations and superﬁcial venous drainage [14]. In short, resection of large, highly eloquent AVMs with deep venous drainage is associated with high surgical morbidity and carries the highest risk for permanent deﬁcits. It is important to note that Spetzler-Martin grade III arteriovenous malformations represent a heterogeneous group of lesions. Lawton and colleagues presented their surgical results for three groups of grade III AVMs in an attempt to

473

474

r.j. komotar et al.

demonstrate that outcomes vary among them [15]. The authors reported surgical risks of new deﬁcit or death to be 2.9% for small AVMs (<3 cm), even with deep venous drainage and eloquent location (group 1); 7.1% for medium-sized (3 to 6 cm) AVMs with deep venous drainage in noneloquent locations (group 2); and 14.8% for medium-sized AVMs without deep venous drainage in eloquent locations (group 3). Thus, group 1, grade III AVMs have a surgical risk similar to that of lowgrade AVMs and can be safely treated with microsurgical resection; group 2, grade III AVMs have intermediate surgical risks and require judicious surgical selection; group 3, grade III AVMs have a surgical risk similar to that of high-grade AVMs and are best managed conservatively. According to these results, the Spetzler-Martin grading scale may not be sufﬁcient for the preoperative assessment of grade III AVMs, as the surgical risks within this category vary substantially. Although the Spetzler-Martin scale is the most widely recognized system for predicting surgical risk and chance of persistent deterioration after surgery, AVM operability should not solely be determined by size, venous drainage, and location. Rather, several other factors may play a role, such as the surgeon’s experience, angiomatous changes, irrigation by deep lenticulostriate perforators, intranidal aneurysms, venous stenosis, and patient preferences. In addition, history of prior hemorrhage should be taken into account, as untreated AVMs that have previously hemorrhaged have been documented to have annual rebleeding and overall mortality rates of 18% and 30%, respectively [16]. Thus, patients presenting with hemorrhage, whenever possible, should be treated surgically for immediate lesion obliteration. The advantage of open surgery is immediate cure and elimination of hemorrhage risk if total resection is achieved. Surgical excision, however, carries the potential for massive bleeding, risk of damage to adjacent neural tissue or ischemic stroke, and risks of infection. Furthermore, perfusion-breakthrough bleeding is a concern. Although open surgery is supported by technological improvements and the use of multimodality therapy in appropriate cases, this approach continues to be invasive, expensive, and carries a higher risk than other modalities of periprocedural hemorrhage and complications resulting in neurologic deﬁcits. Surgical complications have been attributed to judgment error, patient’s preoperative medical and neurologic condition, extensive intraoperative parenchymal injury and hemorrhage, retraction damage, and damage to the visual radiations. Consequently, extensive experience and skill is necessary to obtain favorable patient outcomes using open surgery for AVMs. In the proper hands, however, surgical resection may be the most appropriate treatment for carefully selected patients, as complication rates are minimized, postoperative deﬁcits (frequently the result of dynamic cerebral changes) are usually transient, recurrence is exceedingly rare, and lesion obliteration is obtained immediately, thereby avoiding the risk of future hemorrhage. Regardless of location, surgical extirpation should be strongly considered as the primary mode of therapy for symptomatic small- and medium-sized AVMs (Spetzler-Martin grades I and II lesions). In these cases, retrospective uncontrolled studies have shown open surgery compared with radiosurgery to be associated with signiﬁcantly fewer postoperative hemorrhages, deﬁcits, and deaths, as well as a higher incidence

of obliteration [10, 12, 13, 17–20]. In contrast, asymptomatic patients with small AVMs in eloquent regions should consider radiosurgery and those with medium-size AVMs in difﬁcult locations (Spetzler-Martin grade III lesions) should be treated with combined modalities. Surgical treatment alone is often not recommended for large AVMs located in eloquent regions and/or deep venous drainage (Spetzler-Martin grade IV/V), as the risk of postoperative deterioration is high whether they are treated with resection or multiple modalities. To support these guidelines, Fisher and colleagues developed a model for decision analysis in the treatment of AVMs [20]. The authors assumed the standard yearly hemorrhage rate of 4%, a surgical and radiosurgical morbidity rate of 2%, a surgical and radiosurgical mortality rate of 1%, surgical cure rate of 100%, and radiosurgical cure rates of 90% and 50% for lesions <3 cm and >3 cm, respectively. Based on these statistics, open surgery is recommended for all AVMs grade III or less and radiosurgery is preferred for all AVMs grade III or greater. Although this model is a gross simpliﬁcation of AVM management, it does provide insight for decision making on an individual basis.

Radiosurgery Radiosurgery acts by irradiating endothelial cells lining dysplastic blood vessels. These cells tend to multiply after radiosurgery and become thrombotic. These clots diminish nidal blood ﬂow, thereby reducing the risk of hemorrhage. General indications for radiosurgery, as opposed to open surgical treatment, include surgically inoperable lesions, lesions that have recurred or persisted after surgery, medically inﬁrm patients, and patients who refuse open surgical treatment. History of previous hemorrhage is also important, as such patients are typically treated with open surgery instead radiosurgery. The major criterion for selecting AVMs to be treated by radiosurgery is size, as success is inversely related to lesion diameter [20–28]. For lesions 3 cm or less, the rate of complete occlusion approaches 80% with a less than 1% treatment mortality and less than 3% treatment morbidity. In general, patients with lesions less than 4 cm in diameter are candidates for radiosurgery. For larger AVMs, obliteration rates are lower and either adjuvant embolization or staged radiosurgical treatments may be implemented. It has been suggested that by dividing the AVM into small sections, higher doses of radiation may be administered, resulting in higher rates of obliteration. To avoid radiation-induced injury and allow neuronal DNA repair, sessions are spaced 4 to 6 months apart, but the data to support this technique are sparse. Radiosurgery should be considered only if the overall goal is complete obliteration of the lesion, as partial treatment of a larger lesion with radiosurgery subjects the patient to risks of the procedure without eliminating the risk of hemorrhage. Size, however, is not the only factor in patient selection. The exact location, patient age, presenting symptoms, angiographic anatomy, and patient preference must be considered in this decision process. Compared with open surgery, the primary advantage of stereotactic radiosurgery is that radiosurgery does not carry the risks of conventional surgery (i.e., infection, reperfusion injury, perfusion pressure breakthrough bleeding). Furthermore, AVMs located in eloquent regions may not be safely approached

47.

arteriovenous malformations: surgery perspective

with open surgery, and in these cases, radiosurgery may be the only option. There is little, if any, acute risk of radiosurgical therapy, which allows patients to be treated on an outpatient basis. The long-term risks of radiosurgery, including radiation necrosis, radiation-induced tumors, cyst formation, and cerebral edema, depend on patient age, tumor margin dose, and size and location of the lesion [22]. More speciﬁcally, complications have been shown to be more prevalent in those patients with 12 Gy or greater margin doses and AVMs located in the thalamus, basal ganglia, and brain stem [21]. With appropriate case selection, however, the morbidity rate after therapy may be less than 5%. AVMs located deep within the brain pose a difﬁcult treatment dilemma. The majority of these lesions necessitate treatment due to the high morbidity and mortality rates that accompany hemorrhage in this region. Therapeutic options for these patients are limited, however, as microsurgical resection generally carries unacceptable risks and is not recommended unless accompanied by embolization. To this end, radiosurgery is considered the primary treatment option for such patients, especially young patients with small-volume AVMs. Complication rates, however, are more likely in this population, with one study citing a 5%, 12%, and 27% incidence of permanent radiation-related neurologic deﬁcits at 1, 2, and 5 years, respectively, after radiosurgery in patients with AVMs located in the basal ganglia, thalamus, and brain stem [29]. There are three main disadvantages to treatment of AVM with radiosurgery. First, the biological response to radiation is a slow process that causes gradual shrinkage of the AVM. During this time period, the risk of hemorrhage remains until the AVM is completely obliterated [22]. Although Maruyama and colleagues demonstrated the risk of hemorrhage to be reduced by 54% during the latency period, patients may still suffer a signiﬁcant hemorrhage [30]. Thrombosis usually takes up to 3 years, with about 40% of AVMs becoming obliterated 1 year after treatment and about 80% 3 years after treatment [22, 25, 27, 28, 31]. Of note, the time to obliteration after radiosurgery is related to patient age. For lesions approximately 3 cm in size, the usual time to resolution and occlusion in adults is 2 to 3 years after radiosurgery. For reasons as yet unclear, children have a shorter time to obliteration, commonly occurring in less than 1 year after radiosurgery [32]. In addition to the risk of hemorrhage, the latency period after radiosurgery may have signiﬁcant psychological consequences. In a study by Lai et al. [33], the quality of life (QoL) was measured in 39 patients during the ﬁrst 12 months after radiosurgery. Twenty-nine of these patients were aware of the latency period and associated hemorrhage risk, whereas 10 patients were not. Each patient completed a questionnaire about demographics, AVM-related symptoms, and perceived immediate effects of radiosurgery. In addition, a 63-item health proﬁle was administered, measuring symptom status as well as physical, emotional, and social functioning. The authors found that patients who were knowledgeable regarding the latency period had a better QoL. Also, patients with a history of hemorrhage and subsequent resolution of deﬁcits had a higher QOL. It may be that informed patients know what to expect and, as a result, have less anxiety than noninformed patients. The second main disadvantage to treatment of AVM with radiosurgery is that this modality is not always curative. Even

475

small AVMs (<3 cm in diameter) have a 20% chance of persisting after a single radiosurgery treatment [22, 27, 28, 31]. Successful obliteration is associated with younger patient age, single draining vein, and small hemispheric nidus, as large (>3 cm), deep-seated AVMs have an obliteration rate of only 40% to 60% [24, 26, 29]. Moreover, obliteration rates are signiﬁcantly associated with margin dose delivered, as radiographic cure is obtained in 70%, 80%, and 90% of patient with margin doses of 16 Gy, 18 Gy, and 20 Gy, respectively [21]. Thus, low obliteration rates for deep-seated AVMs, when compared with hemispheric AVMs, can be at least partially attributed to the lower doses delivered in an attempt to avoid substantial damage to critical surrounding structures. Previous hemorrhage or embolization has also been shown to effect outcome, as the hematoma or altered vascularity may distort the nidus so that its actual size is not precisely targeted [31]. Additional factors that can lead to treatment failure include inaccurate target deﬁnition due to poor imaging, lesion re-expansion, arterial recanalization, the development of new nidus after semi- or complete obliteration, and resistance to radiation (especially in basal ganglia and thalamus) [23–26, 28, 31, 34]. In cases of residual lesions, either a second course of stereotactic radiosurgery or open surgery is usually implemented. A study by Steinberg and colleagues assessed the efﬁcacy of treating residual AVM after radiosurgery with surgery [35]. In their cohort, open surgery was able to achieve total resection in 85% of patients with previously incomplete obliteration. Furthermore, 94% of patients had either excellent (able to work with no deﬁcits) or good (able to work/live independently with deﬁcit) outcomes. The authors note that, in their experience, lesions previously treated with radiosurgery are less vascular and are easier to resect with reduced intraoperative blood loss. Therefore, open surgery may be an effective method of treating residual AVMs several years after radiosurgery, as radiotherapy appears to reduce surgical morbidity and improve the outcome in this patient population. Having said this, from our experience the irradiated brain tends to suffer surgical complications less well than the naïve one. The third disadvantage to treatment of AVMs with radiosurgery is that even if an AVM is obliterated with radiosurgery, not all symptoms may disappear [27, 28]. For example, some patients presenting with migrainous headaches or seizures prior to AVM treatment continue to have these symptoms even though their lesion has been obliterated on cerebral angiography. In the case of seizures, symptomatology is believed to result from gliotic tissue that develops around the AVM, during both the process of development and obliteration [36, 37]. Some evidence is starting to emerge as to the role of radiosurgery for treating primary epilepsy, but the data so far are limited.

Multimodality Treatment Multimodality treatment is recommended for treating patients with highly complex AVMs. These include large, deeply located lesions for which resection or radiosurgery alone is not an option. Prior to radiosurgery treatment or surgery in these patients, embolization is required. The goal of this procedure is to diminish the amount of blood ﬂowing into the AVM by ﬁlling the feeding arteries with specially designed particles,

476

r.j. komotar et al.

microcoils, or glue. Through a reduction in nidus size, embolization is believed to facilitate subsequent deﬁnitive treatment by either decreasing the amount of intraoperative bleeding or creating a smaller target for radiation. Despite this theoretical beneﬁt, others have argued that, after radiosurgery, the embolized unradiated portion is at signiﬁcant risk for recanalization over time. As an adjuvant therapy, the advantages of embolization include being minimally invasive with the ability to access lesions that are otherwise inoperable. The risk of embolic stroke causing permanent or transient deﬁcits exists, however. Furthermore, embolization alone is not considered deﬁnitive treatment of an AVM due to low complete obliteration rates, with frequent recurrence and recanalization of feeding vessels [38]. Thus, open surgery and/or radiosurgery are necessary even after successful embolization of an AVM, as patients are typically left with a partially obliterated lesion that is at risk for hemorrhage. A retrospective review by Uno et al. [39] conﬁrmed the extent to which embolization combined with surgical resection improves patient outcome at a single institution. The authors cite that treatment of grade III and IV AVMs prior to the introduction of embolization resulted in rates of permanent and transient complication approaching 18% and 13%, respectively. Upon the advent of embolization, grade III and IV AVMs were treated with combined embolization and surgery, leading to rates of permanent and transient complication of 0% and 23%. Although these ﬁndings support preoperative embolization of complex AVMs (greater than grade III), it is important to note that several factors inﬂuence the success of subsequent AVM resection, including postembolization lesion size, arteriovenous shunting, presence of nonembolized perforators, and development of postembolization leptomeningeal collateral vessels. Another study evaluated the effectiveness of embolization combined with radiosurgery in the treatment of large AVMs (>6 cm) [40]. Embolization resulted in a mean lesion diameter reduction of 37%, and 50% of AVMs were completely obliterated. Without embolization, previously quoted rates of obliteration using radiosurgery alone are 80%, 40%, and 30% for small (<3 cm), medium (3 to 6 cm), and large (>6 cm) AVMs [27, 32]. Thus, this study suggests embolization combined with radiosurgery is more effective than radiosurgery alone in treating large AVMs. Lawton and colleagues reviewed their institutional experience using multimodality therapy for the treatment of deeply located AVMs [41]. Of 32 total patients, 11 underwent surgery, 11 underwent embolization followed by surgery, and 10 underwent embolization followed by radiosurgery. Considering the level of difﬁculty treating lesions in this region, the techniques used were successful, obtaining a 60% rate of complete obliteration with no deaths and a morbidity rate of only 9%. The authors concluded that small, deep lesions could be treated with either surgery or radiosurgery, unless they are inaccessible, in which case they should be treated with radiosurgery and an adjunct modality. In contrast, large, deep lesions are not cured with these treatments. Rather, staged pre-radiosurgery or preoperative embolization is the most effective therapy for this group if that is mandated by intracerebral hemorrhage (ICH). Thus, according to the authors, it appears the ultimate factor to

be considered when determining AVM treatment is surgical accessibility. If the AVM is accessible, surgery can be preceded by embolization to facilitate resection; if inaccessible, the lesion should be embolized and followed by radiosurgery. Any residual AVM should then be treated with radiosurgery. The multiple modalities available for AVM treatment have brought their respective cost-effectiveness into question. In a large retrospective study involving 311 patients, Nussbaum and colleagues compared how surgery compares ﬁnancially with radiosurgery [42]. Compared with observation, the authors estimated that resection of small AVMs saves nearly $8000 and extends life expectancy by 15 years. On the other hand, radiosurgery of similar AVMs costs approximately $6000 less than surgery but extends life expectancy by only 6 years. Thus, according to this study, surgery is more cost-effective and results in increased long-term survival. A study by Berman and colleagues examined the cost of surgery for the treatment of 126 patients with grades II to V AVMs [43]. Ninety-one patients, 38% of who presented with hemorrhage and with higher grades, underwent preoperative embolization, and 35 patients, 77% of who presented with hemorrhage, underwent surgery alone. From the authors’ calculations, the cost per Spetzler-Martin grade was $20,100, as those patients with higher surgical risk stayed in the hospital longer, with an increase of approximately 6 days for each increasing grade. As expected, patients with complex lesions tend to have worse outcomes and consequently utilize a greater amount of hospital resources for a longer period of time, thereby accruing higher costs. More speciﬁcally, patients with new deﬁcits postsurgery stayed in the hospital an average of 15 days and had higher total costs, $68,500, than those that did not sustain any deﬁcits, who were in the hospital for an average of 10 days and accrued $44,700 in medical costs. In the cohort studied, surgery alone cost approximately $50,000 and surgery combined with embolization cost nearly $80,000. It is important to note, however, that although preoperative embolization and high surgical risk are associated with higher costs, embolization, by itself, does not increase length of stay. Thus, the authors conclude that microsurgery for complex AVMs, in possible combination with embolization, is expensive and associated with long lengths of hospitalization. This treatment modality, however, can be considered cost-effective by avoiding delayed morbidity from hemorrhage and resulting deﬁcits.

Conclusion AVMs may be treated by open surgery, radiosurgery, or endovascular embolization. In order to determine the most appropriate therapeutic modality, their relative indications, risks, and beneﬁts must be considered. Open surgery offers immediate cure if total resection is achieved and should be the primary mode of therapy for symptomatic, superﬁcial, small- and medium-sized AVMs. Surgical excision, however, carries operative risks. In contrast, radiosurgery is noninvasive and generally used for small, recurrent lesions or lesions located in eloquent regions that may not be safely approached with open surgery. This modality, however, is not always curative and exposes the patient to persistent hemorrhage risk. These treatments may be

47.

arteriovenous malformations: surgery perspective

used in combination with preoperative embolization to improve outcomes in patients with highly complex AVMs.

References 1. Isoda K, Fukuda H, Takamura N, Hamamoto Y. Arteriovenous malformation of the brain—histological study and micrometric measurement of abnormal vessels. Acta Pathol Jpn 1981; 31(5): 883–893. 2. Gault J, Sarin H, Awadallah NA, et al. Pathobiology of human cerebrovascular malformations: basic mechanisms and clinical relevance. Neurosurgery 2004; 55(1):1–16; discussion 17. 3. Attia W, Tada T, Hongo K, et al. Microvascular pathological features of immediate perinidal parenchyma in cerebral arteriovenous malformations: giant bed capillaries. J Neurosurg 2003; 98(4):823–827. 4. Massoud TF, Vinters HV, Chao KH, et al. Histopathologic characteristics of a chronic arteriovenous malformation in a swine model: preliminary study. AJNR Am J Neuroradiol 2000; 21(7):1268–1276. 5. Wong JH, Awad IA, Kim JH. Ultrastructural pathological features of cerebrovascular malformations: a preliminary report. Neurosurgery 2000; 46(6):1454–1459. 6. Stapf C, Labovitz DL, Sciacca RR, et al. Incidence of adult brain arteriovenous malformation hemorrhage in a prospective population-based stroke survey. Cerebrovasc Dis 2002; 13(1):43–46. 7. Hartmann A, Mast H, Mohr JP, et al. Morbidity of intracranial hemorrhage in patients with cerebral arteriovenous malformation. Stroke 1998; 29(5):931–934. 8. Ondra SL, Troupp H, George ED, Schwab K. The natural history of symptomatic arteriovenous malformations of the brain: a 24year follow-up assessment. J Neurosurg 1990; 73(3):387–391. 9. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg 1986; 65(4):476–483. 10. Batjer HH. Treatment decisions in brain AVMs: the case for and against surgery. Clin Neurosurg 2000; 46:319–325. 11. Hartmann A, Stapf C, Hofmeister C, et al. Determinants of neurological outcome after surgery for brain arteriovenous malformation. Stroke 2000; 31(10):2361–2364. 12. Pik JH, Morgan MK. Microsurgery for small arteriovenous malformations of the brain: results in 110 consecutive patients. Neurosurgery 2000; 47(3):571–575; discussion 575–577. 13. Pikus HJ, Beach ML, Harbaugh RE. Microsurgical treatment of arteriovenous malformations: analysis and comparison with stereotactic radiosurgery. J Neurosurg 1998; 88(4):641–646. 14. Schaller C, Schramm J, Haun D. Signiﬁcance of factors contributing to surgical complications and to late outcome after elective surgery of cerebral arteriovenous malformations. J Neurol Neurosurg Psychiatry 1998; 65(4):547–554. 15. Lawton MT. Spetzler-Martin Grade III arteriovenous malformations: surgical results and a modiﬁcation of the grading scale. Neurosurgery 2003; 52(4):740–748; discussion 748–749. 16. Fults D, Kelly DL Jr. Natural history of arteriovenous malformations of the brain: a clinical study. Neurosurgery 1984; 15(5):658– 662. 17. Soderman M, Andersson T, Karlsson B, Wallace MC, Edner G. Management of patients with brain arteriovenous malformations. Eur J Radiol 2003; 46(3):195–205. 18. Auger RG, Wiebers DO. Management of unruptured intracranial arteriovenous malformations: a decision analysis. Neurosurgery 1992; 30(4):561–569. 19. Chang HS, Nihei H. Theoretical comparison of surgery and radiosurgery in cerebral arteriovenous malformations. J Neurosurg 1999; 90(4):709–719.

477

20. Fisher WS 3rd. Therapy of AVMs: a decision analysis. Clin Neurosurg 1995; 42:294–312. 21. Pollock BE, Meyer FB. Radiosurgery for arteriovenous malformations. J Neurosurg 2004; 101(3):390–392; discussion 392. 22. 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(6):1291– 1296; discussion 1296–1297. 23. Pollock BE, Flickinger JC, Lunsford LD, et al. Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 1998; 42(6):1239–1244; discussion 1244–1247. 24. Pan DH, Guo WY, Chung WY, et al. Gamma knife radiosurgery as a single treatment modality for large cerebral arteriovenous malformations. J Neurosurg 2000; 93(Suppl 3):113–119. 25. Massager N, Regis J, Kondziolka D, et al. Gamma knife radiosurgery for brainstem arteriovenous malformations: preliminary results. J Neurosurg 2000; 93(Suppl 3):102–103. 26. Maruyama K, Kondziolka D, Niranjan A, et al. Stereotactic radiosurgery for brainstem arteriovenous malformations: factors affecting outcome. J Neurosurg 2004; 100(3):407–413. 27. Kondziolka D, Lunsford LD. The case for and against AVM radiosurgery. Clin Neurosurg 2001; 48:96–110. 28. Ellis TL, Friedman WA, Bova FJ, et al. Analysis of treatment failure after radiosurgery for arteriovenous malformations. J Neurosurg 1998; 89(1):104–110. 29. Pollock BE, Gorman DA, Brown PD. Radiosurgery for arteriovenous malformations of the basal ganglia, thalamus, and brainstem. J Neurosurg 2004; 100(2):210–214. 30. Maruyama K, Kawahara N, Shin M, et al. The risk of hemorrhage after radiosurgery for cerebral arteriovenous malformations. N Engl J Med 2005; 352(2):146–153. 31. Kwon Y, Jeon SR, Kim JH, et al. Analysis of the causes of treatment failure in gamma knife radiosurgery for intracranial arteriovenous malformations. J Neurosurg 2000; 93(Suppl 3):104– 106. 32. Levy EI, Niranjan A, Thompson TP, et al. Radiosurgery for childhood intracranial arteriovenous malformations. Neurosurgery 2000; 47(4):834–841; discussion 841–842. 33. Lai EH, Lun SL. Impact on the quality of life of patients with arteriovenous malformations during the latent interval between gamma knife radiosurgery and lesion obliteration. J Neurosurg 2002; 97(5 Suppl):471–473. 34. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multi-institutional analysis of complication outcomes after arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 1999; 44(1):67–74. 35. Steinberg GK, Chang SD, Levy RP, et al. Surgical resection of large incompletely treated intracranial arteriovenous malformations following stereotactic radiosurgery. J Neurosurg 1996; 84(6): 920–928. 36. Chang SD, Shuster DL, Steinberg GK, et al. Stereotactic radiosurgery of arteriovenous malformations: pathologic changes in resected tissue. Clin Neuropathol 1997; 16(2):111–116. 37. Elisevich K, Redekop G, Munoz D, et al. Neuropathology of intracranial arteriovenous malformations following conventional radiation therapy. Stereotact Funct Neurosurg 1994; 63(1–4): 250–254. 38. Wikholm G, Lundqvist C, Svendsen P. The Goteborg cohort of embolized cerebral arteriovenous malformations: a 6-year followup. Neurosurgery 2001; 49(4):799–805; discussion 806. 39. Uno M, Satoh K, Matsubara S, et al. Does multimodality therapy of arteriovenous malformations improve patient outcome? Neurol Res 2004; 26(1):50–54. 40. Mathis JA, Barr JD, Horton JA, et al. The efﬁcacy of particulate embolization combined with stereotactic radiosurgery for

478

r.j. komotar et al.

treatment of large arteriovenous malformations of the brain. AJNR Am J Neuroradiol 1995; 16(2):299–306. 41. Lawton MT, Hamilton MG, Spetzler RF. Multimodality treatment of deep arteriovenous malformations: thalamus, basal ganglia, and brain stem. Neurosurgery 1995; 37(1):29–35; discussion 36.

42. Nussbaum ES, Heros RC, Camarata PJ. Surgical treatment of intracranial arteriovenous malformations with an analysis of costeffectiveness. Clin Neurosurg 1995; 42:348–369. 43. Berman MF, Hartmann A, Mast H, et al. Determinants of resource utilization in the treatment of brain arteriovenous malformations. AJNR Am J Neuroradiol 1999; 20(10):2004–2008.

4 8

Cerebral Arteriovenous Malformations: Endovascular Therapy Perspective Felipe C. Albuquerque, David Fiorella, and Cameron G. McDougall

Introduction Arteriovenous malformations (AVMs) are highly complex vascular lesions that typically present in young patients (ages 20 to 40) with hemorrhage, seizures, headache, or focal neurologic deﬁcits. The most common and compelling reason for treatment is the prevention of hemorrhage. Existing data indicate that only complete eradication of the lesion provides protection from future hemorrhage, and partial treatment is not helpful and may in fact increase the rate of future hemorrhage. To a greater extent than any of the other vascular lesions of the central nervous system (CNS), the treatment of brain AVMs requires a multimodality and multidisciplinary approach. All patients should be evaluated by physicians with expertise in endovascular embolization, microneurosurgical resection, and radiosurgery. After a careful consideration of the clinical data and AVM anatomy, a risk-beneﬁt ratio for treatment can be estimated. Once a treatment plan is agreed upon, all parties must have a clear understanding of their individual roles to facilitate successful treatment.

Formulating a Treatment Strategy The most critical step in the successful management of any patient’s AVM is the formulation of a treatment strategy designed to optimize the risk-beneﬁt ratio. This is predicated upon an understanding of the natural history of the lesion, as well as the morbidity and mortality associated with various treatments. AVMs are relatively uncommon lesions [1]. They are, in most instances, symptomatic at the time of presentation, usually from hemorrhage [2]. As such, the literature describing the natural history of AVMs is limited and composed predominately of retrospective analyses of selected populations (e.g., those not undergoing surgery, patients presenting with symp-

toms other than hemorrhage) yielding biased and relatively variable estimates of the rate of hemorrhage and its associated consequences [3]. Having said this, most estimates approximate a 2% to 4% per year risk of hemorrhage [4, 5]. In the year immediately after a symptomatic hemorrhage, the rehemorrhage risk is generally thought to be considerably higher, of the order 6% to 18% per year; gradually returning toward the 2% to 4% baseline with time [2, 5–7]. The implications of an AVM hemorrhage are not as severe as those for aneurysmal subarachnoid hemorrhage, with most estimates approximating a 10% risk of death and 20% to 30% risk of major disability subsequent to AVM rupture [2]. The risk of surgical intervention is directly related to the angio-architecture of the particular lesion. This relationship is best characterized with the Spetzler-Martin grading system (Table 48-1). In prospective studies, the Spetzler-Martin grade has demonstrated a reliable correlation with surgical outcome. Hamilton and Spetzler [8] reported operative morbidity and mortality for the resection of AVMs grades I and II (<1%) and III (<3%) to be very low. However, much higher morbidity rates were observed for grades IV and V AVMs reaching 31% and 50%, respectively, in the early postoperative period and subsequently improving to 22% and 17%, respectively, at follow-up. Heros et al. [9] reported a similar relationship between SpetzlerMartin grade and outcome. These data form the foundation for most management decisions regarding AVM therapy. In general, for grade I and II AVMs, the risk of hemorrhage far outweighs the risk of surgical resection. As such, these lesions are generally resected surgically. Frequently for grade I lesions, because of the low operative morbidity and mortality, preoperative embolization is not pursued, given that the risk of the embolization procedure may approach or even surpass the risk of surgery. In some instances, stereotactic radiosurgery rather than surgical resection is employed for treatment of a grade II lesion. The most common example would be a small (<4 cm3) grade II AVM in a highly eloquent region.

479

480

f.c. albuquerque et al.

TABLE 48-1. Spetzler-Martin AVM grading system. Description

Size <3 cm 3–6 cm >6 cm Eloquence Yes No Deep venous drainage Yes No

Neuroendovascular Therapy Points

1 2 3 1 0 1 0

Grade III AVMs represent a complex and heterogeneous group, each requiring an individualized assessment. The heterogeneity of this category led Lawton [10] to further stratify these lesions into an additional three angio-architectural subcategories with low (2.9%), intermediate (7.1%), and high (14.8%) risk of postsurgical death or new deﬁcit, respectively. The majority of these lesions are treated with either radiosurgery or preoperative embolization followed by surgical resection. When these lesions are approached surgically, preoperative embolization frequently plays an important role. The surgical resection of grades IV and V AVMs is generally associated with an operative morbidity and mortality that exceeds the risks associated with the natural history of the lesion. Han et al. [11] analyzed outcomes in a series of 73 consecutive patients with grades IV and V AVMs. These authors recommended no treatment for the majority of patients in this group (55/73) and reported a relatively low risk of hemorrhage in these patients (1% per year). In addition, in this series, as well as several additional reports, partial AVM treatment substantially increased the yearly risk for hemorrhage. In accordance with these observations, treatment for grades IV and V AVMs is only recommended in cases of progressive neurologic deﬁcits attributable to repeated hemorrhage or disabling symptoms such as an intractable seizure disorder.

FIGURE 48-1. Preoperative embolization: a 24-year-old man presenting with a seizure disorder. (A) Posteroanterior and (B) lateral projections from a left internal carotid angiogram demonstrates a grade III

The role of neuroendovascular therapy in the management of brain AVMs depends ultimately on the overall treatment plan. In general, ﬁve scenarios comprise the vast majority of rational management strategies (listed from most to least common): 1. Preoperative: Embolization as a precursor to complete curative surgical resection. 2. Targeted therapy: Embolization to eradicate a speciﬁc bleeding source. 3. Pre-radiosurgery: Embolization as a precursor to radiation therapy. 4. Curative: Embolization for attempted cure. 5. Palliative: Embolization to palliate symptoms attributed to shunting.

Preoperative Embolization AVM embolization is most frequently performed as a precursor to curative surgical resection (Fig. 48-1). In this setting, the overall goal of the embolization is to decrease the blood supply to the malformation, thereby decreasing the level of technical difﬁculty and associated morbidity of surgical resection. A successful embolization is effective in reducing the size of the AVM nidus, occluding deep feeding vessels that are difﬁcult to access and control surgically, reducing intraoperative hemorrhage and providing better delineation of a surgical resection plane. The neuroendovascular interventionist must always be cognizant of the surgical complication rate associated with the resection of any particular lesion and make every attempt to ensure that the risks of the embolization do not exceed those of the surgical resection (e.g., preoperative embolization of a grade II AVM, which is associated with a very low operative morbidity). The goal of the vascular neurosurgeon must be to achieve a complete, curative resection of the AVM. Accumulating data suggest that partial AVM resection does not reduce, but rather increases the risk of future hemorrhage. Han et al. [11] observed a hemorrhage rate of 10.4% in patients with grades IV and V AVMs after partial treatment compared with

AVM involving the anterior left temporal lobe. (C) After six pedicle embolizations performed during a single session, the volume of the AVM had been substantially reduced.

48.

cerebral arteriovenous malformations: endovascular therapy perspective

a 1% risk in patients with no previous treatment. Miyamota et al. [12] found an annual risk of hemorrhage of 14.6% in patients who underwent palliative treatment of cerebral AVMs. Wikholm et al. [13] observed an increased rate of hemorrhage and death in patients undergoing partial treatment that resulted in less than 90% nidal obliteration. The efﬁcacy of modern AVM embolization using nbutylcyanoacrylate (n-BCA) has been demonstrated in several clinical studies. Jafar et al. [14] demonstrated that preoperative embolization reduced the operative morbidity of large AVMs to a level similar to that of smaller AVMs that were not embolized preoperatively. DeMerritt et al. [15] reported similar results with preoperative embolization of large AVMs improving postsurgical outcomes in comparison with a control group of smaller AVMs that were not embolized.

481

AVM nidus into a range that is amenable to radiosurgical ablation (Fig. 48-3). In this setting, the use of a permanent embolysate such as n-BCA (see below) is mandatory to avoid recanalization of portions of the AVM that had been embolized but not included in the radiation ﬁeld. Additional goals of pre-radiosurgical embolization would include targeted therapy for components predisposed to hemorrhage (i.e., nidal or feeding vessel aneurysms) and the ablation of large arteriovenous ﬁstulae that are typically more refractory to the effects of radiotherapy.

Targeted Therapy With few exceptions, all treatment strategies for AVM management should be ultimately directed toward the complete eradication of the lesion. However, in some patients with grades IV and V AVMs not amenable to surgical resection, partial treatment targeted to eliminate an identiﬁed bleeding source is undertaken. Aneurysms are identiﬁed in association with AVMs in 7% to 20% of cases [16–19]. These aneurysms may be located on vessels that are remote from the nidus, on a feeding vessel (ﬂow-related aneurysms), or within the nidus itself. In addition, intranidal pseudoaneurysms—composed of an organized hematoma that communicates with the intravascular space—may form after AVM hemorrhage. The presence of an aneurysm represents a risk factor for intracranial hemorrhage in patients with AVMs [20]. Although both intranidal and extranidal aneurysms are risk factors for intracranial hemorrhage in patients with AVMs, the increased risk of hemorrhage in the setting of an extranidal aneurysm may be attributed to aneurysm rupture rather than hemorrhage from the AVM nidus itself [16]. Remote and feeding vessel aneurysms can usually be identiﬁed by conventional angiography. Nidal aneurysms may occasionally be visualized on conventional angiographic views, however, frequently only superselective angiography performed using high frame rates can demonstrate these lesions. Nidal aneurysms are frequently obscured by overlying vessels or other portions of the AVM nidus on conventional angiographic views. As such, when an unresectable AVM hemorrhages one or more times, endovascular exploration for a nidal aneurysm represents a reasonable strategy. In these cases, if the AVM is not to be resected, a targeted embolization may be undertaken to eradicate the aneurysm either with a liquid embolic agent (nidal aneurysm) or coils (proximal, “ﬂow-related” aneurysm or remote aneurysm) (Fig. 48-2).

Pre-Radiosurgery In general, the success of radiotherapy is inversely proportional to the size of the AVM nidus to be treated [21]. AVMs with nidal volumes less than 10 cm3 (<3 cm diameter) are frequently curable by radiosurgery, with rates of cure at 2 years estimated at between 80% and 88% [22, 23]. The theoretical goal of embolization in this setting would be to reduce the size of the

FIGURE 48-2. Targeted therapy: intranidal aneurysm with hemorrhage. (A) CT examination demonstrates a small amount of parenchymal hemorrhage within the right major forceps and marginating the ependymal surface of the right lateral ventricle. (B) Conventional angiography performed through a catheter positioned within the left vertebral artery demonstrates a diffuse AVM nidus within the right medial parietal and occipital lobes with a small central nidal aneurysm. (C) Superselective angiography performed through a microcatheter positioned within a pedicle of the right PCA better deﬁnes the anatomy of the aneurysm (arrow). (D) An unsubtracted image (acquired in the same projection as the superselective angiogram) performed after nBCA infusion demonstrates a glue cast distributed within the proximal aspect of the pedicle. (E) A postembolization angiogram demonstrates complete occlusion of the nidal aneurysm with the residual AVM nidus supplied by small, inaccessible branches of the right PCA. (F) This patient subsequently underwent Gamma Knife radiotherapy with angiographic cure of the AVM demonstrated at follow up angiography.

482

f.c. albuquerque et al.

FIGURE 48-3. Pre-radiotherapy embolization. (A) CT examination of the brain demonstrates diffuse intraventricular hemorrhage. (B) Posteroanterior and (C) lateral projections from a right internal carotid angiogram demonstrate a grade III AVM distributed within the region of the right precentral gyrus and right posterior frontal lobe. Given the high morbidity involved with surgical resection, the plan was to proceed

with embolization to be followed by radiosurgery. After embolization, (D) posteroanterior and (E) lateral projections from a right ICA angiogram demonstrate a substantial reduction in the volume of the AVM with a tiny amount of residual ﬁlling within the deep portion of the nidus (arrow). (F) Posteroanterior and (G) lateral angiograms performed at 3-year follow-up demonstrate angiographic cure of the AVM.

Despite the straightforward rationale for pre-radiosurgical embolization, very little data exists to support this approach. This is in part related to the extended latency period (2 to 3 years) required for radiotherapy to have a deﬁnitive effect. Of the available case series, most were conducted in the late 1980s and early 1990s, and many employed particulate embolysates (e.g., PVA). The utilization of a temporary embolysate for the permanent eradication of a component of AVM is contraindicated at this time, given the availability of more durable agents. The largest series was reported by Gobin et al. [24] who described their experience with 125 patients undergoing embolization (predominately with n-BCA) as a precursor to radiosurgery. These authors were able to achieve total occlusion in 11.2% of AVMs after embolization alone with an additional 76% of lesions reduced sufﬁciently in size to undergo radiotherapy. A 65% rate of total occlusion was observed after radiotherapy in patients undergoing combined treatment. More recently, Henkes et al. [25] reported a series of 30 patients undergoing combined embolization and radiotherapy, observing a less impressive 47% obliteration rate in a series of 30 patients. However, in this study, the majority of the treated AVMs were of very high grade. From the existing data, no compelling evidence exists to justify or refute the utility of preradiosurgical embolization. As a single treatment modality, stereotactic radiosurgery plays a vital role in the management of AVMs. Speciﬁcally, small-volume lesions in surgically difﬁcult regions to access are most amenable to this therapy [26–29]. Such locations include the corpus callosum, basal ganglia, internal capsule, thalamus, and brain stem. Obviously, the risks associated with direct surgical approaches to these areas are higher than those associated

with delayed radiation injury. Nevertheless, the latency period to cure and the overall treatment efﬁcacy of less than 90% are important drawbacks that mandate careful patient selection [30]. The process whereby radiosurgery affects AVMs is similar histologically to that of wound healing and contracture [31]. Light and electron microscopy have delineated this histologic process. Initially, a single, high dose of ionizing radiation triggers the production of granulation tissue [31]. This process is followed by the inﬁltration of spindle cells, which are similar in their cytoskeletal properties to myoﬁbroblasts commonly seen in the scenario of wound healing [31]. These proliferative changes culminate with scar tissue replacement and hyaline degeneration [31]. The resulting hyalinized scar demonstrates no propensity for neovascular growth. Whereas radiosurgical treatment of AVMs is associated with distinct complications, controversy exists as to the predictability and etiology of these events. Seizures, cyst formation, brain injury, arterial stenosis, and AVM rupture constitute the majority of complications [30, 32–35]. Speciﬁcally, larger AVM nidus, lobar location, higher maximal treatment doses, and incomplete lesion obliteration have been implicated as etiologic factors in the occurrence of postoperative complications [33–35]. Posttreatment hemorrhage and its attendant potential for devastating and life-threatening neurologic sequelae remains the most feared complication [33, 34]. Preoperatively, patients harboring smaller lesions in a periventricular location are at the highest risk of AVM rupture [30, 34, 35]. Speciﬁc architectural features associated with hemorrhage after radiosurgery include peri- or intranidal aneurysms, venous stenoses, and arteriovenous ﬁstulae [30, 34, 35]. Although larger treatment doses are

48.

cerebral arteriovenous malformations: endovascular therapy perspective

associated with higher cures rates, failure to cover the margins of large AVMs may promote a higher rate of postoperative hemorrhage [32, 35]. Overall, the risk of post-radiosurgical hemorrhage ranges from 1% to 7%, with 45% occurring within the ﬁrst 6 months after treatment and another 45% occurring more than 3 years after treatment [30, 32–35]. Although hemorrhage after radiosurgery certainly occurs, whether radiosurgery itself increases the likelihood of this complication is not at all clear. Indeed, some authors argue that the risk of AVM rupture is equivalent before and during the latency phase to cure after treatment [33, 34]. Seizures and radiation-induced neurologic deﬁcits constitute the other more frequently encountered complications of stereotactic radiosurgery [30, 32]. Like posttreatment hemorrhage, these complications occur at a variable rate. In fact, some authors contend that seizure frequency may actually decrease after radiosurgery among patients with documented epilepsy prior to therapy [30, 32]. The rate of posttreatment neurologic sequelae varies from 3% to 7% and is closely related to lesion location and treatment dose [30]. The patients optimally treated with stereotactic radiation are those harboring small lesions in eloquent regions of the brain [26–29]. Speciﬁcally, the treatment of lesions involving the corpus callosum, which are typically associated with a high rate of recurrent hemorrhage when left untreated, may convey a protective effect from hemorrhage [26]. Similarly, patients har-

FIGURE 48-4. Curative embolization: a 44-year-old woman who had undergone two prior unsuccessful attempts at AVM resection. (A) Towne and (B) lateral projections from a left vertebral artery angiogram demonstrate a small diffuse AVM involving the right superior cerebellar hemisphere. (C) Superselective angiography performed

483

boring deep, small AVMs involving the brain stem, basal ganglia, and thalamus are optimally treated with radiosurgery [27–29]. The decision to treat such lesions must be weighed against a 19% risk of developing a new neurologic deﬁcit as a result of irradiation [27–29]. Therefore, treatment must be reserved for small, symptomatic AVMs. Unlike cortical AVMs, lesions involving the brain stem are associated with lower rates of obliteration after stereotactic radiosurgery [27–29]. Various studies report a cure rate ranging from 40% to 70%, with younger patients harboring small, spherical lesions having the best chance of cure [27–29]. Overall, one is unable to compare objectively the efﬁcacy of stereotactic radiosurgery versus that of endovascular and microsurgical techniques for the treatment of such eloquently located AVMs as these latter two modalities are infrequently employed in these scenarios.

Curative Therapy Occasionally, a small AVM with a limited number of feeding pedicles can be completely cured through endovascular embolization alone (Fig. 48-4). Although the reported rates of complete endovascular obliteration vary, most estimates are in the range of 10%. If curative therapy is the goal of embolization, it is critical that a permanent agent (e.g., n-BCA) be used. Viñuela et al. [36] reported a 9.7% cure rate for embolization alone. Their cures were achieved exclusively in small

from a branch of the right superior cerebellar artery better demonstrates the anatomy of the nidus. (D) Angiography performed after the embolization of a single pedicle demonstrates the complete obliteration of the AVM nidus.

484

f.c. albuquerque et al.

AVMs with a small number of feeding pedicles. Gobin et al. [24] reported a cure rate of 11.2% (14 patients) in a series of 125 patients undergoing pre-radiosurgical embolization. These authors also reported that the chance of complete obliteration was inversely proportional to AVM volume and the number of feeding pedicles. Wilkholm et al. [37] reported a complete obliteration rate of 13.3%, with success also being heavily dependent upon the size of the AVM nidus—71% with AVMs smaller than 4 cm3, 15% with AVMs of 4 to 8 cm3. Fournier et al. [38] reported a cure rate of 14% with embolization alone. Yu et al. [39] recently reported a 22% cure rate with cyanoacrylate embolization alone in a series of 27 patients. Their success rate was 60% in patients undergoing embolization with the prospective goal of achieving a deﬁnitive cure. In all patients undergoing an attempted curative embolization, the nidus was accessible, less than 3 cm, and was fed by fewer than three arterial pedicles. These authors found that the angiographic obliteration of the AVM was durable at 17 to 32 months with no recurrences. No complications of embolization were reported in the series. Valavanis and Christoforidis [40] reported substantially higher cure rates (40%) in a consecutive series of 387 patients. These authors identiﬁed the presence of direct, dominant feeding arteries, a mono compartmental nidus, and a dominant ﬁstulous component of the nidus without perinidal angiogenesis as being the key characteristics predictive of endovascular cure. These authors did not ﬁnd size or number of feeding pedicles to be an important determinant of the potential for endovascular obliteration.

Palliative Therapy Although controversial, some investigators theorize that large AVMs may cause progressive neurologic deﬁcits, intellectual deterioration, or cause persistent headaches as sequelae of the shunting of blood away from physiologically normal brain (i.e., a steal phenomena) [41, 42]. Given that the lesions responsible for this type of phenomena are large and typically unresectable, some investigators have advocated partial embolization in an attempt to reduce the severity of arteriovenous shunting and improve perfusion pressure in the surrounding functional brain parenchyma [43]. Although no large clinical series exist to support this strategy, several case reports have described success in small numbers of patients [43, 44]. Fox et al. [45] reported improvement in limb weakness in three patients after subtotal embolization of large AVMs located near the motor cortex, attributing the improvement to a reduction in cerebrovascular “steal.” Whereas evidence to support partial embolization in cases of suspected cerebrovascular steal is lacking [46], there is a reasonable amount of evidence that would indicate that partial treatment of large AVMs—either with embolization or surgery—increases the risk of hemorrhage [11, 13, 45]. In addition, whereas studies have indicated diminished cerebral blood ﬂow in brain regions around AVMs in groups of patients with “steal syndromes,” these studies are at the present time not sufﬁcient to prospectively identify patients who will beneﬁt from partial treatment. These factors must be taken into account before initiating the partial embolization of a large, unresectable AVM to control symptoms. From a practical standpoint, when this type of plan is undertaken, it is imperative to proceed

cautiously during the embolization and in particular to avoid producing any restriction of venous outﬂow.

Technique: Preoperative Embolization Goals of Embolization Before initiating the neuroendovascular portion of AVM therapy, it is critical that the interventionist have a complete understanding of the overall plan as well as the goals for the embolization. This understanding is predicated upon maintaining open lines of communication with the vascular neurosurgeon who will be performing the resection (or radiosurgical treatment). The risks of microsurgical resection as deﬁned by the Spetzler-Martin category of the lesion should be clear before the procedure. It is important to weigh these risks with those involved with each catheterization and each embolization. For example, if embolizing a grade II AVM in a noneloquent region, it is critical that the risks of each embolization be minimized as to avoid complicating an otherwise straightforward resection (Fig. 48-5).

Staging The number of embolizations that can be performed during a single session varies with the preference of the interventionist and the anatomy of the lesion. One potential risk of “overembolization” of a large lesion is hemorrhage related to normal perfusion pressure breakthrough—the sequelae of an abrupt reduction in AV shunting and sudden increase in the perfusion pressure of the adjacent normal brain parenchyma, which has impaired autoregulatory capacity (Fig. 48-6). In a patient with a large AVM scheduled for surgical resection on the next morning, we routinely perform between

FIGURE 48-5. Distal catheter position. Unsubtracted ﬁlm obtained during the embolization of a left parietal lobe AVM provides an example of the extent to which the modern generation of ﬂow-directed microcatheters can be efﬁciently and atraumatically manipulated into the most distal of locations within the cerebrovasculature.

48.

cerebral arteriovenous malformations: endovascular therapy perspective

485

FIGURE 48-6. Hemorrhage after aggressive embolization. (A) Posteroanterior and (B) lateral ﬁlms from a right internal carotid angiogram demonstrate a large arteriovenous malformation involving the right temporal and parietal lobes in a 42-year-old woman with a seizure disorder presenting for preoperative embolization. After a single session of embolization, unsubtracted (C) posteroanterior and (D) lateral images depict an extensive n-BCA cast within the nidus. (E, F) Postembolization angiography depicts stasis within several of the embolized arterial branches coursing into the region of the AVM nidus. Only minimal residual ﬂow into the nidus was evident at the conclusion of the embolization (not depicted). There was no evidence of venous outﬂow restriction. The patient emerged from general anesthesia neurologically intact. Forty-eight hours after the embolization, she developed a progressive headache and left hemianopsia. (G) CT was performed emergently, demonstrating a left parietal lobe hematoma.

ﬁve and seven embolysate infusions during a single session. If multiple vascular distributions provide supply to the lesion (e.g., right internal carotid and vertebrobasilar) and multiple sessions are to be performed, it is our preference to embolize

within only one vascular distribution during any given session. In general, for AVMs larger than 3 cm, it is preferable to have at least two sessions of embolization scheduled (Fig. 48-7).

FIGURE 48-7. Staged embolization for a large AVM. (A, B) Posteroanterior and (C, D) lateral images depict a large frontal AVM in a 46year-old woman with headaches. (E, F, G) Unsubtracted images obtained after each of three successive stages of embolization depict

the progression of the n-BCA cast. (H) Posteroanterior and (I) lateral images after the ﬁnal stage of embolization demonstrate a marked reduction in the size of the AVM nidus as well as the volume of arterial to venous shunting.

486

f.c. albuquerque et al.

FIGURE 48-8. Inadvertent venous compromise during embolization. (A–D) Cerebral angiography performed in a 57-year-old woman presenting with headaches demonstrates a grade III AVM of the right frontal lobe. The late arterial phase images demonstrate shunting into a large curvilinear venous pouch that subsequently supplies multiple additional tributaries that efﬂux over the right cerebral convexity (B, D, arrows). (E) Subtracted and (F) unsubtracted images from control angiography after the ﬁrst n-BCA infusion demonstrate a large collection of n-BCA within the dominant draining vein. Later phase images depicted a change in the ﬂow dynamics through the AVM with more persistent opaciﬁcation of the nidus and slow outﬂow into the venous

system. Because of the venous outﬂow compromise produced by the initial n-BCA injection, the initial plan to proceed with a staged embolization was abandoned and a more aggressive embolization was pursued with urgent surgical resection to be performed the next morning. (G, H) Four additional pedicles were embolized resulting in near complete obliteration of the AVM nidus. The patient was maintained under general anesthesia after the procedure with the systolic blood pressure held under 90 mm Hg. Surgical resection performed the next day was uneventful, and the patient emerged from the procedures neurologically intact.

48.

cerebral arteriovenous malformations: endovascular therapy perspective

Postoperative Care Most of our AVM embolizations are completed in a single session with surgical resection to be performed the next morning. Larger lesions are occasionally embolized over multiple sessions. In most cases, heparinization is reversed at the conclusion of the procedure. In adult patients, the arterial sheath is frequently left in place in anticipation of intraoperative angiography to be performed the next day after resection. Occasionally, heparinization is not reversed or the patient is placed on a heparin drip if the venous outﬂow appears sluggish on the postembolization angiogram or if an important component of the venous outﬂow has been obviously compromised. This is frequently the case when a large AVM nidus with associated venous varices is near totally obliterated. In these cases, sluggish ﬂow in the varices could lead to preoperative venous thrombosis with subsequent hemorrhage from the remaining nidus. After embolization, particularly if large AV shunts or a large volume of the nidus has been occluded, the theoretical possibility of normal perfusion pressure breakthrough hemorrhage exists. For this reason, we attempt to maintain a low systolic blood pressure (<100 to 120 mm Hg) after the procedure with the level determined empirically by the amount of shunting that has been reduced, the patient’s baseline blood pressure, and the presence of any additional vascular lesions (e.g., ﬂowrelated stenoses, carotid atheromatous stenoses, etc.). We prefer a nicardipine or nitroprusside drip for this purpose supplemented by intravenous medications as needed. If a very large AVM has been near completely embolized during a single session, we pursue this hypotensive strategy even more aggressively. In these cases, the patient is maintained under continuous general anesthesia (usually using an agent such as propofol) after the procedure. The systolic blood pressure is maintained at <90 mm Hg. These conditions are maintained through the time of surgical resection on the next day. Head computed tomography (CT) is performed immediately after the completion of the embolization and 4 hours later. Any increases in blood pressure that occur during the postembolization period represent a potential harbinger of intracranial hemorrhage and also require immediate evaluation by CT. We have found this strategy to be effective when inadvertent venous occlusion occurs early in the stage of the embolization of a large lesion (Fig. 48-8). If surgical resection cannot be carried out on the same day after completion of the embolization, we proceed to embolize all accessible arterial feeders during the same session and then maintain the patient in this hypotensive anesthetized state through the time of resection the next day.

Complications Our experience with AVM embolization at the BNI is similar to that reported in the literature. From 1995 through 2003, we have performed 262 embolizations in 178 patients with 10 ischemic complications (1 major, 3 quadrantanopsia, 6 temporary deﬁcits with full recovery) and 7 hemorrhages (1 death, 2 major morbidity)—overall per patient yielding an 8.9% rate of neurologic morbidity (3.4% permanent) and a 0.6% death rate.

487

Taylor et al. [47] reported a 2% death rate and 9% permanent neurologic deﬁcit rate in 201 patients undergoing 339 embolization procedures during an 11-year period. In this series, embolization procedures were performed using polyvinyl alcohol particles, n-butylcyanoacrylate, detachable coils, and/or the liquid polymer Onyx. Debrun et al. [48] reported a 5.6% rate of neurologic morbidity and 2 deaths (3.7%) in 54 patients undergoing n-BCA embolization. Viñuela et al. [49] reported a morbidity rate of 13% and a single death in a series of 101 patients. In Gobin’s [24] series of 125 patients, permanent complications occurred in 12.8% (minor deﬁcits in 5.6%, moderate deﬁcits in 4.8%, major deﬁcits in 2.4%), and there were 2 deaths (1.6% mortality rate). In the initial study by Wallace et al. [50], comparing n-BCA and PVA, 4 of 22 patients who underwent particle embolization had ischemic neurologic complications (one of which was major) and 4 of 23 patients who underwent acrylic embolization suffered minor neurologic deﬁcits. In a subsequent trial, the n-BCA investigators reported an overall death rate of 3.9% in 101 patients undergoing either n-BCA or PVA embolization. Jahan et al. [51] reported four adverse events and no deaths in a series of 23 patients undergoing embolization with Onyx. Of the adverse events, only one resulted in permanent morbidity (4%). Hartman et al [52] reported a 14% rate of new deﬁcits, a 2% rate of permanent disability, and a 1% death rate in 233 patients undergoing 545 sessions of embolization. Frizzel and Fisher [53] compiled data from the medical literature available from 1969 through 1993 and calculated a 10% rate of temporary morbidity, 8% permanent morbidity, and 1% death rate in 1246 patients with brain AVMs undergoing embolization. The authors found no signiﬁcant difference in morbidity or mortality when comparing series published before with those reported after 1990.

Conclusion Neuroendovascular embolization continues to be a critical component of the multidisciplinary, multimodality management of cerebral arteriovenous malformations. Safe and effective embolization may only be performed in the context of a well-designed, rational treatment plan that is fundamentally based on a clear understanding of the natural history of the lesion as well as the cumulative risks of multimodality treatment.

References 1. Berman MF, Sciacca RR, Pile-Spellman J, et al. The epidemiology of brain arteriovenous malformations. Neurosurgery 2000; 47: 389–397. 2. Itoyama Y, Uemura S, Ushio Y, et al. Natural course of unoperated arteriovenous malformations: study of 50 cases. J Neurosurg 1989; 71:805–809. 3. Al-Shahi R, Warlow CP. Quality of the evidence for the management of cerebral arteriovenous malformations. Lancet 2002; 360:1022–1023. 4. Crawford PM, West CR, Chadwick DW, Shaw MDM. Arteriovenous malformations of the brain. Natural history in unoperated patients. J Neurol Neurosurg Psychiatry 1986; 49:1–10. 5. Mast H, Young WL, Koennecke HC, et al. Risk of spontaneous hemorrhage after diagnosis of cerebral arteriovenous malformation. Lancet 1997; 350:1065–1068.

488

f.c. albuquerque et al.

6. Graf CJ, Perrett GE, Torner JC. Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg 1983; 58:331–337. 7. Jane JA, Kassell NF, Torner JC, et al. The natural history of aneurysms and arteriovenous malformations. J Neurosurg 1985; 62:321–323. 8. Hamilton MG, Spetzler RF. The prospective application of a grading system for arteriovenous malformations. Neurosurgery 1994; 34:2–7. 9. Heros RC, Korosue K, Diebold PM. Surgical excision of cerebral arteriovenous malformations: late results. Neurosurgery 1990; 26(4):570–577. 10. Lawton MT. Spetzler-Martin Grade III arteriovenous malformations: surgical results and a modiﬁcation of the grading scale. Neurosurgery 2003; 52: 740–748. 11. Han PP, Ponce FA, Spetzler RF. Intention-to-treat analysis of Spetzler-Martin grade IV and V arteriovenous malformations: Natural history and treatment paradigm. J Neurosurg 2003; 98: 3–7. 12. Miyamota S, Hashimoto N, Nagata I, et al. Posttreatment sequelae of palliatively treated cerebral arteriovenous malformations. Neurosurgery 20004; 6:589–595. 13. Wikholm G, Lundqvist C, Svendsen P. The Göteborg cohort of embolized cerebral arteriovenous malformations: a 6-year followup. Neurosurgery 2001; 9:799–806. 14. Jafar JJ, David AJ, Berenstein A, et al. The effect of embolization with n-butylcyanoacrylate prior to surgical resection of cerebral arteriovenous malformations. J Neurosurg 1993; 78:60–69. 15. DeMerritt JS, Pile-Spellman J, Moohan MN, et al. Outcome analysis of preoperative embolization with n-butylcyanoacrylate in cerebral arteriovenous malformations. Am J Neuroradiol AJNR 1995; 16:1801–1807. 16. Kim EJ, Halim AX, Dowd CF, et al. The relationship of coexisting extra-nidal aneurysms to intracranial hemorrhage in patients harboring brain arteriovenous malformations. Neurosurgery 2004; 54(6):1349–1358. 17. Lasjunias P, Piske R, Terbrugge K, et al. Cerebral arteriovenous malformations and associated cerebral aneurysms. Acta Neurochir 1988; 91:29–36. 18. Marks MP, Lane B, Steinberg G, Chang PJ. Hemorrhage in intracerebral arteriovenous malformations. Radiology 1990; 176: 807–813. 19. Redekop G, TerBrugge K, Montanera W, Willinsky R. Arterial aneurysms associated with cerebral arteriovenous malformations. J Neurosurg 1998; 89:539–546. 20. Brown RD, Wiebers DO, Forbes GS. Unruptured intracranial aneurysms and arteriovenous malformations: Frequency of intracranial hemorrhage and relationship of lesions. J Neurosurgery 1990; 73:859–863. 21. Kwon Y, Jeon SR, Kim JH, et al. Analysis of the causes of treatment failure in gamma knife radiosurgery for intracranial arteriovenous malformations. J Neurosurg 2000; 93(Suppl 3):104– 106. 22. Lundsford LD, Kondziolka D, Flickinger JI, et al. Stereotactic radiosurgery for arteriovenous malformations of the brain. J Neurosurg 1991; 75:512. 23. Steiner L, Lindquist C, Adler JR, et al. Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 1992; 77:1. 24. Gobin YP, Laurent A, Merienne L, et al. Treatment of brain arteriovenous malformations by embolization and radiosurgery. J Neurosurg 1996; 85:19–28. 25. Henkes H, Nahser HC, Berg-Dammer E, et al. Endovascular therapy of brain AVMs prior to radiosurgery. Neurol Res 1998; 20(6):479–492.

26. Maruyama K, Shin M, Tago M, et al. Gamma knife surgery for arteriovenous malformations involving the corpus callosum. J Neurosurg 2005; 102:49–52. 27. Andrade-Souza YM, Zadeh G, Scora D, et al. Radiosurgery for basal ganglia, internal capsule, and thalamus arteriovenous malformation: clinical outcome. Neurosurgery 2005; 56:56–64. 28. Pollock BE, Gorman DA, Brown PD. Radiosurgery for arteriovenous malformations of the basal ganglia, thalamus, and brainstem. J Neurosurg 2004; 100:210–214. 29. Maruyama K, Kondziolka D, Niranjan A, et al. Stereotactic radiosurgery for brainstem arteriovenous malformations: factors affecting outcome. J Neurosurg 2004; 100:407–413. 30. Shin M, Maruyama K, Kurita H, et al. Analysis of nidus obliteration rates after gamma knife surgery for arteriovenous malformations bases on long-term follow-up data: the University of Tokyo experience. J Neurosurg 2004; 101:18–24. 31. Szeifert GT, Major O, Kemeny AA. Ultrastructural changes in arteriovenous malformation after gamma knife surgery: an electron microscopic study. J Neurosurg 2005; 102:289–292. 32. Schaubles B, Cascino GD, Pollack BE, et al. Seizure outcomes after stereotactic radiosurgery for cerebral arteriovenous malformations. Neurology 2004, 63(4):683–687. 33. Izawa M, Hayashi M, Chernov M, et al. Long-term complications after gamma knife surgery for arteriovenous malforamtions. J Neurosurg 2005; 102:34–37. 34. Nataf F, Ghossoub M, Schlienger M, et al. Bleeding after radiosurgery for cerebral arteriovenous malformations. Neurosurgery 2004; 55:298–306. 35. Zipfel GJ, Bradshaw P, Bova FJ, Friedman WA. Do the morphological characteristics of arteriovenous malformations affect the results of radiosurgery? J Neurosurgery 2004; 101:393–401. 36. Viñuela F, Duckwiler G, Guglielmi G. Contribution of interventional neuroradiology in the therapeutic management of brain arteriovenous malformations. J Stroke Cerebrovasc Dis 1997; 4: 268–271. 37. Wilkholm G, Lundqvist C, Svendsen P. Embolization of cerebral arteriovenous malformations: part I. Technique, morphology and complications. Neurosurgery 1996; 3:448–459. 38. Fournier D, TerBrugge KG, Willinsky R, et al. Endovascular treatment of intracerebral arteriovenous malformations. J Neurosurg 1991; 75:228–233. 39. Yu SCH, Chan MSY, Lam JMK, et al. Complete obliteration of intracranial arteriovenous malformation with endovascular cyanoacrylate embolization: initial success and rate of permanent cure. AJNR 2004; 25:1139–1143. 40. Valavanis A, Christoforidis G. Endovascular management of cerebral arteriovenous malformations. Neurointerventionist 1999; 1:34–40. 41. Batjer HH, Devous MD Sr, Seibert GB, et al. Intracranial arteriovenous malformation: relationships between clinical and radiographic factors and ipsilateral steal severity. Neurosurgery 1988; 23:322–328. 42. Marks MP, Lane B, Steinberg G, et al. Vascular characteristics of intracerebral arteriovenous malformations in patients with clinical steal. AJNR 1991; 12:489–496. 43. Kusske JA, Kelly WA. Embolization and reduction of the “steal” syndrome in cerebral arteriovenous malformations. J Neurosurg 1974; 40:313–321. 44. Luessenhop AJ, Mujica PH. Embolization of segments of the circle of Willis and adjacent branches for management of certain inoperable cerebral arteriovenous malformations. J Neurosurg 1981; 54:573–582. 45. Fox AJ, Girvin JP, Vinuela F, Drake CG. Rolandic arteriovenous malformations: improvement in limb function by IBC embolization. AJNR 1985; 6(4):575–582.

48.

cerebral arteriovenous malformations: endovascular therapy perspective

46. Mast H, Mohr JP. Osipov A, et al. “Steal” is an unestablished mechanism for the clinical presentation of cerebral arteriovenous malformations. Stroke 1995; 26:1215–1220. 47. Taylor CL, Dutton K, Rappard G, et al. Complications of preoperative embolization of cerebral arteriovenous malformations. J Neurosurg 2004; 100(5):810–812. 48. Debrun GM, Aletich V, Ausman JI, et al. Embolization of nidus of brain arteriovenous malformations with n-butyl cyanoacrylate. Neurosurgery 1997; 40:112–121. 49. Viñuela F, Dion JE, Duckwiler GR, et al. Combined endovascular embolization and surgery in the management of cerebral arteriovenous malformations. J Neurosurg 1991; 75:856– 864.

489

50. Wallace RC, Flom RA, Khayata MH, et al. The safety and effectiveness of brain arteriovenous malformation embolization using acrylic and particles. Neurosurgery 1995; 37:606–618. 51. Jahan R, Murayama Y, Gobin YP, et al. Embolization of arteriovenous malformations with Onyx: clinicopathological experience in 23 patients. Neurosurgery 2001; 48(5):984–995. 52. Hartmann A, Pile-Spellman J, Stapf C, et al. Risk of endovascular treatment of brain arteriovenous malformations. Stroke 2002; 33: 1816–1820. 53. Frizzel RT, Fisher WS. Cure, morbidity, and mortality associated with embolization of brain arteriovenous malformations: a review of 1246 patients in 32 series over a 35 year period. Neurosurgery 1995; 37:1031–1040.

4 9

Cavernous Malformations and Other Vascular Diseases Ajay Niranjan, David Mathieu, Douglas Kondziolka, John C. Flickinger, and L. Dade Lunsford

Intracranial Cavernous Malformations Intracranial cavernous malformations are angiographically occult developmental malformations of the vascular bed. These abnormal vascular connections may enlarge over time. The lesions can occur on a familial basis. Patients may be asymptomatic, although they often present with headaches, seizures, or parenchymal hemorrhages. Substantial advances in the cellular and molecular biology of the vasculogenesis, angiogenesis, and cardiovascular physiology of these anomalies along with ﬁndings from detailed clinicopathologic and clinicoradiologic retrospective and prospective longitudinal studies have led to a better understanding of these vascular malformations as a whole.

Pathophysiology Cavernous malformations can be found in any part of the brain because they can occur at any location along the vascular bed. Most (80% to 90%) of the lesions are supratentorial, and the frontal and temporal lobes are the most common sites. The deep cerebral white matter, corticomedullary junction, and basal ganglia are common supratentorial sites, whereas the pons and cerebellar hemispheres are common posterior fossa sites. Cavernous malformations are considered to be congenital vascular hamartomas composed of closely approximated endothelial-lined sinusoidal collections without signiﬁcant amounts of interspersed neural tissue. The lack of intervening neural tissue distinguishes these lesions from capillary telangiectasias. De novo pathogenesis may occur spontaneously or in association with initiating events, such as biopsy, a preexisting venous malformation, or perhaps after radiation therapy. Pathologically, nearly all cavernous malformations show evidence of recent and remote hemorrhage, in the form of hemosiderin-

laden macrophages, cholesterol crystals, and hemosiderinstained parenchymal tissues. Cavernous malformations are discrete multilobulated lesions and grossly resemble small mulberries. Clots and blood products in various stages of evolution can be seen within these lesions. Calciﬁcation and gliosis are also often seen. The architecture of the component vessels consists of a single layer of endothelium and varying quantities of subendothelial ﬁbrous stroma. Smooth muscle and elastic ﬁbers are absent. Re-endothelialization of the hemorrhagic cavities, growth of new blood vessels, and proliferation of granulation tissue may account for the apparent growth of some cavernous malformations.

Epidemiology Cavernous malformations represent approximately 1% of intracranial vascular lesions and 15% of cerebrovascular malformations although these numbers may be underestimations. In early studies of major autopsy reports, the calculated prevalence was 0.02% to 0.53%. With the advent of magnetic resonance imaging (MRI), cavernous malformations are currently the most commonly identiﬁed brain vascular malformations. The detection of previously unidentiﬁed asymptomatic lesions by MRI has recently raised the estimated overall prevalence to 0.45% to 0.9% [1]. Multiple lesions are seen in approximately 15% to 33% of spontaneous cases, although one series reported an incidence as high as 50%. Multiple lesions are more common in the familial form, occurring in as many as 84% of patients [2]. The familial form of the disorder is inherited as an autosomal dominant trait with variable expression. Cavernous malformations are the most common CNS vascular malformation subtype in patients with mixed vascular lesions. Associated developmental venous anomalies (DVAs) are present in approximately 10% to 30% of patients with cavernous malformations.

491

492

a. niranjan et al.

Molecular Basis Although most cavernous malformations are believed to be sporadic, many familial cases have been observed over the past two decades. These cases exhibit an autosomal dominant pattern of inheritance and seem to affect the Hispanic population in particular. Recent research has demonstrated at least three separate genes (CCM1, CCM2, and CCM3) related to the familial form of the disease. Linkage analyses using autosomal dominant families manifesting CCMs have identiﬁed three different causative loci on chromosomes 7q21.2 (CCM1), 7p13 (CCM2), and 3q25.2-q27 (CCM3). Mutations in the gene Krit1 are responsible for CCM1, mutations in the gene MGC4607 are responsible for CCM2, and mutations in the gene PDCD10 were recently reported to be responsible for CCM3. Of familial cavernous malformations, 40% can be linked to a protein created by the mutated CCM1 gene. This is the gene responsible for most of the cases of familial multiple cavernous malformation in Mexican-American families and in a number of other families. CCM1 is responsible for creating KRIT1 protein, or Krev interaction-trapped 1 protein. The exact function of KRIT1 protein is not known. If both copies of the CCM1 gene mutate, the KRIT1 protein cannot function and cavernous malformations form. CCM2 controls the production of a protein named malcavernin. Of familial cavernous malformations, 20% can be linked to a CCM2 mutation [3–5].

Clinical Presentation Cavernous malformations can occur at any age, but they are most likely to become clinically apparent in patients aged 20 to 40 years. Patients with cavernous malformations may remain asymptomatic, but they most often present with headache or neurologic symptoms after a hemorrhage or repeated hemorrhage. Because of the extruded blood products and the fact that some malformations can enlarge slowly, the lesions may also produce seizures and a variety of neurologic ﬁndings similar to those expected of intracranial tumors. Once patients become symptomatic, 40% to 50% present with seizures, 20% present with focal neurologic deﬁcits, and 10% to 25% present with hemorrhage. Symptoms may progress rapidly, remain stable for years, or wax and wane. Headache is estimated to be a symptom in as many as 25% of patients. Acute headaches may result from parenchymal irritation secondary to gross or repeated extralesional hemorrhage. Chronic headaches could be the result of mass effect in slow-growing larger lesions as a result of multiple intralesional hemorrhages. The clinical symptoms usually result from the location of the lesion and, at times, their slow expansion. Infratentorial location and previous gross hemorrhage are associated with increased risk of subsequent and progressive neurologic disability. Large hemorrhages can cause both obstructive and nonobstructive hydrocephalus. Any hemorrhage found on computed tomography (CT) scans in a relatively young patient should be investigated further, and cavernous malformation must be considered a possible etiology. In the workup of a patient with a seizure disorder, cavernous malformation must be considered as a potential cause, especially if the patient is aged 20 to 40

years. The bleeding rate is estimated to be 0.1% to 2.5% per lesion-year and 0.25% to 16.5% per patient-year [6]. This risk is increased in patients with established prior hemorrhage. A study at the University of Pittsburgh concluded that in patients who had one prior hemorrhage, the subsequent annual risk of hemorrhage was 33.9%. After the ﬁrst bleed, annual hemorrhage rates in years 1 through 5 were 52%, 35%, 39%, 24%, and 32%, respectively [7]. In addition, women are at a slightly higher risk for hemorrhage, especially those in the ﬁrst trimester of pregnancy. The hemorrhage is rarely life-threatening. Hemorrhages in critical locations can have more severe effects, and thus, they are more likely to produce symptoms. Progressive neurologic deﬁcits are more often associated with cavernous malformations in the infratentorial space and with lesions that demonstrate slow enlargement because of rebleeding episodes.

Radiologic Features Although cavernous malformations can be diagnosed using CT scans, MRI is the modality of choice for the long-term followup of patients as well as the assessment of family members in whom similar lesions are suspected. Cavernous malformations can be divided into three components: the peripheral pseudocapsule composed of gliotic hemosiderin-laden tissue, the irregular intersecting connective tissue septa separating the sinusoidal spaces, and the central vascular area composed of slow-ﬂowing sinusoidal spaces [8]. MRI ﬁndings of parenchymal cavernous malformations demonstrate typical, popcornlike, well-circumscribed, well-delineated lesions. The core is formed by multiple foci of mixed signal intensities, which represents hemorrhage in various stages of evolution. Acute hematoma containing deoxyhemoglobin is isointense on T1weighted images and markedly hypointense on T2-weighted images. Subacute hematoma, which contains extracellular methemoglobin, displays hyperintensity on both T1- and T2weighted images because of the paramagnetic effect of the methemoglobin. The interspersed ﬁbrous-containing elements demonstrate mild hypointensity on both T1- and T2-weighted images because they contain a combination of calciﬁcation and hemosiderin. The heterogeneous core typically is surrounded completely by a low-signal-intensity hemosiderin rim on T1weighted images. The hypointensity of this rim becomes more prominent on T2-weighted images because of the magnetic susceptibility effects. Smaller cavernous malformations may appear as focal hypointense nodules with both T1- and T2weighted sequences. Typically, cavernous malformations are not associated with mass effect or edema and do not demonstrate a feeding artery or draining vein, except when associated with other vascular malformations. Cavernous malformations can be associated with DVAs, which typically demonstrate a prominent medullary vein with the classic caput medusae pattern of drainage. MRI has largely replaced conventional angiography in the diagnosis of cavernous malformations. However, when the lesions occur in combination with other types of vascular malformations, as they do in as many as 30% of patients, MRI characteristics become more complicated and less speciﬁc. In these patients, angiography can help further deﬁne

49.

493

cavernous malformations and other vascular diseases

the lesions. Because of the extremely slow ﬂow of blood through these lesions, cerebral angiograms are often normal. Most cavernous malformations (37% to 48%) are avascular masses on conventional angiograms. The avascular appearance is the result of compression or destruction of vascular channels by hemorrhage, thrombosis, and generalized slow ﬂow because of the small size of the connecting sinusoidal vessels with the peripheral normal parenchymal vessels. When lesions are smaller and not associated with hematomas, 20% to 27% of angiograms demonstrate normal ﬁndings. Capillary blush, which is not a speciﬁc ﬁnding, is demonstrated at 12% to 20% [9].

Management Options When a patient is diagnosed with a cavernous malformation, the reason that led to the detection of the lesion must be factored into the ﬁnal decision as to what, if any, treatment should be offered. Severe complications can occur after treatment of lesions in critical locations in the brain and brain stem. The rate of these complications must be carefully balanced against the natural history of the lesion itself (Table 49-1) [1, 2, 6, 10–13].

It is important to ascertain the presence of DVAs, as these are composed of functional venous channels that should not be resected or radiated because of the risk for venous infarction. Increasingly, DVAs are identiﬁed in association with cavernous malformations [14]. In older patients with lesions that are difﬁcult to reach surgically, observation with periodic imaging may be appropriate. When patients present with recurrent hemorrhage, progressive neurologic deterioration, or intractable epilepsy, then treatment in the form of surgery or radiosurgery should be considered. The management option for a patient with a cavernous malformation must be based on the exact location of the lesion and its surgical accessibility. Surgery has the advantage of immediate elimination of the lesion with stabilization of symptoms if a complete resection can be performed. However, attempted resection of deep-seated lesions can lead to serious morbidity (Table 49-2) [15–20]. For younger patients with more accessible lesions, removal of lesions may help control epilepsy, improve neurologic deﬁcits, and prevent any subsequent hemorrhage. Radiosurgery is a minimally invasive option for patients presenting with recurrent hemorrhages from deepseated cavernous malformations.

TABLE 49-1. Natural history of cavernous malformations. Patients (lesions)

Follow-up length

Del Curling (1991) [11]

32 (76)

NS

Robinson (1991) [1]

66 (76 )

26 months

Zabramski (1994) [2]

31 (128)

2.2 years

92% Supratentorial 8% Infratentorial

Aiba (1995) [10]

110

2.39–7.98 years

Kondziolka (1995) [12]

122 S 80% M 20%

34 months

Supratentorial Superﬁcial: 65% Deep: 11% Cerebellar: 6% Brain stem: 18% 17% basal ganglia or thalamus 35% brain stem

Labauge (2000) [6]

40 (232) M 93%

3.2 years

176 supratentorial 30 cerebellar lesions 26 brain-stem lesions

Kupersmith (2001) [13]

37 patients

4.9 years

All brain stem 12 midbrain 18 pons 7 medulla

First author

HRG, hemorrhage.

Lesion location

Clinical presentation

Annual bleeding risk

Comment

72% supratentorial 16% infratentorial 12% both Supratentorial Superﬁcial: 55 Deep: 4 Cerebellar: 9 Brain stem: 8

50% seizures 34% headaches 9% HRG Seizures: 34 patients Focal deﬁcit: 30 patients Headaches: 20 patients Incidental: 9 patients HRG: 6 patients 61% symptomatic 39% incidental

0.25% per patient 0.10% per lesion

Review of brain MRI (1986–1989)

0.7% per lesion Not related to size and age of patients

HRG common in female and in infratentorial lesions

Symptomatic: 6.5% per patient 1.1% per lesion Overall: 13% per patient 2% per lesion If initial hemorrhagic presentation: 22.9% per year per lesion

29% of patients developed new lesions over time

1.3% prior to study 2.6% during study 0.6% (no prior HRG) 4.5% (after prior HRG)

No inﬂuence of location, sex, and number on HRG risk

11% per patient 2.5% per lesion Only 1/3 of bleeds clinically signiﬁcant

27.5% of patients developed new lesions during follow-up

2.5% per patient 5.1% rebleeding risk

Young patient with lesion >10 mm had higher risk of HRG

HRG: 62 patients Seizures: 25 patients Incidental: 23 patients

50% HRG 41%: 1 episode 7%: 2 2%: 3 23% seizures 15% headaches HRG: 19 patients Seizures: 12 patients Focal signs: 4 patients Asymptomatic: 5 patients 73% HRG 22% mass effect (without bleed) 5% asymptomatic

Younger and female patients had higher HRG risk

24 (16 had surgery)

56 (63 operations)

36 patients

68 (29 had surgery)

137 (141 brain-stem lesions)

52

Zimmerman (1991) [20]

Steinberg (2000) [18]

Samii 2001 [17]

Mathiesen (2003) [16]

Wang (2003) [19]

Ferroli (2005) [15]

HRG, hemorrhage.

Patients

First author

Medulla: 7 Pontomedullary: 3 Pons: 31 Pontomes: 3 Midbrain: 6 Mesencephalo-thalamic: 2

Basal ganglia: 11 Thalamus: 12 Mesencephalon: 5 Pons: 31 Medulla: 5 Cerebellar peduncle: 4 Midbrain: 29 Pons: 85 Medulla: 19 Cerebellar peduncle: 8

Pons: 28 Ponto-mesencephalic: 4 Medulla: 4

Midbrain: 11 Pons: 9 Medulla: 4 Brain stem: 42 Thalamus: 5 Basal ganglia: 10

Location

1 bleed: 45 patients 2 bleeds: 58 3 bleeds: 23 >3 bleeds: 11 6% annual risk 60% annual rebleeding 1 bleed: 32 patients >2 bleeds: 18 No bleed: 2 patients 3.8% annual bleeding risk 34.7% annual rebleeding

Varied according to lesion location Surgically treated patients had 1 or 2 prior bleed

Bleeding 1 episode: 20 2 episodes: 10 3 episodes: 4 4 episodes: 2

Average of 2.1 prior bleeding episodes (1 to 6)

Average of 2 prior bleed for the surgical patients

Presentation

TABLE 49-2. Summary of reports on resection of deep-seated cavernous malformations.

Complete resection in all (4 needed repeat surgery) 29 patients stable 23 patients worse

131/137 total resection 72.3% stable or improved

25/29 had complete resection

Complete resection of all lesions No surgery-related death

16% improved 55% stable 29% worse No death

12 patients stable or improved

Immediate postoperative outcome

23 patients worsened 13 transient 10 permanent 1 death

27.7% new deﬁcits or deterioration (28 patients)

69% had immediate postoperative neurologic deterioration Only 2 (5%) permanent

29% with deﬁcit 2 patients Ondine’s curse 2 patients left with hemiparesis and 1 with quadriparesis 47% cranial nerve deﬁcits 8% motor deﬁcits 33% sensory deﬁcits 14% INO

4 patients with transient deﬁcits

Morbidity

Long-term outcome

19% permanent deﬁcits

KPS >90: 44% KPS 80: 20.5% KPS 70: 20.5% KPS 50–60: 12% KPS 40: 3% Re-HRG in all 4 partial resection 20 patients improved (10 patients intact) 4 stable, 5 worse 1 died of rebleed 3 re-HRG needed second resection (1 death) 7.8% living dependently

1 patient died at 6 months (shunt infection) Others stable or improved 43% improved 52% stable 5% worse 4 patients had re-HRG

494 a. niranjan et al.

49.

cavernous malformations and other vascular diseases

The Role of Radiosurgery The successful management of cerebral AVMs with stereotactic radiosurgery prompted the exploration of its role in the management of cavernous malformations. The beneﬁts of radiosurgery for cavernous malformations are difﬁcult to assess because of its unclear natural history and the lack of an imaging technique that can document a “cure.” The role of radiosurgery in this disease is still considered controversial by many physicians. However, the lack of options for surgically inaccessible cavernous malformations has made radiosurgery a possible alternative to conservative management for lesions with a high risk of symptomatic bleeding (Table 49-3) [7, 21–28].

Dose Planning Technique for Cavernous Malformations Radiosurgery All patients must have stereotactic MRI. We prefer contrastenhanced MRI using gradient recalled acquisitions (1-mm slice thickness) supplemented with T2-weighted sequences. The malformation is deﬁned as the region characterized by mixed signal intensity within an outer hemosiderin ring, typiﬁed by low signal intensity. Hematoma eccentric from the malformation is excluded from dose planning. The prescription isodose covers the mixed signal intensity malformation but not the low signal intensity region surrounding it. This low signal intensity region contains hemosiderin pigment, which has iron in it. It is hypothesized that presence of iron could make this region more sensitive to radiation. This enhanced radiation sensitivity could cause adverse radiation effects if this area is included in prescription isodose. Single or multiple isocenter dose plans are used to construct a conformal irradiation volume that matches the malformation margin (Fig. 49-1). The 50% or greater isodose is used to prescribe margin dose. Associated developmental venous anomalies are excluded from dose planning (Fig. 49-2). Selection of the radiosurgery dose is dependent upon malformation volume and brain location. Typically, the dose is less than that used for AVMs of similar volumes. The margin dose to the cavernous malformation margin varies from 12 to 20 Gy, which is dependent upon the brain volume and location. Follow-up imaging studies and clinical evaluations are requested at 6-month intervals for the ﬁrst 2 years after radiosurgery and then annually (Fig. 49-3).

Results of Cavernous Malformations Radiosurgery The early results of radiosurgery indicated that radiosurgery might be an option for some patients with deep-seated cavernous malformations. Kida et al. reported 20 cases of symptomatic angiographically occult vascular malformations treated using radiosurgery [29]. Among 20 lesions, 14 were located supratentorially, 4 in the brain stem, and 2 in the cerebellar hemispheres. The lesions were treated using 15 to 20 Gy at the lesion margins. These investigators reported signiﬁcant reduction of rebleeding risk, as well as improvement in seizure control. Adverse effects were generally mild and well controlled by medication. Thus, the preliminary results indicated a certain usefulness of radiosurgery in the treatment of symptomatic cavernous malformations. Kondziolka et al. reported the results of radiosurgery in 47 patients who harbored a hemorrhagic malformation in a

495

critical intraparenchymal location [30]. Of these, 44 patients had experienced at least two hemorrhages before radiosurgery. At a mean follow-up of 3.6 years after radiosurgery, these authors reported signiﬁcant reductions in the proportion of patients with hemorrhage and the mean number of hemorrhages per patient. The annual hemorrhage rate in the ﬁrst 2 years after radiosurgery was 8.8%, which further decreased to 1.1% in the 2- to 6-year interval after radiosurgery. After radiosurgery, 12 (26%) patients sustained neurologic worsening, of which 8 improved on medications, 2 underwent surgical resection and died, and 2 had new permanent deﬁcits. This study provided further support for the role of radiosurgery in selected patients. Amin-Hanjani et al. analyzed 95 patients treated with proton beam therapy at the Harvard Cyclotron [21]. The average follow-up was 5.4 years. Their analysis revealed a reduction in annual hemorrhage rates from 17.3% per lesion per year before treatment to 4.5% per lesion per year after a latency period of 2 years. Seizure control was also superior after radiosurgery. These authors reported a 3% mortality rate and a 16% incidence of permanent neurologic deﬁcit. Mitchel et al. treated 18 patients with cerebral cavernous malformations with cobalt-60 source stereotactic radiosurgery [31]. The mean follow-up was 4.5 years. The annual hemorrhage rate fell from 13% before to 3.7% after radiosurgery. Three patients developed complications of radiosurgery, and two of them recovered fully. Pollock et al. evaluated the efﬁcacy and safety of radiosurgery for patients with cavernous malformations [28]. Seventeen patients underwent radiosurgery for high-surgical-risk cavernous malformations that were located in eloquent regions (thalamus/basal ganglia, brain stem, and corpus callosum). All patients had at least two documented hemorrhages before undergoing radiosurgery. The median margin radiation dose was 18 Gy, and the median maximum dose was 32 Gy. The annual hemorrhage rate during the 51 months preceding radiosurgery was 40.1% compared with 8.8% in the ﬁrst 2 years after radiosurgery and 2.9% thereafter. These investigators reported a high (41%) permanent radiation-related morbidity rate with 18 Gy as margin dose. Hasegawa et al. studied outcome of Gamma Knife radiosurgery for 82 symptomatic patients who had imagingconﬁrmed hemorrhages for which resection was believed to be associated with high risk [7]. Most patients had multiple hemorrhages from brain-stem or diencephalic cavernous malformations. During an observation of 354 patient-years prior to radiosurgery, 202 hemorrhages were observed, for an annual hemorrhage rate of 33.9%, excluding the ﬁrst hemorrhage. After radiosurgery, during a total of 401 patient-years, 19 hemorrhages were identiﬁed, of which 17 occurred in the ﬁrst 2 posttreatment years, and only two after 2 years had elapsed. The annual hemorrhage rate was 12.3% per year for the ﬁrst 2 years after radiosurgery. The annual hemorrhage rate dropped to 0.76% per year after 2 years. Eleven (13.4%) patients developed new neurologic symptoms without hemorrhage after radiosurgery. The symptoms were minor in six of these patients and temporary in ﬁve. These investigators concluded that radiosurgery conferred a reduction in the risk of hemorrhage for high-risk cavernous malformations. This reduction was most pronounced after 2 years [7].

Margin: 16 Gy (9–36) Vol.: 0.9 cm3 (0.06–12.5) HRG in 59 patients Seizures in 40 patients Neuro deﬁcit 51patients 7 patients had prior partial resection

112 patients 48 months

Liscak (2005) [26], Gamma Knife

75% seizure cessation without medication after a mean of 31.3 months Only one rebleed after Gamma Knife (occurred 13 months later) Rebleed rate of 1.6%/ year 2 deaths from HRG 33% of patients with neuro deﬁcits improved

Margin: 14.55 Gy (10–25)

28.6% seizures 26.6% bleed 45.2% focal deﬁcits

42 patients 29.6 months

Kim (2005) [25], Gamma Knife

HRG rate 6.5% per year after Gamma Knife 10.3% ﬁrst 2 years 3.3% after year 2

Margin: 12.1 Gy (9–20) Vol.: 3.12 cm3 (0.032–25.9)

HRG: 112 patients (45 with more than 1 bleed, HRG rate 29.2% per year) Seizures: 28 patients

125 patients 5.4 years

Liu (2005) [27], Gamma Knife

HRG rate reduced to 1.55%/year after SRS

Margin: 16.1 Gy (8–24) Vol.: 1.42 cm3 (0.09–4.8)

22 patients 47 months (linac) 29 months (Gamma Knife)

Kim (2002) [24], Gamma Knife and linac

HRG: 20 patients (1 HRG, 12; 2 HRG, 6; 3 HRG, 2) HRG rate: 35.5% Seizures: 3 patients

82 patients 4.89 years

Hasegawa (2002) [7], Gamma Knife

HRG risk reduced to 8.8% the ﬁrst 2 years postGamma Knife, and 2.9% thereafter HRG rate reduced after Gamma Knife: 12.3% for the ﬁrst 2 years 0.76% years 2 to 12

Margin dose: 18 Gy (16–18) Vol.: 2.1 cm3 (0.5–7.9) Margin: 16.2 Gy (12–20) Vol.: 1.85 cm3 (0.12–6.98)

HRG, hemorrhage; Vol., volume; ARE, adverse radiation effect; T, transient; P, permanent.

Brain stem: 33 Cortical: 54 Thalamus: 9 Basal ganglia: 8 Cerebellum: 8

Supratentorial Superﬁcial: 16 Deep: 13 Brain stem: 52 Cerebellum: 1 Brain stem: 10 Cerebellum: 4 Corpus callosum: 1 Intern. caps.: 1 Cerebral hemisphere: 6 Brain stem: 49 Basal ganglia/thalamus: 14 Cortical: 39 Cerebellum: 10 Multiple: 13 Cortical: 21 Brain stem: 6 Basal ganglia: 5 Cerebellum: 3 Multiple: 7

Thalamus/basal ganglia: 4 Brain stem: 12 Corpus callosum: 1

17 patients 51 months

Pollock (2000) [28], Gamma Knife

Margin dose: 7–40 CGE (proton) 12–20 Gy (linac) Vol.: 2.25 cm3 (0.08–15.2)

HRG in all 8 patients had prior partial resection 5 patients had prior VP shunts 4 patients also suffered form seizures Multiple HRG in all patients HRG rate: 6.4% Rebleeding: 24.8% HRG in all patients (1 to 7 bleeds) Annual HRG rate: 33.9% Prior surgery: 20 patients

Brain stem: 31 Thalamus: 13 Basal ganglia: 7 Cortical: 5 Hypothalamus: 1

57 patients 7.5 years

Chang (1998) [22], proton beam and linac

Post-Gamma Knife HRG rate: 8%/year Years 1–2: 10% Years 3–4: 12% Years 5–6: 5% Years 7–8: 7% Surgery needed in 6 patients 20 patients rebled (2 died) Annual risk: 9.4% (ﬁrst 3 years) 1.6% after 3 years Surgery needed in 8 patients

Margin: 18 Gy (9–35) Volume not speciﬁed

HRG: 16 patients Epilepsy: 6 patients 1 patient had prior partial resection

Cerebral hemisphere: 8 Deep cerebral: 7 Brain stem: 6 Cerebellum: 1

22 patients 6.9 years

Karlsson (1998) [23], Gamma Knife

HRG risk reduced to 10.4% 22.4% (ﬁrst 2 years) 4.5% (after 2 years) 3 deaths due to HRG 5 patients needed surgery

Margin: 15 Gy (6.5–30) Vol.: 3.1 cm3 (0.2–38.8)

HRG: 74% (17.4%/year) Seizures: 18%

Lobar: 24 (25%) Deep: 18 (18%) Brain stem: 56 (57%)

95 patients 98 lesions total 5.44 ± 2.97 years

Amin-Hanjani (1998) [21], proton beam

Outcome

Dose/volume

Presentation

Location

Patients/follow-up

First author

TABLE 49-3. Summary of radiosurgical series of cavernous malformations.

Edema on MR in 30 patients (27%) Symptomatic 17 patients 16 patients recovered P 0.9%

5 patients with symptomatic edema with subsequent improvement

Radiation effect on MRI in 13.1% Symptomatic in only 2.5% of patients

ARE: 6 patients (4 T, 2 P) 5 to 12 months after SRS Tended to occur more after linac

ARE: 11 patients (13.4%) (5 T, 6 P.) None after 1992

ARE: (59%) 5–16 months after Gamma Knife 7 (41%) P

Symptomatic edema 7% (4 patients) T in 3 patients Radiation necrosis 2% Increased seizure activity 2% (1 patient)

ARE: 26.5% 61.5% P 26.9% T 3 patients dead Occurred within 2 years in 85% ARE: 6 patients (1 T, 5 P) Onset 6–41 months after Gamma Knife (mean 16)

Morbidity

496 a. niranjan et al.

49.

cavernous malformations and other vascular diseases

497

FIGURE 49-1. Gamma Knife radiosurgery dose plan for cavernous malformation of brain stem. Fifty percent isodose line (white line) projected on axial contrast-enhanced MR images with sagittal and

coronal reconstruction covers the malformation but stays within low signal region.

Kim et al. evaluated the efﬁcacy of radiosurgery in symptomatic cavernous malformations for which the surgical risk was thought to be unacceptable. These investigators treated the ﬁrst 11 cases with linac radiosurgery using CT-based planning and the next 11 cases with Gamma Knife radiosurgery using magnetic resonance (MR)-based planning. The volume of the lesion ranged from 0.09 cm3 to 4.8 cm3 (mean, 1.42 cm3) and the mean marginal dose was 16.1 Gy (range, 8 to 24 Gy). The median follow-up period after radiosurgery was 38.3 months. In the group with prior hemorrhage, the bleeding rate of cavernous malformation after radiosurgery (1.55%/year) was lower than that during the pre-radiosurgical period (35.5%/ year). Six patients showed neurologic deterioration after radiosurgery; however, the neurologic deﬁcits persisted in only two of the patients treated with linac. These authors suggested a possible correlation of linac radiosurgery with radiation-induced neurologic deﬁcits. The size of the lesion decreased in 11 patients, increased in 1, and did not change in 10 cases. A perilesional hyperintense signal on T2-weighted MR images was seen in nine patients and correlated with the treatment modality used, being more frequent after linac radiosurgery [24]. Kim et al. evaluated 42 cavernous malformation patients treated with Gamma Knife surgery with 14.55 Gy as mean margin dose [25]. The mean follow-up period after radiosurgery was 29.6 months (range, 5 to 93 months). The tumor decreased in size in 29 cases, was unchanged in 12, and increased in size in 1. In the seizure group, seizures were controlled without anticonvulsant

medication in nine cases (81.8%) after 31.3 months. Rebleeding occurred in only one (2.3%) case. On T2-weighted imaging, changes were seen in 11 (26.2%) cases, and neurologic deterioration was correlated with imaging changes in 3 (7.1 %). These authors concluded that patients who receive a marginal dose below 15 Gy had better outcome than those treated with higher doses [25]. Liscak et al. followed 107 patients treated with Gamma Knife radiosurgery using an average of 16 Gy margin dose [26]. After a median of 48 months, lesion regression was observed in many patients. These authors reported a transient morbidity (rebleeding and edema) rate of 20.5% and permanent morbidity rate of 4.5% [26].

Radiobiological Considerations Post-radiosurgery AVM obliteration is likely due to endothelial cell proliferation leading to vessel wall hyalinization, ﬁbrosis, and myoﬁbroblast-induced vessel wall contracture. We believe that the response of cavernous malformation to irradiation is similar. It is reasonable to expect a latency interval of several years for hemorrhage protection as occurs before AVM obliteration. At present, there are no good histologic studies of irradiated cavernous malformations in patients who were thought to be successfully managed. Histologic studies of irradiated patients who later underwent cavernous malformations resection could not draw any conclusion regarding the mechanism of efﬁcacy [32]. However, a recent case report suggests

498

a. niranjan et al.

FIGURE 49-2. Gamma Knife radiosurgery dose plan for cavernous malformation of left basal ganglia. Fifty percent isodose line (white line) projected on axial contrast-enhanced MR images with sagittal and

coronal reconstruction covers the malformation. An associated developmental venous anomaly (arrow) is noted.

that the histopathologic ﬁndings after irradiation on these lesions may be similar to those described in arteriovenous malformations after Gamma Knife surgery [33]. A signiﬁcant decrease in the symptomatic hemorrhage rate after stereotactic radiosurgery of cerebral cavernous malforma-

tions suggests that this technique may be an effective management strategy for patients with hemorrhagic malformations in high-risk brain locations. The patients that were selected for radiosurgery harbored malformations with a hemorrhage rate much greater than that typically seen. There is a subset of

FIGURE 49-3. A 28-year-old woman presented with a history of two hemorrhagic episodes from cavernous malformation in her anterior right pons. Axial MR gadolinium-enhanced MRI (left) showed pontine cavernous malformation at the time of radiosurgery. A margin dose of

16 Gy was delivered to 50% isodose line (middle). A 4-year follow-up MRI shows regression in cavernous malformation with stable appearance of brain stem.

49.

cavernous malformations and other vascular diseases

patients who seem to bleed more frequently than others [10, 12, 34]. The reduced hemorrhage rate achieved after a 2-year latency from radiosurgery (1%) approaches the low risk (0.6%) identiﬁed in many natural history studies for nonhemorrhagic cavernous malformations. All reports on cavernous malformation radiosurgery are based on clinical outcome because cavernous malformation obliteration cannot be documented using any of the available imaging modality. Without this documentation, patients cannot be declared as “cured” after radiosurgery. Although patients may be encouraged by a reduction in hemorrhage risk, they cannot be told that this risk is eliminated. Although radiosurgery signiﬁcantly lowers the risk of hemorrhage, there is potential for complications and continued lesion progression after radiosurgery. These risks and beneﬁts must be carefully balanced against the natural history of untreated lesions if the use of radiosurgery is considered.

Radiosurgery for Other Vascular Abnormalities Dural Arteriovenous Fistulas Dural arteriovenous ﬁstulas (AVFs) represent approximately 10% to 15% of all intracranial vascular abnormalities [35]. They are direct connections between the meningeal arterial system and dural venous sinuses or cortical veins. Most are thought to be acquired lesions, secondary to dural venous channel occlusion leading to recruitment of collateral blood supply. Trauma, infection, and hypercoagulable states are risk factors for the development of dural AVFs. The majority involve the transverse, sigmoid, and cavernous sinuses. These are generally low-ﬂow lesions, which present with bruits and pulsatile tinnitus. Cavernous sinus lesions can also present with chemosis, ophthalmoplegia, and eventually lead to loss of vision. Other notable locations include the superior sagittal sinus and tentorium; these tend to follow a more aggressive course. Regardless of their locations, dural AVFs can cause subarachnoid or intraparenchymal hemorrhage. The bleeding risk is related primarily to the presence of leptomeningeal venous drainage, dural sinus occlusion, and aneurysmal venous dilation [36]. Dural AVFs can be classiﬁed based on the pattern of venous drainage and patency of dural sinus involved [37]. Type I ﬁstulas drain in an anterograde fashion in a patent dural sinus. Type IIa ﬁstulas drain retrogradely in an obstructed sinus, and type IIb drain retrogradely into leptomeningeal veins through a venous sinus. Type IIa + b ﬁstulas combine the features of both previous categories. Type III ﬁstulas involve direct connections between meningeal arteries and cortical veins, without intervening dural sinus. Type IV is the same as type III, with added variceal dilation of the cortical veins. Type V represents a ﬁstula draining into perimedullary veins. The risk of bleeding increases directly with the ﬁstula grade for types I through IV. Type V ﬁstula usually presents with myelopathy. Treatment of dural AVFs varies according to the location and grade of the lesion. Endovascular embolization through a transarterial or transvenous route is the mainstay of treatment for simple, low-ﬂow lesions involving the transverse-sigmoid or cavernous regions, with high cure rates [38]. High-risk lesions are usually treated with multimodality management comprising

499

embolization and microsurgery [39]. Stereotactic radiosurgery has recently emerged as a minimally invasive adjunct or alternative management modality for AVFs (Fig. 49-4). Link et al. reported the results of Gamma Knife radiosurgery for 29 patients with dural AVFs [40]. Fourteen patients had cortical venous drainage and 14 had dural sinus occlusion. Five patients had prior hemorrhage. One patient had previously failed surgical ligation, and two had failed prior embolization. Nidus volume ranged from 0.35 to 12.4 cm3, with a mean of 3.3 cm3. Mean dose delivered to the margin was 19.2 Gy (range, 18 to 20 Gy). Seventeen patients had subsequent embolization within 48 hours after radiosurgery. Clinically, 52% of patients improved and 31% had stable nondisabling symptoms. Only one patient had treatment-related morbidity, a transient facial paresis related to embolization. No further bleeding episode was seen. Follow-up angiogram was obtained in 18 patients and demonstrated ﬁstula obliteration in 13 patients and shrinkage in 5 patients. Obliteration occurred within 1 to 3 years. Koebbe et al. described 18 dural AVF patients managed with Gamma Knife radiosurgery [41]. Three patients presented with intracerebral hemorrhage. Embolization was performed in 10 patients (prior to radiosurgery in 9 patients and after in 2 patients). Margin dose ranged from 15 to 30 Gy (mean, 20 Gy). The mean nidus volume was 2.16 cm3. Complete resolution of symptoms was seen in nine patients, with the nine others experiencing signiﬁcant symptom reduction. Follow-up imaging demonstrated ﬁstula obliteration in 12 patients (8 on angiography and 4 on MRA) and decrease in size in 3 patients. Three patients did not have follow-up imaging. No further hemorrhage was seen after the Gamma Knife procedure. One patient experienced transient symptomatic radiation-induced edema, and two patients with cavernous sinus ﬁstulas had a permanent complication from embolization (facial numbness and sixth nerve palsy). Clinical outcomes were better for patients treated with radiosurgery alone compared with the combined treatment. Friedman et al. reported the outcome of 23 patients with transverse-sigmoid ﬁstulas treated with combined radiosurgery followed by transarterial embolization in most cases [42]. Symptoms completely resolved in 87% of patients, with 9% also experiencing signiﬁcant relief. Out of 17 patients who had follow-up angiography, 7 (41%) were cured, 4 (24%) had more than 90% obliteration, and 6 (35%) had more than 50% obliteration of the ﬁstula. No patient had radiation-induced complication, and one patient suffered a transient embolization-related ischemic episode. Two patients had ischemic events after diagnostic angiography (transient in one case and permanent in the other). Pan et al. published the results of Gamma Knife radiosurgery as primary treatment of patients (n = 20) with transversesigmoid AVFs [43]. Tinnitus and headache were the most common symptoms. Three patients had previously bled. Two patients had failed prior surgery and embolization. Marginal doses ranged from 16.5 to 19 Gy. Complete obliteration associated with resolution of symptoms was seen in 58%, with an additional 16% having near-total occlusion after median followup of 19 months. Temporary hair loss over the mastoid region was the only adverse effect of irradiation, seen in 10 patients. Guo et al. reported the outcome of patients who had radiosurgery for cavernous sinus dural AVFs [44]. Fifteen patients were

500

a. niranjan et al.

FIGURE 49-4. (A) A 46-year-old woman presented with a 3-week history of high-pitched pulsatile tinnitus in the left ear. Cerebral angiogram revealed a dural vascular malformation of the jugular bulb. This malformation was treated with radiosurgery followed by embolization. Radiosurgery dose planning was performed on MR images supple-

mented with cerebral angiograms. GammaPlan shows dose plan projected on axial MR poster with sagittal and coronal reconstruction. A margin dose of 17 Gy was delivered to the 50% isodose line. (B) Gamma Knife dose plan projected on digital cerebral angiograms.

irradiated with a maximum dose ranging from 22 to 38 Gy (mean, 28 Gy). The amount was selected with the goal of keeping the optic pathway dose below 8 Gy. In 12 (80%) patients, complete obliteration was seen, with the other three having only small shunts remaining. No patient had worsening of symptoms or complication after radiosurgery. Pollock et al. managed 20 patients with cavernous sinus ﬁstulas using radiosurgery alone (7 patients) or followed by embo-

lization (13 patients) [45]. Clinical symptoms improved in 95% of patients. Follow-up angiogram was obtained for 15 patients and demonstrated total obliteration in 13 and near-total in 1 patient. No subsequent hemorrhage occurred. Morbidity was seen in two patients after embolization procedures. Finally, Lewis et al. reported the results of transarterial embolization followed by linac radiosurgery (seven patients) or surgery (two patients) in a cohort of patients with high-risk tentorial dural

49.

cavernous malformations and other vascular diseases

AVFs [46]. Five patients had experienced prior hemorrhage. In the radiosurgical group, dose delivered ranged from 8 to 20 Gy (mean, 15.6 Gy). Four of seven patients showed complete thrombosis of the lesion 24 months after the procedure, and the others had signiﬁcant occlusion. One patient suffered from transient radiation injury to the brain stem 14 months after radiosurgery. The outcome data suggests that stereotactic radiosurgery provides signiﬁcant beneﬁt in the management of dural AVFs. The obliteration rate is high with minimal radiation-induced morbidity. It is a valuable addition in the multimodality management of high-risk dural AVFs and also as primary treatment of low-ﬂow dural AVFs of the transverse, sigmoid, and cavernous regions, with or without adjunct embolization. Radiosurgery followed by embolization appears to be a sound strategy for patients with bothersome symptoms such as tinnitus. Radiosurgery should be performed ﬁrst because the whole nidus can be better deﬁned and treated prior to embolization. It is followed by embolization, which can provide immediate relief from acute symptoms.

Vein of Galen Malformations Vein of Galen malformations (VGM; also known as vein of Galen aneurysms) are rare vascular abnormalities that are usually found in the pediatric population. They constitute approximately 1% of all intracranial vascular malformations [47]. They are direct ﬁstulous connections between the cerebral arterial vasculature and the galenic venous system, often associated with straight sinus stenosis or occlusion and persistence of a falcine sinus. Clinical presentation is dependent on the age at which it manifests. Neonates usually show signs of overt highoutput cardiac failure, and older patients can present with hydrocephalus or intracranial hemorrhage. Yasargil classiﬁed VGM into four different types based on the arterial supply of the ﬁstula [48]. Type I malformations are direct anastomosis between branches of the distal posterior cerebral artery or the pericallosal artery and the vein of Galen. Type II malformations constitute connections between thalamoperforating arteries and the vein of Galen. Type III malformations combine the features of type I and II. Type IV malformations are classic parenchymal AVMs with venous drainage in the galenic system and are not true VGM. Treatment of VGM usually consists of staged endovascular embolization, with the intent of progressively reducing the arteriovenous shunt and avoiding complications of cerebral venous congestion. Neonates with cardiac decompensation need to be medically stabilized and have urgent treatment, but elective embolization can be done for stable patients. Microsurgery is technically demanding and is best reserved for type I and simple type III malformations, which are not amenable to endovascular procedures [49]. There have been few reports of radiosurgical management of VGM. Watban et al. reported three infantile cases that were successfully obliterated by transarterial embolization 18 to 24 months after having failed radiosurgery [50]. Margin doses used varied from 20 to 25 Gy. Payne et al. studied the outcomes of Gamma Knife radiosurgery in a series of nine patients with VGM [51]. The population consisted of 8 children (aged from 4 to 14 years) and 1 adult. Four patients presented with hemor-

501

rhage, two with increased head circumference, two with headaches, and one with hydrocephalus. Three patients had a Yasargil type I lesion, one type II, two type III, and three type IV. Four had failed prior endovascular treatment. Three patients required two Gamma Knife procedures. The mean marginal dose used was 20.5 Gy (range, 15 to 25 Gy). Out of eight patients who underwent follow-up cerebral angiogram, four had complete obliteration and four showed signiﬁcant obliteration (50%, 80%, 80%, and 90%) with a follow-up ranging from 15 months to 8 years. One patient had transient symptomatic radiation-induced edema after two Gamma Knife procedures, and MRI showed evidence of edema in another patient who remained asymptomatic. These promising results demonstrate that stereotactic radiosurgery can be of beneﬁt in the management of stable VGM patients who can tolerate the time interval needed for obliteration.

References 1. Robinson JR, Awad IA, Little JR. Natural history of the cavernous angioma. J Neurosurg 1991; 75(5):709–714. 2. Zabramski JM, Wascher TM, Spetzler RF, et al. The natural history of familial cavernous malformations: results of an ongoing study. J Neurosurg 1994; 80(3):422–432. 3. Liquori CL, Berg MJ, Siegel AM, et al. Mutations in a gene encoding a novel protein containing a phosphotyrosine-binding domain cause type 2 cerebral cavernous malformations. Am J Hum Genet 2003; 73(6):1459–1464. 4. Liquori CL, Berg MJ, Squitieri F, et al. Low frequency of PDCD10 mutations in a panel of CCM3 probands: potential for a fourth CCM locus. Hum Mutat 2006; 27(1):118. 5. Verlaan DJ, Laurent SB, Rochefort DL, et al. CCM2 mutations account for 13% of cases in a large collection of kindreds with hereditary cavernous malformations. Ann Neurol 2004; 55(5): 757–758. 6. Labauge P, Brunereau L, Levy C, Laberge S, Houtteville JP. The natural history of familial cerebral cavernomas: a retrospective MRI study of 40 patients. Neuroradiology 2000; 42(5):327–332. 7. Hasegawa T, McInerney J, Kondziolka D, et al. Long-term results after stereotactic radiosurgery for patients with cavernous malformations. Neurosurgery 2002; 50(6):1190–1197; discussion 1197– 1198. 8. Ide C, De Coene B, Baudrez V. MR features of cavernous angioma. JBR-BTR 2000; 83(6):320. 9. Hallam DK, Russell EJ. Imaging of angiographically occult cerebral vascular malformations. Neuroimaging Clin N Am 1998; 8(2): 323–347. 10. Aiba T, Tanaka R, Koike T, et al. Natural history of intracranial cavernous malformations. J Neurosurg 1995; 83(1):56–59. 11. Del Curling O Jr, Kelly DL Jr, Elster AD, Craven TE. An analysis of the natural history of cavernous angiomas. J Neurosurg 1991; 75(5):702–708. 12. Kondziolka D, Lunsford LD, Kestle JR. The natural history of cerebral cavernous malformations. J Neurosurg 1995; 83(5):820– 824. 13. Kupersmith MJ, Kalish H, Epstein F, et al. Natural history of brainstem cavernous malformations. Neurosurgery 2001; 48(1):47– 53; discussion 54. 14. McLaughlin MR, Kondziolka D, Flickinger JC, et al. The prospective natural history of cerebral venous malformations. Neurosurgery 1998; 43(2):195–200; discussion 201. 15. Ferroli P, Sinisi M, Franzini A, et al. Brainstem cavernomas: longterm results of microsurgical resection in 52 patients. Neurosurgery 2005; 56(6):1203–1212; discussion 1212–1214.

502

a. niranjan et al.

16. Mathiesen T, Edner G, Kihlstrom L. Deep and brainstem cavernomas: a consecutive 8-year series. J Neurosurg 2003; 99(1):31– 37. 17. Samii M, Eghbal R, Carvalho GA, Matthies C. Surgical management of brainstem cavernomas. J Neurosurg 2001; 95(5):825– 832. 18. Steinberg GK, Chang SD, Gewirtz RJ, Lopez JR. Microsurgical resection of brainstem, thalamic, and basal ganglia angiographically occult vascular malformations. Neurosurgery 2000; 46(2):260– 270; discussion 270–271. 19. Wang CC, Liu A, Zhang JT, et al. Surgical management of brainstem cavernous malformations: report of 137 cases. Surg Neurol 2003; 59(6):444–454; discussion 454. 20. Zimmerman RS, Spetzler RF, Lee KS, et al. Cavernous malformations of the brain stem. J Neurosurg 1991; 75(1):32–39. 21. Amin-Hanjani S, Ogilvy CS, Candia GJ, et al. Stereotactic radiosurgery for cavernous malformations: Kjellberg’s experience with proton beam therapy in 98 cases at the Harvard Cyclotron. Neurosurgery 1998; 42(6):1229–1236; discussion 36–38. 22. Chang SD, Levy RP, Adler JR Jr, et al. Stereotactic radiosurgery of angiographically occult vascular malformations: 14-year experience. Neurosurgery 1998; 43(2):213–220; discussion 220–221. 23. Karlsson B, Kihlstrom L, Lindquist C, et al. Radiosurgery for cavernous malformations. J Neurosurg 1998; 88(2):293–297. 24. Kim DG, Choe WJ, Paek SH, et al. Radiosurgery of intracranial cavernous malformations. Acta Neurochir (Wien) 2002; 144(9):869–878; discussion 878. 25. Kim MS, Pyo SY, Jeong YG, et al. Gamma knife surgery for intracranial cavernous hemangioma. J Neurosurg 2005; 102(Suppl): 102–106. 26. Liscak R, Vladyka V, Simonova G, et al. Gamma knife surgery of brain cavernous hemangiomas. J Neurosurg 2005; 102(Suppl): 207–213. 27. Liu KD, Chung WY, Wu HM, et al. Gamma knife surgery for cavernous hemangiomas: an analysis of 125 patients. J Neurosurg 2005; 102(Suppl):81–86. 28. Pollock BE, Garces YI, Stafford SL, et al. Stereotactic radiosurgery for cavernous malformations. J Neurosurg 2000; 93(6):987– 991. 29. Kida Y, Kobayashi T, Tanaka T. Treatment of symptomatic AOVMs with radiosurgery. Acta Neurochir Suppl 1995; 63:68–72. 30. Kondziolka D, Lunsford LD, Flickinger JC, Kestle JR. Reduction of hemorrhage risk after stereotactic radiosurgery for cavernous malformations. J Neurosurg 1995; 83(5):825–831. 31. Mitchell P, Hodgson TJ, Seaman S, et al. Stereotactic radiosurgery and the risk of haemorrhage from cavernous malformations. Br J Neurosurg 2000; 14(2):96–100. 32. Gewirtz RJ, Steinberg GK, Crowley R, Levy RP. Pathological changes in surgically resected angiographically occult vascular malformations after radiation. Neurosurgery 1998; 42(4):738–742; discussion 742–743. 33. Nyary I, Major O, Hanzely Z, Szeifert GT. Histopathological ﬁndings in a surgically resected thalamic cavernous hemangioma 1 year after 40-Gy irradiation. J Neurosurg 2005; 102(Suppl): 56–58. 34. Tung H, Giannotta SL, Chandrasoma PT, Zee CS. Recurrent intraparenchymal hemorrhages from angiographically occult vascular malformations. J Neurosurg 1990; 73(2):174–180.

35. Houser OW, Baker HL Jr, Rhoton AL Jr, Okazaki H. Intracranial dural arteriovenous malformations. Radiology 1972; 105(1):55– 64. 36. Awad IA, Little JR, Akarawi WP, Ahl J. Intracranial dural arteriovenous malformations: factors predisposing to an aggressive neurological course. J Neurosurg 1990; 72(6):839–850. 37. Cognard C, Gobin YP, Pierot L, et al. Cerebral dural arteriovenous ﬁstulas: clinical and angiographic correlation with a revised classiﬁcation of venous drainage. Radiology 1995; 194(3):671– 680. 38. Halbach VV, Higashida RF, Hieshima GB, David FD. Endovascular therapy of dural ﬁstulas. In: Vinuela F, Halbach VV, Dion J, eds. Interventional Neuroradiology: Endovascular Therapy of the Central Nervous System. New York: Raven, 1992:29–50. 39. Cawley CM, Barrow DL, Dion JE. Treatment of lateral-sigmoid and sagittal sinus dural arteriovenous malformations. In: Winn HR, ed. Youmans Neurological Surgery, 5th ed. Philadelphia: Saunders, 2004:2283–2291. 40. Link MJ, Coffey RJ, Nichols DA, Gorman DA. The role of radiosurgery and particulate embolization in the treatment of dural arteriovenous ﬁstulas. J Neurosurg 1996; 84(5):804–809. 41. Koebbe CJ, Singhal D, Sheehan J, et al. Radiosurgery for dural arteriovenous ﬁstulas. Surg Neurol 2005; 64(5):392–398; discussion 398–399. 42. Friedman JA, Pollock BE, Nichols DA, et al. Results of combined stereotactic radiosurgery and transarterial embolization for dural arteriovenous ﬁstulas of the transverse and sigmoid sinuses. J Neurosurg 2001; 94(6):886–891. 43. Pan DH, Chung WY, Guo WY, et al. Stereotactic radiosurgery for the treatment of dural arteriovenous ﬁstulas involving the transverse-sigmoid sinus. J Neurosurg 2002; 96(5):823–829. 44. Guo WY, Pan DH, Wu HM, et al. Radiosurgery as a treatment alternative for dural arteriovenous ﬁstulas of the cavernous sinus. AJNR Am J Neuroradiol 1998; 19(6):1081–1087. 45. Pollock BE, Nichols DA, Garrity JA, et al. Stereotactic radiosurgery and particulate embolization for cavernous sinus dural arteriovenous ﬁstulae. Neurosurgery 1999; 45(3):459–466; discussion 66–67. 46. Lewis AI, Tomsick TA, Tew JM Jr. Management of tentorial dural arteriovenous malformations: transarterial embolization combined with stereotactic radiation or surgery. J Neurosurg 1994; 81(6):851–859. 47. Drake CG. Cerebral arteriovenous malformations: considerations for and experience with surgical treatment in 166 cases. Clin Neurosurg 1979; 26:145–208. 48. Yasargil MG. Microneurosurgery. New York: Thieme, 1988. 49. Mickle JP, Mericle RA, Burry MV, Williams LS. Vein of Galen malformations. In: Winn HR, ed. Youmans Neurological Surgery, 5th ed. Philadelphia: Saunders, 2004:3433–3445. 50. Watban JA, Rodesch G, Alvarez H, Lasjaunias P. Transarterial embolization of vein of Galen aneurysmal malformation after unsuccessful stereotactic radiosurgery. Report of three cases. Childs Nerv Syst 1995; 11(7):406–408. 51. Payne BR, Prasad D, Steiner M, et al. Gamma surgery for vein of Galen malformations. J Neurosurg 2000; 93(2):229–236. 52. Kim DS, Park YG, Choi JU, et al. An analysis of the natural history of cavernous malformations. Surg Neurol 1997; 48(1):9–17; discussion 18.

5 0

Cerebral Cavernous Malformations: Surgical Perspective Robert L. Dodd and Gary K. Steinberg

Introduction Cerebral cavernous malformations (CCMs) are rare vascular lesions [1, 2]; however, because hemorrhage of CCMs can result in signiﬁcant morbidity and mortality [3, 4], a great deal of effort has been devoted to their detection and treatment. Although previously thought to be solely congenital, it is now well recognized that many cavernous malformations are acquired [5], arising de novo or occasionally after radiotherapy. The natural history of cavernous malformations is only now being elucidated with the widespread use of magnetic resonance imaging (MRI). Preliminary reports suggest the annual risk of clinically signiﬁcant hemorrhage to be less than 1% [1–4], although many factors may increaHse this risk. In general, the results of surgical removal of accessible symptomatic cavernomas are excellent, with improved control of medically intractable seizures, restoration of neurologic function, and decreased risk of future hemorrhage. Surgically inaccessible lesions remain a difﬁcult clinical challenge.

Epidemiology The prevalence of CCMs remains undetermined; however, most reports estimate their occurrence in approximately 0.3% to 0.9% of the general population [1, 2]. These estimates have been made on the basis of autopsy results and prospective cohort studies using MRI, which suggest cavernous malformations comprise 8% to 15% of all vascular lesions [6, 7]. The incidence of these lesions is approximately 0.5 to 0.6 per 100,000 people per year [1, 2], of which more than 40% are clinically symptomatic, most of whom present with seizures. The age distribution of patients with clinically symptomatic cavernous malformations is bell shaped, with the highest incidence between the third and ﬁfth decades [8]. New presentation of cavernous malformations in infants or in adults over the age of 80 years, though documented, is rare. Whereas a few published reports have found slight male or female preponderance, most studies

with large patient populations concur that there is no difference in the sex incidence of CCMs. Cavernous malformations are deﬁned as an assembly of thin-walled vascular sinusoids without intervening brain parenchyma, lined by a single layer of endothelium, and devoid of mature vessel elements such as smooth muscle and elastin. In contrast with other vascular lesions, cavernous malformations rarely bleed into the subarachnoid space or into the ventricular system; more commonly, these lesions bleed into the sinusoidal space of the malformation itself, resulting in acute enlargement or intraparenchymal hemorrhage. Although these vascular malformations are low-ﬂow and undergo continual microhemorrhaging, clinically signiﬁcant intralesional hemorrhage or gross hemorrhage beyond the lesion is rare, with an estimated risk between 0.25% to 6% per year [3, 9–14]. Hypertension, cigarette smoking, oral contraceptives, alcohol consumption, and cocaine use are all known risk factors for intracerebral hemorrhage in general, but no study to date has shown increased risk in patients who harbor cavernous malformations. The most widely cited risk factor for clinically signiﬁcant hemorrhage, apart from family history, is prior hemorrhage [3, 11, 13, 14]. Interestingly, whereas the high frequency, rapidity, and gravity of hemorrhagic recurrences after the ﬁrst intracranial hemorrhage has been stressed by some, other series have shown little inﬂuence. Several authors have found a female preponderance in bleeding risk, suggesting that endocrine factors may inﬂuence hemorrhage tendencies, particularly as some of the bleeding episodes occurred during pregnancy [3, 13, 15]. These observations are supported by the detection of estrogen receptors in a few cavernous malformations obtained from women [16], as well as by speculation that hormonal ﬂuctuations during pregnancy increase endothelial cell proliferation in these lesions and their propensity to hemorrhage [13]. The occurrence of cavernous malformations in more than one family member is uncommon; however, a familial form of the disease has been ﬁrmly established, with identiﬁcation of three autosomal dominant loci mapped to human chromosomes 7q21–22 (CCM1), 7p13–15 (CCM2), and 3q25.2–27 (CCM3)

503

504

r.l. dodd and g.k. steinberg

[8]. Most patients (50% to 80%) with cavernous malformation are however sporadic. More than 42 different single-gene (KRIT1) mutations have been found in CCM1-linked families, which are believed to cause cavernous malformation induction in 40% of familial cases, with genes at CCM2 and CCM3 accounting for the remaining 60% [8]. Reports of cavernous malformations in familial series indicate the majority of these individuals had multiple lesions (up to 93%), were clinically symptomatic, and presented at a younger age.

Anatomic Distribution Cavernous malformations can be found throughout the central nervous system including every region of the brain, brain stem, spinal cord, cranial nerves, and ventricles. The cerebral hemispheres are the most frequently affected areas (in one study 21% frontal, 16% parietal, 15% temporal, and 3.5% occipital), with most being subcortical in location [17]. Recently, more cases of lesions in the basal ganglia, corpus callosum, thalamus, and third ventricle have been reported because of the interest generated by their therapeutic complexity. Between 10% and 23% of cavernous malformations are located infratentorially, where the pons is the preferred site [16–20]. Solitary lesions are more common than multiple lesions, and multiplicity should always raise the suspicion of familial form of the condition. Multiple intracranial cavernous malformations are sometimes associated with similar vascular lesions in other organs, suggesting the presence of a wider and more systemic condition.

Natural History Because decisions on further treatment recommendations are based to a great extent on the estimated risk of further morbidity, knowledge regarding the natural history of cavernous malformations is an important consideration in clinical practice. Despite recognition of this point, our current knowledge is incomplete because of a number of factors: the low frequency of occurrence limits the size of study populations; the relatively recent introduction of MRI technology required for accurate diagnosis limits the length of clinical follow-up; the paucity of prospective studies analyzing the natural history of cavernous malformations; and confusion in the literature as to what constitutes a clinically signiﬁcant hemorrhage. Nevertheless, several recent studies have provided some estimates for risk of bleeding and growth [8, 12, 21, 22]. The risk of clinically relevant hemorrhage after detection of a cavernous malformation depends on the presenting symptoms and on the location of the lesion. Preliminary reports indicate that the frequency of clinical hemorrhage among patients with supratentorial nonfamilial cavernous malformations who present either with incidental diagnosis or with seizures is approximately 0.25% to 2% per year [2, 3, 11, 13]. The percentage appears to be signiﬁcantly higher (6.5% per patient per year; 1.3% per lesion per year) in patients with the familial form of the disease [14]. Likewise, worsening of neurologic

signs and symptoms is more frequent in the familial group, occurring in more than 15% per year [14]. Among patients with nonfamilial cavernomas who present with symptomatic hemorrhage, the annual recurrent hemorrhage rate may be higher (about 4% to 5%) in the next year than in patients with no prior hemorrhage (0.6%) [2]. Barker et al. provide evidence for temporal clustering of hemorrhages, suggesting rebleed rates as high as 2.1% per month for the ﬁrst 2.5 years after the initial hemorrhage before decreasing to 1% per month [21]. The risk of hemorrhage may also depend somewhat on location, with some reports of patients with deep lesions having initial annual clinical hemorrhage risk of 4.1% compared with 0.4% among those with superﬁcial lesions [2]. Younger patients may also have a higher risk of a second hemorrhage than older patients. Cavernous malformations are often surprisingly dynamic, with changes in size, number, and MRI signal characteristics. The size of these lesions in the literature has been reported to vary between less than 1 mm up to more than 10 cm in diameter [18]. The mean cavernoma size in several large series was approximately 15 mm to 19 mm, with the majority of symptomatic lesions being greater than 1 cm in diameter [10, 13, 18]. Clinical symptoms, however, are related more to location than size, as very small tectal cavernomas can cause ophthalmoplegia, whereas giant frontal lobe lesions may be asymptomatic. Clatterbuck et al. prospectively followed 114 lesions in 68 patients over 352.9 patient-years and demonstrated that most lesions either increased or decreased in size with less than 22% being stable in volume [23]. Three de novo lesions were observed during this period and may represent the growth of very small lesions without symptomatic hemorrhage. In addition to changes in volume, they also demonstrated a stereotypical progression of changes in appearance. The mechanisms purported to cause growth and MRI signal changes include (1) progressive ectasia of vascular channels and budding of capillaries from cavernous spaces, resulting in growth of the vascular nidus; (2) recurrent “external” microhemorrhages or red blood cell diapedesis with thrombosis, followed by organization, ﬁbrosis, and calciﬁcation; (3) acute gross hemorrhage, resulting in acute size increase; and (4) cyst formation “internal” hemorrhage that ruptures septae between adjacent sinusoids. Further growth and perifocal edema may be related to secondary microhemorrhages and ﬂuid exudation from the neocapillary membrane in the cyst capsule. Because our understanding of bleeding and growth rates is far from exact, and indeed many patients who experience bleeding from cavernous malformations may show good clinical recovery, some authors have encouraged conservative treatment, particularly for brain-stem cavernous malformations. Kupersmith et al. report that the majority of 23 nonsurgically treated patients with symptomatic brain-stem cavernous malformations followed over a mean of almost 5 years either improved (78%) or remained clinically stable (17.4%) [24]. Other larger series of similar patients, however, detail a much more ominous outlook, with a 3.5% perioperative mortality rate, an 8% overall mortality rate during follow-up after surgery, and a 20% mortality rate for nonsurgically treated patients after 35.7 months of follow-up [16]. Such variability demonstrates the sometimes unpredictable clinical courses of

50.

cerebral cavernous malformations: surgical perspective

these lesions and highlights deﬁciency in our knowledge of how they behave. Furthermore, survival and successful treatment from a cavernous malformation that has hemorrhaged may not always provide a cure for the disease. Clinical reports have documented the de novo formation of cavernomas for both familial and sporadic forms of the disease [5, 14, 23]. In patients with genetic predispositions, the new cavernoma formation has been estimated to have an incidence of 0.2 to 0.4 lesions per patient-year [14, 25]. The incidence of de novo lesions in patients without known genetic predisposition is unclear, but one report described multiple cavernomas occurring in 14 of 55 (25.4%) sporadic cases [23].

Clinical Presentation The clinical symptomatology of cavernous malformations is highly variable, ranging from an asymptomatic incidental ﬁnding to discovery in autopsy after fatal hemorrhage. Typically, patients come to medical attention because of seizures, focal neurologic deﬁcits, or headaches, though there have been an increasing number of patients with malformations found incidentally after imaging for other purposes such as trauma. All of these symptoms are believed secondary to either lesion hemorrhage or mass effect on neural structures. In one series of 565 cases of symptomatic cavernous malformations, 217 (38.4%) presented with seizures, 191 (33.8%) with tumor-like symptoms, 66 (11.6%) presented with headache, and 101 (17.8%) with overt hemorrhage [26]. Numerous other reports have conﬁrmed that epileptic seizures constitute the most frequent clinically presenting symptom. Between 35% and 70% of symptomatic cavernomas are associated with recurrent seizures, which are drug-resistant in 40% of cases [27]. All seizure types including simple seizures, complex partial, and generalized seizures [8] have been known to present in patients with supratentorial cavernous malformations. The pathogenesis of seizures is debated, but most believe they are related to the presence of iron products after red cell breakdown secondary to multiple microhemorrhages. Most patients with seizures present with lesions in their frontal or temporal lobes. The estimated risk of developing seizure is 1% to 2% per personyear exposure, and mean age at time of ﬁrst seizure is 42 ± 3.78 years [9, 15]. The second most frequent presentation is that of focal neurologic deﬁcits, which correspond with lesion location and are usually produced by intralesional or perilesional hemorrhage. They account for 35% to 50% of reported cases [28] Vertigo, diplopia, ataxia, hemiparesis, associated sensory disturbances, and changes in level of consciousness are typical of infratentorial cavernous malformations, which occasionally may even produce signs of increased intracranial pressure due to obstructive hydrocephalus. Aphasia and apraxias can be produced by supratentorial lesions; visual loss with lesions within the optic nerve or chiasm and symptoms of trigeminal neuralgia from extraaxial cavernous malformations of the middle fossa have also been reported. Neurologic deﬁcits may be transient, progressive, or ﬁxed. In rare instances, cavernous malformations may simulate multiple sclerosis due to ﬂuctuating progressive

505

neurologic deﬁcits such as repeated exacerbation of complaints and alternating periods of remission. Cavernous malformations have also been found in the spinal cord where they present with an acute or subacute myelopathy. Headache accompanies cavernous malformations in many patients and may have prompted diagnostic evaluation uncovering the lesion. Although this is a controversial presenting complaint because of its nonspeciﬁc and nonlocalizing nature, it represents a major symptom in 25% to 30% of reported cases [28]. The prevalence of asymptomatic lesions in the general population is unknown, but recent evidence suggests it is much higher than previously suspected. Studies among consecutive patients undergoing MRI have detected asymptomatic cavernous malformations in 14% to 19% of patients [28]. Other reviews of the literature have found as many as 21% of patients to be without symptoms at ﬁrst discovery [3, 17]. The incidence of cavernous malformations that present with clinical hemorrhage is uncertain. This is because of variability in hemorrhage patterns of cavernous malformations as well as variability in the literature of what deﬁnes a clinically signiﬁcant event. Some have described clinically signiﬁcant hemorrhages as those that are sizable and that present with acute symptoms. The prevalence of this type of hemorrhage in large series ranges between 6% to 30% of cases [9, 10, 15, 17, 23]. The clinical presentation of hemorrhage is acute, but usually not as dramatic as with hemorrhages occurring from intracranial aneurysms or arteriovenous malformations (AVMs). Sudden-onset headache followed by a focal neurologic deﬁcit is typical, depending on the location and size of the hematoma.

Diagnostic Imaging The most sensitive and therefore the most important diagnostic tool is MRI (Figs. 50-1, 50-2, and 50-3). The high sensitivity of gradient-echo sequences are particularly helpful in detecting both acute and chronic hemorrhage that may not be seen on conventional spin-echo techniques. Indeed, the widespread use of MRI has led to the detection of asymptomatic cavernous malformation, therefore increasing previous estimates regarding the prevalence of these lesions. The MRI appearance of cavernous malformations has been detailed by many authors including Zabramski et al., who distinguish four types of MR characteristics [14]. Type I lesions are hyperintense on T1- and T2-weighted images and indicate subacute hemorrhage. Type II lesions show a mixed signal intensity on T1- and T2-weighted imaging with a surrounding hemosiderin ring. Type III lesions are hypo- to isointense on T1- and T2-weighted images and most probably indicate chronic hemorrhage. Type IV lesions are poorly visualized except on gradient-echo sequences. Histopathologic correlations with these MRI patterns have been conﬁrmed. In addition to establishing the diagnosis, MRI also deﬁnes the exact size, location, and extent of cavernous malformations and provides information about multiplicity, hydrocephalus, and the condition of the adjacent brain parenchyma.

506

r.l. dodd and g.k. steinberg

FIGURE 50-1. Selected (A) axial FLAIR, (B) coronal T1 postcontrast, and (C) sagittal T1 MRI scans of a 31-year-old female medical student who presented with four clinical hemorrhages resulting in headaches, gait instability, and symptoms of hypothalamic dysfunction including hypersomnolence, loss of appetite, and memory difﬁculties. A cavernous malformation is centered in the midline involving the massa intermedia, hypothalamus, midbrain, third ventricle, and extending to the suprasellar cistern.

FIGURE 50-2. Selected immediately postoperative (A) axial T1, (B) coronal T1 postcontrast, and (C) sagittal T1 MRI scans demonstrate gross total resection of the cavernous malformation seen in Fig. 50-1. (D) Microsurgical removal was performed via an interhemispheric transcallosal approach to the third ventricle. The asterisk (*) is on the corpus callosum, through which a small incision was made that exposed an enlarged foramen of Monro. The laser pointer is focused on the superior aspect of the cavernous malformation. (E) A 3.5-cm cavernous

malformation was resected piecemeal from the third ventricle, diencephalon, and midbrain. There was acute, subacute, and chronic hemorrhage within the malformation, which was composed of thin-walled, fragile vascular channels. No attempt was made to remove the surrounding hemosiderin-stained brain. Postoperatively, the patient quickly recovered to her neurologic baseline and has continued to improve, though some memory difﬁculties still remain.

50.

cerebral cavernous malformations: surgical perspective

507

FIGURE 50-3. (A) Sagittal T1, (B) coronal T1 postcontrast, and (C) axial T2 and (D) T1 MRI scans illustrate a 1-cm, right midpontine cavernous malformation with subacute hemorrhage surrounded by a rim of hemosiderin and surrounding brain-stem edema (arrow in C). Although this lesion did not present to a pial or ventricular surface, three clinically signiﬁcant hemorrhages were observed over a 7-month period, resulting in diplopia, facial weakness, sensory loss, and mild

hemiparesis. Microsurgical resection of the cavernous malformation was therefore performed. (E) A right subtemporal approach was used as illustrated, aided by the intraoperative surgical navigation. Brain-stem functional mapping was employed prior to making the fewmillimeter incision in the lateral surface of the pons (right upper panel, E). Postoperatively, the patient’s hemiparesis transiently worsened, but he recovered within 1 week and currently has no focal deﬁcits.

Treatment

do not bleed much at surgery. The most difﬁcult task in resection is often intraoperative localization and choosing an appropriate surgical corridor. This has been facilitated with the use of stereotaxic MRI computer-guidance systems for neuronavigation, which allows planning of surgical trajectories to avoid functional cortex, major blood vessels, as well as target localization to within 1 to 2 mm. The risk of open microsurgery is generally related to lesion size, location, and proximity to critical structures. In addition, advanced age, medical comorbidities, and poor neurologic condition are all negative prognostic factors for a good surgical result. Several surgical series have reported excellent results in removal of superﬁcial lesions of both eloquent and noneloquent areas. More than 75% of all cavernous malformations are located supratentorially [17], most of which are located in cortical and subcortical regions that are readily accessible with modern surgical techniques. Even prior to MRI, surgical series have documented the successful removal of cavernous malformations with low risk. Vaquero et al. report successful treatment of 19 supratentorial cavernous malformations located in the cerebral hemispheres from 1977 to 1987. Of the 19 patients, 17 had excellent results (no seizures without medication) and 2 were classiﬁed as good outcomes (occasional seizures with medication) [29]. There were no poor outcomes or deaths in this cohort. Similarly, excellent results have been documented in the post-MRI era.

Given the excellent results of microsurgical resection of cavernous malformations, it is generally easy to endorse the philosophy of operative management for most of these lesions. However, increasing evidence suggesting a relatively benign natural history of some lesions has indicated the need for reevaluation of management decisions. Particularly when dealing with a patient harboring several lesions, a family affected by a hereditary form of this condition, or deep-seated lesions in the brain stem, it becomes apparent that an aggressive approach is not always advisable. The only clear indications for surgery are for medically intractable epilepsy, documented recurrent hemorrhage, and progressive neurologic deﬁcit. Patients with established diagnosis of CCM who present without clinical hemorrhage, seizures, or other speciﬁc symptoms are usually candidates for clinical observation and repeat imaging. Conservative management is also appropriate for many patients with purely incidental lesions or for malformations located deeply within functional areas that do not present to a pial or ventricular surface.

Microsurgical Treatment of Supratentorial Lesions Surgical resection of most cavernous malformations is straightforward because these lesions have no major arterial supply and

508

r.l. dodd and g.k. steinberg

McCormick reviewed his experience with removal of supratentorial cavernous malformations; in 63 of 65 patients complete removal was obtained, two patients experienced worsening of an existing deﬁcit, a third patient developed a new deﬁcit, and there was only one postoperative death [30]. When performed for medically intractable epilepsy, several published results conﬁrm a greater than 90% likelihood of a seizure-free outcome with surgical extirpation of cavernous malformations and the surrounding epileptogenic tissue [27]. The surgical approach is deﬁned by the lesion location, and in general the craniotomy is centered over the lesion. The use of neuronavigation allows precise targeting that minimizes the size of the craniotomy and risk to exposed brain (Fig. 50-3). Lesions that involve the anterior corpus callosum near the genu and cingulate gyrus can be approached through a small frontal craniotomy. Lesions that are more basal in the subfrontal or septal regions often require lower frontal craniotomy through a bicoronal incision or an interhemispheric approach. Posterior parasagittal approaches are generally used for lesions in the body of the corpus callosum, cingulate gyrus, and midline posterior frontal, parietal, or occipital lesions (Fig. 50-4). Medial anterior temporal lesions involving the amygdala, the hippocampus, or the uncus are approached through the sylvian ﬁssure via a standard pterional craniotomy. Superﬁcial cortical lesions are often easy to identify because of overlying cortical discoloration, whereas small subcortical lesions often require stereotaxis. Both transsulcal and transgyral approaches can be employed depending on lesion geometry and the location of cortical vasculature. Once identiﬁed, a well-deﬁned gliotic plane typically allows for easy separation of the lesion from surrounding tissue so that complete lesion resection can be achieved. When surgery is indicated for seizures, the hemosiderin-laden gliotic tissue should be removed because of its role in seizure generation. Intraoperative electrocorticography may be helpful for conﬁrming the extent of abnormal electrical activity, particularly for lesions near the motor or sensory cortexes. For dominant hemisphere lesions, awake intraoperative speech mapping may be required to preserve language function.

Microsurgical Treatment of Infratentorial Lesions Lesions located deep within the brain are difﬁcult to remove and represent special challenges (Figs. 50-1 to 50-4), though recent series describe improved outcomes with image-guided techniques. Approximately 10% to 30% of all intracranial cav-

ernous malformations are located in the posterior fossa [16–20], and some reports suggest these lesions may be more frequently symptomatic [16]. Surgical removal is indicated for symptomatic lesions located in the cerebellum or superﬁcially in the brain stem when eloquent parenchyma can be spared. Several surgical approaches have been detailed in the literature including suboccipital, infratentorial supracerebellar, occipital transtentorial, far lateral, transtemporal, subtemporal transtentorial, and combined petrosal. Furthermore, much emphasis has been placed on the selection of an approach that optimizes exposure to the lesion with minimum disruption to surround structures. In general, lesions of the cerebellar vermis, medial cerebellar hemispheres, ﬂoor of the fourth ventricle, and dorsal medulla are easily approached through a standard suboccipital approach. Lesions of the tectum and pineal region are often best approached via infratentorial supracerebellar access. The occipital transtentorial approach provides exposure of the superior cerebellar peduncles and vermis, the anterior medullary velum, posterior third ventricle, the splenium, and the quadrigeminal plate. The far lateral approach provides excellent exposure of the anterior and lateral medulla and cervicomedullary junction. Transtemporal approaches allow access to lesions in and around the internal auditory meatus such as pontomedullary junction cavernous malformations and can be employed in patients without serviceable hearing. Cavernous malformations involving the anterior midbrain–interpeduncular fossa region can be approached through a subtemporal or orbitozygomatic craniotomy, and occasional lesions located in the upper twothirds of the brain stem require a combined petrosal approach. Cerebrospinal ﬂuid drainage for brain relaxation, electrophysiologic monitoring, and functional mapping are all important for minimizing complications and reducing the risks of permanent neurologic deﬁcit. Unlike supratentorial cavernous malformations, deep lesions can be extremely adherent to the normal parenchyma and are sometimes removed in a piecemeal fashion after circumferential separation and devascularization from the surrounding gliosis. All perforating arterial vessels must be dissected meticulously and preserved. Furthermore, the hemosiderin staining surrounding the malformation should be spared from surgical removal. An association between cavernous malformations and venous malformation has been described and may be as high as 16% [31], and whereas the goal of surgery is complete excision of the cavernoma, interruption of a venous malformation is to be avoided because of the risk of venous

FIGURE 50-4. Selected 3-year postoperative (A) axial, (B) coronal, and (C) sagittal T1 MRI scans demonstrate gross total resection of the cavernous malformation seen in Fig. 50–3.

50.

cerebral cavernous malformations: surgical perspective

509

infarction. The risk of open microsurgery for superﬁcial cerebellar lesions is low, similar to that for most supratentorial cavernous malformations. The risks associated with surgical exploration of deep-brain and brain-stem cavernous malformation is considerably higher, as discussed below.

Radiosurgical Treatment of Cavernous Malformations A number of groups have used stereotactic focused radiosurgery with heavy charged particles (protons or helium ions) or photons (Gamma Knife or linac radiosurgery) in attempts to improve the natural history of deep, difﬁcult to resect cavernous malformations [32–41]. Although there has been no prospective randomized study, recent reports from several groups suggest that, for certain patients, radiosurgery may decrease the risk of clinical hemorrhage compared with the statistically derived natural history [32, 33, 37]. Our combined Stanford University–University of California, Berkeley, program previously treated 57 cavernous malformations in the brain stem, thalamus, and basal ganglia using Bragg peak helium ion radiosurgery (47 lesions) or linac radiosurgery (10 lesions), with mixed results [33]. Eighteen (32%) patients experienced symptomatic bleeding (20 hemorrhaging episodes) after radiosurgery. Sixteen hemorrhaging episodes occurred within the 36 months after radiosurgery (9.4% annual bleeding rate; 16 hemorrhaging episodes/171 patient-years), and 4 hemorrhaging episodes occurred more than 36 months after treatment (1.6% annual bleeding rate; 4 hemorrhaging episodes/257 patient-years) (p < 0.001). The radiosurgery group at Harvard University reported a decrease in the annual hemorrhage rate from 17.3% before radiosurgery to 4.5% after a latency period of 2 years after radiosurgery [32]. Similar symptomatic hemorrhage rates after Gamma Knife radiosurgery for cavernous malformations were reduced from 8.8% during the ﬁrst 2 years after radiosurgery to 1.1% thereafter [37]. However, a recent analysis showed the hemorrhage rate from untreated cavernous malformations using statistical modeling appeared to demonstrate temporal clustering of hemorrhages such that the rehemorrhage rate from untreated cavernous malformations was found to be high initially (2% monthly rehemorrhage rate during the ﬁrst 2.5 years after hemorrhage; and 14% cumulative incidence of a second hemorrhage during the ﬁrst year) and decreased 2 to 3 years after a previous hemorrhage (to less than 1% per month) [21]. These results therefore question whether the data from radiosurgical series represent protective effects of treatment or simply the natural bleeding history of these lesions. Additionally, in 10 patients who underwent microsurgical resection of their cavernous malformations 1 to 10 years after radiosurgical treatment, none of the cavernous malformations were found to be thrombosed on pathology examination [42]. Furthermore, in our series of patients with cavernous malformations that underwent radiosurgery, 7% experienced symptomatic radiation edema and 2% experienced radiation necrosis [33] (Fig. 50-5). In a series of 47 brain-stem cavernous malformations treated with Gamma Knife radiosurgery, Kondziolka et al. reported that 26% of their patients experienced symptomatic radiation induced injury that was either temporary (17%), permanent (4.5%), or contributed to two deaths (4.5%) [37]. Another group has documented a 16% incidence of permanent neuro-

FIGURE 50-5. (A) Axial proton-density MRI of a 35-year-old woman who initially presented with generalized seizures shows a 1-cm vascular malformation located in the left thalamus adjacent to the posterior limb of the internal capsule. She was ﬁrst treated with stereotactic heavycharged-particle Bragg peak radiosurgery. Nine months after radiosurgery, she experienced worsening right hemiparesis; (B) axial T2 imaging at that time demonstrated extensive radiation-induced changes extending into the basal ganglia and thalamus. She subsequently underwent stereotactically guided microsurgical resection via a small left posterior parietal, interhemispheric transcallosal approach through the pulvinar. The patient recovered extremely well and was discharged 2 days after surgery without new deﬁcits. (C) Hematoxylin and eosin–stained specimen from a surgically removed brain-stem cavernous malformation 5 years after stereotactic heavy-charged-particle radiosurgery. Patent vascular channels (arrow) are observed. (D) Surgical specimen from tissue adjacent to a cavernous malformation removed from another patient with delayed radiation-induced injury 3 years after stereotactic radiosurgery of a right parietal cavernous malformation. Histopathologic changes consisted of thick-walled hyalinized vessels with ﬁbrinoid necrosis (arrow), postradiation gliosis, and reactive astrocytosis and necrosis (arrowhead), characteristic of radiation effects.

logic deﬁcits and 3% mortality rate secondary to radiationinduced complications [32].

Brain-Stem Cavernous Malformations Brain-stem cavernomas, though very rare, deserve special mention because their clinical presentation, natural history, and treatment decisions often differ from those of cavernous

510

r.l. dodd and g.k. steinberg

malformations in other locations. Such lesions account for 9% to 35% of all cavernomas and for 18% to 22% of all intracranial cavernomas [16, 18, 20]. Fifty percent to 70% of these lesions occur in the pons, 20% to 35% in the midbrain, and 15% to 25% in the medulla. Symptoms vary from cranial nerve deﬁcits, sensory dysfunction, paresis/plegia, ataxia, dysmetria, speech difﬁculty, to headaches and decreased levels of consciousness. The clinical presentation can often be confusing, and many patients who have experienced episodic clinical deterioration secondary to hemorrhaging have been diagnosed as having multiple sclerosis. There is evidence that the natural history of brain-stem cavernous malformations differs from that of other locations, suggesting that this subgroup has a higher propensity for bleeding. Porter et al. reported nearly all of their 100 brain-stem cavernomas presented with evidence of clinically signiﬁcant hemorrhage, 56% with multiple bleeds, and 22% with more than two [16]. Although some patients with hemorrhage from brain-stem cavernous malformations make good recoveries without intervention, it is also clear that multiple bleeding episodes increase the likelihood of a persistent deﬁcit. Furthermore, when multiple bleeding episodes do occur, the time interval between hemorrhages is increased and the mortality rate for each hemorrhage may be as high as 20%. Surgical treatment of brain-stem cavernous malformations is considerably more risky than operative intervention in other less eloquent areas of the nervous system. Difﬁcult assess, narrow surgical corridors, and the sensitivity to injury of critical structure near these lesions present unique obstacles (Fig. 50-3). Despite the high risk of permanent morbidity and technical challenges of microsurgical excision, several centers have demonstrated expertise in removing these lesions with good outcomes. Steinberg et al. reported a series of 56 patients with 57 deep-brain cavernous malformations located in the brain stem, thalamus, and basal ganglia, which were microsurgically resected [20]. Neuronavigation, mild hypothermia (32°C to 33°C), electrophysiologic monitoring, and cranial nerve mapping were used to assist in the safe removal of these lesions. Over a 9-year period with a mean follow-up of 4.7 years, the overall results were excellent in 52%, good in 43%, and poor in 5%; 3% of patients died. Ninety-three percent of these lesions were completely resected, 95% of the patients were either neurologically improved or unchanged after 6 months, and the long-term neurologic morbidity rate was only 5%. In 92 patients undergoing microsurgical resection of 93 brain-stem cavernous malformations at Stanford University Medical Center from 1990 to 2005, although one third of patients were immediately worse after surgery, in the long-term follow-up (>6 months) 91% of patients were neurologically unchanged or improved, whereas 9% worsened. Porter et al. reported their experience with 100 patients with brain-stem cavernous malformation, of which 86 underwent microsurgical resection [16]. Over a 12-year period, 87% were the same or better, 10% were worse, and 4% died; the long-term neurologic morbidity rate was 12%. Bertalanffy et al. reviewed their series of 24 brain-stem cavernous malformations over a 5-year period [18]. Seventy percent had Karnofsky scores of 90 or better at follow-up, 21% had scores of 60 to 70, and 4% had scores less than 60. Their rate of long-term morbidity was 5.5%, and there were no deaths. Therefore, whereas the immediate risks of surgical intervention for brain-stem and

deep-brain cavernous malformation are not low, the long-term results are good, and the natural history of such lesions is likely worse than that of cavernous malformations at other CNS sites. We recommend that surgical resection be considered for symptomatic brain-stem and deep-seated lesions that present to a pial or ependymal surface and, in certain circumstances, even for those malformations that do not present to a pial/ependymal surface, if they have bled repetitively.

Conclusion Surgical resection of symptomatic cavernous malformations provides an excellent treatment option, and when completely excised is curative for most lesions. When cavernous malformations are removed in patients with medically refractory epilepsy, improved seizure control is typical, and many centers report high rates of “seizure-free” patients. In general, the surgical morbidity associated with the resection of superﬁcial lesions in the cerebral and cerebellar cortices is minimal, and mortality is considered unusual. Resection of lesions in and around eloquent cortex requires electrophysiologic monitoring, and often functional mapping, but can be performed safely. Deep-seated cavernous malformations in the thalamus, basal ganglia, and brain stem are associated with higher surgical risks; however, for symptomatic lesions that have demonstrated recurrent hemorrhage and present to a pial or ventricular surface, resection by an experienced neurosurgeon is well reported and should be considered. The role of radiosurgery for unresectable malformations remains uncertain and is associated with signiﬁcant risks of radiation-induced injury without any proven change in the natural history of hemorrhaging. Until better evidence of its efﬁcacy can be established, we do not recommend radiosurgery for treatment of cavernous malformations.

References 1. Al-Shahi R, Bhattacharya JJ, Currie DG, et al. Prospective, population-based detection of intracranial vascular malformations in adults: the Scottish Intracranial Vascular Malformation Study (SIVMS). Stroke 2003; 34:1163–1169. 2. Brown RD Jr, Flemming KD, Meyer FB, et al. Natural history, evaluation, and management of intracranial vascular malformations. Mayo Clin Proc 2005; 80:269–281. 3. Aiba T, Tanaka R, Koike T, et al. Natural history of intracranial cavernous malformations. J Neurosurg 1995; 83:56–59. 4. Brown RD Jr, Wiebers DO, Torner JC, O’Fallon WM. Frequency of intracranial hemorrhage as a presenting symptom and subtype analysis: a population-based study of intracranial vascular malformations in Olmsted Country, Minnesota. J Neurosurg 1996; 85:29–32. 5. Detwiler PW, Porter RW, Zabramski JM, Spetzler RF. De novo formation of a central nervous system cavernous malformation: implications for predicting risk of hemorrhage. Case report and review of the literature. J Neurosurg 1997; 87:629–632. 6. Giombini S, Morello G. Cavernous angiomas of the brain. Account of fourteen personal cases and review of the literature. Acta Neurochir (Wien) 1978; 40:61–82. 7. Sarwar M, McCormick WF. Intracerebral venous angioma. Case report and review. Arch Neurol 1978; 35:323–325.

50.

cerebral cavernous malformations: surgical perspective

8. Raychaudhuri R, Batjer HH, Awad IA. Intracranial cavernous angioma: a practical review of clinical and biological aspects. Surg Neurol 2005; 63:319–328; discussion 328. 9. Del Curling O Jr, Kelly DL Jr, Elster AD, Craven TE. An analysis of the natural history of cavernous angiomas. J Neurosurg 1991; 75:702–708. 10. Kim DS, Park YG, Choi JU, et al. An analysis of the natural history of cavernous malformations. Surg Neurol 1997; 48:9–17; discussion 17–18. 11. Kondziolka D, Lunsford LD, Kestle JR. The natural history of cerebral cavernous malformations. J Neurosurg 1995; 83:820– 824. 12. Pollock BE, Garces YI, Stafford SL, et al. Stereotactic radiosurgery for cavernous malformations. J Neurosurg 2000; 93:987– 991. 13. Robinson JR, Awad IA, Little JR. Natural history of the cavernous angioma. J Neurosurg 1991; 75:709–714. 14. Zabramski JM, Wascher TM, Spetzler RF, et al. The natural history of familial cavernous malformations: results of an ongoing study. J Neurosurg 1994; 80:422–432. 15. Moriarity JL, Clatterbuck RE, Rigamonti D. The natural history of cavernous malformations. Neurosurg Clin N Am 1999; 10:411–417. 16. Porter RW, Detwiler PW, Spetzler RF, et al. Cavernous malformations of the brainstem: experience with 100 patients. J Neurosurg 1999; 90:50–58. 17. Hsu F, Rigamonti D, Huhn S. Epidemiology of cavernous malformations. In: Awad IA, Barrow DL, eds. Cavernous Malformations. Park Ridge, IL: AANS, 1993:13–23. 18. Bertalanffy H, Benes L, Miyazawa T, et al. Cerebral cavernomas in the adult. Review of the literature and analysis of 72 surgically treated patients. Neurosurg Rev 2002; 25:1–53; discussion 54–55. 19. Sandalcioglu IE, Wiedemayer H, Secer S, et al. Surgical removal of brain stem cavernous malformations: surgical indications, technical considerations, and results. J Neurol Neurosurg Psychiatry 2002; 72:351–355. 20. Steinberg GK, Chang SD, Gewirtz RJ, Lopez JR. Microsurgical resection of brainstem, thalamic, and basal ganglia angiographically occult vascular malformations. Neurosurgery 2000; 46:260– 270; discussion 270–261. 21. Barker FG 2nd, Amin-Hanjani S, Butler WE, et al. Temporal clustering of hemorrhages from untreated cavernous malformations of the central nervous system. Neurosurgery 2001; 49:15–24; discussion 24–15. 22. Barrow DL, Krisht A. Cavernous malformations and hemorrhage. In: Awad IA, Barrow DL, eds. Cavernous Malformations. Park Ridge, IL: AANS, 1993:65–80. 23. Clatterbuck RE, Moriarity JL, Elmaci I, et al. Dynamic nature of cavernous malformations: a prospective magnetic resonance imaging study with volumetric analysis. J Neurosurg 2000; 93:981– 986. 24. Kupersmith MJ, Kalish H, Epstein F, et al. Natural history of brainstem cavernous malformations. Neurosurgery 2001; 48:47– 53; discussion 53–44. 25. Labauge P, Brunereau L, Levy C, et al. The natural history of familial cerebral cavernomas: a retrospective MRI study of 40 patients. Neuroradiology 2000; 42:327–332.

511

26. Rigamonti D, Hsu F, Huhn S. Angiograhically occult vascular malformations. In: Carter LP, Spetzler RF, eds. Neurovascular Surgery. New York: McGraw-Hill, 1994:521–540. 27. Awad IA, Robinson JR. Cavernous malformations and epilepsy. In: Awad IA, Barrow DL, eds. Cavernous Malformations. Park Ridge, IL: AANS, 1993:49–63. 28. Robinson JR. Clinical spectrum and natural history. In: Awad IA, Barrow DL, eds. Cavernous Malformations. Park Ridge, IL: AANS, 1993:25–36. 29. Vaquero J, Salazar J, Martinez R, et al. Cavernomas of the central nervous system: clinical syndromes, CT scan diagnosis, and prognosis after surgical treatment in 25 cases. Acta Neurochir (Wien) 1987; 85:29–33. 30. McCormick PC. Management of intracranial cavernous and venous malformations. In: Intracranial Vascular Malformations: Neurosurgical Topics. Park Ridge, IL: AANS, 1990:197–218. 31. Wascher TM, Spetzler RF. Cavernous malformations of the brain stem. In: Carter LP, Spetzler RF, eds. Neurovascular Surgery. New York: McGraw-Hill, 1994:541–555. 32. Amin-Hanjani S, Ogilvy CS, Candia GJ, et al. Stereotactic radiosurgery for cavernous malformations: Kjellberg’s experience with proton beam therapy in 98 cases at the Harvard Cyclotron. Neurosurgery 1998; 42:1229–1236; discussion 1236–1228. 33. Chang SD, Levy RP, Adler JR Jr, et al. Stereotactic radiosurgery of angiographically occult vascular malformations: 14-year experience. Neurosurgery 1998; 43:213–220; discussion 220–211. 34. Fabrikant JI, Levy RP, Steinberg GK, et al. Stereotactic chargedparticle radiosurgery: clinical results of treatment of 1200 patients with intracranial arteriovenous malformations and pituitary disorders. Clin Neurosurg 1992; 38:472–492. 35. Kida Y, Kobayashi T, Tanaka T. Treatment of symptomatic AOVMs with radiosurgery. Acta Neurochir Suppl 1995; 63:68– 72. 36. Kondziolka D, Lunsford LD, Coffey RJ, et al. Stereotactic radiosurgery of angiographically occult vascular malformations: indications and preliminary experience. Neurosurgery 1990; 27: 892–900. 37. Kondziolka D, Lunsford LD, Flickinger JC, Kestle JR. Reduction of hemorrhage risk after stereotactic radiosurgery for cavernous malformations. J Neurosurg 1995; 83:825–831. 38. LeDoux MS, Aronin PA, Odrezin GT. Surgically treated cavernous angiomas of the brain stem: report of two cases and review of the literature. Surg Neurol 1991; 35:395–399. 39. Levy RP, Fabrikant JI, Frankel KA, et al. Charged-particle radiosurgery of the brain. Neurosurg Clin N Am 1990; 1:955– 990. 40. Seo Y, Fukuoka S, Takanashi M, et al. Gamma Knife surgery for angiographically occult vascular malformations. Stereotact Funct Neurosurg 1995; 64(Suppl)1:98–109. 41. Steinberg GK, Levy RP, Fabrikant JI, et al. Stereotactic helium ion Bragg peak radiosurgery for angiographically occult intracranial vascular malformations. Stereotact Funct Neurosurg 1991; 57:64–71. 42. Gewirtz RJ, Steinberg GK, Crowley R, Levy RP. Pathological changes in surgically resected angiographically occult vascular malformations after radiation. Neurosurgery 1998; 42:738–742; discussion 742–743.

5 1

Cavernous Malformations and Other Vascular Abnormalities: Observation-Alone Perspective Sepideh Amin-Hanjani and Frederick G. Barker II

Introduction Observation alone as a reasonable option in management of cavernous malformations (CMs) or other vascular abnormalities can only be justiﬁed based on a favorable comparison between the natural history of the lesion and the expected efﬁcacy and risks of treatment. For each lesion type in question, we will review existing knowledge regarding natural history in comparison with the effects of radiosurgical treatment.

Cavernous Malformations Natural History The natural history of CMs was ﬁrst addressed systematically in two large consecutive series in the early 1990s [1, 2], both based on magnetic resonance imaging (MRI) scan results. Subsequent reports have addressed both familial and sporadic cases. Reported annual hemorrhage rates vary widely, ranging from as low as 0.25% to as high as 25% [2, 3]. Some of the variability appears to reﬂect patient-based risk factors, such as a positive family history of CMs, history of prior hemorrhage from a speciﬁc lesion, and lesion location [3–6]. Variation between studies can also result from differing assumptions about the period at risk for hemorrhage (the denominator in the equation of events observed divided by time spent at risk). The calculated hemorrhage risk differs markedly depending on whether the risk interval is assumed to start at birth, at ﬁrst presentation (whether incidental or symptomatic), or after a ﬁrst hemorrhage. The annual hemorrhage risk can also vary depending on the criteria used to deﬁne hemorrhage. Clinical deterioration in CM patients is not always accompanied by imaging-detectable overt hemorrhage from the lesion, and conversely, acute hemorrhage on imaging can on occasion be clinically silent. Whether only radio-

graphically conﬁrmed events are considered as a hemorrhage or any acute clinical deterioration is also counted can lead to differing estimates of hemorrhage risk [5]. Despite these challenges in study methodology, certain aspects of the behavior of CMs have become evident. It is clear that hemorrhage from CMs is rarely catastrophic. However, repeated small hemorrhages can result in progressive deterioration, and Robinson et al. described a strong association between hemorrhage and neurologic disability [7]. The long-term risk of hemorrhage therefore is of importance in decision making even in incidental or minimally symptomatic lesions, particularly in young patients. The current available estimates of the risk of hemorrhage from CMs indicate a low overall annual hemorrhage rate, in the range of 1% to 3%. Speciﬁcally, prospective follow-up studies have reported hemorrhage rates from 0.7% to 2.5% per lesion per year [1, 4, 8] and 1.6% to 3.1% per person per year [5, 9]. The annual hemorrhage risk may be higher in deep or brainstem CMs. In a cohort of 110 prospectively followed patients, Porter et al. found a 10-fold higher hemorrhage rate among infratentorial lesions (3.8% per year) compared with supratentorial CMs (0.4% per year) [5]. This difference may reﬂect the eloquence of the surrounding tissue, with even small brain-stem hemorrhages more likely to be clinically manifest, rather than a true difference in the rates of actual hemorrhage from lesions in the two locations. Porter et al. emphasized in their study the importance of neurologic deterioration from CMs even without imaging-detectable hemorrhage. Deteriorations without overt hemorrhage comprised nearly half of new neurologic deﬁcits in that study, with no better rate of recovery from symptoms than for symptomatic overt hemorrhagic events [5]. After an initial hemorrhage, the risk of rebleeding appears to be higher. In a cohort of 122 patients followed for 34 months, Kondziolka et al. noted a low 0.6% per person-year risk of hemorrhage in those with no prior history of bleed, whereas

513

514

s. amin-hanjani and f.g. barker ii

Monthly risk of rehemorrhage

0.030

80% confidence interval for model cutpoint 95% confidence limits for hazard rates

0.025 0.020 0.015 0.010 0.005 0.0 0

2

4 6 8 10 Years after preceding hemorrhage FIGURE 51-1. Step model superimposed over a smoothly ﬁtted local likelihood hazard model for risk of rehemorrhage from an untreated CM after a previous hemorrhage. The model indicated that the hazard for hemorrhage decreases approximately 28 months after prior hemorrhage. (From Barker FG 2nd, Amin-Hanjani S, Butler WE, et al. Temporal clustering of hemorrhages from untreated cavernous malformations of the central nervous system. Neurosurgery 2001; 49(1):15–24; discussion 25. Used with permission.)

those with a prior hemorrhage had a 4.5% per person-year risk of rebleeding [6]. Higher rates of recurrent hemorrhage compared with initial hemorrhage have subsequently been substantiated by other reports [3, 10, 11]. Barker et al. documented an increased hemorrhage rate during the ﬁrst 2.5 years after a preceding hemorrhage compared with thereafter (Fig. 51-1); this pattern accounts for the “clustering” of multiple hemorrhages from CMs noted by several observers [7, 12–15]. Giving consideration to the low risk of initial hemorrhage, and that bleeding from CMs is not a devastating or unrecoverable event in the great majority of cases, observation would be the preferred management strategy (or at minimum an option to be carefully considered) for most asymptomatic lesions, even those in critical locations. For symptomatic CMs, surgical removal will be preferable for most patients.

Radiosurgery The efﬁcacy of radiosurgery in reducing hemorrhage rates from CMs is not universally accepted. There has been difﬁculty in establishing valid end points for success of therapy. Unlike angiographic obliteration, used as a surrogate end point for radiosurgery efﬁcacy in many studies of arteriovenous malformation (AVM) treatment, there are no radiographic criteria by which to judge the efﬁcacy of radiosurgery for CMs. Consequently, claims of radiosurgery efﬁcacy have usually been based on comparisons of pre- and posttreatment hemorrhage rates [16–18]. This type of comparison has the inherent ﬂaw of assuming a constant risk of hemorrhage over time, an assumption that has been contradicted by natural history studies. First, the risk of hemorrhage has been documented to be higher after the initial bleed, as discussed in the prior section [3, 6, 10, 11]. Second, CM hemorrhages have often been noted to occur in clusters, with a temporarily elevated risk of rebleeding that then reverts to a lower rate. A phenomenon of elevated hemorrhage risk lasting approximately 2.5 years after a previous bleed, with a spontaneous decline to lower hemorrhage risk thereafter, has

FIGURE 51-2. Axial T2-weighted magnetic resonance image (A) prior to and (B) 6 years after proton beam radiosurgery for a posterior frontal CM, demonstrating radiation-induced injury that resulted in progressive hemiparesis. (From Amin-Hanjani S, Ogilvy CS, Candia GJ, et al. Stereotactic radiosurgery for cavernous malformations: Kjellberg’s experience with proton beam therapy in 98 cases at the Harvard Cyclotron. [see comment]. Neurosurgery 1998; 42(6):1229–1236; discussion 1236–1238. Used with permission.)

been documented in untreated CMs [12]. Because most radiosurgical series assume a 2- to 3-year latency period before a therapeutic reduction in hemorrhage risk could be expected, analogous to AVM treatment, the natural decline in hemorrhage risk that occurs after a cluster could easily be misattributed to a beneﬁcial effect from radiosurgery. Balanced against the uncertain efﬁcacy of radiosurgery is the risk of radiation-related complications. Rates of radiationinduced injury vary from about 10% to 40%, depending on the location of the lesion and the treatment doses used [16–22] (Fig. 51-2). It is clear that the risk of radiation injury is substantially higher after radiosurgery for CMs compared with AVMs of similar size and location treated with the same dose and treatment volume parameters [22, 23] (Table 51-1). Although the basis for this heightened risk is unclear, a role for the potentially radiosensitizing properties of the hemosiderin ring around CMs has been proposed [24]. Neither the highest safe dose nor the lowest effective dose has been established for radiosurgical treatment of CMs, leaving the therapeutic ratio undeﬁned.

TABLE 51-1. Comparison of radiosurgery risk for CMs and AVMs (lesions <10 cm3). Barker et al., 1998 [23]

Size (cm3) Marginal dose (Gy) Percent located in thalamus/brain stem (%) Complication rate (%) (permanent deﬁcit)

Pollack et al., 2000 [22]

AVM (n = 293)

CM (n = 75)

AVM (n = 36)

CM (n = 17)

4.3 16 39

2.8 16.5 71

3.4 18 86

2.1 18 94

5.5

23*

10

41

* Odds ratio = 5.2 (95% conﬁdence interval, 2.3 to 12; p < 0.001) for radiationrelated complication in CMs compared with AVMs. Comparison adjusted for dose, volume, and location in multivariate analysis.

51.

cavernous malformations and other vascular abnormalities: observation-alone perspective

Even in the face of reduced hemorrhage rates, progressive neurologic disability from CMs after radiosurgical treatment has been reported [16]. Therefore, overall quality of life may worsen after radiosurgery despite any beneﬁt in terms of hemorrhage reduction. Regardless of the speciﬁc etiology, worsening neurologic condition represents a failure of radiosurgery to convey a deﬁnitive beneﬁt over expectant management.

Observation Versus Radiosurgery For symptomatic CMs that are surgically accessible, operative intervention remains the treatment of choice. Surgery leads to better outcomes than radiosurgery, not only in preventing rehemorrhage but also in seizure control [25]. The dilemma of expectant management versus radiosurgery arises principally for symptomatic lesions in eloquent locations that are not readily approached operatively without extensive morbidity. Such lesions are fortunately uncommon. To date, no randomized trial in patients with CMs comparing radiosurgery versus observation has been completed. Retrospective studies and prospective case series have thus far failed to provide convincing evidence of a reduction in hemorrhage rates with radiosurgery in light of the potential clustering of bleeds and the absence of plausible surrogate end points. Furthermore, the risk of radiation-related complications appears to be signiﬁcant. Given the current state of knowledge, which reﬂects uncertain efﬁcacy in the face of tangible risk, radiosurgery should be used rarely, if at all, in the management of CMs outside the context of prospective clinical trials.

Dural Arteriovenous Fistulas Natural History The behavior of dural arteriovenous ﬁstulas (AVFs) ranges from benign to aggressive. Their natural history is best considered by categorizing lesions as high risk or low risk based on the propensity for potentially life-threatening hemorrhage or progressive neurologic decline. Dural AVFs can produce other symptoms and signs appropriate to their location, such as pain, cranial neuropathy, and tinnitus, but these are far outweighed by the potential consequences of lesions that follow an aggressive course. The most important factor predicting a lesion’s probable behavior is its pattern of venous drainage. Dural AVFs with cortical or leptomeningeal venous drainage are clearly associated with a worse natural history than those with drainage conﬁned to major dural venous sinuses [26–28]. Although dural AVFs in certain locations are less likely to manifest cortical venous drainage (such as those associated with the cavernous sinus or transverse and sigmoid sinus), drainage into cortical or leptomeningeal veins is an ominous sign regardless of a lesion’s location [27]. A study of 20 patients harboring dural AVFs with cortical venous drainage who were followed prospectively over a mean of 4.3 years revealed an annual mortality of 10.4% and an annual hemorrhage rate of 8.1% [28]. In contrast, follow-up of 117 patients with low-risk lesions (those without cortical venous drainage) over more than 2 years demonstrated that they have generally benign presentations, such as headache or bruit, and

515

a benign disease course without hemorrhage, neurologic decline, or death [29]. There is a small risk of developing cortical venous drainage over time, approximately 2% per year, necessitating clinical follow-up and repeat imaging if any deterioration in the patient’s condition occurs. In summary, lesions classiﬁed as high risk based on pattern of venous drainage mandate prompt and effective treatment. Low-risk lesions can be managed safely with observation alone, although treatment may be considered to palliate bothersome symptoms such as bruits or ophthalmologic sequelae.

Radiosurgery Treatment for dural AVFs typically consists of endovascular or surgical obliteration or a combination of the two modalities. Experience with radiosurgery for dural AVFs is limited, but there have been a few reports regarding this modality of therapy. Pan et al. reported results of Gamma Knife treatment of 20 patients with dural AVFs involving the junction of the transverse and sigmoid sinuses; several of these were classiﬁed as having some degree of cortical venous drainage [30]. They reported obliteration of the ﬁstula in 58% of patients, based primarily on MRI (less than half of obliterations were conﬁrmed with angiography at a mean of 20 months). All four patients with a dural AVF classiﬁed as Cognard type I (no cortical venous drainage, and anterograde ﬂow in the sinus) were conﬁrmed obliterated by angiography, compared with four of nine patients with cortical venous drainage (Cognard type IIb and IIa + b). Higher overall obliteration rates of 80% to 100% have been reported in several series of low-ﬂow cavernous sinus dural AVFs [31–33], with improvement of symptoms within 1 to 3 months. Combined endovascular and radiosurgical treatment has also been reported as a treatment paradigm [34–36]. Link et al. used a strategy of radiosurgery in 29 patients with dural AVFs in a variety of locations followed by embolization in a selected group of 17 patients, with the rationale of palliating symptoms or reducing risk of hemorrhage during the expected latency interval [34]. Follow-up angiograms obtained in 18 patients at 1 to 3 years revealed a 72% obliteration rate. A staged approach of radiosurgery and embolization was reported by Friedman et al. in 25 transverse-sigmoid dural AVFs, the majority of which had no cortical venous drainage (Cognard type I and IIa); follow-up angiography in 17 patients revealed complete obliteration in only 41%, although symptom improvement or resolution was reported in 96% [36]. Although the comparative numbers are very small, total obliteration was more prevalent in the most benign classiﬁcation, Cognard type I. In another series from the same group, 20 cavernous sinus dural AVFs were treated with staged radiosurgery and embolization; lesion obliteration was seen in 87% of 15 patients with follow-up imaging [35]. Radiosurgery alone was effective in this group, although it was believed that improvement took longer than in patients with adjunctive embolization. No cases of internal carotid artery occlusion or stenosis or optic or cranial nerve deﬁcit related to radiation were seen in this series, but the cohort was small, and long-term follow-up is lacking. Overall, success rates for radiosurgical treatment of dural AVFs appear to be highest in low-ﬂow cavernous dural AVFs. Results with radiosurgery alone or in conjunction with

516

s. amin-hanjani and f.g. barker ii

embolization for other dural AVFs are less convincing, although this strategy may be reasonable in patients whose lesions lack high-risk features on angiography or who are elderly and have signiﬁcant risk with other treatment strategies.

Observation Versus Radiosurgery Established treatment with surgical or endovascular obliteration is the treatment of choice for lesions presenting with hemorrhage or with cortical venous drainage indicating a high risk for bleeding as discussed above. Unlike parenchymal AVMs, which carry a relatively low yearly risk of hemorrhage, the annual risk of bleeding appears to be considerably higher for dural AVFs and mandates an intervention that obliterates the lesion quickly and effectively. In such high-risk lesions, observation alone is not appropriate, and radiosurgery, even if effective, carries a latency period prior to obliteration that makes it an inappropriate choice for most patients. In low-risk lesions, the available natural history data would support observation alone. If bothersome symptoms occur in an otherwise benign lesion, then radiosurgery alone or in conjunction with embolization may offer an additional treatment option, especially for low-ﬂow cavernous sinus dural AVFs (where the best radiosurgery results have been seen) or for low-risk lesions in other locations. However, the latency for treatment effect and relative lack of knowledge regarding longterm risks makes radiosurgery less attractive as an initial treatment modality, even under these limited circumstances.

ﬁed through screening studies while asymptomatic. Typically, each vHL patient harbors several hemangioblastomas scattered throughout the cerebellum, brain stem, and spinal cord, and resection of all of the lesions is not generally feasible. What little is known about the natural history of untreated hemangioblastomas derives from two studies of observation in vHL patients, and the applicability of these results to sporadic lesions is uncertain. Wanebo et al. followed 160 consecutive vHL patients with cranial and spinal MRI; they harbored a total of 655 individual hemangioblastoma lesions [37]. Among 88 patients with serial imaging over a period of 6 months or longer, 44% of solid tumor components increased in size, compared with 67% of associated tumor cysts. A saltatory growth pattern was seen, with periods of growth interspersed with apparently quiescent phases. No tumor or cyst showed spontaneous decrease in size. About 10% of patients showed formation of a new tumor cyst associated with a previously solid tumor over the follow-up period, and most patients developed at least one new lesion. Lesions associated with cysts were much more commonly symptomatic. In a much smaller series, Slater et al. reported very similar results (28 tumors in 8 patients) [38]. In a series that focused on hemangioblastoma treatment, Conway et al. found a similar rate of development of new hemangioblastomas in vHL patients; about three quarters of new lesions were identiﬁed in the presymptomatic state. These investigators did not comment on progression rates in existing lesions [39].

Radiosurgery

Hemangioblastoma Natural History Hemangioblastomas are rare, benign tumors of the central nervous system that occur most commonly in the cerebellar hemispheres, dorsal brain stem, and dorsal spinal cord. Histologically, they display many characteristics common to benign neoplasms, such as a low mitotic rate and bland cytology. However, because of their critical location within the nervous system, hemangioblastomas can cause rapidly progressive symptoms. Grossly, many hemangioblastomas take the form of solid mural nodules with associated cysts, and symptoms from mass effect usually appear to be more directly related to the cyst than to the smaller solid portion of the tumor. A substantial fraction of cerebellar hemangioblastomas, and nearly all hemangioblastomas located in the dorsal brain stem or spinal cord, are associated with a familial inherited syndrome known as von Hippel–Lindau disease (vHL). vHL patients suffer from cerebellar and retinal hemangioblastomas, renal cell carcinomas, cystic lesions within many abdominal organs, and other rarer tumors. vHL has an autosomal dominant inheritance pattern with high penetrance, and many affected family members are identiﬁed in the presymptomatic state through noninvasive screening images. Because of their critical location, frequent presentation with symptoms from mass effect, and uncertainty about diagnosis, most sporadic hemangioblastomas are resected as soon as they are detected. vHL patients, however, are often identi-

Six case series of small to moderate size of radiosurgical treatment of hemangioblastomas supply the bulk of published experience with this tumor [40–45]. The aggregate number of lesions treated was 157 in 99 patients, most of whom had vHL. Followup periods for most patients were short, and use of widely varying marginal doses further complicates interpretation of results. Most reports agree that control of treated solid tumor nodules is better (70% to 97%) than control of associated cysts (40% to 70% control, with some cysts shrinking after up to 1 year).

Observation Versus Radiosurgery For most sporadic hemangioblastomas and symptomatic hemangioblastomas in vHL patients, surgical excision is the preferred treatment. For patients who are not medically suitable for operation, for those with unresectable lesions, and perhaps for enlarging asymptomatic lesions, radiosurgery is a relatively promising treatment option whose role cannot yet be said to be well deﬁned. Further study is needed. When a large tumor cyst is present, radiosurgery does not provide short-term relief of mass effect or reliable prevention of further cyst growth, and surgery will often be indicated.

References 1. Robinson JR, Awad IA, Little JR. Natural history of the cavernous angioma. J Neurosurg 1991; 75(5):709–714.

51.

cavernous malformations and other vascular abnormalities: observation-alone perspective

2. Del Curling O Jr, Kelly DL Jr, Elster AD, Craven TE. An analysis of the natural history of cavernous angiomas. J Neurosurg 1991; 75(5):702–708. 3. Aiba T, Tanaka R, Koike T, et al. Natural history of intracranial cavernous malformations. J Neurosurg 1995; 83(1):56–59. 4. Zabramski JM, Wascher TM, Spetzler RF, et al. The natural history of familial cavernous malformations: results of an ongoing study. J Neurosurg 1994; 80(3):422–432. 5. Porter PJ, Willinsky RA, Harper W, Wallace MC. Cerebral cavernous malformations: natural history and prognosis after clinical deterioration with or without hemorrhage. J Neurosurg 1997; 87(2):190–197. 6. Kondziolka D, Lunsford LD, Kestle JR. The natural history of cerebral cavernous malformations. J Neurosurg 1995; 83(5): 820–824. 7. Robinson JR Jr, Awad IA, Magdinec M, Paranandi L. Factors predisposing to clinical disability in patients with cavernous malformations of the brain. Neurosurgery 1993; 32(5):730–735; discussion 735–736. 8. Labauge P, Brunereau L, Levy C, et al. The natural history of familial cerebral cavernomas: a retrospective MRI study of 40 patients. Neuroradiology 2000; 42(5):327–332. 9. Moriarity JL, Clatterbuck RE, Rigamonti D. The natural history of cavernous malformations. Neurosurg Clin N Am 1999; 10(3): 411–417. 10. Kim DS, Park YG, Choi JU, et al. An analysis of the natural history of cavernous malformations. Surg Neurol 1997; 48(1):9–17; discussion 18. 11. Kupersmith MJ, Kalish H, Epstein F, et al. Natural history of brainstem cavernous malformations.[see comment]. Neurosurgery 2001; 48(1):47–53; discussion 54. 12. Barker FG 2nd, Amin-Hanjani S, Butler WE, et al. Temporal clustering of hemorrhages from untreated cavernous malformations of the central nervous system. Neurosurgery 2001; 49(1):15– 24; discussion 25. 13. Pozzati E, Acciarri N, Tognetti F, et al. Growth, subsequent bleeding, and de novo appearance of cerebral cavernous angiomas. Neurosurgery 1996; 38(4):662–669; discussion 669–670. 14. Tung H, Giannotta SL, Chandrasoma PT, Zee CS. Recurrent intraparenchymal hemorrhages from angiographically occult vascular malformations. [see comment]. J Neurosurg 1990; 73(2): 174–180. 15. Awad I. Steretoactic radisourgery for cavernous malformations (comment). Neurosurgery 1998; 42:1237. 16. Amin-Hanjani S, Ogilvy CS, Candia GJ, et al. Stereotactic radiosurgery for cavernous malformations: Kjellberg’s experience with proton beam therapy in 98 cases at the Harvard Cyclotron. [see comment]. Neurosurgery 1998; 42(6):1229–1236; discussion 1236–1238. 17. Hasegawa T, McInerney J, Kondziolka D, et al. Long-term results after stereotactic radiosurgery for patients with cavernous malformations. Neurosurgery 2002; 50(6):1190–1197; discussion 1197– 1198. 18. Kondziolka D, Lunsford LD, Flickinger JC, Kestle JR. Reduction of hemorrhage risk after stereotactic radiosurgery for cavernous malformations.[see comment]. J Neurosurg 1995; 83(5): 825–831. 19. Chang SD, Levy RP, Adler JR Jr, et al. Stereotactic radiosurgery of angiographically occult vascular malformations: 14-year experience. Neurosurgery 1998; 43(2):213–220; discussion 220–221. 20. Karlsson B, Kihlstrom L, Lindquist C, et al. Radiosurgery for cavernous malformations. J Neurosurg 1998; 88(2):293–297. 21. Kim DG, Choe WJ, Paek SH, et al. Radiosurgery of intracranial cavernous malformations. Acta Neurochir 2002; 144(9):869–878; discussion 878.

517

22. Pollock BE, Garces YI, Stafford SL, et al. Stereotactic radiosurgery for cavernous malformations. J Neurosurg 2000; 93(6):987–991. 23. Barker F 2nd, Amin-Hanjani S, Butler W, et al. Complication rates after radisourgery for cavernous malformations. Neurosurgery 1998; 43:670. 24. St George EJ, Perks J, Plowman PN. Stereotactic radiosurgery XIV: the role of the haemosiderin “ring” in the development of adverse reactions following radiosurgery for intracranial cavernous malformations: a sustainable hypothesis. Br J Neurosurg 2002; 16(4):385–391. 25. Shih YH, Pan DH. Management of supratentorial cavernous malformations: craniotomy versus gammaknife radiosurgery. Clin Neurol Neurosurg 2005; 107(2):108–112. 26. Awad IA, Little JR, Akarawi WP, Ahl J. Intracranial dural arteriovenous malformations: factors predisposing to an aggressive neurological course. J Neurosurg 1990; 72(6):839–850. 27. Davies MA, TerBrugge K, Willinsky R, et al. The validity of classiﬁcation for the clinical presentation of intracranial dural arteriovenous ﬁstulas. J Neurosurg 1996; 85(5):830–837. 28. van Dijk JM, terBrugge KG, Willinsky RA, Wallace MC. Clinical course of cranial dural arteriovenous ﬁstulas with long-term persistent cortical venous reﬂux. Stroke 2002; 33(5):1233–1236. 29. Satomi J, van Dijk JM, Terbrugge KG, et al. Benign cranial dural arteriovenous ﬁstulas: outcome of conservative management based on the natural history of the lesion.[see comment]. J Neurosurg 2002; 97(4):767–770. 30. Pan DH, Chung WY, Guo WY, et al. Stereotactic radiosurgery for the treatment of dural arteriovenous ﬁstulas involving the transverse-sigmoid sinus.[see comment]. J Neurosurg 2002; 96(5): 823–829. 31. Barcia-Salorio JL, Soler F, Barcia JA, Hernandez G. Radiosurgery of carotid-cavernous ﬁstulae. Acta Neurochir Suppl 1994; 62:10–12. 32. Guo WY, Pan DH, Wu HM, et al. Radiosurgery as a treatment alternative for dural arteriovenous ﬁstulas of the cavernous sinus. AJNR Am J Neuroradiol 1998; 19(6):1081–1087. 33. Onizuka M, Mori K, Takahashi N, et al. Gamma knife surgery for the treatment of spontaneous dural carotid-cavernous ﬁstulas. Neurol Med Chir (Tokyo) 2003; 43(10):477–482; discussion 482–483. 34. Link MJ, Coffey RJ, Nichols DA, Gorman DA. The role of radiosurgery and particulate embolization in the treatment of dural arteriovenous ﬁstulas. J Neurosurg 1996; 84(5):804–809. 35. Pollock BE, Nichols DA, Garrity JA, et al. Stereotactic radiosurgery and particulate embolization for cavernous sinus dural arteriovenous ﬁstulae. Neurosurgery 1999; 45(3):459–466; discussion 466–467. 36. Friedman JA, Pollock BE, Nichols DA, et al. Results of combined stereotactic radiosurgery and transarterial embolization for dural arteriovenous ﬁstulas of the transverse and sigmoid sinuses. J Neurosurg 2001; 94(6):886–891. 37. Wanebo JE, Lonser RR, Glenn GM, Oldﬁeld EH. The natural history of hemangioblastomas of the central nervous system in patients with von Hippel-Lindau disease. J Neurosurg 2003; 98(1):82–94. 38. Slater A, Moore NR, Huson SM. The natural history of cerebellar hemangioblastomas in von Hippel-Lindau disease. AJNR Am J Neuroradiol 2003; 24(8):1570–1574. 39. Conway JE, Chou D, Clatterbuck RE, et al. Hemangioblastomas of the central nervous system in von Hippel-Lindau syndrome and sporadic disease. Neurosurgery 2001; 48(1):55–62; discussion 63. 40. Chang SD, Meisel JA, Hancock SL, et al. Treatment of hemangioblastomas in von Hippel-Lindau disease with linear accelerator-based radiosurgery. Neurosurgery 1998; 43(1):28–34; discussion 35.

518

s. amin-hanjani and f.g. barker ii

41. Jawahar A, Kondziolka D, Garces YI, et al. Stereotactic radiosurgery for hemangioblastomas of the brain. Acta Neurochir 2000; 142(6):641–644; discussion 644–645. 42. Niemela M, Lim YJ, Soderman M, et al. Gamma knife radiosurgery in 11 hemangioblastomas. J Neurosurg 1996; 85(4):591–596. 43. Pan L, Wang EM, Wang BJ, et al. Gamma knife radiosurgery for hemangioblastomas. Stereotact Funct Neurosurg 1998; 70 (Suppl 1):179–186.

44. Patrice SJ, Sneed PK, Flickinger JC, et al. Radiosurgery for hemangioblastoma: results of a multiinstitutional experience. Int J Radiat Oncol Biol Phys 1996; 35(3):493–499. 45. Rajaraman C, Rowe JG, Walton L, et al. Treatment options for von Hippel-Lindau’s haemangioblastomatosis: the role of gamma knife stereotactic radiosurgery. Br J Neurosurg 2004; 18(4):338– 342.

5 2

Trigeminal Neuralgia Lawrence S. Chin, Shilpen Patel, Thomas Mattingly, and Young Kwok

Introduction Trigeminal neuralgia (TN) is a facial pain syndrome characterized by a severe electrical, shooting pain lasting seconds to minutes that is conﬁned to one or more distributions of the trigeminal nerve. More recently, the International Headache Society has deﬁned TN pain as the following: (a) paroxysmal attacks, lasting from a second to 2 minutes, affecting one or more divisions of the trigeminal nerve; (b) pain that is intense, sharp, superﬁcial or stabbing, precipitated from trigger areas or factors; (c) attacks stereotyped in the individual patient; (d) no clinically evident neurologic deﬁcit; and (e) pain not attributed to another disorder [1]. The pain, typically unilateral, may be bilateral in approximately 5% of cases and may occur either spontaneously or be triggered by innocuous sensations or activities such as a gust of wind, chewing, drinking, brushing teeth, shaving, or washing the face. Severely affected patients may lose weight from not eating. The median age of presentation is in the sixties, and it affects women more often than men [2]. The diagnosis is based on history. Typically, the neurologic exam is normal, unless the syndrome is secondary to another process such as a tumor in the posterior fossa or multiple sclerosis. Imaging studies obtained in the workup include magnetic resonance imaging (MRI) to rule out structural causes, but usually this is normal. An atypical form of TN exists and is characterized by pain of longer duration that may be associated with a burning sensation and facial numbness [3]. Occasionally, patients with typical TN may develop atypical features over time. It is not unusual for TN to have periods of remissions and exacerbations; seasonal variation is also common with winter being a particularly difﬁcult time for most sufferers. Patients with TN usually note a progression of their symptoms with time: the pain becomes more intense and the remissions more rare. The pathophysiology of this process is somewhat controversial. As a result of surgical observations in the posterior fossa, Dandy postulated that vascular compression, usually from a branch of the superior cerebellar artery, at the trigeminal root was the primary cause of TN [4]. This idea was advanced by Gardner, Jannetta, and others who developed microvascular decompression (MVD) via retromastoid craniectomy as a treatment [5]. It is postulated that arterial pulsation of the trigeminal nerve root entry zone results in

demyelination and subsequent ephaptic transmission [6]. However, not all patients with typical TN have vascular compression found at MVD, and vascular compression of the trigeminal roots has been observed in patients without TN. In addition, patients with multiple sclerosis (MS) and brain-stem infarcts may develop TN, indicating that TN may be generated by central mechanisms as well [7].

Treatment Options Patients with TN have a number of treatment options. The ﬁrst line is medical therapy. Carbamazepine (Tegretol) is the drug of choice, but other anticonvulsants such as gabapentin (Neurontin), phenytoin (Dilantin), oxcarbazepine (Trileptal), lamotrigine (Lamictal), and topiramate (Topamax) can also be effective. Baclofen (Kemstro) has also been used though it has a different mechanism of action; it acts by inhibiting monosynaptic and polysynaptic spinal reﬂexes. In severe cases, two or more drugs may be required. Narcotic pain medications are typically not recommended because most trigeminal attacks do not last for a signiﬁcant duration to justify routine use of such medications. Medication may only help in approximately 75% of patients, and side effects such as drowsiness, cognitive dysfunction, bone marrow suppression, and hepatic dysfunction can be signiﬁcant. In addition, patients will frequently become refractory to medication over time [8]. If medical therapy fails or is not tolerated, there exist two primary surgical therapies: MVD and ablative percutaneous procedures. Performing an MVD requires general anesthesia and carries a risk of stroke, infection, hemorrhage, cerebrospinal ﬂuid (CSF) leak, facial numbness, weakness, and hearing loss. For patients over the age of 70 years and for those with surgical contraindications, MVD may not be an ideal treatment. The percutaneous approaches (radiofrequency ablation, glycerol injection, and balloon compression) to the trigeminal ganglion are performed under local anesthesia but are associated with a signiﬁcant risk of trigeminal dysfunction. Stereotactic radiosurgery (SRS) provides an alternative treatment that is minimally invasive yet effective and causes few side effects. SRS was ﬁrst described in 1951 by Lars Leksell as an adaptation of his arc-centered stereotactic frame with a radiation source replacing the lesioning electrode. In his initial attempts

519

520

l.s. chin et al.

TABLE 52-1. Results of linac-based or CyberKnife SRS in idiopathic TN.

Chen et al. [11] (Kaiser Los Angeles) Richards et al. [12] (Wisconsin) Lim et al. [13] CyberKnife (Stanford) Smith et al. [14] (UCLA)

No. of patients

Median follow-up (months)

BNI I–III relief (%)

Time to pain relief

Median dose (Gy)

Recurrence rate (%)

Complication rate (%)

32

8

78

1.5 months

NR

NR

6

28

12

75

1 month

80

46

14

41

11

93

7 days

78

16

51

60

23

72

2.7 months

90

26

25

NR, not reported.

at SRS, Leksell coupled a dental X-ray tube to his stereotactic arc to irradiate the trigeminal ganglion [9]. After trying several different radiation sources, he subsequently reported the successful treatment of TN in two patients with Gamma Knife stereotactic radiosurgery (GK-SRS) using cobalt-60 [10]. With recent improvements in technology and imaging, linear accelerator (linac)-based and CyberKnife SRS for TN have become more common. Table 52-1 compares some of the major series reported in the literature. The range of patients achieving good relief (Barrow Neurological Institute [BNI] I to III) is 72% to 93%. The median time to pain relief varies from 1 week to 2.7 months. An analysis of recurrence revealed rates that ranged from 6% to 51% [11–14]. Ma from the University of Maryland performed a comparative analysis of linac and GK-SRS for TN therapy. Speciﬁcally, the dose fall-off characteristics and set-up error tolerance of the two modalities were examined. This study found equivalent dose fall-off properties between Gamma Knife and linac-based SRS provided only if a sufﬁcient number of arcs (>7) and small intra-arc error (<0.5 mm) were satisﬁed for linac-based deliveries. Furthermore, because of the large number of arcs that are required, the treatment time for linac-based SRS compared with Gamma Knife is signiﬁcantly increased [15]. At this time, robust safety and long-term efﬁcacy data exist only for Gamma Knife treatment of TN; however, the clinical experience in this disease with other forms of radiosurgery is growing and will lead to greater patient options.

Mechanism of GK-SRS Effect The relationship between postprocedure numbness and efﬁcacy suggests that GK-SRS works by blocking axonal transmission. As predicted by models of radiation injury, both the time to effective pain relief and numbness are delayed, although pain relief frequently occurs many months before any side effects are experienced. Kondziolka et al. used a primate model to explore the effects of 80 or 100 Gy to trigeminal nerves and observed a combination of axonal degeneration and edema [16]. Necrosis was seen in nerves that received the higher dose, and both myelinated and unmyelinated ﬁbers were equally affected. Why the functional improvement is seen in patients before these histologic changes are seen is unknown, but an effect of GK-SRS on ephaptic transmission provides a possible mechanism.

Targeting, Dose, and Dose Rate Early SRS treatments for TN targeted the trigeminal ganglion because the trigeminal cistern in the Meckel cave could be inferred from landmarks on skull ﬁlms, a practice later reﬁned by stereotactic cisternography. In fact, glycerol rhizotomy was discovered incidentally during the use of glycerol as a carrier for tantalum dust in outlining the trigeminal cistern [17]. In 1991, Lindquist et al. reported 46 cases of GK-SRS directed at the trigeminal ganglion with 22 patients localized using cisternography [18]. They did not report prescription doses. The results were disappointing as only 4 of 22 patients had prolonged pain relief. Rand et al. reported a series of 12 patients treated by GK-SRS doses ranging from 57 to 74 Gy with only mild success and concluded that the ganglion was not the appropriate target [19]. Based on these ﬁndings and with the development of high-resolution MRI, which made it possible to reliably target the trigeminal nerve anywhere along its course from the pons to the Meckel cave, the optimal target moved toward the trigeminal nerve root entry zone. In 1996, Kondziolka et al. published the results of a multicenter trial of GK-SRS that standardized the target [20]. A single 4-mm isocenter was applied to the trigeminal nerve 2 to 4 mm anterior to the junction of the nerve at the pons as seen on axial and coronal MRI scans. This location has since been conﬁrmed effective by several large series [21]. The 20% isocenter line typically contacts the brain stem using this protocol, which is well tolerated at the current maximum doses of 70 to 80 Gy to the isocenter. Evaluation of brain-stem dose as well as evidence of compression was evaluated at the University of Maryland. Doses to the brain stem ranged from 10% to 20%, and evidence of compression was seen in 12% of patients. Both brain-stem dose as well as evidence of compression did not have any effect on treatment response, medication use, or development of facial numbness [22]. Beam shaping is accomplished by plugging some of the 201 sources during a treatment and can be used to accomplish a ﬂattening of the lower isodose curves such that the 10% line is parallel to the brain stem (Fig. 52-1). Theoretically, this should decrease the risk of radiation necrosis to the brain stem. Some centers have also tried using two isocenters to irradiate a greater length of nerve. A randomized, prospective study reported by Flickinger et al. found no increase in efﬁcacy but did show a statistically signiﬁcant increase in trigeminal dysfunction [23]. Efﬁcacy was reported to be 67.7 ± 5.1% in both arms, and facial numbness and mild and severe

52.

trigeminal neuralgia

521

Case 52-1 This 42-year-old man with a 10-year history of well-controlled multiple sclerosis began experiencing typical right V3 trigeminal neuralgia 4 years prior to treatment. He had been placed on increasing doses of Tegretol and Neurontin with decreasing effectiveness. The increased doses of medication also worsened his MS symptoms. He had typical lancinating pain with no numbness that was triggered by touching the face and talking. At the time of GK-SRS, he had T1- and T2-weighted MRI scans done. Because the T1 scans did not show the nerve, we used the T2 scan to target the nerve (Fig. 52-2A, B). A 4-mm shot was used with a plugging pattern to shape the 10% and 20% isodose lines. A dose of 75 Gy was delivered to the 100% line. One month after GK-SRS, he noticed the sudden improvement in his pain. At 2 months after treatment, he was weaned off both medications. He remains pain-free for 2 years.

FIGURE 52-1. Gamma Knife treatment plan for a patient with right trigeminal neuralgia. Isodose coverage of the nerve is seen in the right upper panel. The 20% line touches the brain stem, and the 10% line cuts into it. Beam shaping is accomplished by plugging 32 beams as shown in the left lower panel. The result, seen in the right lower panel, is a ﬂattened 10% and 20% isodose line such that the brain stem receives 10% only along the edge.

paresthesias developed in 19%, 12%, and 2% in two isocenter patients versus 7%, 9%, and 0 in one isocenter patient, respectively. In the multivariate analysis, the length of nerve treated was the only signiﬁcant factor predicting for facial numbness. In Leksell’s original 1971 paper, he described successful treatment at maximum doses of 16.5 Gy and 22 Gy [10]. Patients studied in the 1996 multicenter trial received doses ranging from 60 to 90 Gy [20]. Patients who received at least 70 Gy fared signiﬁcantly better, but no difference was seen by increasing the dose to 90 Gy. At these doses, the edge of the brain stem receives no more than 16 Gy, which is well tolerated. At the University of Maryland, MRI imaging is performed with gadolinium-enhanced short-TR sequences (T1-weighted) with axial volume acquisitions using 512 × 216 matrices at 1.5mm slice intervals. Occasionally, it is difﬁcult to visualize the affected prepontine trigeminal nerve on the axial MRI view. This was evaluated and it was found that MRI quality did not affect treatment response, medication use, and development of facial numbness [22]. Based on this study, it is recommended that if the trigeminal nerve is difﬁcult to image, additional axial long-TR sequences (T2-weighted) or use of the coronal and sagittal views are useful (Case 52-1). A single 4-mm isocenter shot is placed just anterior to the entry of the trigeminal nerve into the pons with the 20% isocenter barely touching the pons surface. Coronal MRI scans are examined to ensure that the 50% line surrounds the nerve, and 90% and 99% lines are evaluated to make sure the center of the nerve receives the maximum dose. A customized plugging pattern is then applied that elongates the lower isodose curves and minimizes the volume of brain stem covered by the 10% line. Finally, a dose of 75 Gy is delivered to the 100% isocenter (37.5 Gy to the 50%).

Patients with a pacemaker or other MRI contraindications present a dilemma. Provided the patient is not on anticoagulants, computed tomography (CT) contrast cisternography is used for localizing the trigeminal nerve (Case 52-2). In rare circumstances, an MRI scan in patients with a pacemaker is required. A cardiologist is present in these cases in the MRI suite with a deﬁbrillator, and the pacemaker is programmed before the scan to avoid extraventricular pacing and is then reset immediately after the MRI. Radiobiologically for gamma rays, dose rate is one of the principal factors that determine the biologic consequences of a given absorbed dose. As the dose rate is lowered and the exposure time extended, the biologic effect of a given dose in generally reduced. The University of Maryland examined this effect in a series of patients treated the year prior to a source change (median dose rate, 162 cGy/min; range, 151 to 179 cGy/min) compared with patients treated in the year immediately after a source change (median dose rate, 343 cGy/min; range, 321 to 366 cGy/min) and found the difference in pain relief was 61% compared with 83%, respectively (p = 0.05). These data suggest that if dose rate falls below 179 cGy/min, consideration should be made for dose escalation in the treatment of TN [24].

FIGURE 52-2. MRI images from the Gamma Knife treatment of a patient with right trigeminal neuralgia. (A) T1-weighted MRI scan that shows the left nerve (arrow) but not the right. (B) T2-weighted MRI scan at the same level illustrates the right nerve well (arrow).

522

l.s. chin et al.

Case 52-2 This 60-year-old man had a 5-year history of typical trigeminal neuralgia that initially responded to glycerol rhizotomy. His pain recurred and he required treatment with Tegretol, which was becoming ineffective. One year prior to GK-SRS, he had a cardiac deﬁbrillator placed for arrhythmias. A cisternogram was performed through a C1-2 puncture. Axial CT scans were obtained that showed the right trigeminal nerve (Fig. 52-3). The nerve was targeted according to protocol with a 4-mm shot of radiation with a total dose of 75 Gy delivered to the isocenter. Three weeks after treatment, his pain disappeared. He remains pain-free 1 year after treatment.

Treatment Response Assessing positive treatment responses in TN is challenging because the symptoms are subjective. The Barrow Neurological Institute (BNI) pain intensity score combines severity of pain with medication use (Table 52-2) [25]. Patients able to stop all medications are classiﬁed as BNI I if they have complete pain relief or BNI II if they have occasional tolerable pain. Patients remaining on medication are BNI III if they have good pain relief, but are BNI IV if they are not adequately controlled. Patients receiving no relief are classiﬁed BNI V. Most patients are BNI V before their treatment. For comparison with other series, we consider BNI I to III to be a good outcome. The

TABLE 52-2. Barrow neurological institute pain intensity score. I II III IV V

No trigeminal pain, no medications Occasional pain that is well tolerated, no medications Occasional pain that requires medication to be tolerated Pain that is not adequately controlled with medication Severe pain without relief

McGill pain score is useful in the preoperative patient assessment (Table 52-3). Patients are asked to rate their TN pain and their worst non-TN headache using a range from mild (I) to excruciating (V). A comparison of selected major series reported in the literature is shown in Table 52-4. In general, there is good correlation between the different series. The range of patients achieving good relief (BNI I to III) is 69% to 94% with 22% to 76% achieving complete pain relief on no medications (BNI I). The median time to pain relief varies from 2 to 4 weeks. An actuarial analysis of recurrence has shown rates of 23%, 33%, and 39% at 1, 2, and 3 years, respectively [26]. Quality of life assessments have documented that most patients believe their GK-SRS treatments were successful even when the relief was temporary. The lack of side effects appears signiﬁcant in this regard, and a high proportion of patients who fail initial GK-SRS elect for an additional treatment. A variety of prognostic factors have been found to correlate with treatment outcome. A universal ﬁnding in all forms of surgical therapy for TN is that the patient’s initial treatment is the most effective; the results with GK-SRS are no different. Most studies ﬁnd that the absence of a prior surgical treatment correlates with a better outcome [21, 26, 28, 31]. Other preoperative prognostic factors for treatment success include a shorter duration of symptoms and a patient description (McGill pain scale) of their TN pain as severe and their non-TN headaches as mild [26]. The implications of this ﬁnding are unclear. Patients with more classic TN pain may have fewer secondary gain issues or conﬂicting social issues that may impair their treatment. A prognostic ﬁnding that has had conﬂicting results is the presence of atypical TN symptoms. Whereas some studies indicate atypical symptoms carry an unfavorable prognosis, others have not conﬁrmed this [20, 21, 25, 26, 28]. A minimum treatment dose of 70 Gy appears necessary for a good outcome, but doses of 90 Gy and above do not appear to provide additional pain relief. The issue of dose escalation, however, remains an open question for a number of reasons. The possibility of a type II error in the published clinical series may mask a small beneﬁt from a higher prescription dose. Also, the development of posttreatment facial numbness has been shown to correlate with good results, and because facial numbness is dependent on the dose, an increased dose should result in an improved or more durable outcome [25, 31]. The developTABLE 52-3. McGill pain assessment score.

FIGURE 52-3. CT cisternogram of a patient with right trigeminal neuralgia and a cardiac deﬁbrillator. CT contrast in the CSF cisterns around the trigeminal nerve highlights it well (arrow).

I II III IV V

Mild Discomforting Distressing Horrible Excruciating

52.

523

trigeminal neuralgia

TABLE 52-4. Results of initial GK-SRS in idiopathic TN. No. of patients

Kondziolka et al. [20] (University of Pittsburgh) Maesawa et al. [21] (University of Pittsburgh) Rogers et al. [25] (Barrow Neurological Institute) Petit et al. [26] (University of Maryland) Pollock et al. [27] (Mayo Clinic) Young et al. [28] (Northwest Gamma Knife Center) Sheehan et al. [29] (University of Virginia) Brisman et al. [30] (Columbia)

Median follow-up

BNI I relief (%)

BNI I–III relief (%)

Time to pain relief (weeks)

Recurrence rate (%)

Median dose

Complication rate (%)

50

18 months

58

94

4

70 Gy to the 100%

6

10

220

22 months

40

69

8

80 Gy to the 100%

13.6

7.7

54

12 months

35

89

2

35 Gy prescribed to the 50%

21

9

96

30 months

42

75

3

75 Gy to the 100%

29

7.3

117

26 months

58

75

3

80 Gy to the 100%

23

110

19.8 months

76

88

2

70 Gy to the 100%

34

2.7

151

19 months

47

70

3.5

80 Gy to the 100%

27

9

293

4.6 years

22

76

NR

76.8 Gy to the 100%

24

12

37

NR, not reported.

ment of more dysesthesias with higher doses, however, may temper the desire to use higher doses.

reported but has been observed in our patients is the temporary exacerbation of trigeminal pain in the initial month after GKSRS. In no patients has this become permanent.

Complications The distinguishing characteristic of GK-SRS for TN is the very low morbidity and absence of mortality. Similar to the percutaneous ablative procedures, the primary side effect is trigeminal dysfunction. There is no risk to any cranial nerves other than cranial nerve V. The reported risk of posttreatment numbness or paresthesias varies from 2.7% to 37% (Table 52-4) and is not bothersome for the majority. The primary risk factor for developing this complication is receiving a dose of 90 Gy. Only one case of deafferentation pain/pain due to loss of sensory input into the central nervous system, the most feared complication because it is so difﬁcult to treat, has been reported [21]. Corneal numbness, another dangerous complication because of the potential risk of permanent damage to the eye, has been reported in only three patients, all treated with 90 Gy [27]. No mastication motor deﬁcits were seen in any study. As expected, patients who undergo a repeat GK-SRS treatment to the same trigeminal nerve have a higher incidence of facial numbness, but there has not been an increase of dysesthesias or symptomatic radiation necrosis reported. A complication that is not

Treatment Options for Recurrent Trigeminal Neuralgia Although the initial response to GK-SRS is often quite good, the response may not be durable. As seen in patients treated by MVD and the percutaneous procedures, recurrence rates increase with time. Using an actuarial analysis, this recurrence rate for GK-SRS may be as high as 39% by 3 years after treatment. Petit found that absence of a prior surgical treatment, duration of TN less than 50 months before GK-SRS, and a BNI I response to GK-SRS correlated with smaller risk of recurrence [26]. Most patients choose to undergo a repeat GK-SRS at ﬁrst recurrence. Data from four institutions on repeat GK-SRS for recurrent TN are presented in Table 52-5. Although there are variations in technique, primarily related to the dose given at the second treatment, the results of repeat GK-SRS are generally quite good. The University of Maryland reported 18 patients from a total series of 112 patients who underwent repeat GKSRS for TN. BNI I result was seen in 45% and BNI I to III was

TABLE 52-5. Results of repeat GK-SRS for idiopathic TN. N

Herman [32] (University of Maryland) Hasegawa et al. [33] (University of Pittsburgh) Shetter [34] (Arizona Oncology Services) Pollock [35] (Mayo Clinic)

Follow-up (months)

BNI class I (%)

BNI class I–III (%)

First dose (Gy)

Second dose (Gy)

Trigeminal dysfunction (%)

18/112

28

45

78

75

70

11

27/387

20.4

19

85

75

64

11

19/240

13.5

53

85

78

46.6

42

10/100

15

80

80

70

90

80

524

l.s. chin et al.

seen in 33%. Repeat GK-SRS was more effective for patients that experienced a longer period of pain relief after their initial GK-SRS. Patients that had no response to initial GK-SRS also failed to respond to their second treatment. Only 11% experienced increased numbness in this experience. The target for the second treatment is usually not different from the initial treatment, although Hasegawa advocated placing the second lesion just anterior to the ﬁrst one to reduce the brain-stem dose [33]. The optimum dose to be used at second treatment has not been established. Some centers use a lower dose at repeat treatment, whereas others have advocated increasing the second dose to 90 Gy. The University of Maryland recently presented data comparing treatment outcomes for patients treated with repeat radiosurgery for refractory or recurrent TN with a low dose (70 to 75 Gy) performed at the University of Maryland and a high dose (90 Gy) performed at the University of Kentucky. Similar rates of pain control were seen in both groups. The higher dose was associated with an increased probability of discontinuing all medications. It is worth noting though that 25% of these patients had increased or new-onset numbness or dysesthesias [36]. Not unexpectedly, using the higher doses results in 80% of patients experiencing trigeminal dysfunction, but the percentage of patients having an excellent result (BNI I) is also increased. Because there is no precedence for performing a third GK-SRS on the same nerve, the risk of radiation necrosis of the brain stem, which is presumed to be high, is unknown. The alternative to a repeat GK-SRS is either MVD or a percutaneous procedure. Pollock et al. found that an excellent outcome was seen in 75% of patients undergoing MVD and in 71% undergoing glycerol rhizotomy [31]. Interestingly, ﬁve of eight patients were observed to have a region of presumed radiation injury on the segment of the superior cerebellar artery that contacted the trigeminal nerve. In four of these ﬁve cases, the patient either received a high dose (90 Gy) or had a repeat GK-SRS treatment. Our anecdotal experience with MVD after failed GK-SRS is that arachnoidal adhesions do form on the trigeminal nerve and can complicate mobilization of arterial loops. Good outcomes, however, are still possible with MVD in this setting [37].

Secondary Trigeminal Neuralgia Patients with secondary TN experience typical TN symptoms caused by the compressive effects of a tumor. The two most common tumors are meningiomas and trigeminal schwannomas, but other tumor types such as metastases, chordomas, and epidermoids are also possible. These tumors are usually located in the cavernous sinus, the Meckel cave, petroclival region, or tentorial notch. In addition to TN, patients may present with facial numbness, weakness, diplopia, visual loss, and hemiparesis. Treatment with GK-SRS in these patients is both effective at controlling tumor growth and TN symptoms. Regis et al. reported 46 patients with tumors amenable to GK-SRS were treated with a multiple-shot dose plan to a mean dose of 14 Gy (range, 8 to 25 Gy) [38]. Forty-ﬁve patients were followed for a median time of 55 months. Initially, 80% had complete pain relief (BNI I) and 16% were improved. Recurrent pain was

experienced by 6 (13.3%) patients, and slight hypesthesias were seen in 2 (4.4%). In comparison with published microsurgical series, GK-SRS for secondary TN has a much lower rate of postoperative trigeminal dysfunction (4% vs. 17% to 45%), and essentially no risk of other cranial nerve deﬁcit or motor weakness. In addition, patients avoid the customary complications of craniotomy and general anesthesia. The mechanism of pain relief is not known because these benign tumors do not change signiﬁcantly in size after GK-SRS, and the doses given are signiﬁcantly smaller than typical doses for idiopathic TN. Compensating for the lower doses may be the fact that the prescription isodose lines in the tumor cases are generally much closer to the brain stem indicating that the brain stem at the root entry zone may receive a dose similar to patients treated by the standard 4-mm shot protocol. These results raise the interesting hypothesis that the effectiveness of GK-SRS in TN is related more to the brain-stem dose than to the nerve dose [39]. Further studies will be needed to clarify this issue.

Multiple Sclerosis, Post-Herpetic Neuralgia, and Atypical Facial Pain TN is a well-recognized symptom of MS, and its manifestations are indistinguishable from classic, idiopathic TN. It occurs in 1% to 2% of patients with MS but rarely as the presenting symptom. Patients with MS present at an earlier age and frequently have bilateral pain [39]. Pharmacological treatment of MS-TN patients is complicated by their greater sensitivity to gait imbalance and other side effects of medications. Both MVD and percutaneous procedures have a lower rate of effectiveness for patients with MS, but GK-SRS has been shown to be effective in up to 80% of MS-TN patients [40]. The treatment protocol for these patients is unchanged with respect to dose and target. Because MS patients comprise a small portion of the major series published to date, their sensitivity to GKSRS compared with typical TN patients is unknown [26, 28]. An increase in trigeminal complications has not been seen [40]. The use of GK-SRS for atypical facial pain syndromes such as post-herpetic neuralgia is controversial. Pharmacological therapy is frequently ineffective, and MVD and percutaneous treatments are contraindicated because they may worsen the pain. Only nucleus caudalis dorsal root entry zone (DREZ) lesions have been documented to show some beneﬁt. Urgosik reported on 16 patients who received GK-SRS and found a successful result (excellent, very good, and good) in only 44% of patients. Pain relief occurred after a median interval of 1 month, and no radiation-related side effects have been observed in these patients [41]. The University of Maryland experience with atypical facial pain syndromes is similar. In this study, approximately half of the patients presented with atypical TN and half initially presented with classic TN that then progressed to atypical TN. Seventy-two percent of patients reported pain relief with a median time to relief of 5.8 weeks (range, 0 to 24 weeks). There was no signiﬁcant difference between those that presented with atypical pain and those that progressed to atypical TN. Approximately 60% of patients were able to discontinue or decrease the use of pain medication. Bothersome numbness was reported in ﬁve (27%) patients. Of the patients

52.

trigeminal neuralgia

with sustained pain relief, quality of life improved an average of 82%, and of the patients in whom pain returned, quality of life improved an average of 63% [42]. The standard treatment protocols are used, and an increase in complications or worsened pain has not been noted.

Conclusion GK-SRS for TN is effective and should be considered a standard alternative to conventional surgical treatments for idiopathic TN. Initial results are comparable with other treatments, but a signiﬁcant recurrence rate is observed. It is, however, an extraordinarily safe procedure particularly for elderly patients and those wishing to avoid surgery or have surgical contraindications.

References 1. International Headache Society. International Classiﬁcation of Headache Disorders, 2nd ed. Oxford: Blackwell Publishing, 2003. 2. Wilkins R. Trigeminal neuralgia: introduction. In: Wilkins R, Rengachary SS, eds. Neurosurgery. New York: McGraw-Hill, 1996:3921–3929. 3. Hughes B. Atypical trigeminal neuralgia. Br Dent J 1950; 89(11):243–249. 4. Dandy WE. Concerning the cause of trigeminal neuralgia. Am J Surg 1934; 24:447–455. 5. Lunsford LD, Apfelbaum RI. Choice of surgical therapeutic modalities for treatment of trigeminal neuralgia: microvascular decompression, percutaneous retrogasserian thermal, or glycerol rhizotomy. Clin.Neurosurg 1985; 32:319–333. 6. Gardner WJ. Trigeminal neuralgia. Clin Neurosurg 1968; 15:1– 56. 7. Loeser JD, Calvin WH, Howe JF. Pathophysiology of trigeminal neuralgia. Clin Neurosurg 1977; 24:527–537. 8. Canavero S, Bonicalzi V. Drug therapy of trigeminal neuralgia. Expert Rev Neurother 2006; 6(3):429–440. 9. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316–319. 10. Leksell L. Sterotaxic radiosurgery in trigeminal neuralgia. Acta Chir Scand 1971; 137:311–314. 11. Chen JCT, Girvigian M, Greathouse H, et al. Treatment of trigeminal neuralgia with linear accelerator radiosurgery: initial results. J Neurosurg 2004; 101(Suppl 3):346–350. 12. Richards GM, Bradley KA, Tomé WA, et al. Linear accelerator radiosurgery for trigeminal neuralgia. Neurosurgery 2005; 57:1193–1200. 13. Lim M, Villavicencio AT, Burneikiene S, et al. CyberKnife radiosurgery for idiopathic trigeminal neuralgia. Neurosurg Focus 2005; 18:E9. 14. Smith ZA, De Salles AA, Frighetto L, et al. Dedicated linear accelerator radiosurgery for the treatment of trigeminal neuralgia. J Neurosurg 2003; 99(3):511–516. 15. Ma L, Kwok Y, Chin LS, et al. Comparative analyses of linac and Gamma Knife radiosurgery for trigeminal neuralgia treatments. Phys Med Biol 2005; 50:5217–5227. 16. Kondziolka D, Lacomis D, Niranjan A, et al. Histological effects of trigeminal nerve radiosurgery in a primate model: implications for trigeminal neuralgia radiosurgery. Neurosurgery 2000; 46:971– 976.

525

17. Hakanson S. Transovale trigeminal cisternography. Surg Neurol 1978; 10:137–144. 18. Lindquist C, Kihlstrom L, Hellstrand E. Functional neurosurgery—a future for the Gamma-Knife. Stereotact Funct Neurosurg 1991; 57:72–81. 19. Rand RW, Jacques DB, Melbye RW, et al. Leksell Gamma Knife treatment of tic douloureux. Stereotact Funct Neurosurg 1993; 61(Suppl 1):93–102. 20. Kondziolka D, Lunsford LD, Flickinger JC, et al. Stereotactic radiosurgery for trigeminal neuralgia: a multi-institutional study using the gamma unit. J Neurosurg 1996; 84:940–945. 21. Maesawa S, Salame C, Flickinger JC, et al. Clinical outcomes after stereotactic radiosurgery for idiopathic trigeminal neuralgia. J Neurosurg 2001; 94:14–20. 22. Cheuk AV, Chin LS, Petit JH, et al. Gamma Knife surgery for trigeminal neuralgia: outcome, imaging, and brainstem correlates. Int JRadiat Oncol Biol Phys 2004; 60(2):537–541. 23. Flickinger JC, Pollock BE, Kondziolka D, et al. Does increased nerve length within the treatment volume improve trigeminal neuralgia radiosurgery? A prospective double-blind, randomized study. Int J Radiat Oncol Biol Phys 2001; 51:449–454. 24. Patel S. Evaluating the inﬂuence of dose-rate on outcome with Gamma-Knife stereotactic radiosurgery in the treatment of trigeminal neuralgia. Int J Radiat Oncol Biol Phys 2005; 63(S1): S429. 25. Rogers CL, Shetter AG, Fiedler JA, et al. Gamma knife radiosurgery for trigeminal neuralgia: the initial experience of The Barrow Neurological Institute Int J Radiat Oncol Biol Phys 2000; 47: 1013–1019. 26. Petit JH, Herman JM, Nagda S, et al. Radiosurgical treatment of trigeminal neuralgia: evaluating quality of life and treatment outcomes Int J Radiat Oncol Biol Phys 2003; 56:1147–1153. 27. Pollock BE, Phuong LK, Foote RL, et al. High-dose trigeminal neuralgia radiosurgery associated with increased risk of trigeminal nerve dysfunction. Neurosurgery 2001; 49:58–62. 28. Young RF, Vermeulen S, Posewitz A. Gamma Knife radiosurgery for the treatment of trigeminal neuralgia. Stereotact Funct Neurosurg 1998; 70:192–199. 29. Sheehan J, Pan H-C, Stroila M, et al. Gamma Knife surgery for trigeminal neuralgia: outcomes and prognostic factors. J Neurosurg 2005; 102:434–441. 30. Brisman R. Gamma Knife surgery with a dose of 75 to 76.8 Gray for trigeminal neuralgia. J Neurosurg 2004; 100:848–854. 31. Pollock BE, Phuong LK, Gorman DA, et al. Stereotactic radiosurgery for idiopathic trigeminal neuralgia. J Neurosurg 2002; 97:347–353. 32. Herman JM. Repeat gamma knife radiosurgery for refractory or recurrent trigeminal neuralgia: treatment outcomes and quality of life assessment. Int J Radiat Oncol Biol Phys 2004; 59(1):112– 116. 33. Hasegawa T, Kondziolka D, Spiro R, et al. Repeat radiosurgery for refractory trigeminal neuralgia. Neurosurgery 2002; 50:494– 500. 34. Shetter AG. Gamma Knife: radiosurgery for recurrent trigeminal neuralgia. Neurosurgery 2002; 97(5 Suppl):536–538. 35. Pollock BE. Results of repeated gamma knife radiosurgery for medically unresponsive trigeminal neuralgia. J Neurosurg 2000; 93(3):162–164. 36. Dutta P. Comparison of repeat GK-SRS for refractory or recurrent trigeminal neuralgia: does dose matter. Int J Radiat Oncol Biol Phys 2005; 60(S1):S547. 37. Shetter AG, Zabramski JM, Speiser BL. Microvascular decompression after gamma knife surgery for trigeminal neuralgia: intraoperative ﬁndings and treatment outcomes. J Neurosurg 2005; 102(Suppl):259–261.

526

l.s. chin et al.

38. Regis J, Metellus P, Dufour H, et al. Long-term outcome after gamma knife surgery for secondary trigeminal neuralgia. J Neurosurg 2001; 95:199–205. 39. Brisman R. Trigeminal neuralgia and multiple sclerosis. Arch Neurol 1987; 44:379–381. 40. Rogers CL, Shetter AG, Ponce FA, et al. Gamma knife radiosurgery for trigeminal neuralgia associated with multiple sclerosis. J Neurosurg 2002; 97:529–532.

41. Urgosik D, Vymazal J, Vladyka V, et al. Treatment of postherpetic trigeminal neuralgia with the gamma knife. J Neurosurg 2000; 93:165–168. 42. Dhople A. Efﬁcacy and quality of life in patients with atypical trigeminal neuralgia treated with gamma knife stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2005; 63(S2): S106.

5 3

Trigeminal Neuralgia: Surgical Perspective David B. Cohen, Michael Y. Oh, and Peter J. Jannetta

Introduction Trigeminal neuralgia, characterized by intermittent sharp, lancinating pain in the distribution of one or more branches of the trigeminal nerve, is the most common facial pain syndrome in the United States, with approximately 15,000 cases diagnosed each year [1]. Although a minority of cases are caused by spaceoccupying posterior fossa lesions such as tumors or vascular malformations, it is now well accepted that in most cases, the disorder results from vascular compression of the trigeminal nerve root [2–5]. The majority of patients respond, at least initially, to medical management, with surgery reserved for patients who do not respond to medications. Surgical options range from percutaneous, destructive procedures (glycerol or radiofrequency rhizotomy, balloon compression) to the invasive but nondestructive microvascular decompression. This chapter will review each of these surgical methods of treatment and compare and contrast them with stereotactic radiosurgery in the broader context of clinical decision-making.

Glycerol Rhizotomy Glycerol rhizotomy was initially described by Håkanson and colleagues [6]. He described a series of 75 patients with a mean follow-up of 17 months. Eighty-six percent of the patients became pain-free after 1 to 2 injections, with the pain recurring during the follow-up period in 18%.

Technique Modiﬁcations have been made over time by various surgeons, but the basic technique is still similar to the original description [6]. The procedure is generally carried out under local anesthesia. Brieﬂy, in the supine position with the patient’s head rotated about 15° away from the side of pain, a 22-gauge spinal needle is inserted 2.5 to 3 cm lateral to the corner of the mouth and then guided into the foramen ovale under ﬂuoroscopic control. Mental visualization of the trajectory of the needle can be aided by imagining two points, one on the lower eyelid at

the medial edge of the pupil, and another about 3 cm anterior to the external auditory meatus. The coronal plane passing through the point anterior to the external meatus intersects the sagittal plane passing through the medial edge of the pupil at the medial aspect of the foramen ovale. Entry into the foramen is conﬁrmed by spontaneous cerebrospinal ﬂuid (CSF) ﬂow upon removal of the stylet in the spinal needle. Often, entry into the foramen is signaled by a contraction of the masseter muscle. With the patient in the sitting position, glycerol is slowly injected through the needle. Injection continues until the patient begins to feel paresthesias in the distribution of the trigeminal branch closest to the needle, with generally at least 0.15 mL being required at this point. The patient’s facial sensation is then assessed and injection continued until the patient notices a signiﬁcant decrease compared with the other side or until a decrease in the corneal reﬂex is noted. In total, about 0.2 to 0.4 mL of glycerol is used [7]. The patient is then maintained in a sitting position with the head ﬂexed to prevent the glycerol from escaping too rapidly from the cistern [6].

Results Several authors have published series of trigeminal neuralgia patients treated with glycerol rhizotomy. Lunsford et al. [8] reported 67% of patients with complete relief and 23% with improvement, and Fujimaki et al. [9] demonstrated a 26% rate of long-lasting relief. Seventeen percent of Lunsford’s patients eventually had a second procedure, and 72% of Fujimaki’s patients had a recurrence within 54 months of follow-up. North et al. [10], in a series of 85 patients, noted a median time to recurrence of pain requiring further treatment of 2 years. Slettebø [11] treated 60 patients with a median 7-year follow-up and noted 93% of patients to have initial pain relief. The halflife of the treatment effect was 47 months, with 53% of patients eventually experiencing recurrence of their pain. 75% of patients initially had facial numbness, with numbness persisting at follow-up in 33%. Thirty-eight percent of patients also noted long-lasting dysesthesias, an effect elaborated on further by Burchiel [12].

527

528

d.b. cohen et al.

Burchiel [12] performed glycerol rhizotomy on 60 patients with 1 year of follow-up. Eighty percent of patients obtained relief initially, with life-table analysis indicating that at 18 months, 50% of patients had persistent or recurrent pain minor complications were noted in 23% and major morbidity in 1.6%. Seventy-two percent of patients had facial sensory loss for at least 1 month, 15% had corneal hypesthesia, and 7% had corneal anesthesia. Importantly, in this study, the rate of success was directly related to the production of facial sensory loss. Glycerol is an effective procedure for the short-term treatment of facial pain. The recurrence rate is signiﬁcant as patients are followed for longer periods of time. It is a destructive procedure, with the rate of success correlating with the degree of damage inﬂicted on the nerve [12]. The procedure is fairly well tolerated with low rates of side effects other than trigeminal nerve–related effects and represents a good option for patients who refuse or are not good candidates for a posterior fossa craniotomy.

Yoon et al. [16] followed 81 patients after RF rhizotomy for between 6 and 11 years. The initial success rate was 87%. The probability of being pain-free successively decreased from 65% at 1 year to 49% at 2 years and 26% at 11 years. Likewise, Taha et al. [17] studied 154 patients, of whom 100 were prospectively followed for 15 years. Ninety-nine percent initially were relieved of their pain, and 23% experienced dysesthesias. The 14-year recurrence rate was 25% but was greater in patients with less dysesthesias. Ischia et al. [18] noted a 67% rate of complete relief after an average follow-up period of 3.7 years in 124 patients, with a recurrence rate of 28.2%. Complications included anesthesia dolorosa in 3%, dysesthesias in 3%, and paresthesias in 17%. Radiofrequency rhizotomy is another effective percutaneous procedure, but one that is destructive and whose success depends on the degree of destruction of the nerve. The severity of the possible side effects should not be underestimated. Radiofrequency rhizotomy is another viable option for patients who are not candidates for microvascular decompression because of medical comorbidities or patient preference.

Radiofrequency Rhizotomy Radiofrequency (RF) rhizotomy is another percutaneous, destructive procedure for the treatment of face pain. A brief overview of the technique follows.

Technique The procedure is performed under local anesthesia. Needle placement into the foramen ovale is similar to that described with glycerol rhizotomy. As with the glycerol procedure, entry into the foramen is conﬁrmed by spontaneous CSF ﬂow when the needle stylet is removed. Either a straight or curved electrode is then inserted through the needle, with the curved electrode offering the ability to selectively lesion a particular root and thereby limit side effects. V2 is located at the clivus, with V3 proximal to it and V1 distal. Low-intensity stimulation is then used to ensure placement at the appropriate root, and then the permanent lesion is made by heating the electrode to 65°C for 60 seconds. The patient is then examined for facial sensory loss, as with the glycerol procedure, with additional lesioning performed if the sensory deﬁcit is not extensive enough or if pain persists [13].

Results Published series indicate consistent results from RF rhizotomy, with the suggestion of a lower recurrence rate and longer duration of beneﬁt than glycerol rhizotomy. Sweet [14] described his technique and results in 214 patients with trigeminal neuralgia, reporting a rate of immediate pain relief of 91%. In 125 patients who were followed for between 2.5 and 6 years, the recurrence rate was 22% and tended to increase with time. Notably, 28 patients experienced an anesthetic cornea, with one ultimately losing sight due to corneal scarring. Broggi et al. [15] reported a large series of 1000 patients who underwent RF rhizotomy, with a follow-up period of 9.3 years. He obtained pain relief in 95% of patients, with masseter weakness occurring in 10.5% and a recurrence of pain in 18.1%. Similar to the glycerol procedure, he noted a clear correlation between the production of a sensory deﬁcit and chance of cure.

Balloon Compression Balloon compression of the trigeminal ganglion is a third percutaneous option for trigeminal neuralgia patients. It, too, like glycerol and radiofrequency rhizotomy, relies on partial destruction of nervous tissue to accomplish its goal. The original description was provided by Mullan and Lichtor [19], who carried out the procedure on 50 patients and followed them for between 0.5 and 4.5 years, noting a 12% recurrence rate during the follow-up period and an anticipated 20% recurrence rate at 5 years.

Technique Balloon compression is commonly performed under general anesthesia with an external pacemaker attached before beginning the procedure in the event of bradycardia occurring. In about two thirds of patients, a depressor response (bradycardia and brief hypotension) occurs during compression of the ganglion. The procedure begins in the same fashion as with the glycerol and RF procedures, with the needle being advanced into the foramen ovale under ﬂuoroscopic guidance. Direction of the needle in the medial-lateral direction can select one of the trigeminal divisions for treatment. The balloon is then inserted into the needle and inﬂated. When the balloon is properly positioned within the porus, it inﬂates in a characteristic “pear” shape. The balloon is insufﬂated with dye until an intraluminal pressure of 1100 to 1500 mm Hg is obtained, causing the surrounding tissue to be compressed with a pressure of 650 to 950 mm Hg. The balloon remains inﬂated for 1 to 1.5 minutes and is then deﬂated and withdrawn [20].

Results Skirving and Dan [21] performed 531 balloon compression procedures in 496 patients and followed the patients for an average of 10.7 years. Prompt pain relief occurred in all patients except for one. The recurrence rate was 19.2% at 5 years and 31.9%

53.

trigeminal neuralgia: surgical perspective

over the entire study period. Complications included 3.8% with symptomatic dysesthesias, 3.4% with symptomatic masseter weakness, and 1.6% with diplopia. Brown et al. [22] reported on 50 patients who were followed for 3 years. Ninety-four percent obtained initial relief and 86% reported that they were satisﬁed or very satisﬁed. Pain recurred in 26% with a mean time to recurrence of 1.5 years, and 62% of those with a recurrence received another compression procedure. Half of those receiving a repeat compression then had another recurrence. Mild numbness persisted in 74% of patients. Brown and Pilitsis [23] then performed another study to evaluate the relationship between compression pressure and morbidity rates. They performed 65 procedures in 56 patients with a mean follow-up of 17 months. The mean compression pressure was limited to 1160 mm Hg with a compression duration of 1.15 min, which is on the shorter end of the reported range of pressure. The successful pain relief rate of their earlier study was replicated, with initial relief in 92%. Morbidity rates, though, were lower than those typically reported in the literature. Immediate numbness was found in 83%, but only 17% reported persistent mild numbness at their most recent evaluation. Additionally, corneal function was preserved in all patients. Twenty-four percent of patients had masseter weakness, but it resolved in all patients within 1 year. The recurrence rate was 16% at 13 months. The authors concluded that morbidity rates can be lowered with compromising pain relief by monitoring the balloon pressure.

Microvascular Decompression Microvascular decompression (MVD) of the ﬁfth cranial nerve has been shown to be the most effective and durable treatment method for trigeminal neuralgia. The senior author (P.J.J.) has performed more than 7000 such operations in his career and recently offered some helpful hints regarding operative technique [24]. A description of the technical aspects of the procedure follows.

Technique Surgery is performed in the lateral position with the head placed in a head holder and with the neck mildly ﬂexed and rotated toward the affected side and the vertex tilted down. A short arcuate incision is made in the retromastoid region with the concave side facing the ear, centered over the region of the burr hole. The burr hole is made just posterior and caudal to the junction of the transverse and sigmoid sinuses. The position of the transverse sinus can be estimated by drawing an imaginary line from the inion to the external auditory canal. The transverse sinus will lie under this line. The digastric groove is a landmark for the sigmoid sinus. A craniectomy the size of a quarter is then made using a perforator drill and bone rongeurs. It is vital that the junction of the transverse and sigmoid sinuses is exposed by visualizing the bluish edges of the sinuses through the dura. The dura is then opened in a T-shaped fashion, with the base of the two resulting ﬂaps being on the sigmoid sinus, and the ﬂaps are retracted out of the way with a suture. Next, a rubber dam and cottonoid are carefully advanced over the cerebellar surface and a self-retaining malleable retractor is used to begin to gently elevate the cerebellum. Any bridg-

529

FIGURE 53-1. Intraoperative photographs demonstrating arterial (top left), venous (top right), and both arterial and venous (bottom right) compression of the trigeminal nerve.

ing veins between the cerebellar surface and the dura are coagulated and divided. As the cerebellum is elevated and the cisterns entered, CSF begins to drain, which relaxes the cerebellum and provides for greater exposure. If necessary, the more caudal cisterns can be entered to drain more CSF. Exposure continues until the trigeminal nerve is found, rostral and deep to the seventh and eighth nerve complex. The exposure, as well as the decompression, is done under constant monitoring of the brain-stem auditory-evoked response to monitor for any compromise to the patient’s hearing. Once the nerve is exposed, it is then examined from the brain-stem surface out to the Meckel cave in order to identify any and all offending vessels (Fig. 53-1). Typically, isolated V3 pain is caused by rostral compression (usually from the superior cerebellar artery [SCA]), isolated V2 pain is caused by a vessel distal on the nerve (commonly a vein), and isolated V1 pain results from caudal compression of the nerve. Teﬂon felt is then placed between the vessel and the nerve, or small veins can be coagulated and divided, keeping in mind that most early recurrences are probably the result of these veins recollateralizing. Prior to beginning the closure, the Valsalva maneuver is performed to check for venous oozing. If present, venous oozing responds well to 1 mL of hydrogen peroxide followed by saline irrigation. The dura is then closed in a watertight fashion with running suture. A titanium mesh cranioplasty is performed, which the senior author has found to be useful in preventing chronic headache. The fascia is closed tightly and the rest of the wound is closed in the usual manner. Postoperatively, patients are observed overnight in the neurosurgical intensive care unit and are typically discharged from the hospital within 2 to 3 days. More recently, we have begun allowing some patients to go directly to the regular nursing ﬂoor after an extended observation period in the postanesthesia unit [24, 25].

Results Many authors have published their results with large numbers of patients, including the senior author. Apfelbaum [26] followed 273 patients after MVD for an average of 55 months and

530

d.b. cohen et al.

found an initial rate of complete pain relief of 94.1%, with an additional 4% noting a reduction in their pain. At follow-up, 70.4% of patients had either complete pain relief or only a rare twinge of pain. Similarly, Klun [27] followed 178 patients for 5.2 years and found immediate pain relief in 96%, with eventual recurrence in 6%. Klun did note an association of better outcomes with arterial, as opposed to venous, compression, as did Burchiel et al. and Kolluri and Heros [28, 29]. Burchiel et al. [28] had a 47% recurrence rate in 36 patients followed for 8.5 years, with better results obtained in patients who had only arterial compression, and Kolluri and Heros [29] also noted an increased recurrence rate with venous compression. Burchiel et al. found a rate of recurrence of 3.5% annually for major recurrences and 1.5% annually for minor recurrences. Olson et al. [30] found similar results in his series of 156 MVDs over 25 years, with his patients having a 93% probability of becoming pain-free and a 94% chance of becoming medication-free. The overall recurrence rate was 18% over the 25-year period, with most recurrences within the ﬁrst 2 years after surgery. After that, the recurrence rate was 2% to 3.5% per year. Interestingly, 30% of what were classiﬁed as “recurrences” in this series were on the opposite side. Bederson and Wilson [31] found recurrence in 12% at an average of 1.9 years, with a recurrence rate of only 2% per year after that. In the senior author’s largest published series [32], 1155 patients had at least 1 year of follow-up after MVD, with an overall median follow-up period of 6.2 years. Initially, 82% of patients obtained complete relief of their pain (excellent result), and 16% had partial relief (good result), overall yielding 98% of patients with a successful result. At 1 year, 75.2% of patients still had complete relief and 8.9% partial relief, and at 10 years 70% were pain-free and another 4% had only occasional pain not requiring long-term medications. In all, 30% of patients had a recurrence, with most of the recurrences in the early postoperative period. At 10 years, the annual rate of recurrence was <1%. Complication rates were low, with death in 0.2%, brainstem infarction in 0.1%, and hearing loss in 1%. Recurrence of pain after MVD, then, occurs at a relatively low rate after the initial 1- to 2-year period after surgery. Theodosopoulos et al. [33] analyzed their data on 420 patients who underwent either an MVD or partial sensory rhizotomy (PSR) in order to determine risk factors for recurrence and to develop a predictive model. They obtained similar operative results to the aforementioned series, with 72% of patient having no pain and 93% noting a signiﬁcant improvement in their pain at a mean follow-up of 56.3 months. They estimated a recurrence rate of 34% at 8 years and found the following factors to be predictive of recurrence: age <53 at the time of surgery, preoperative symptom duration of >11.5 years, female sex, and leftsided pain in men. After putting these factors into a statistical model and separating patients into various risk categories, the authors found that the chance of remaining pain-free at 4 years was 89% for the low-risk group, 80% for the moderate-risk group, and 58% for the high-risk group. Surgeon experience has also been shown to have an impact on operative success and complication rates. Klun [27] observed that the rate of negative explorations (posterior fossa craniotomies during which no offending vessel was found) decreased from 14% to 3% over the course of his series. Kondo [34] divided his overall series of MVDs into two groups: 127 opera-

tions between 1976 and 1986, and 154 operations from 1987 to 1991. His postoperative cure rate increased from 92.9% to 96.7% and patient satisfaction rate also increased from 80.3% to 82.5% between the two time periods. Likewise, the rate of recurrence dropped from 10.2% to 6.5% and rate of hearing dysfunction decreased from 7.1% to 4.5%. Similarly, McLaughlin et al. [24] evaluated rates of complications for MVDs performed by the senior author for several indications (trigeminal neuralgia, hemifacial spasm, etc.) and broke the overall series of 4415 into two groups: 2420 cases before 1990, and 1995 cases after 1990. Prior to 1990, the rates of cerebellar injury, hearing loss, and CSF leak were 0.87%, 1.98%, and 2.44%, respectively. After 1990, the corresponding rates were 0.45%, 0.8%, and 1.85%. Finally, the nationwide inpatient sample (NIS) database was used in one study [35] to examine the relationship between outcome and volume of procedures performed. Among 1326 patients undergoing MVD for trigeminal neuralgia, outcomes at discharge were determined to be better in those patients operated on by higher-volume surgeons (and also at highervolume hospitals).

Overview of Treatment Options Clearly, no single procedure is appropriate for all patients with face pain, and the choice of which procedure to recommend to a particular patient can be difﬁcult at times. The recent emergence of stereotactic radiosurgery (SRS), ﬁrst described by Leksell [36] as another treatment possibility has served to increase this complexity. Although radiosurgery is reviewed in detail in another chapter, a short summary of recent ﬁndings seems appropriate here. In the mid-1990s, the report of a multi-institutional trial was published by Kondziolka et al. [37]. This preliminary study enrolled 50 patients at 5 centers treated with the Gamma Knife, with a wide dose range of between 60 and 90 Gy. At 2 years, 54% of patients were pain-free and 88% had 50% to 100% relief. This study established that a maximum dose of at least 70 Gy was associated with a signiﬁcantly higher rate of complete pain relief, a dose that virtually all subsequent studies have adhered to as a minimum dose. This study was followed by many others by the same group and others [38–47], obtaining similar results for the Gamma Knife. The speciﬁc outcome measures used vary from study to study, making comparison between studies or between SRS and other modalities difﬁcult at times. Also, many of the SRS studies have considered patients to have a good outcome even if they continue to take regular medications, which would put them into the failure category in most surgical series. Rates of complete or partial pain relief combined tend to range between 45% and 89% at follow-up intervals of 1 to 5 years. The latency period from treatment until pain relief occurs tends to be around 1 month. As with the percutaneous destructive procedures, many studies of SRS have identiﬁed new facial numbness after the procedure as an important predictive factor for pain relief [39, 41, 44, 46]. Rates of other complications with SRS appear to be low. More recently, other radiosurgery devices have been reported to provide similar results to the Gamma Knife when treating trigeminal neuralgia patients, including linac systems [48–50] and CyberKnife [51]. Linac has been demonstrated to have equiva-

53.

trigeminal neuralgia: surgical perspective

lent dose fall-off properties to the Gamma Knife if more than 7 arcs are used and intra-arc errors are small (<0.5 mm) [52]. Lopez and colleagues [53] thoroughly reviewed the studies of SRS using strict methodologic inclusion criteria. Out of 38 studies identiﬁed, overall data quality was poor, with only 4 studies able to be used to evaluate pain relief on a yearly basis, 2 to determine actuarial rates of pain relief, 7 to determine latency of pain relief, and 18 to determine complication rates. Overall, the studies tend to show that complete pain relief occurs in about 75% of patients initially but drops to about half this number at 3 years. At most, 50% of patients permanently stop taking their medications. Recurrence was observed in 21% at a median 6.7 months and 36% to 40% at 2 years, if recurrence is deﬁned as any decrease from maximum pain relief. How, then, to decide what option is appropriate for each patient? The literature does offer some assistance in the form of head-to-head comparisons between the various treatments. Tronnier et al. [54] compared the results of 206 patients undergoing RF rhizotomy with 225 patients after MVD. Strikingly, pain recurred at 2 years in 50% of the RF patients, whereas at 20 years 64% of MVD patients were still pain-free. Additionally, patients who did not have a sensory deﬁcit after MVD remained pain-free signiﬁcantly longer than those who had such a deﬁcit, refuting the hypothesis that trauma to the trigeminal nerve during MVD plays a role in pain relief. Barker et al. [55] and Pollock [56] also provided evidence along this line, demonstrating that unlike the percutaneous procedures, numbness after MVD is not predictive of pain relief. Erbay et al. [57], in a study of preoperative MRI scans of patients about to have Gamma Knife treatment, found that neurovascular contact on the preprocedure MRI predicts response to SRS. All this evidence substantiates the theory of vascular compression as the cause of idiopathic trigeminal neuralgia and supports decompression as the optimal treatment, instead of a procedure that traumatizes the nerve. A report comparing glycerol rhizotomy to SRS with the Gamma Knife supported glycerol rhizotomy as a better treatment for immediate pain relief, due to the latency of radiation treatment, but otherwise the Gamma Knife provided superior outcomes [58]. Pollock [56] retrospectively compared 49 patients after MVD with 28 patients after SRS with a mean follow-up

531

of 25.5 months and found MVD patients to have signiﬁcantly higher rates of being pain-free without medications at both 1 and 3 years (75% vs. 59% at 1 year and 72% vs. 59% at 3 years). Deciding which option to recommend for each patient depends on multiple factors (Table 53-1) [59–61]. The percutaneous procedures are highly efﬁcacious for short-term relief but are not durable and rely on partial destruction of the nerve to achieve their intended effect. Lopez and co-workers [62] systematically reviewed the published data for the percutaneous procedures. Better data is needed to decide between the various percutaneous options, but the available data seems to indicate that RF rhizotomy yields higher rates of complete pain relief compared with glycerol rhizotomy and SRS but also is associated with the highest rate of complications. The therapeutic effect of glycerol decreases rapidly after 24 months and is the least beneﬁcial of the techniques at 36 months, and SRS may have the lowest complication rate of these options, according to the authors. More high-quality data was deemed to be needed with regard to balloon compression [62]. There is scant evidence in the literature with regard to the cost-effectiveness of the various treatments, but one recent study did examine this issue [1]. The authors determined that MVD provided greatly superior outcomes to glycerol rhizotomy and SRS, but that glycerol rhizotomy was more cost-effective than the other two treatments. For older patients with signiﬁcant medical comorbidities who are unlikely to need many additional treatments over time, glycerol rhizotomy may be a more cost-effective treatment. For patients who are unable to tolerate a posterior fossa operation due to medical comorbidities or who refuse surgery, though, these procedures are an important option. MVD, on the other hand, has been shown through published experience of thousands of patients over the past several decades to be an extremely effective, durable treatment with a low rate of side effects, especially when performed by highly trained personnel. MVD has been shown to be safe in the elderly, with no difference reported in the postoperative length of stay or in long-term recurrence rates between older and younger patients [63]. Patient satisfaction is high with MVD as well, with one study reporting a satisfaction rate of 89% after the procedure and 80% of patients judging the ﬁnal outcome to be better than they expected [64].

TABLE 53-1. Relative advantages and disadvantages of the percutaneous methods, stereotactic radiosurgery, and microvascular decompression. Method

Percutaneous techniques

Stereotactic radiosurgery

Microvascular decompression

Advantages

• • • • • • • • •

Safe Possible in patients in poor medical condition Same-day discharge Repeated easily when necessary Safe in inﬁrm patients Noninvasive Same-day discharge Easily repeated Low procedural risks

• • • •

Treats cause of pain Not destructive to nerve Low recurrence rate Low complication rates in experienced hands, even in elderly

Disadvantages

• Do not treat cause of pain • Destructive; rely on alteration of facial sensation for pain relief • Efﬁcacious only in short-term • Destructive; new facial numbness correlates with pain relief • Latency period for pain relief unacceptable for patients with excruciating pain • Lack of long-term follow-up • Efﬁcacious only in short-term • Risks of craniectomy • Longer hospital stay

532

d.b. cohen et al.

Conclusion The traditional percutaneous procedures (glycerol rhizotomy, RF rhizotomy, and balloon compression) may be best suited for patients who cannot or refuse to undergo posterior fossa surgery. Although effective initially, these treatments are not durable in the great majority of patients. SRS, although newer than the other percutaneous techniques, has now accumulated enough data to show that it, too, may be most useful for the same patient population. SRS can also be used in patients with coagulopathy. The advantage of SRS is its lack of invasiveness, which may make it more popular in some patients’ and clinicians’ eyes. MVD remains the deﬁnitive procedure for longterm relief with a low rate of recurrence and should be offered to patients in good health by an experienced surgeon.

References 1. Pollock BE, Ecker RD. A prospective cost-effectiveness study of trigeminal neuralgia surgery. Clin J Pain 2005; 21:317–322. 2. Jannetta PJ. Arterial compression of the trigeminal nerve at the pons in patients with trigeminal neuralgia. J Neurosurg 1967; 26:159–162. 3. Jannetta PJ. Observations on the etiology of trigeminal neuralgia, hemifacial spasm, acoustic nerve dysfunction and glossopharyngeal neuralgia. Deﬁnitive microsurgical treatment and results in 117 patients. Neurochirurgia (Stuttg) 1977; 20:145–154. 4. Jannetta PJ. Microsurgery of cranial nerve cross-compression. Clin Neurosurg 1979; 26:607–615. 5. Jannetta PJ. Neurovascular compression in cranial nerve and systemic disease. Ann Surg 1980; 192:518–525. 6. Håkanson S. Trigeminal neuralgia treated by the injection of glycerol into the trigeminal cistern. Neurosurgery 1981; 9:638–646. 7. Young RF. Percutaneous trigeminal glycerol rhizotomy. In: Rengachary SS, Wilkins RH, eds. Neurosurgical Operative Atlas, Volume 1. Chicago: American Association of Neurological Surgeons, 1991:117–123. 8. Lunsford LD, Bennett MH. Percutaneous retrogasserian glycerol rhizotomy for tic douloureux: part 1; technique and results in 112 patients. Neurosurgery 1984; 14:424–430. 9. Fujimaki T, Fukushima T, Miyazaki S. Percutaneous retrogasserian glycerol injection in the management of trigeminal neuralgia: long-term follow-up results. J Neurosurg 1990; 73:212–216. 10. North RB, Kidd DH, Piantadosi S, et al. Percutaneous retrogasserian glycerol rhizotomy: predictors of success and failure in treatment of trigeminal neuralgia. J Neurosurg 1990; 72:851–856. 11. Slettebø H, Hirschberg H, Lindegaard K-F. Long-term results after percutaneous retrogasserian glycerol rhizotomy in patients with trigeminal neuralgia. Acta Neurochir (Wien) 1993; 122:231– 235. 12. Burchiel KJ. Percutaneous retrogasserian glycerol rhizolysis in the management of trigeminal neuralgia. J Neurosurg 1988; 69:361– 366. 13. Fick J, Tew JM Jr. Percutaneous radiofrequency rhizolysis for trigeminal neuralgia. In: Rengachary SS, Wilkins RH, eds. Neurosurgical Operative Atlas, Volume 1. Chicago: American Association of Neurological Surgeons, 1991:379–390. 14. Sweet WH, Wepsic JG. Controlled thermocoagulation of trigeminal ganglion and rootlets for differential destruction of pain ﬁbers. Part I: trigeminal neuralgia. J Neurosurg 1974; 40:143–156. 15. 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; 26:783–787.

16. Yoon KB, Wiles JR, Miles JB, et al. Long-term outcome of percutaneous thermocoagulation for trigeminal neuralgia. Anaesthesia 1999; 54:803–808. 17. Taha JM, Tew JM, Buncher CR. A prospective 15-year follow up of 154 consecutive patients with trigeminal neuralgia treated by percutaneous stereotactic radiofrequency thermal rhizotomy. J Neurosurg 1995; 83:989–993. 18. Ischia S, Luzzani A, Polati E, et al. Percutaneous controlled thermocoagulation in the treatment of trigeminal neuralgia. Clin J Pain 1990; 9:96–104. 19. Mullan S, Lichtor T. Percutaneous microcompression of the trigeminal ganglion for trigeminal neuralgia. J Neurosurg 1983; 59:1007–1012. 20. Brown JA, Gouda JJ. Percutaneous balloon compression for the treatment of trigeminal neuralgia. In: Rengachary SS, Wilkins RH, eds. Neurosurgical Operative Atlas, Volume 7. Chicago: American Association of Neurological Surgeons, 1998:107–116. 21. Skirving DJ, Dan NG. A 20-year review of percutaneous balloon compression of the trigeminal ganglion. J Neurosurg 2001; 94:913–917. 22. Brown JA, McDaniel MD, Weaver MT. Percutaneous trigeminal nerve compression for treatment of trigeminal neuralgia: results in 50 patients. Neurosurgery 1993; 32:570–573. 23. Brown JA, Pilitsis JG. Percutaneous balloon compression for the treatment of trigeminal neuralgia: results in 56 patients based on balloon compression pressure monitoring. Neurosurg Focus 2005; 18:E10. 24. McLaughlin MR, Jannetta PJ, Clyde BL, et al. Microvascular decompression of cranial nerves: lessons learned after 4400 operations. J Neurosurg 1999; 90:1–8. 25. Jannetta PJ, McLaughlin MR, Casey KF. Technique of microvascular decompression: technical note. Neurosurg Focus 2005; 18:E5. 26. Apfelbaum RI. Surgery for tic douloureux. Clin Neurosurg 1983; 31:351–68. 27. Klun B. Microvascular decompression and partial sensory rhizotomy in the treatment of trigeminal neuralgia: personal experience with 220 patients. Neurosurgery 1992; 30:49–52. 28. Burchiel KJ, Clarke H, Haglund M, et al. Long-term efﬁcacy of microvascular decompression in trigeminal neuralgia. J Neurosurg 1988; 69:35–38. 29. Kolluri S, Heros RC. Microvascular decompression for trigeminal neuralgia: a ﬁve-year follow-up study. Surg Neurol 1984; 22:235– 240. 30. Olson S, Atkinson L, Weidmann M. Microvascular decompression for trigeminal neuralgia: recurrences and complications. J Clin Neurosci 2005; 12:787–789. 31. Bederson JB, Wilson CB. Evaluation of microvascular decompression and partial sensory rhizotomy in 252 cases of trigeminal neuralgia. J Neurosurg 1989; 71:359–367. 32. Barker FG, Jannetta PJ, Bissonette DJ, et al. The long-term outcome of microvascular decompression for trigeminal neuralgia. N Engl J Med 1996; 334:1077–1083. 33. Theodosopoulos PV, Marco E, Applebury C, et al. Predictive model for pain recurrence after posterior fossa surgery for trigeminal neuralgia. Arch Neurol 2002; 59:1297–1302. 34. Kondo A. Microvascular decompression surgery for trigeminal neuralgia. Stereotact Funct Neurosurg 2001; 77:187–189. 35. Kalkanis SN, Eskandar EN, Carter BS, et al. Microvascular decompression surgery in the United States, 1996 to 2000: mortality rates, morbidity rates, and the effects of hospital and surgeon volumes. Neurosurgery 2003; 52:1251–1262. 36. Leksell L. Stereotaxic radiosurgery in trigeminal neuralgia. Acta Chir Scand 1971; 137:311–314. 37. Kondziolka D, Lunsford LD, Flickinger JC, et al. Stereotactic radiosurgery for trigeminal neuralgia: a multi-institutional study using the gamma unit. J Neurosurg 1996; 84:940–945.

53.

trigeminal neuralgia: surgical perspective

38. Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for the treatment of trigeminal neuralgia. Clin J Pain 2002; 18:42–47. 39. Rogers CL, Shetter AG, Fiedler JA, et al. Gamma Knife radiosurgery for trigeminal neuralgia: the initial experience of the Barrow Neurological Institute. Int J Radiat Oncol Biol Phys 2000; 47:1013–1019. 40. Shetter AG, Rogers CL, Ponce F, et al. Gamma Knife radiosurgery for recurrent trigeminal neuralgia. J Neurosurg 2002; 97(Suppl 5):536–538. 41. Pollock BE, Phuong LK, Gorman DA, et al. Stereotactic radiosurgery for idiopathic trigeminal neuralgia. J Neurosurg 2002; 97:347–353. 42. Urgošík D, Vymazal J, Vladyka V, et al. Gamma Knife treatment of trigeminal neuralgia: clinical and electrophysiological study. Stereotact Funct Neurosurg 1998; 70(Suppl 1):200–209. 43. Young RF, Vermuelen S, Posewitz A. Gamma Knife radiosurgery for the treatment of trigeminal neuralgia. Stereotact Funct Neurosurg 1998; 70(Suppl 1):192–199. 44. McNatt SA, Yu C, Giannotta SL, et al. Gamma Knife radiosurgery for trigeminal neuralgia. Neurosurgery 2005; 56:1295– 1303. 45. Sheehan J, Pan H-C, Stroila M, et al. Gamma Knife surgery for trigeminal neuralgia: outcomes and prognostic factors. J Neurosurg 2005; 102:434–441. 46. Tawk RG, Duffy-Fronckowiak M, Scott BE, et al. Stereotactic Gamma Knife surgery for trigeminal neuralgia: detailed analysis of treatment response. J Neurosurg 2005; 102:442–449. 47. Drzymala RE, Malyapa RS, Dowling JL, et al. Gamma Knife radiosurgery for trigeminal neuralgia: the Washington University initial experience. Stereotact Funct Neurosurg 2005; 83:148–152. 48. Chen JCT, Girvigian M, Greathouse H, et al. Treatment of trigeminal neuralgia with linear accelerator radiosurgery: initial results. J Neurosurg 2004; 101(Suppl 3):346–350. 49. Richards GM, Bradley KA, Tomé WA, et al. Linear accelerator radiosurgery for trigeminal neuralgia. Neurosurgery 2005; 57: 1193–1200. 50. Kubicek GJ, Hall WA, Orner JB, et al. Long-term follow-up of trigeminal neuralgia treatment using a linear accelerator. Stereotact Funct Neurosurg 2004; 82:244–249. 51. Lim M, Villavicencio AT, Burneikiene S, et al. CyberKnife radiosurgery for idiopathic trigeminal neuralgia. Neurosurg Focus 2005; 18:E9.

533

52. Ma L, Kwok Y, Chin LS, et al. Comparative analyses of linac and Gamma Knife radiosurgery for trigeminal neuralgia treatments. Phys Med Biol 2005; 50:5217–5227. 53. Lopez BC, Hamlyn PJ, Zakrzewska JM. Stereotactic radiosurgery for primary trigeminal neuralgia: state of the evidence and recommendations for future reports. J Neurol Neurosurg Psychiatry 2004; 75:1019–1024. 54. Tronnier VM, Rasche D, Hamer J, et al. Treatment of idiopathic trigeminal neuralgia: comparison of long-term outcome after radiofrequency rhizotomy and microvascular decompression. Neurosurgery 2001; 48:1261–1268. 55. Barker FG II, Jannetta PJ, Bissonette DJ, et al. Trigeminal numbness and tic relief after microvascular decompression for typical trigeminal neuralgia. Neurosurgery 1997; 40:39–45. 56. Pollock BE. Comparison of posterior fossa exploration and stereotactic radiosurgery in patients with previously nonsurgically treated idiopathic trigeminal neuralgia. Neurosurg Focus 2005; 18: E6. 57. Erbay SH, Bhadelia RA, Riesenburger R, et al. Association between neurovascular contact on MRI and response to Gamma Knife radiosurgery in trigeminal neuralgia. Neuroradiology 2006; 48:26–30. 58. Henson CF, Goldman HW, Rosenwasser RH, et al. Glycerol rhizotomy versus Gamma Knife radiosurgery for the treatment of trigeminal neuralgia: an analysis of patients treated at one institution. Int J Radiat Oncol Biol Phys 2005; 63:82–90. 59. Elias WJ, Burchiel KJ. Trigeminal neuralgia and other neuropathic pain syndromes of the head and face. Curr Pain Headache Rep 2002; 6:115–124. 60. Elias WJ, Burchiel KJ. Microvascular decompression. Clin J Pain 2002; 18:35–41. 61. Liu JK, Apfelbaum RI. Treatment of trigeminal neuralgia. Neurosurg Clin N Am 2004; 15:319–334. 62. Lopez BC, Hamlyn PJ, Zakrzewska JM. Systematic review of ablative neurosurgical techniques for the treatment of trigeminal neuralgia. Neurosurgery 2004; 54:973–983. 63. Ashkan K, Marsh H. Microvascular decompression for trigeminal neuralgia in the elderly: a review of the safety and efﬁcacy. Neurosurgery 2004; 55:840–850. 64. Zakrzewska JM, Lopez BC, Kim SE, et al. Patient reports of satisfaction after microvascular decompression and partial sensory rhizotomy for trigeminal neuralgia. Neurosurgery 2005; 56:1304– 1312.

5 4

Trigeminal Neuralgia: Medical Management Perspective Neil C. Porter

Introduction

Neuropathic Pain

Trigeminal neuralgia is a well-recognized neuropathic pain syndrome characterized by brief paroxysms of facial pain in the distribution of the trigeminal nerve. The syndrome can occur secondary to a number of medical conditions including multiple sclerosis, basilar artery aneurysm, and brain tumor [1]. The disorder can also be idiopathic, the latter instance being known as tic douloureux. Trigeminal neuralgia can strike persons of any age but typically affects older individuals. Treatment options include a number of neuropathic pain medications as well as neurosurgical and radiosurgical interventions. The mainstay of treatment for most individuals, however, remains medication.

Trigeminal neuralgia is a condition characterized by neuropathic pain. Other notable conditions that also involve neuropathic pain include painful diabetic neuropathy, post-herpetic neuralgia, and complex regional pain syndrome. In contrast with nociceptive pain, whereby the nervous system detects some unpleasant stimulus alerting the body to potential damage, in neuropathic pain the nervous system generates the impulses that are appreciated as pain by the patient. Neuropathic pain can be constant or intermittent in nature. It may be characterized as burning, aching, stabbing, or shooting. The pain may occur spontaneously, or, as with trigeminal neuralgia, be elicited by tactile stimulation.

Clinical Features Trigeminal neuralgia is characterized by recurrent facial pain in the distribution of the trigeminal nerve. The trigeminal nerve is divided into three branches, V1, V2, and V3, providing sensory innervation to the forehead, cheek, and chin respectively. The pain of trigeminal neuralgia most frequently occurs in the distribution of V2 or V3. The pain is usually lancinating in nature, occurring as ﬂeeting shocks. This pain can be extremely intense, producing a great deal of distress in sufferers. The pain is typically triggered by chewing, ingesting cold foods or liquids, cold winds, or simple tactile stimulation on the face.

Pathophysiology The pathophysiology and pathogenesis of trigeminal neuralgia are poorly understood. A number of theories abound, however [2, 3]. Given that the condition occurs in multiple sclerosis, some investigators have argued that demyelination at the entry zone of the nerve root is a common feature to both primary and secondary trigeminal neuralgia [4]. Others have focused on structural concerns with the “vascular compression theory” [5]. In this latter circumstance, the pain is believed to be the consequence of compression on the trigeminal nerve by some vascular anomaly. In any event, the major focus in treating the patient is the elimination of the subjective pain.

Neuropathic Pain Agents: Overview Medications used to treat neuropathic pain can be conveniently organized into a number of classes including antiepileptic drugs, gamma-aminobutyric acid (GABA)-related medications, antidepressants, and local anesthetics or topical agents. Although not traditionally viewed as neuropathic pain agents, narcotics are also effective in treating neuropathic pain. Particular beneﬁts and drawbacks are associated with each class of agent as well as speciﬁc agents within each class. Most of the agents are well tolerated, but cognitive side effects may limit the utility of many of the antiepileptic drugs and antidepressants in certain cases. For trigeminal neuralgia, the antiepileptic medications dominate the treatment strategy followed by the GABA agonists. Antidepressants have little role, and narcotics provide only adjunctive therapy.

Antiepileptic Drugs Carbamazepine Carbamazepine is clearly the drug of choice for trigeminal neuralgia. Known mostly for treatment of partial and secondarily generalized seizures, carbamazepine is also a well-known neuropathic pain agent. Structurally similar to the tricyclic

535

536

n.c. porter

antidepressants, carbamazepine’s effectiveness for seizures and neuropathic pain is believed to occur through its action on sodium channels. Carbamazepine acts to block sodium channel function in a rate-related fashion. This allows the medication to affect dysfunctioning neurons greater than those ﬁring normally. Carbamazepine was ﬁrst reported to be effective in treating trigeminal neuralgia in 1962 [6]. This agent was proved efﬁcacious in a number of randomized placebo-controlled clinical trials [7–9]. In each trial, initial response rates were as high as 70% to 90%. Long-term efﬁcacy, however, diminished to approximately 50%. Carbamazepine has a relatively short half-life, necessitating a dosing schedule of 3 to 4 times per day. Because it autoinduces its own metabolism, a patient cannot be loaded with carbamazepine. Instead, carbamazepine must be started at a relatively low dose and then slowly titrated upwards. The patient should be started at 100 mg twice daily (b.i.d.) or three times daily (t.i.d.). The dose can be increased every 3 to 5 days by 200 mg total up to 200 to 400 mg t.i.d. Some refractory patients may require doses as high as 2400 mg/day (800 mg t.i.d.). Carbamazepine’s major advantage is its efﬁcacy in most patients. The medication is relatively well tolerated with only minor side effects in the majority of patients. Additionally, carbamazepine is relatively inexpensive, being available in generic form. Unfortunately, carbamazepine is not without its shortcomings. Given its relatively short half-life, the medication requires relatively frequent dosing that may affect compliance. Carbamazepine is associated with a dose-related leukopenia but in rare circumstances has been linked to irreversible aplastic anemia (risk at 1 : 40,000). Additional side effects include dizziness and hyponatremia. Drug interactions are also problematic given that carbamazepine affects the levels of other medications that are metabolized by the P450 system. In summary, carbamazepine is the mainstay of medical treatment for trigeminal neuralgia, with proven efﬁcacy in randomized clinical trials. Efﬁcacy initially ranges from 70% to 90%, but response rates may diminish over time to 50%. The medication must be started cautiously and titrated relatively slowly.

Oxcarbazepine Oxcarbazepine is a newer antiepileptic drug structurally related to carbamazepine. Like its structural analogue, oxcarbazepine is a sodium channel blocker. Because of structural modiﬁcations, however, this newer drug has a number of theoretical and practical advantages over carbamazepine. Oxcarbazepine has much less effect on the P450 system and therefore is associated with far fewer drug interactions. Furthermore, oxcarbazepine cannot be metabolized to an epoxide derivative that is responsible for some of the side effects seen with carbamazepine. Obviously, though, given its lesser track record, oxcarbazepine is less well established than carbamazepine as a treatment for trigeminal neuralgia. Although oxcarbazepine has not been proved effective for trigeminal neuralgia in a randomized, placebo-controlled trial, it has been shown to have similar efﬁcacy as carbamazepine in two randomized trials. Response rates in these trials approached

90% [10, 11]. Furthermore, the drug appeared to be better tolerated than carbamazepine. Oxcarbazepine is typically started at 150 to 300 mg b.i.d., with the lower dosing associated with fewer side effects. Typical target dosages range from 300 to 600 mg b.i.d. with some refractory patients requiring up to 1200 mg b.i.d. Oxcarbazepine appears to possess several advantages over carbamazepine, including better tolerability and fewer drug interactions. One disadvantage, however, is cost. Being a newer agent, oxcarbazepine is signiﬁcantly more expensive than the older carbamazepine. Additionally, oxcarbazepine has been associated with a number of side effects similar to other sodium channel blockers, including dizziness, headache, gait instability, nausea, and vomiting. In summary, oxcarbazepine holds great promise to potentially replace carbamazepine as the ﬁrst-line treatment for trigeminal neuralgia. Although both medications have similar efﬁcacy, oxcarbazepine appears to be safer and better tolerated.

Gabapentin Gabapentin has assumed a major role as an antiepileptic drug used for neuropathic pain. A number of trials have proved its efﬁcacy in treating painful diabetic neuropathy and postherpetic neuralgia. Despite this, no clinical trials have proved gabapentin’s role in trigeminal neuralgia. Nonetheless, gabapentin has been advocated for trigeminal neuralgia as well as other disorders characterized by neuropathic pain. Gabapentin was developed as an antiepileptic drug with structural similarity to GABA. The medication, however, appears to have no action at the GABA receptor. Instead, gabapentin is believed to exert its effect through the α-2-δ calcium channel. Gabapentin has been proved efﬁcacious in randomized clinical trials for neuropathic pain syndromes such as painful diabetic neuropathy and post-herpetic neuralgia but not speciﬁcally for trigeminal neuralgia. The drug has been reported, however, to provide adequate pain relief for trigeminal neuralgia in small, uncontrolled series [12, 13]. Like many of the other neuropathic pain agents, gabapentin is best tolerated when started at a very low dose. Patients may be started at 300 mg daily with their dose being increased by 300 mg every 3 days up to 600 mg t.i.d. Studies have failed to show superiority of higher doses of gabapentin, but some individuals may require doses as high as 900 to 1200 mg t.i.d. The most attractive aspect of gabapentin is its apparent safety. Gabapentin is extremely well tolerated with no appreciable serious medical side effects and negligible drug interactions. Furthermore, overdosage has not been seen, given the lack of dose-limiting side effects. Given the large dosages required by patients for adequate pain relief, gabapentin can be quite costly even as a generic. Furthermore, problematic side effects include dizziness, lower extremity edema, weight gain, and headaches. Lastly, a withdrawal syndrome has been reported with rapid cessation of the medication. In short, gabapentin is a very popular drug for neuropathic pain because of its favorable side-effect proﬁle. Evidence to date supports its use in trigeminal neuralgia after trials of the better-established medications.

54.

trigeminal neuralgia: medical management perspective

Lamotrigine Lamotrigine is a newer antiepileptic drug that acts through a novel mechanism. It is believed to exert its effect through the stabilization of the inactive form of the sodium channel. Although lamotrigine has a relatively short track record with regard to neuropathic pain, this agent has already been shown to be efﬁcacious in treating patients with trigeminal neuralgia. In a number of uncontrolled series, lamotrigine conveyed adequate pain relief to a majority of patients [14, 15]. In one randomized, double blind, placebo-controlled crossover study, lamotrigine again performed well, being superior to placebo [16]. Lamotrigine must be started at a low dose and increased slowly to minimize the risk of a serious rash. The medication is typically started at 25 mg daily (q.d.) and increased by 25 mg every 3 to 7 days up to 200 mg q.d. The major side effect with lamotrigine is the aforementioned rash; otherwise, the medication is safe and well tolerated. In summary, lamotrigine is another promising agent whose use in trigeminal neuralgia should increase over time. The medication is safe when used correctly, apparently effective and well tolerated.

Topiramate Topiramate is a novel antiepileptic agent that is believed to work by blocking sodium channels, blocking certain glutamate receptors, and enhancing activity of GABA via a nonbenzodiazepine site of the GABA receptor. Although it has never been studied in randomized trials, a number of small series have demonstrated its efﬁcacy in trigeminal neuralgia [17, 18]. Topiramate is typically started at 25 mg b.i.d. and increased by 50 mg at weekly intervals up to 200 mg b.i.d. On the positive side, topiramate is safe and reasonably well tolerated, but it does have some notable drawbacks. As a new drug, it is relatively expensive. Additionally, topiramate appears to cause mental clouding, including word-ﬁnding difﬁculty. Lastly, it is associated with kidney stone formation. To summarize, topiramate may be a reasonable thirdline agent that is safe and reasonably well tolerated. Future studies are needed to speciﬁcally justify its use in trigeminal neuralgia.

537

Pregabalin appears to be safe and well tolerated, similar to gabapentin. Its greater potency, however, allows for lower dosages and therefore lower pill counts than its predecessor. Furthermore, cost is similar to that of gabapentin. Pregabalin is associated with similar side effects as gabapentin. Additionally, though, thrombocytopenia and arthralgias have been reported. In summary, pregabalin may be an important medication in the future for trigeminal neuralgia. Further studies may help support its use for this indication.

Phenytoin Although reported as the ﬁrst antiepileptic drug effective in trigeminal neuralgia, phenytoin’s role has greatly diminished in recent years. Like carbamazepine, phenytoin is a sodium channel blocker that works in a rate-related fashion. Phenytoin was ﬁrst reported effective in treating trigeminal neuralgia in 1942 [19]. Although a number of reports have been published since that time [20, 21], no clinical trials demonstrating efﬁcacy of phenytoin have surfaced. Nonetheless, phenytoin may be a reasonable choice for selected patients with trigeminal neuralgia. Phenytoin has a relatively long half-life, allowing for daily or twice daily dosing. Although twice-daily dosing is optimal, daily dosing may improve compliance. Dosing of the medication in a three-times-daily fashion is not necessary even though it is commonplace. Phenytoin can be loaded, but this may not be necessary in patients with trigeminal neuralgia. Typical dosing schedules consist of 100 to 200 mg b.i.d. with daily doses around 6 mg/kg per day. Phenytoin’s primary strong suit is its low cost. Its major disadvantages are its modest efﬁcacy and its accompanying cosmetic side effects. Phenytoin is believed to cause “coarsening” of facial features, gum hypertrophy, and hirsutism in women. Furthermore, like other older antiepileptic agents, long-term use of phenytoin has been associated with osteoporosis. In summary, phenytoin may be a reasonable option in selected patients in whom multiple other medications have failed or who cannot afford the newer agents and who do not wish to undergo invasive procedures. Cosmetic side effects and modest efﬁcacy markedly limit its role in trigeminal neuralgia.

Pregabalin Pregabalin is a compound designed to be similar to gabapentin. It has been proved efﬁcacious in treating painful diabetic neuropathy but is not speciﬁcally approved for trigeminal neuralgia. The mechanism of action is believed to be similar to that of gabapentin, namely via action at the α-2-δ calcium channel. Because of its relatively recent release, experience with pregabalin is more limited than with many of the other agents. Although no speciﬁc studies using pregabalin for trigeminal neuralgia have been performed, the medication’s affect on neuropathic pain in general makes it a reasonable consideration once the more established agents have been exhausted. The target dose of pregabalin is 300 mg daily divided b.i.d. The medication is typically started at 75 mg b.i.d. and increased up to 150 mg b.i.d. after approximately 1 week.

GABA-Related Agents Baclofen Baclofen, an agent used predominately to diminish spasticity in patients with disturbances of the central nervous system, has historically been used as a second-line agent behind carbamazepine in treating patients with trigeminal neuralgia. Given the advent of newer agents, however, the role of baclofen for trigeminal neuralgia has become less certain. Baclofen was ﬁrst reported as a treatment for trigeminal neuralgia in the 1980s [22]. Similar to a number of other agents used to treat trigeminal neuralgia, baclofen has never been proved efﬁcacious in a randomized clinical trial. The drug was shown to be effective in a double-blind crossover trial, though [23].

538

n.c. porter

Baclofen is typically dosed in a three-times-daily fashion, starting at 5 to 10 mg t.i.d. and increasing moderately slowly up to 20 mg t.i.d. Baclofen is safe and well tolerated, with dizziness, drowsiness, and gait instability being the predominant side effects. In summary, baclofen is a safe and reasonably effective medication for trigeminal neuralgia. Because of its safety proﬁle and long-term track record, baclofen should be considered early for the treatment of trigeminal neuralgia.

In summary, capsaicin appears to be a reasonable option in refractory cases of trigeminal neuralgia. Patients should be cautioned against exposure of the agent to sensitive areas. Additionally, patients should be forewarned of the potential for extreme burning pain in areas exposed to the medication.

Narcotic Agents Oral Agents

Clonazepam Clonazepam is a long-acting benzodiazepine that has been reported efﬁcacious at high doses in some patients with trigeminal neuralgia [24]. Like other benzodiazepines, clonazepam acts by binding to GABA receptors and potentiating the effects of that neurotransmitter. In past years, clonazepam was used as a third-line agent. With the availability of more effective neuropathic pain agents in recent years, however, its use has become negligible. In contrast with dosing for epilepsy and mood disorders, relatively high doses of clonazepam are required to treat trigeminal neuralgia. The major advantage of clonazepam is its modest cost and long-term safety. Although clonazepam is well tolerated at lower doses, the high doses required to treat trigeminal neuralgia are associated with more profound side effects such as sedation and cognitive impairment. Furthermore, dependence must be a consideration, although this concern is usually less prominent given the long half-life of the drug. In summary, clonazepam like phenytoin may have a role in extremely selected patients with trigeminal neuralgia averse to invasive procedures and refractory to the numerous neuropathic agents that are generally considered more effective.

Local Anesthetics Capsaicin Although it has not been widely used in trigeminal neuralgia, capsaicin has been shown to be efﬁcacious in some uncontrolled series. Capsaicin is a compound derived from chili peppers. The commercially available formulation consists of a cream containing the active substance as well as lidocaine, to buffer the burning caused by the medication. Capsaicin acts by depleting substance P, a neurotransmitter involved in pain pathways, from nerve terminals. At least two uncontrolled series have been published documenting the apparent efﬁcacy of capsaicin in trigeminal neuralgia. Adequate pain control was obtained in approximately 30% [25] of patients in one study and 60% in another [26]. Pain relief was sustained despite only transient use of the medication. Capsaicin should be applied to the affected area 2 to 4 times daily. Duration of use should be tailored to the patient. Caution must be used to avoid exposure to the eye. Capsaicin is extremely safe with no systemic side effects. Additionally, cost is highly reasonable. The major drawback to capsaicin, however, is tolerability. Prior to causing analgesia, the agent may cause severe, intolerable burning in the areas to which the medication is applied.

Oral narcotics of varying potency may ameliorate excruciating episodes of pain due to trigeminal neuralgia. Although narcotics are not typically considered neuropathic pain agents, they are nonetheless effective in treating this type of pain. Narcotics act at opioid receptors, providing relatively nonspeciﬁc pain reduction. Short-acting agents may be useful for “breakthrough pain.” None, however, has been studied speciﬁcally with regard to trigeminal neuralgia. Dosing of oral narcotics must be individualized to the patient. The key to management is ﬁnding the minimally effective dose. Used judiciously, narcotics can be very useful. Obvious concerns are tolerance and dependence.

Fentanyl Patch Long-acting narcotics may provide more sustained pain relief than the short-acting agents. Unfortunately, given the degree of bad press surrounding some oral agents, many physicians are reluctant to prescribe such medications. With this in mind, controlled-release formulations such as the fentanyl patch represent attractive solutions to the problem. Fentanyl is a short-acting narcotic. The patch is composed of a matrix that allows slower release of fentanyl into the skin over many days. Thus, these patches are less likely to produce any euphoria that may be the reinforcing aspect of many oral and parenteral products. Fentanyl patches are applied to the skin and changed every 3 days. The patches come in varying strengths. For those individuals naive to narcotics, the starting dosage should be 25 μg/h, but the dose can be increased as tolerated up to 100 μg/h. As with other narcotics, the major initial side effects include sedation and nausea. The most concerning long-term side effect remains dependence.

Conclusion In conclusion, trigeminal neuralgia is a serious medical condition that can be debilitating when individuals are severely affected. A number of medical treatment options exist for the disorder, including various antiepileptic drugs, GABA-mimetic agents, topical anesthetics, and even narcotics. Selection of any speciﬁc agent should be dictated by the medication’s relative efﬁcacy and side-effect proﬁle as well as the patient’s medical history and personal preferences. In refractory cases, nonpharmacological treatments such as glycerol injections, neurosurgical manipulation of the trigeminal nerve, and local irradiation serve as attractive alternatives to medication. For the foreseeable future, though, medical management will remain the ﬁrstline treatment for trigeminal neuralgia.

54.

trigeminal neuralgia: medical management perspective

References 1. Rozen TD. Trigeminal neuralgia and glossopharyngeal neuralgia. Neurol Clin N Am 2004; 22:185–206. 2. Joffroy A, Levivier M, Massager N. Trigeminal neuralgia: pathophysiology and treatment. Acta Neurol Belg 2001; 101:20– 25. 3. Rappapor ZH, Devor M. Trigeminal neuralgia: the role of selfsustaining discharge in the trigeminal ganglion. Pain 1994; 56:127–138. 4. Love S, Coakham HB. Trigeminal neuralgia: pathology and pathogenesis. Brain 2001; 124:2347–2360. 5. Kaye AH. Trigeminal neuralgia: vascular compression theory. Clin Neurosurg 2000; 46:499–506. 6. Blom S. Trigeminal neuralgia: its treatment with a new anticonvulsants drug. Lancet 1962; 1:839–840. 7. Campbel FG, Graham JG, Zikha KJ. Clinical trial of carbamazepine (Tegretol) in trigeminal neuralgia. J Neurosurg Neurol Psychiatry 1966; 29:265–267. 8. Nicol C. A four-year double-blind randomized study of Tegretol in facial pain. Headache 1969; 9:54–57. 9. Rockliff BW, Davis EH. Controlled sequential trials of carbamazepine in trigeminal neuralgia. Arch Neurol 1966; 15:129– 136. 10. Beydoun A. Safety and efﬁcacy of oxcarbazepine: results of randomized, double-blind trials. Pharmcotherapy 2000; 20:152S– 158S. 11. Carrazana E, Mikoshiba I. Rationale and evidence for the use of oxcarbazepine in neuropathic pain. J Pain Symptom Manage 2003; 25:S31–S35. 12. Sist T, Filadora V, Miner M, Lema M. Gabapentin for idiopathic trigeminal neuralgia: reported to cases. Neurology 1997; 48: 1467. 13. Khan OA. Gabapentin relieves trigeminal neuralgia in multiple sclerosis patients. Neurology 1998; 51:611–614.

539

14. Lunardi G, Leandri M, Albano C, et al. Clinical effectiveness of Lamotrigine and plasma levels in essential and symptomatic trigeminal neuralgia. Neurology 1997; 48:1714–1717. 15. Canavero S, Bonicalzi V. Lamotrigine control of trigeminal neuralgia: an expanded study. J Neurol 1997; 244:527–532. 16. Zakrzewska JM, Chaudhry Z, Nurmikko TJ, et al. Lamotrigine (Lamictal) in refractory trigeminal neuralgia: results from a doubleblind placebo-controlled crossover trial. Pain 1997; 73:223–230. 17. Zvartau-Hind M, Din MU, Gilani A, et al. Topiramate relieves refractory trigeminal neuralgia in MS patients. Neurology 2000; 55:1587–1588. 18. Solaro C, Messmer MM, Brichetto G, et al. Topiramate relieves idiopathic and symptomatic trigeminal neuralgia. J Pain Symptom Manage 2001; 21:367–368. 19. Bergouignan M. Cures heureuses de nevralgies facials essentielles par diphenyl-hidantoinate de soude. Rev Laryngol Otol Rhinol 1942; 63:34–41. 20. Raskin NH. Facial pain. In: Headache. New York: Churchill Livingstone, 1988:333–374. 21. Zakrzewska JM. Trigeminal neuralgia. In: Medical Management. London: W.B. Saunders, 1955:80–107. 22. Fromm GH, Terrence CF, Chatta AS, Glass JD. Baclofen in trigeminal neuralgia: its effect on the spinal trigeminal nucleus: a pilot study. Arch Neurol 1980; 37(12):768–771. 23. Fromm GH, Terrence CF, Chatta AS. Baclofen in the treatment of trigeminal neuralgia: double-blind study and long-term followup. Arch Neurol 1984; 15(3):240–244. 24. Court JE, Kase CS. Treatment of tic douloureaux with a new anticonvulsant (clonazepam). J Neurol Neurosurg Psychiatry 1976; 39:297–299. 25. Epstein JB, Marcoe JH. Topical application of capsaicin for treatment of oral neuropathic pain and trigeminal neuralgia. Oral Surg Oral Med Oral Pathol 1994; 77:135–140. 26. Fusco BM, Alessandri M. Analgesics effect of capsaicin in idiopathic trigeminal neuralgia. Anesth Analg 1992; 74:375–377.

5 5

Movement Disorder Sangjin Oh, Ajay Niranjan, and William J. Weiner

Introduction In 1951, Lars Leksell introduced noninvasive neurosurgery with narrow beams of radiation, which allowed neurosurgeons to lesion intracranial areas unreachable or ill-suited for open surgery. A semicircular arc targets gamma rays to destroy deepbrain structures without the risk of bleeding and infection associated with open surgery [1, 2]. This technological development opened new possibilities for treating disease in patients who were poor candidates for open surgery due to anticoagulation use, respiratory or cardiac disease, and advanced age. Gamma Knife (GK) ablative technique may provide a safer alternative to conventional neurosurgical ablative procedures. Open neurosurgical techniques are associated with higher mortality and morbidity especially in the debilitated patients. There is morbidity associated with the opening of the cranium as well as risk of hemorrhage when electrodes are passed to deep-brain structures. Perioperative infections including meningitis are also more common with open procedures. There has been a resurgence of interest in the neurosurgical approach to the treatment of movement disorders, especially Parkinson disease (PD). The reported success of pallidotomy in treating advanced PD led to even more innovative surgical approaches to effect deep targets. An alternative to the conventional lesioning technique, and favored by clinicians currently, is deep-brain stimulation (DBS). Benabid et al. [3] reported that chronic stimulation of the ventralis intermedius (VIM) nucleus of the thalamus is better than thalamotomy for PD and essential tremor (ET). Other investigators also showed favorable clinical outcomes (Table 55-1) [3–6]. The advantages of DBS include reversibility of stimulation effect, adaptability (stimulation setting can be changed), and ability to perform bilateral operations without increased morbidity [5]. The disadvantages of DBS include expense related to cost of the equipment, time needed to program the stimulator, implantation of foreign material, possible equipment failure (e.g., lead fractures), and need to replace the battery [7]. In addition, there is a lack of head-to-head studies comparing ablative lesion therapy to stimulation therapy for movement disorders. The initial success in treating pain with GK procedures involved ablating the central median region of the thalamus in patients with advanced cancer. This led investigators to attempt to treat various neurologic problems including trigeminal neuralgia, severe anxiety, obsessive-compulsive neurosis, focal sei-

zures, tremors, vascular malformations, and tumors [8]. In the past decade, with the advent of better targeting with magnetic resonance imaging (MRI), interest in GK for the treatment of movement disorders has increased. There are many case reports and several case series concerning the use of GK for movement disorders, but a formal controlled study is lacking. Interpretation of these case series is exceedingly difﬁcult. The documentation of therapeutic response to the procedure is variable, and the follow-up data is inconsistent and often poor. Moreover, the GK radiation dosage, the desired localization, and the size of the lesions vary from surgeon to surgeon. The usefulness of some of the published case series should be interpreted with caution especially when published in journal supplements. Supplements do not usually undergo a peer-review process and lack the scientiﬁc scrutiny of published journal articles. This chapter will review the current literature and examine the role of GK in movement disorders.

Parkinson Disease The classic signs of parkinsonism include resting tremor, bradykinesia, cogwheel rigidity, and postural instability. The most common cause of parkinsonism is Parkinson disease. Although James Parkinson’s initial description of the disorder ﬁrst noted in An Essay on the Shaking Palsy (1817) remains quite accurate, it has become evident that there are clinical clues to differentiate PD from parkinsonism (Table 55-2). PD is a progressive degenerative neurologic disorder primarily characterized by degeneration of the dopaminergic neurons in the substantia nigra pars compacta [9]. Initial treatment was limited to alleviating symptoms of excessive sweating and drooling with belladonna alkaloids. In the 1950s, synthetic anticholinergics were introduced to reduce tremors. Along with medical treatment for PD, neurosurgical techniques were reﬁned with the development of the stereotaxic atlas by Spiegel and Wycis [10] in 1952. However, a milestone in PD treatment occurred in the 1967 with the successful use of high-dose levodopa (l-dopa) [11]. Although it was initially hoped that levodopa would alter the degenerative process of PD, it soon became apparent that despite its potent symptomatic effect, it neither accelerated nor slowed disease progression. Side effects of treatment include motor ﬂuctuations and dyskinesia.

541

542

s. oh et al.

TABLE 55-1. Favorable clinical outcomes. Investigation

Patients improved/ total patients

Blond et al., 1992 [4] Benadid et al., 1996 [3] Koller et al., 1997 [5] Ondo et al., 1998 [6]

4/4 20/20 27/29 14/14

TABLE 55-3. Neurosurgical procedures for movement disorder. 1. Morbidity

None None None None

2.

3. 4.

In the ensuing 30 to 40 years, multiple new therapeutic agents were introduced for symptomatic treatment of PD and to attempt to alter disease progression. These agents include dopamine agonists (bromocriptine, pergolide, pramipexole, ropinirole, cabergoline, apomorphine), MAO inhibitors (selegiline, rasagiline), amantadine, COMT inhibitors (tolcapone, entacapone), and anticholinergics. These newer agents all play signiﬁcant roles in the symptomatic management of PD. However, no protective or disease-modifying strategies evolved. Certain symptoms of PD (resting tremor, bradykinesia, and rigidity) have been shown to be particularly amenable to dopaminergic therapy [12], whereas other symptoms and signs (dementia, balance and falling, autonomic symptoms) either do not respond or respond very little to therapy [13]. This balance between dopaminergic and nondopaminergic responsive symptoms is important when considering which patient is a surgical candidate.

Radiosurgical Thalamotomy The role of surgical intervention for PD has historically been reserved for patients who fail pharmacologic therapy and who develop disabling motor complications (motor ﬂuctuations and dyskinesias) from medical therapy. In the 1930s, Putnam sectioned the pyramidal tracts in the spinal cord [14], while Bucy lesioned the motor and premotor cortex [15] to treat tremors while causing paresis. Meyers, in 1939, resected the head of the caudate and anterior limb of the internal capsule resulting in relief of tremor, rigidity, and gait without causing paresis [16]. This surgical approach, improved by lesioning the pallidofugal ﬁbers, was reported to improve bradykinesia and posture [17]. In 1952, accidental ligation of the anterior choroidal artery in patients with PD resulted in improvement in parkinsonian symptoms [18]. At about the same time, with the development

TABLE 55-2. Symptoms of parkinsonism. Types of parkinsonism

Clinical symptoms

Idiopathic Parkinson disease Progressive supranuclear palsy Multiple system atrophy

Tremor, rigidity, bradykinesia, postural instability Rigidity, bradykinesia, postural instability, vertical ophthalmoparesis, early falls Rigidity, bradykinesia, autonomic instability, cerebellar signs Rigidity, akinesia, postural instability, dystonia, markedly unilateral ﬁndings, apraxia Rigidity, bradykinesia, postural instability, early dementia

Corticobasal degeneration Lewy body dementia

Ablative procedure Thalamotomy Pallidotomy Subthalamotomy Deep-brain simulation Thalamus Pallidum Subthalamic nucleus Brain grafting Striatum Focal radiation with Gamma Knife Thalamus Pallidum Subthalamic nucleus

of the stereotaxic frame [10], Leksell and colleagues used radiofrequency electrocoagulation to lesion the globus pallidus [19]. Although lesioning the anterodorsal region provided inconsistent results, targeting the posterior portion of the globus pallidus provided more consistent outcomes. Hassler and Reichert [20] in 1954 used thalamotomy to treat tremor in PD patients with more reliable relief of symptoms. Although posterior pallidotomy showed favorable results, thalamotomy gained favor and was used extensively until the introduction of levodopa in 1967. In the 1970s, interest in surgery for PD reemerged because disease progression and treatment-associated adverse effects continued to disable patients. Increased understanding of the functional neurophysiology and neuroanatomy of the basal ganglia and their connections have made surgery a more effective treatment modality (Table 55-3). Otsuki et al. [21] reported one patient with a 5-year history of hemiparkinsonian tremor and levodopa-induced hallucinations who underwent GK thalamotomy. The patient experienced complete resolution of tremors. Subsequently, the patient’s levodopa was decreased, which eliminated the hallucinations. The patient was reevaluated clinically at the 11th and 18th postoperative months and was found to have absence of tremor and rigidity. However, no formal objective measures were used to evaluate the patient. Friehs et al. [22] reported 10 PD patients with caudatotomy and 2 with thalamotomy. The caudatotomy patients underwent the procedure because of their unresponsiveness to medication and bradykinesia (10 patients). Five patients were Hoehn and Yahr stage 5 and ﬁve stage 3 or 4 with duration of parkinsonism of at least 3 years. The results of these caudatotomy patients were difﬁcult to analyze. The patients were followed from 1 to 12 months with the Uniﬁed Parkinson’s Disease Rating Scale (UPDRS) score. Seven patients showed “marked improvement,” whereas three patients were either slightly worse or no better compared with preoperative levels. Three months postoperatively, results were available for four patients, and three of these patients showed improvement compared with preoperative level. Only two patients were followed for 12 months. One showed no improvement whereas the other showed improvement in the UPDRS score. This same pattern of improvement in UPDRS score was reported in the two thalamotomy patients, however, no actual scores were reported. The thalamotomies were performed on patients with disease duration between 5 and 8 years to treat tremor and rigidity. The

55.

movement disorder

authors concluded that 11 of the 12 patients showed clear beneﬁt from the procedure with moderate to excellent outcome with regard to tremor. However, the few actual scores reported with regard to the overall beneﬁt of the procedure do not seem to support much beneﬁt. A larger case series included 34 patients with disabling tremor due to PD that was refractory to medical therapy [23]. A total of 38 thalamotomies (four patients received bilateral procedures) were performed. Patients were initially evaluated for invasive radiofrequency thalamotomies but were considered poor surgical or anesthetic candidates due to age, anticoagulation use, mild dementia, and/or severe cardiopulmonary disease. In addition, 18 of the 34 patients preferred a less invasive treatment. Patients were evaluated using the UPDRS tremor score and subjective symptom relief scale with a mean follow-up of 28 months. The median time to symptom improvement was 2 months. Thirty-four (89%) of 38 thalamotomies resulted in decrease or complete resolution of tremors. Fifty-ﬁve percent of the patients improved their UPDRS tremor score by two to three grades and 24% experienced complete tremor relief. There are other small series of PD patients treated with GK thalamotomy that report good functional outcome. Pan et al. [24] reported eight patients with PD who either stopped responding or who experienced serious side effects from medications who underwent the procedure. The duration of disease in the patients ranged from 2 to 10 years. Tremor was the dominant symptom in seven patients, and rigidity was the dominant symptom in one. Follow-up was available for six patients, and their clinical examination showed three patients with complete resolution of tremors with varying improvement in the other three. Rigidity improved in four, but none of the patients had complete resolution of symptoms. Another series reported six patients with PD who underwent GK radiosurgery [25]. Three of the patients were treated with GK thalamotomy as the primary procedure for their symptoms. The other three patients had prior stereotactic coagulation of the contralateral thalamus and then the GK thalamotomy. The major symptom in these patients was tremor, and the symptom response time after the procedure varied from 3 months to more than a year. Follow-up data was available for four of the six patients in this series. In two patients who underwent GK thalamotomy as the primary procedure, one experienced marked reduction and eventual cessation of resting tremor, while the other patient died of a “cerebrovascular disorder.” In the two patients with prior stereotactic coagulation of the thalamus and no residual tremor in the contralateral limb, the target for the GK radiosurgery was the contralateral thalamus. Both patients were free of tremor by 12 months without complications. Another single case report [26] discussed one patient with tremor who showed improvement 3 months after GK thalamotomy. At 6 months, the patient had greatly diminished tremor without complications.

543

had relief of tremor and 35 (92%) had relief of rigidity and hypokinesia. The authors also noted improvement in both levodopa-induced dyskinesia and “off” period painful dystonia in their patients. The effectiveness of pallidotomy on dyskinesia has been demonstrated by subsequent experience of various centers and surgeons [28–35]. Direct observations, subjective disability rating by patients, and UPDRS scores were used to evaluate patients and showed improvements after pallidotomy. De Bie et al. [36] published the ﬁrst randomized, single-blind, multicenter trial of unilateral pallidotomy versus medical therapy. This trial found signiﬁcant beneﬁt from pallidotomy in the off-period of PD patients with regard to UPDRS motor assessment and levodopa-induced dyskinesias. However, the efﬁcacy and reporting of GK pallidotomies in PD patients with dyskinesia is not consistent in the literature. Rand et al. [37] reported eight PD patients who underwent GK ventral medial pallidotomy for rigidity and bradykinesia. Four of these patients improved. However, only one patient had dyskinesia. The patient’s symptom improved after the procedure, but the medication dose was also reduced. Friedman et al. [38] documented four patients, two with bradykinesia nonresponsive to levodopa and two with motor ﬂuctuations, treated by GK posteroventral pallidotomies. Only one of four patients showed improvement in dyskinesia, but that was accompanied by worsening of UPDRS motor score. In addition, this same patient experienced transient disorientation, behavioral changes, and visual hallucinations with paranoid, bizarre delusions. Friedman and colleagues [39] also reported two patients with PD and dyskinesia treated with radiosurgical pallidotomy. Outcome measures included subjective measures and a thorough neurologic evaluation 3 months after the procedure. One patient had complete resolution, and the other had “mild” improvement of dyskinesia. However, information regarding patients’ PD medication changes after the procedure was not included. Young and colleagues [40] performed GK pallidotomies in 28 patients with rigidity, bradykinesia, or levodopa-induced dyskinesia. Fourteen of 28 patients had dyskinesia and in 12 (85.7%) patients dyskinesias improved. In addition, 18 (64.3%) of 28 patients showed improvement in rigidity and bradykinesia. However, the UPDRS scores in these patients did not show signiﬁcant change after the procedure. Further study of these patients compared 29 GK pallidotomies (one additional patient from prior case series) to 22 standard radiofrequency pallidotomies for dyskinesia, bradykinesia, and rigidity [41]. The authors report 85% relief of dyskinesia with both GK and radiofrequency methods. About two thirds of patients in both groups improved in bradykinesia and rigidity. Like the prior report, neither group showed improvement in UPDRS scores. It is difﬁcult to interpret these results because improvement in UPDRS scores would be expected to occur if bradykinesia and rigidity improved.

Radiosurgical Pallidotomy In the 1990s, interest in pallidotomy was renewed after the publication by Laitinen and colleagues [27]. The authors performed stereotactic pallidotomies on 38 patients with PD whose main symptom was hypokinesia unresponsive to medication. Postoperative examination revealed 34 (89%) of the patients

Radiosurgical Subthalamotomy Another anatomic target in the therapy of PD is the subthalamic nucleus (STN). Both ablative and DBS technique have demonstrated the capacity to ameliorate PD [42, 43]. Limousin and colleagues [44] showed that bilateral stimulation of the

544

s. oh et al.

STN in 20 patients signiﬁcantly improved all motor features of PD, including bradykinesia, rigidity, gait, balance, dyskinesia, and tremor. The success with either radiofrequency ablation or DBS of the STN led to trials of GK STN procedures. Keep et al. [45] reported the one and only GK subthalamotomy on a patient with a 20-year history of PD. Right radiofrequency pallidotomy was performed 27 months prior to GK subthalamotomy for control of tremor and dyskinesia on the left side. The prior right pallidotomy improved left-sided tremor, dyskinesia, motor ﬂuctuations, tone, motor control, balance, and overall mobility. However, the disease progressed over 2 years and the patient required additional medications. The patient experienced recurrent motor ﬂuctuations predominately on the right, deterioration of balance and mobility, and increased rigidity mainly in the right limbs. Left GK subthalamotomy was performed, and the patient was able to reduce levodopa dosage with better mobility and minimal dyskinesia. At 1-year follow-up, the patient had normal tone on the right with mild increased tone and cogwheel rigidity on the left. Rapid alternating and ﬁne ﬁnger movements were better on the right. The patient was assessed using the Parkinson’s Disease Disability Rating and was improved from a 28 to a 16 after the procedure. Tremor or abnormal movements were not observed, and the patient moved with “ease.” Until more clinical experience and a clinical trial is performed, the role of GK surgery targeting the STN in PD is unclear.

Essential Tremor Essential tremor (ET) is estimated to affect 0.4% to 3.9% of the population [46]. ET is deﬁned as an involuntary, rhythmic tremor, most typically in the hands and arms. ET is most prominent in action (kinetic tremor) or with arms held against gravity (postural tremor). It is differentiated from PD by the absence of resting tremor, symmetrical onset, absence of cogwheel rigidity, bradykinesia, postural reﬂex impairment, and the presence of a kinetic and postural tremor. Medications provide relief from tremor in approximately 50% to 60% of patients and include beta-blockers (propranolol), primidone, and benzodiazepines (diazepam) [47]. Surgical intervention for ET is reserved for patients who do not respond to medical therapy and who have signiﬁcant functional disability. Surgical therapies include ablative procedures and stimulation of the VIM nucleus of the thalamus. Young et al. [48] performed unilateral GK thalamotomies in 52 patients with ET refractory to medical treatment. The lesion was placed in the VIM nucleus contralateral to the side of the patient’s more severe tremor. Ninety-two percent of the patients were fully or nearly tremor free, usually within 2 to 3 months after the procedure, and 88% remained tremor-free after 4 years. Seventeen patients were followed for 4 years or longer and continued to show signiﬁcant improvement in the Clinical Rating Scale for tremor. Niranjan et al. [49] reported similar ﬁndings in eight patients. These patients underwent GK thalamotomy for tremor refractory to medication. Tremor was graded by performing ﬁnger to nose testing on a scale from zero (no tremor) to four (severe tremor). Handwriting and drawing (name, sentence,

and Archimedes’ spiral) were also used to assess functional disability. After the procedure, six of the eight patients experienced complete resolution of their tremor and two experienced more than 50% improvement. All patients showed improvement in activities of daily living including eating, drinking, holding, and using their hands. Jawahar and colleagues [50] described an 80-year-old patient with medically refractory essential tremor. The tremor began as “head bobbing” but after 21 years involved the upper extremities and hampered daily function. Unilateral GK thalamotomy was performed with reduction in tremor 14 months postprocedure. Archimedes’ spiral drawing and handwriting improved, and the patient was completely satisﬁed with the result. The literature is limited regarding the effectiveness of GK thalamotomy for ET. Whether DBS of the VIM thalamus or GK thalamotomy is more effective with fewer complications is unknown. However, GK thalamotomy may be an effective option in the treatment of ET patients refractory to medications who are not suitable for DBS procedure.

Other Applications of Gamma Knife Procedure There are isolated reports of GK ablative procedures used for other movement disorders. Niranjan et al. [49] performed GK thalamotomies in three patients with tremor secondary to multiple sclerosis. These tremors did not respond to medical therapy and caused severe disability. Patients were objectively assessed by grading the tremor and by handwriting and drawing tests. Other surgical options (DBS, radiofrequency ablation) were not pursued because patients were judged to be unable to participate in stimulation and recording protocols during an open procedure. All three patients noticed improvement in activities of daily living. Handwriting grade improved in one patient, but two were not able to write or draw secondary to MS-related neurologic deﬁcits despite improvement in tremor. Kwon and Whang [51] reported the use of GK pallidotomy to treat a patient with left hemidystonia caused by an old infarct in the right caudate nucleus, globus pallidus, and putamen. The patient had a cerebral angiogram, which did not show vascular anomalies. She was treated with trihexyphenidyl and clonazepam without improvement in symptoms. GK pallidotomy was performed, and the patient experienced marked improvement in hand movement and reduced stiffness in the left leg over 16 months. Two patients with focal dystonia underwent radiosurgery ablation of the anterior portion of the ventrolateral nucleus and had excellent clinical improvement [52]. One patient preoperatively had severe right lower-limb dystonia, and the other had right upper-limb torsion dystonia. Both patients had clinical improvement in their symptoms 2 to 3 months after the GK procedure. Torticollis [53], dystonia [53, 54], and tremors [53] are occasionally associated with arteriovenous malformations (AVM). Surgical excision of vascular malformation in the putamen, which caused hemichorea-hemiballism, resulted in cessation of movements [55]. This anecdotal evidence led to GK ablation of an AVM in the left caudate and putamen in a patient with right hemiballism. The patient experienced cessation of abnormal

55.

545

movement disorder

movements after 6 months [56]. The evidence is limited but movement disorders caused by vascular malformations may be amenable to GK ablative therapy.

Movement Disorder Radiosurgery: Technical Considerations Radiosurgery Team Movement disorder radiosurgery should only be performed by surgeons experienced in stereotactic functional neurosurgery using radiofrequency or deep-brain stimulation. Target planning for functional radiosurgery is operator dependent. The target selection is entirely based on the stereotactic imaging and surgeon’s experience. In radiosurgical ablation, there is no opportunity to conﬁrm target selection because electrophysiologic feedback is not available.

The Radiosurgical Device

These coordinates are projected on fast inversion recovery axial images to see the relation of the target with internal capsule. The target can be ﬁne-tuned by moving it in x, y, or z dimensions. We place the 50% isodose line of the 4-mm collimator at the edge of the contralateral internal capsule. Other centers may use different strategies to localize the VIM nucleus. The dose proﬁle of the current Gamma Knife unit (model 4-C) is elliptical with its maximum diameter in x- and y-dimensions (Fig. 55-1A). This dose fall-off can be made steeper if a plug pattern is used to block beams coming from lateral directions. A change in Gamma angle can also inﬂuence the dose proﬁle. The Gamma angle of 125 instead of 90 might produce a better dose proﬁle for VIM target (Fig. 55-1B). In the model 4-C Gamma Knife unit, the target coordinates can also be obtained using Multiview software, which allows surgeons to identify AC and PC and then allows them to use a formula that provides the target coordinates. It also has the capability to register a computerized atlas, which allows the user to project the calculated target on the registered atlas (Fig. 55-2). It is a nice feature, but the target coordinates calculated by an experienced surgeon are preferred over the target shown in the atlas.

To date, there is no reported clinical experience using linac technique for the treatment of movement disorders. In Gamma Knife radiosurgery, once the patient is in position and treatment begins, neither the radiation source nor the patient (or target) move until the completion of treatment. This eliminates any inaccuracy because of patient or source movement. At the target, Gamma Knife creates a well-circumscribed lesion measuring 5 to 6 mm at the 50% isodose line. The dose proﬁle and precision of dose delivery makes Gamma Knife the perfect tool for noninvasive radiobiological effect on aberrant neural pathways.

Magnetic Resonance Imaging and Target Planning In this section, the technique of radiosurgical VIM thalamotomy is described, as this is the most preferred functional procedure performed using Gamma Knife radiosurgery. Magnetic resonance imaging (MRI) is the imaging modality of choice. However, computed tomography (CT) can be used if MRI is contraindicated. The choice of magnetic resonance (MR) sequences may vary at different centers. At the University of Pittsburgh, SPGR (spoiled gradient recalled acquisition in steady state) sequence with 1-mm-thick axial images is used for planning. The target is localized in relation to AC-PC line. This is supplemented by anatomic information gathered from highresolution, short-TI inversion recovery (STIR) MR images and subjective surgeon correlation with the Schaltenbrand atlas. Computer atlases referenced to the size of the patient’s brain and AC-PC line can also be helpful. The x, y, and z coordinates can be determined for the target using either SurgiPlan or GammaPlan (Elekta Corp, Georgia, USA) software. We calculate the initial target using GammaPlan software with the following formula: x coordinate: one-half of width of third ventricle + 11 mm y coordinate: one-quarter of AC-PC distance + 1 or 2 mm from PC z coordinate: 2 to 3 mm above AC-PC level

FIGURE 55-1. Gamma Knife unit.

546

s. oh et al.

FIGURE 55-2. Computerized atlas.

There are no studies deﬁning safe radiation doses for GK thalamotomy and pallidotomy. Case reports have ranged from 120 to 200 Gy with variable responses and complications. In addition, delayed development of neurologic complications secondary to radiation necrosis is another drawback to this procedure. Radiosurgical pallidotomy has been associated with higher complication rates. The explanation for high complication rate in pallidotomy series could be the variability and unpredictability of radiation effect in the globus pallidus. This unpredictability and variability was not seen in the VIM thalamotomy series and likely represents anatomic susceptibility to very small vessel venous or arterial infarction in the area of GPi. It is also hypothesized that there is a differential sensitivity to radiation between thalamus and pallidum. Historically, the pallidum is sensitive to relative hypoxia. In addition, the pallidum is known to contain high levels of iron, and these levels typically rise with age. It has been hypothesized that the presence of iron within this structure may catalyze free-radical reactions causing toxicity to the aging brain. The complication rate can possibly be reduced by lowering the prescription dose.

Dose Selection Early studies reported radiosurgical necrotic lesions using maximum doses of 160 to 200 Gy. More modern reports describe single-fraction doses varying between 110 and 160 Gy. The selection of dose has been guided by a dose-volume analysis of prior experience. In most centers with a large experience, a central target dose of 130 to 140 Gy has become the norm for movement disorder radiosurgery.

Radiosurgery Adverse Events One concern with the use of GK ablation procedures for deep cerebral structures is lesion location and lesion size. It is difﬁcult to control both of these features. There is also concern regarding the latency period between procedure and desired clinical effect. The complication rate after GK procedures is unknown, but there are numerous case reports describing isolated events. A patient who underwent Gamma Knife thalamotomy for essential tremor developed complex progressive movements approximately 6 months after the procedure. The movement is described as dystonic and choreoathetotic [57]. Another patient developed homonymous hemianopsia and transient hemiparesis after GK pallidotomy [58]. Additional reports include a patient who developed a “neuropsychological disturbance” after pallidotomy and another patient with a GK ablative lesion, which increased in size even after 1 year [23, 40]. In a larger series of eight patients after GK thalamotomy or pallidotomy, uncontrollable laughter, hypophonia, dysarthria, headache, facial droop, weakness, gait abnormalities, and visual ﬁeld deﬁcits were noted [59]. In these eight patients, the lesion was off target on average 1.5 mm (range, 0 to 4 mm) when compared with MRI. Localization errors of several millimeters can be caused by speciﬁc distortions on MRI. These localization errors of several millimeters may result in ablation of the internal capsule and/or the optic tract. The lesion size ranged from 6 to 22 mm in diameter [40] and volume ranged from undetectable to 4000 mm3 [23]. Radio-necrosis of tissue can result in unpredictable responses and contributes to variability in lesion size.

Conclusion Gamma Knife ablative therapy has been used for movement disorder patients who are medically refractory and medically unstable. This noninvasive procedure is attractive for patients who cannot tolerate an open neurosurgical procedure. Gamma Knife radiosurgical thalamotomy is a safe and effective alternative to invasive radiofrequency or DBS techniques for highsurgical-risk patients. This does not seem to be the case with radiosurgical pallidotomy where higher complication rates have been reported. Case reports and uncontrolled series provide some information regarding its effectiveness in treating different movement disorders, but there is a need for long-term follow-up studies to evaluate the ultimate consequence of GK thalamotomy and its effectiveness compared with other surgical approaches. Until these studies are reported, the role of GK procedures in the treatment of patients with movement disorders remains unclear. At the present time with the limited information available, GK ablative therapy in PD and ET should be used only in those patients whose medical conditions impose an unacceptable risk for more deﬁnitive procedures. GK ablative surgery for other movement disorders remains an open question.

References 1. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316–319. 2. Leksell L. Cerebral radiosurgery. Acta Chir Scand 1968; 134:585– 595. 3. Benabid AL, Pollak P, Gao D, et al. Chronic electrical stimulation of the ventralis intermedius nucleus of the thalamus as a treatment of movement disorders. J Neurosurg 1996; 84:203–214. 4. Blond S, Caparros-Lefebvre D, Parker F, et al. Control of tremor and involuntary movement disorders by chronic stereotactic stimulation of the ventral intermediate thalamic nucleus. J Neurosurg 1992; 77:62–68. 5. Koller W, Pahwa R, Busenbark K, et al. High-frequency unilateral thalamic stimulation in the treatment of essential and parkinsonian tremor. Ann Neurol 1997; 42:292–299.

55.

movement disorder

6. Ondo W, Jankovic J, Schwartz K, et al. Unilateral thalamic deep brain stimulation for refractory essential tremor and Parkinson’s disease tremor. Neurology 1998; 51:1063–1069. 7. Pahwa R, Lyons K, Koller WC. Surgical treatment of essential tremor. Neurology 2000; 54(Suppl 4):S39–S44. 8. Lindquist C, Kihlstrom L, Hellstrand E. Functional neurosurgery—a future for the gamma knife? Stereotact Funct Neurosurg 1991; 57:72–81. 9. Weiner WJ, Lang AE. Movement Disorders: A Comprehensive Review. New York: Futura Publishing Co, 1989. 10. Spiegel EA, Wycis HT, Marks M, et al. Stereotaxy apparatus for operations on the human brain. Science 1947; 106:349–350. 11. Cotzias CG, Papavasilou PS, Gallene R. Modiﬁcation of parkinsonism—chronic treatment with L-dopa. N Engl J Med 1969; 280:337–345. 12. Klawans HL. Individual manifestations of Parkinson’s disease after ten or more years of levodopa. Mov Disord 1986; 1(3): 187–192. 13. Weiner WJ, Shulman LM, Oertel WH. Do we need more drugs to treat Parkinson’s disease. Adv Neurol 1999; 80:481–483. 14. Putnam TJ. Treatment of unilateral paralysis agitans by section of the lateral pyramidal tract. Arch Neurol Psychiatr 1940; 44:950. 15. Bucy JC. Cortical extirpation in the treatment of involuntary movement. Arch Neurol Psychiatr 1942; 21:551. 16. Meyers R. Surgical interruption of the pallidofugal ﬁbres: its effect on the syndrome paralysis agitans and technical considerations in its application. N Y State J Med 1942; 42:317–325. 17. Meyers R. The modiﬁcation of alternating tremors, rigidity and festination by surgery of the basal ganglia. Assoc Nerv Ment Dis 1942; 20:602–665. 18. Cooper IS. Litigation of the anterior choroidal artery for involuntary movements of parkinsonism. Arch Neurol 1956; 75:36– 48. 19. Svennilson E, Torvik A, Lowe R, Leksell L. Treatment of parkinsonism by stereotactic thermo lesions in the pallidal region. A clinical evaluation of 81 cases. Acta Psychiat Neurol Scand 1960; 35:358–377. 20. Hassler R, Reichert T. Indikationen und lokalisations: methode de gezeilten hirnoperationen. Nervenarzt 1954; 25:441–447. 21. Otsuki T, Jokura H, Takahashi K, et al. Stereotactic gammathalamotomy with a computerized brain atlas: technical case report. Neurosurgery 1994; 35(4):764–768. 22. Friehs GM, Ojakangas CL, Pachatz P, et al. Thalamotomy and caudatotomy with the gamma knife as a treatment for Parkinsonism with a comment of lesion sizes. Stereotact Funct Neurosurg 1995; 64(Suppl 1):209–221. 23. Duma CM, Jacques DB, Kopyov OV, et al. Gamma knife radiosurgery for thalamotomy in parkinsonian tremor: a ﬁve-year experience. J Neurosurg 1998; 88:1044–1049. 24. Pan L, Dai JZ, Wang BJ, et al. Stereotactic Gamma thalamotomy for the treatment of parkinsonism. Stereotact Funct Neurosurg 1996; 66(Suppl 1):329–332. 25. Ohye C, Shibazaki T, Hirato M, et al. Gamma thalamotomy for Parkinsonian and other kinds of tremor. Stereotact Funct Neurosurg 1996; 66(Suppl 1):333–342. 26. Hirato M, Ohye C, Shibazaki T, et al. Gamma knife thalamotomy for the treatment of functional disorders. Stereotact Funct Neurosurg 1995; 64(Suppl 1):164–171. 27. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992; 76:53–61. 28. Baron MS, Vitek JL, Bakay RA, et al. Treatment of advanced Parkinson’s disease by posterior GPi pallidotomy: 1-year results of a pilot study. Ann Neurol 1996; 40:355–366. 29. Kishore A, Turnbull IM, Snow BJ, et al. Efﬁcacy, stability and predictors of outcome of pallidotomy for Parkinson’s disease: six-

30.

31.

32.

33.

34.

35.

36.

37.

38. 39.

40.

41.

42.

43. 44.

45.

46.

47. 48.

49.

50.

547

month follow-up with additional 1-year observations. Brain 1997; 120:729–737. Kopyov O, Jacques D, Duma C, et al. Microelectrode-guided posteroventral medial radiofrequency pallidotomy for Parkinson’s disease. J Neurosurg 1997; 87:52–59. Fazzini E, Dogali M, Sterio D, et al. Stereotactic pallidotomy for Parkinson’s disease: a long-term follow-up of unilateral pallidotomy. Neurology 1997; 48:1273–1277. Uitti RJ, Wharen RE Jr, Turk MF, et al. Unilateral pallidotomy for Parkinson’s disease: comparison of outcome in younger versus elderly patients. Neurology 1997; 49:1072–1077. Samuel M, Caputo E, Brooks DJ, et al. A study of medial pallidotomy for Parkinson’s disease: clinical outcome, MRI location and complications. Brain 1998; 121:59–75. Scott R, Gregory R, Hines N, et al. Neuropsychological, neurological and functional outcome following pallidotomy for Parkinson’s disease: a consecutive series of eight simultaneous bilateral and twelve unilateral procedures. Brain 1998; 121:659–675. Shannon KM, Penn RD, Kroin JS, et al. Stereotactic pallidotomy for the treatment of Parkinson’s disease. Efﬁcacy and adverse effects at 6 months in 26 patients. Neurology 1998; 50:434–438. de Bie RM, de Haan RJ, Nijssen PC, et al. Unilateral pallidotomy in Parkinson’s disease: a randomised, single-blind, multicentre trial. Lancet 1999; 354:1665–1669. Rand RW, Jacques DB, Melbye RW, et al. Gamma knife thalamotomy and pallidotomy in patients with movement disorders: preliminary results. Stereotact Funct Neurosurg 1993; 61(Suppl 1): 65–92. Friedman JH, Epstein M, Sanes JN, et al. Gamma knife pallidotomy in advanced Parkinson’s disease. Ann Neurol 1996; 39:535–538. Friedman DP, Goldman HW, Flanders AE, et al. Stereotactic radiosurgical pallidotomy and thalamotomy with the gamma knife: MR imaging ﬁndings with clinical correlation-preliminary experience. Radiology 1999; 212:143–150. Young RF, Shumway-Cook A, Vermeulen SS, et al. Gamma knife radiosurgery as a lesioning technique in movement disorder surgery. J Neurosurg 1998; 89:183–193. Young RF, Vermeulen S, Posewitz A, Shumway-Cook A. Pallidotomy with the gamma knife: a positive experience. Stereotact Funct Neurosurg 1998; 70(Suppl 1):218–228. Barlas O, Hanagasi HA, Imer M, et al. Do unilateral ablative lesions of the subthalamic nucleu in parkinsonian patients lead to hemiballism? Mov Disord 2001; 16:300–310. Gill SS, Heywood P. Bilateal dorsolateral subthalamotomy for advanced Parkinson’s disease. Lancet 1997; 350:1224. Limousin P, Krack P, Pollak P, et al. Electrical stimulation of the subthalamic nucleus in advanced Parkinson’s disease. N Engl J Med 1998; 339:1105–1111. Keep MF, Mastrofrancesco L, Erdman D, et al. Gamma knife subthalamotomy for Parkinson disease: the subthalamic nucleus as a new radiosurgical target. Case report. J Neurosurg 2002; 97(Suppl 5):592–599. Vaillancourt DE, Sturman MM, Verhagen Metman L, et al. Deep brain stimulation of the VIM thalamic nucleus modiﬁes several features of essential tremor. Neurology 2003; 61(7):919–925. Zesiewicz TA, Elble R, Louis ED, et al. Practice parameter: therapies for essential tremor. Neurology 2005; 64:2008–2020. Young RF, Jacques S, Mark R, et al. Gamma knife thalamotomy for treatment of tremor: long-term results. J Neurosurg 2000; 93(Suppl 3):128–135. Niranjan A, Kondziolka D, Baser S, et al. Functional outcomes after gamma knife thalamotomy for essential tremor and MSrelated tremor. Neurology 2000; 55:443–446. Jawahar A, Cardenas RJ, Zwieg RM, et al. A case report of complete disappearance of essential tremor after gamma knife radiosurgery. J La State Med Soc 2004; 156:140–142.

548

s. oh et al.

51. Kwon Y, Whang CJ. Stereotactic gamma knife radiosurgery for the treatment dystonia. Stereotact Funct Neurosurg 1995; 64(Suppl 1):222–227. 52. Pan L, Zhang N, Dai JZ, Wang EM. Gamma knife radiosurgery for torsion spasm: report of two cases. J Neurosurg 2000; 93 (Suppl 3):189–190. 53. Labo-Antunes J, Yahr MD, Hilal SK. Extrapyramidal dysfunction with cerebral arteriovenous malformations. J Neurol Neurosurg Psychiatr 1974; 37:259–268. 54. Friedman DI, Jankovic J, Rolak LA. Arteriovenous malformation presenting as hemidystonia. Neurology 1986; 36:1590–1593. 55. Tamaoka A, Sakuta M, Yamada H. Hemichorea-hemiballism caused by arteriovenous malformations in the putamen. J Neurol 1987; 234:124–125.

56. Kurita H, Kawamoto S, Sasaki T, et al. Relief of hemiballism from a basal ganglia arteriovenous malformation after radiosurgery. Neurology 1999; 52:188–190. 57. Siderowf A, Gollump SM, Stern MB, et al. Emergence of complex, involuntary movements after gamma knife radiosurgery for essential tremor. Mov Disord 2001; 16(5):965–967. 58. Bonnen JG, Iacono RP, Lulu B, et al. Gamma knife pallidotomy: case report. Acta Neurochirurg 1997; 139(5):442–445. 59. Okun MS, Stover NP, Subramanian T, et al. Complications of gamma knife surgery for Parkinson disease. Arch Neurol 2001; 58:1995–2002.

5 6

Movement Disorders: Deep-Brain Stimulation Perspective John Y.K. Lee, Joshua M. Rosenow, and Ali R. Rezai

Introduction The introduction and technical reﬁnements of deep-brain stimulation (DBS) for the treatment of movement disorders over the past 15 years have resulted in a renaissance in the ﬁeld of functional neurosurgery. For patients with medically refractory movement disorders, DBS procedures have largely replaced neuroablative techniques in the treatment of patients with movement disorders. This shift away from ablative procedures is multifactorial, but among the more important factors are its proven efﬁcacy, the dramatic reversibility of the effects of neurostimulation, and the perceived minimally invasive nature of DBS compared with destructive procedures. Opposing this trend, however, has been the rise in the past decade in the use of Gamma Knife (GK) and occasionally linear accelerator (linac)-based radiosurgical units to perform thalamotomies and pallidotomies. The beneﬁts of radiosurgical ablations include the fact that it is even less invasive than traditional surgical approaches, thus virtually eliminating the risk of hemorrhage, infection, and hardware-related complications. In addition, Gamma Knife radiosurgery is less time-consuming than microelectrode-guided radiofrequency lesioning and can be performed in patients who are at higher surgical risks for open procedures, such as patients with medical comorbidities or patients who cannot tolerate an awake procedure. This article reviews the evidence for DBS versus the evidence for radiosurgical lesioning in the treatment of patients with medically refractory movement disorders. Based on the evidence, this paper attests to the greater evidence of efﬁcacy and safety with DBS procedures and also deﬁnes the role of radiosurgery in the treatment of patients with movement disorders.

Thalamic DBS Versus Radiosurgical Thalamotomy Thalamic DBS Stimulation of the ventralis intermedius (VIM) nucleus of the thalamus for the relief of medically intractable parkinsonian or

essential tremor is considered to be the forerunner of all other modern DBS procedures. In his initial 1987 report, Benabid performed chronic stimulation in patients with a contralateral thalamotomy [1]. This subset of patients was chosen because of the recognition of cognitive complications in patients who had undergone bilateral thalamotomies [2, 3]. In this initial report, all patients achieved some relief of tremor, but there was greater relief in the side of the thalamotomy as opposed to the side of stimulation. Benabid attributed this ﬁnding to a technical limitation of the implanted stimulators at that time, as the frequency limit of the implanted generators was 130 Hz as opposed to the optimal frequency goal of approximately 200 Hz. With this approach, Benabid initiated the ﬁeld of DBS with the VIM as the nuclear target and tremor as the primary symptom. Over the past few years, multiple practitioners in multiple centers in both North America and Europe have conﬁrmed and deﬁned the beneﬁts of DBS of the VIM nucleus in the treatment of parkinsonian tremor. In Parkinson disease (PD), Benabid et al. have followed approximately 100 patients for up to 8 years and have documented an 88% improvement in tremor [4, 5]. Only a minority of patients developed tolerance to DBS (e.g., a diminished effect over time requiring greater voltages and higher battery use). Thalamic DBS appeared to improve parkinsonian tremor but not to improve the other two cardinal features of Parkinson disease, bradykinesia and rigidity nor the levodopa-induced dyskinesias [6, 7]. In another study, Koller et al. reported the results of a double-blind, North American, multicenter study (four centers) in 24 PD patients [8]. They randomized unilateral thalamic DBS against placebo, and double-blinded assessments demonstrated a signiﬁcant reduction in resting, postural, and action tremor. However, DBS of VIM again did not improve the rigidity and bradykinesia of PD patients, and hence VIM DBS in parkinsonian patients is currently reserved for the rare unilateral tremor-predominant elderly patient. The majority of patients with medically refractory PD will undergo STN DBS at this time. DBS of the VIM nucleus for the treatment of essential tremor is highly effective with more than 90% of patients

549

550

j.y.k. lee et al.

demonstrating a satisfactory result [5, 9–13]. It is the surgical treatment of choice for patients with essential tremor who experience diminished quality of life despite maximal medical therapy. As many as 50% of patients have complete abolition of upper extremity as well as head tremor with signiﬁcant improvements in quality of life [14]. In the North American and European multicenter trials, DBS of the VIM thalamus was effective, and beneﬁcial results were maintained after several years [8, 15]. Adverse effects from VIM DBS are infrequent and usually mild. Benabid reported mild contralateral paresthesias, limb dystonia, and cerebellar dysmetria, which were all controlled by adjustments of stimulation parameters [6]. Dysarthria and gait disequilibrium were uncommon and nearly always limited to patients receiving bilateral stimulation or who had undergone prior thalamotomy. Signiﬁcant cognitive deﬁcits have not been observed after thalamic stimulation, although mild deﬁcits in verbal ﬂuency have been documented [14]. The great beneﬁt of DBS compared with permanent lesioning procedures is the ability to control multiple parameters, including voltage, pulse width, frequency, and contacts. Hence, the therapeutic index (e.g., the difference between beneﬁcial effect and adverse effect) can be navigated through careful programming [16]. An important complication of DBS pertinent to all DBS cases whether targeting VIM, globus pallidus interna (GPi), or STN is the risk of intraoperative/procedural hemorrhage. Benabid reported a microhemorrhage rate of six incidences out of a total of 177 operations, of which three were asymptomatic [5]. The intracranial hemorrhage rate in the North American and European trials was 2% and 5%, respectively [8, 15]. Starr recently reported a hemorrhage rate of 3.3%, but only a subset of these were symptomatic, thus resulting in a symptomatic hemorrhage rate of 0.6%. In addition, there has been some retrospective study evidence that the hemorrhage rate may increase with the aggressive use of microelectrode recording (MER). Hariz et al., in a meta-analysis of reported series, noted that non-MER techniques were at least ﬁve times less likely to have hemorrhagic complications [17]. For example, Palur et al. demonstrated a slightly higher rate of hemorrhage in cases performed with MER versus macrostimulation alone (0.4% vs. 0.2%) [18]. The DBS Study Group reported a relationship between the number of microelectrode tracks and the risk of hemorrhage: patients without hemorrhage had a mean of 2.9 ± 1.8 passes compared with 4.1 ± 2.0 among those who had hemorrhage (p = 0.05) [19]. There are no prospective, randomized studies of MER versus non-MER strategies, and there is a large variability in MER techniques used at different centers with regard to number of passes and equipment employed. In all studies noted, however, the overall small risk of bleeding in both MER and non-MER groups indicates the overall relative safety of both techniques [18–20]. The risk of hemorrhage is not negligible, but not all hemorrhages result in obvious clinical deﬁcit [21], and the risk of hemorrhage must be considered in relation to the beneﬁt obtained in patients with medically refractory movement disorders. Another speciﬁc complication of DBS is related to the need to implant permanent hardware into the body. At the

University of Toronto, Oh et al. documented a hardwarerelated complication rate of 25% of a total of 79 patients. These included lead fractures, lead migrations, short/open circuits, skin erosions/infections, foreign body reactions, and cerebrospinal ﬂuid leak [22]. The hardware-related complication rate per electrode-year was 8.4% with many skin erosions/infections occurring greater than a year after surgery. Kondziolka et al. documented a similar hardware-related complication rate in Pittsburgh: 27% of 66 patients developed hardware complications [23]. In contrast with these two reports, other studies have documented lower rates of hardware-related complications. In Koller’s series of 53 patients with unilateral thalamic stimulators, the incidence of infection was 3.8% with a 1.9% malfunction rate [8]. The Grenoble group’s large series of 197 patients contained 3 patients who experienced infections and 5 with scalp erosions leading to exposed hardware, for a total rate of 2.5% [24]. The European Multicenter Study yielded an erosion/infection rate of 2.7% and the North American trial reported 2.9% [15]. None of these large studies reported data on lead migration or fracture, and in both the mean follow-up time was 12 months or less. Among the 143 patients in the multicenter prospective series of patients undergoing GPi or STN DBS, there were 5 leads that migrated, 4 infected leads, 2 broken leads, 1 scalp erosion, and 1 incidence of equipment malfunction [19]. With improved electrode and hardware design as well as improvements in surgical technique, hardware-related complication rates appear to decrease. Nevertheless, hardware failure contributes to morbidity in patients who undergo the placement of permanent foreign-body, deep-brain stimulators. This complication is obviously avoided in patients who undergo a radiosurgical lesioning procedure, and the relative risks must be weighed for each patient. In summary, thalamic DBS for the treatment of tremor is effective in Parkinson disease and essential tremor as demonstrated in multiple trials that include double-blinded, prospective studies. The risk of intraoperative/procedural hemorrhage is approximately 3% with less than one-third of these being symptomatic. The risk of hardware-related complications ranges from 3% to 27%, but improvements in technique and hardware development will limit these problems. It is the opinion of most practitioners in the ﬁeld that the dramatic reversibility and remarkable control and ability to ﬁne-tune side effects warrant the risks of hardware failure, infection, and hemorrhage.

Radiosurgical Thalamotomy In order to discuss the merits of radiosurgical thalamotomy, it is instructive to consider the role of conventional radiofrequency thalamotomy in the treatment of patients with tremor. Tasker retrospectively studied patients who had radiofrequency thalamotomy and thalamic stimulation [25]. He noted tremor recurrence in 15% of thalamotomy patients, as opposed to 5% of DBS patients. Importantly, 23% of the thalamotomy patients required repeat procedures for tremor control. In addition, thalamotomy resulted in a decline in writing performance in 23% of patients, as opposed to no patients in the

56.

movement disorders: deep-brain stimulation perspective

DBS group. Whereas all of the DBS patients achieved better than 50% improvement in dexterity, writing, and drinking ability, only half of the thalamotomy cohort obtained that level of beneﬁt. With respect to complications, Tasker demonstrated that chronic stimulation had a lower incidence of ataxia and dysarthria (common complications of bilateral thalamotomy) [25]. Moreover, when these complications occurred in the DBS group, simply adjusting the stimulation parameters served to abolish these problems. In another study that was a prospective, blinded study, Schuurman et al. randomized 68 patients (45 with PD, 13 with ET, and 10 with multiple sclerosis) to radiofrequency thalamotomy versus VIM DBS [26]. Both groups of patients had excellent suppression of tremor, but the VIM DBS group had greater improvements in activities of daily living, and more importantly the VIM DBS group had fewer complications. Patients who underwent thalamotomy were more likely to develop somnolence, cognitive deterioration, and dysarthria, as well as gait or balance disturbance after thalamotomy. Thus, Tasker and Schuurman et al. have both conﬁrmed the higher complication rates associated with thalamotomy compared with VIM DBS, and it is in this context that the use of radiosurgical thalamotomy must be considered. Gamma Knife thalamotomy differs from radiofrequency thalamotomy in some important aspects. Gamma Knife thalamotomy is much less invasive as it does not require a skin incision, burr hole, or any brain parenchymal penetrations. Hence, the risk of complications is potentially lower than that of radiofrequency thalamotomy. On the other hand, targeting for a Gamma Knife thalamotomy relies purely on imagingdeﬁned anatomic targeting. The physiologic methods of macrostimulation and microelectrode recording are not available in purely imaging-based procedures such as Gamma Knife radiosurgical ablation. In addition, the method of lesion generation is delayed and as has been suggested in some recent publications, the lesion can be variable in size [27]. Hence, Gamma Knife thalamotomy has both advantages and disadvantages that may give it a role in the treatment of tremor. Compared with the large multicenter studies of deep-brain stimulation for tremor, there are only a few published case series of patients treated by Gamma Knife radiosurgical thalamotomy for the treatment of tremor. Ohye was one of the ﬁrst to advocate radiosurgical lesioning of the thalamus in patients who had a contralateral radiofrequency lesion or in patients who had a previously mapped lesion [28]. At present, Ohye has reported radiosurgical thalamotomy in 70 patients with PD and has reported tremor suppression of at least two thirds in more than 80% of patients with no loss of effect even after 10 years in some patients [27, 29]. Duma et al. has reported his series of 38 radiosurgical thalamotomies over a 5-year period. In his series of patients, 24% had complete relief of tremor, 55% had “excellent or good” improvement in their tremor, and 21% of patients had only “mild” or no relief of tremor. Median follow-up in this series was 28 months [30, 31]. Niranjan et al. reported favorable results in 12 patients who were not candidates for open surgical procedures [32]. Eight of their 12 patients with essential tremor had improvement in tremor. The largest series of patients undergoing Gamma Knife thala-

551

motomy has been published by Young et al. who retrospectively reviewed 158 patients who underwent radiosurgical thalamotomy [33]. Of the 74 patients with parkinsonian tremor followed at least 4 years, 79.7% were still tremor-free while 70.6% of the 16 patients with essential tremor followed for this period were tremor-free. In this series, only one patient sustained a transient complication and two patients sustained permanent complications including hemiparesis and mild facial paresthesias. No cognitive changes were reported in this series. As can be seen in the review of the world experience of Gamma Knife thalamotomy for tremor, results can be quite good; however, the follow-up has generally not been conducted by blinded observers, and no study has been multicentered. Hence, Gamma Knife radiosurgical thalamotomy does not have the demonstrated efﬁcacy compared with the multicenter, double-blinded studies that have been conducted in the evaluation of radiofrequency thalamotomy and DBS procedures. Thus, it is difﬁcult to compare directly radiosurgical thalamotomy to deep-brain stimulation or even radiofrequency thalamotomy. An important distinction between Gamma Knife radiosurgical thalamotomy and radiofrequency thalamotomy is the time it takes for radiosurgical lesioning to take effect. Beneﬁts and complications from radiosurgical procedures are delayed from 1 month to 12 months [27, 30, 32]. This delay is related to the time it takes for the radiation to functionally damage or to destroy the tissue targeted. The time to lesion effect can be accelerated with increased doses [34]; however, the optimal technique for Gamma Knife radiosurgical lesioning has not been established. Peak central doses can range from 120 Gy to 200 Gy. Duma et al. reported a difference in clinical outcome with different doses, suggesting an improvement with higher doses [31]. Despite this correlation, most surgeons at this time now perform radiosurgical lesioning for functional disorders at lower doses (e.g., 140 Gy) [32]. This decision is partially based on the series of eight patients all treated at the same institution who developed complications published by Okun et al. [35]. In that series, all except one of the thalamotomy procedures were performed at exceptionally high doses using 200 Gy (one patient was treated at 150 Gy). Unfortunately, a systematic study of dose-response-complication rate has not been performed to date, but it is the choice of prudent neurosurgeons at this time to perform radiosurgical thalamotomy at doses <50 Gy (D. Kondziolka, personal communication, 2005). Because of the delay in effect after GK thalamotomy, an accurate documentation of complications requires vigilant follow-up. In Young’s series, 3 of 74 patients developed a complication [33]. Other reports are single-center reports of complications without a denominator to deﬁne the total number of procedures or patients who underwent Gamma Knife thalamotomy. Siderowf et al. reported one case of a complex, involuntary movement after Gamma Knife radiosurgical thalamotomy [36]. Unfortunately, speciﬁc dose parameters were not speciﬁed in that case report. Okun et al. reported a series of eight patients who developed multiple complications, including hemiplegia, homonymous visual ﬁeld deﬁcit, hand weakness, dysarthria, hypophonia, aphasia, arm and face numbness,

552

j.y.k. lee et al.

and pseudobulbar laughter [35]. These complications all developed after treatment at one Gamma Knife center, and it does not appear that the same rate of complications have developed at all radiosurgical centers. Nevertheless, the true incidence of complications after radiosurgical thalamotomy is not well-known. One particular explanation of the complications associated with radiosurgical thalamotomy may be related to the variability in lesion size. Animal studies in rodents and non-human primates have demonstrated the ability to create necrotic lesions in animal brain parenchyma [34, 37, 38]. Dose, volume, and time are the three key factors that determine the nature of the functional ablation [34]. Higher doses can create necrosis faster as can the use of a larger-volume collimator, although almost all movement disorder GK lesions employ a 4-mmdiameter collimator. This dose-volume-time response curve can roughly be correlated in humans; however, despite this apparent consistency in dose delivery and tissue response, Friehs et al. report variable size on follow-up imaging after GK radiosurgical ablation for functional procedures. Friehs et al. examined follow-up imaging after 140 Gamma Knife radiosurgical lesions produced in the treatment of patients with PD, pain, and ET [39]. In all cases, the 4-mm collimator was used for the Gamma Knife, but the postoperative scans demonstrated lesion sizes that ranged from undetectable to 4000 mm3. Friehs correlated the higher doses with greater lesion volume and suggested that doses of 160 Gy and above were associated with inordinately large volumes. Thus, although lesion size appears to be correlated with radiation dose, there appears to be individual patient to patient variability in response to Gamma Knife radiosurgical lesioning. In addition, Ohye reports two different types of lesions after radiosurgical lesions using the single 4-mm collimator isocenter with central doses of 130 Gy: “circumscribed round high signal area of approximately 7- to 8-mm diameter surrounding a smaller low signal area. The other is characterized by an irregular-shaped high signal zone extending into the medial thalamic area and/or internal capsule and often accompanied by streaking along the thalamocapsular border” [27, 31]. Ohye speculates that the delivery time may inﬂuence lesion size. Since the cobalt sources of the Gamma Knife need to be replaced over time, delivery of the same 143-Gy dose may require double the length of time with old sources compared with new sources [27]. “After reloading, the restricted lesion was more frequent and the lesion volume was smaller.” In conclusion, both DBS of the VIM or GK thalamotomy have been employed in the treatment of tremor. Several wellconducted trials have demonstrated the efﬁcacy and complication rate of DBS of the VIM thalamic nucleus in the treatment of both parkinsonian tremor and essential tremor. In contrast, large, well-conducted trials of Gamma Knife radiosurgical thalamotomy have not been performed. Thus, it is difﬁcult to compare directly the efﬁcacy and risks of GK thalamotomy of the VIM nucleus for the treatment of tremor. Nevertheless, lesioning of the thalamus is an alternative treatment for tremor, and Gamma Knife ablation may be effective when properly performed. Hence, the ﬁrst-line surgical therapy in the treatment of patients with medically refractory tremor is VIM DBS, and Gamma Knife thalamotomy remains an option for patients who cannot undergo an open surgical procedure. For example,

patients with severe disabling essential tremor and coexisting medical conditions such as a mechanical mitral valve necessitating anticoagulation may be candidates for Gamma Knife radiosurgical thalamotomy. Gamma Knife radiosurgery of the VIM thalamus, however, should only be performed by neurosurgeons who have experience in both deep-brain stimulation and radiofrequency thalamic procedures.

Pallidal DBS Versus Radiosurgical Pallidotomy Pallidal DBS Since its reintroduction by Laitinen and colleagues in 1992, many successful pallidotomies have been performed throughout the world in the treatment of Parkinson disease [40]. The most consistent effect of pallidal ablation appears to be the relief of contralateral l-dopa–induced dyskinesias, whereas the amount of improvement in contralateral bradykinesia, rigidity, and tremor differs among the many studies [41]. In contrast with the number of studies documenting the effects of ablation of the posteroventral globus pallidus interna (GPi), however, there are remarkably few studies that extend the initial observation of Siegfried and Lippitz that chronic highfrequency stimulation of the internal pallidum may effectively treat PD [42]. GPi DBS is currently not as common a procedure as STN DBS for the treatment of Parkinson disease. Despite the current enthusiasm for STN DBS for Parkinson disease, DBS of the internal segment of the globus pallidus can be quite effective as well. The most consistent effect produced by pallidal DBS is a marked reduction of contralateral l-dopa– induced dyskinesias [42–52]. Volkmann et al. demonstrated a 54% improvement in the “off” period UPDRS motor score at 1-year follow-up, with signiﬁcant improvement in bradykinesia, tremor, posture, and gait. “On” period motor symptoms did not improve signiﬁcantly after surgery except for dyskinesias, which were reduced by 83% at 1-year follow-up [52]. It appears that GPi DBS may directly reduce levodopa-induced dyskinesias, whereas STN DBS reduces levodopa-induced dyskinesias indirectly via reduction of overall medication intake. Loher et al. reported 1-year results of 16 patients showing a 38% improvement in medication-off UPDRS motor scores and a 33% improvement in the ADL score in patients receiving unilateral stimulation [53]. Bilateral stimulation led to a slight improvement in these results. Some improvement in the medication-on state was also noted. A larger series of 36 patients [54] demonstrated that bilateral pallidal stimulation results in a median motor improvement of 37% and an increase from 28% to 64% of the day without disabling involuntary movements [19]. Many reports demonstrate the beneﬁcial effects of GPi stimulation on dyskinesias, on-off ﬂuctuations, and tremor [47, 49, 52, 55, 56]. Also, pallidal stimulation has been validated in those patients who have previously undergone contralateral pallidotomy. In a cohort of four patients who underwent GPi stimulation contralateral to a prior pallidotomy, motor scores improved by almost 50% while bradykinesia was decreased by 37% and tremor by 93% without serious adverse cognitive or motor effects [57]. Despite the beneﬁts of GPi DBS, most centers currently favor STN DBS in the treatment of Parkinson disease.

56.

movement disorders: deep-brain stimulation perspective

Dystonia is one movement disorder whose primary surgical target is the GPi as opposed to the STN or VIM; however, the treatment of dystonia with DBS remains in its infancy. Unlike the rapid improvement seen in PD, GPi DBS in dystonia results only in gradual improvement over several months. Preliminary results suggest that more mobile, ﬂuid abnormal posturing may respond more quickly, whereas ﬁxed tonic abnormal posturing may be slower to improve [58–60]. Given these encouraging results, there have been more patients with both primary and secondary dystonia who have undergone bilateral GPi DBS. The results of these studies remain to be seen. The complication rate of GPi DBS is similar to that of VIM DBS. The risk of hemorrhage, infection, and hardware-related complications are similar. A particular consideration of GPi DBS, however, is the risk of cognitive deterioration. Although there are no direct studies that compare the incidence of cognitive complications after bilateral pallidal DBS, the overall reported complication rate after bilateral DBS does not appear to be as high as the rate after bilateral pallidal destructive lesions. Global scores of cognitive function exhibit little change after unilateral or bilateral pallidal stimulation, although subtle worsening of frontal lobe scores including verbal ﬂuency has been described [47, 52, 61, 62]. Risk factors for cognitive deterioration include preoperative l-dopa dosages and advanced age [61]. Despite the lack of direct comparison between the efﬁcacy of STN DBS versus GPi DBS, most neurologists/neurosurgeons have moved away from GPi DBS to STN DBS. On theoretical grounds, the STN appears to be positioned in such a manner as to inﬂuence both output nuclei of the basal ganglia, GPi, and substantia nigra pars reticulate (SNpr), and hence, this may represent the reason why most surgeons have opted for the STN as the ﬁnal target for DBS. In addition, another powerful inﬂuence may be the less consistent success rate after pallidal DBS and the lack of an incentive for groups who are achieving satisfactory results with STN DBS to change the target to GPi. Hence, the choice of GPi as a target for DBS at the present time remains limited primarily to patients with dystonia, although this too may be challenged in the future, as some groups are pursuing randomized, controlled trials to compare the beneﬁts of STN DBS versus GPi DBS in the treatment of patients with Parkinson disease.

Radiosurgical Pallidotomy The pallidotomy is a time-honored surgical procedure that was performed in the 1950s by many surgeons, including Guiot, Spiegel and Wycis, Talairach, Riechert, and Leksell. The reintroduction of Leksell’s posteroventral pallidotomy by Laitinen in 1992 (after its disappearance, which coincided with the beneﬁts of l-dopa therapy) initiated a resurgence of interest in the surgical treatment of movement disorders. One of the most consistent beneﬁts of radiofrequency pallidotomy is improvement in contralateral drug-induced dyskinesias. Nearly every study has documented improvement in l-dopa dyskinesias from 61% to 82% [63–71]. Tremor appears to respond by 33% to 90% at 6 months, although tremor alone is not an indication for pallidotomy. The results for the symptoms of parkinsonian rigidity and bradykinesia are not as consistent. Some studies demon-

553

strate no beneﬁt and others demonstrate some beneﬁt. In addition, there have only been minimal beneﬁts in gait. Given the results described above, conventional pallidotomy is only recommended for patients who are severely disabled by asymmetrical l-dopa dyskinesias and who are not candidates for implanted hardware systems. Bilateral pallidotomies are not considered a good option for most patients given the high rates of cognitive complications after bilateral destructive lesions in the globus pallidus [47]. Gamma Knife radiosurgical pallidotomy differs from conventional radiofrequency pallidotomy primarily in its less invasive nature. It can theoretically expect to provide excellent relief of levodopa-induced dyskinesias and tremor with fewer complications than conventional pallidotomies. The evidence supporting the use of Gamma Knife pallidotomy is even more sparse than the evidence supporting the use of Gamma Knife thalamotomy. Only case reports and case series are available for study. Indeed, the last published case series of patients treated with Gamma Knife pallidotomy dates back to 1998 [72], which most likely reﬂects the established success of STN DBS in the treatment of Parkinson disease, and the high complication rates of GK pallidotomy. Friedman et al. reported four patients in 1995 who were treated with 180 Gy to the right posteroventral pallidum for advanced PD [73]. One patient had improvement in dyskinesias but also became transiently psychotic and demented. The three other patients neither improved nor suffered a complication. In a later publication, Friedman et al. reported two patients who underwent radiosurgical pallidotomy with lower doses of 120 to 140 Gy [74]. They chose to perform the procedure only in patients who had undergone a prior radiofrequency pallidotomy with physiologic mapping. They created a mirror-image lesion, assuming symmetry between the two sides. Both of these patients, however, developed complications: hemiparesis in one patient and hypophonia and swallowing difﬁculties in the other patient. In addition, Friedman et al. reported a stroke in one patient as a complication from GK pallidotomy [75]. This was attributed to vascular hyalinization and thrombosis. The largest series of GK pallidotomies comes from Young et al. who reported a series of 29 patients who underwent Gamma Knife pallidotomy [72]. He reports that two thirds of his patients demonstrated improvements in bradykinesia and rigidity. Only one of these 29 patients developed a complication—hemianopsia 9 months after the procedure. UPDRS scores were not consistently evaluated by blinded observers in this study. With respect to complications, Okun et al. reported three complications after Gamma Knife pallidotomy performed in one center [35]. Three patients who had Gamma Knife pallidotomy had been treated with doses of 200 Gy, 150 Gy, and 100 Gy. All three experienced visual ﬁeld loss and/or hemiparesis and dysarthria. Compared with Gamma Knife thalamotomy, there appears to be a dearth of published case series of patients treated with GK pallidotomy for the treatment of Parkinson disease. This most likely represents lack of enthusiasm on the part of the treating surgeons and referring neurologists. There are few studies that truly document the beneﬁt of GK pallidotomy in a blinded and unbiased fashion, and there are several reports of complications after GK pallidotomy. Hence, GK pallidotomy cannot be recommended as a viable method of lesioning in the treatment of PD without further careful study.

554

j.y.k. lee et al.

FIGURE 56-1. T2-weighted MRI scan with a superimposed reformatted Schaltenbrand and Wahren atlas and bilateral electrode tracks for placement in the STN. Direct targeting serves to conﬁrm formula-derived coordinates.

Subthalamic DBS Versus Radiosurgical Subthalamotomy Subthalamic DBS At the present time, DBS of the subthalamic nucleus is the most widely practiced surgical treatment for patients with idiopathic Parkinson disease (Figs. 56-1, 56-2, and 56-3). Benabid et al. was the ﬁrst group to demonstrate the efﬁcacy of STN DBS in the treatment of PD in the mid-1990s [76]. One-year follow-up of 24 patients implanted with bilateral STN DBS demonstrated that UPDRS ADL and motor scores improved by 60% in the off-medication state [77]. The UPDRS subscores for akinesia (56%), tremor (80%), rigidity (68%), and gait (55%) also improved, in contrast with the results seen with VIM stimulation [5]. However, the effect in the on-medication state was not as pronounced, with only a 10% improvement in the motor score noted. Levodopa drug use was cut in half. These results have held up at 5-year follow-up [78]. Kumar et al. performed a prospective, randomized, double-blinded study comparing STN DBS with the stimulator on and off [79]. They found a 65% reduction in off-period motor UPDRS scores, a 40% reduction in on-period motor UPDRS scores, and an 85% reduction in l-dopa–induced dyskinesias in the seven patients who were evaluated. Limousin et al. found a 60% improvement in off-period UPDRS scores but only a 10% improvement in on-period UPDRS scores [77]. In summary, STN DBS is quite effective in the treatment of Parkinson disease. Levodoparesponsive motor features typically respond well to STN DBS with the possible exception of tremor. The axial features of PD such as dysarthria, postural instability, and freezing of gait may

FIGURE 56-2. Postoperative anteroposterior and lateral skull X-rays demonstrating the implanted STN DBS electrodes.

not reliably respond to STN DBS. Disease-related progression of these features over time may compromise overall postsurgical beneﬁt despite continuing response of limb motor function. Preoperative levodopa-responsiveness (by varying deﬁnitions) is often reported as a predictive factor for a positive response to surgery [24, 80–87]. A positive response to a levodopa challenge and correlation with good outcome may be more evident with STN DBS than with GPi DBS [88, 89]. The mean percent improvement in motor UPDRS during the levodopa test ranges from 40% to 70% in published studies. As the levodopa response is the main predictor of outcome after STN DBS [80, 82, 88, 90], the beneﬁt will be greatest in those patients who have a high off-drug score and a low on-drug score. In other words, the “best on” may be a better predictor for functional outcome than the numerical magnitude of the response. The hardware complications of STN DBS are similar to those listed in the section on VIM DBS as are the issues related to hemorrhage rates (e.g., approximately 3% hemorrhage rate with less than one third being symptomatic). Complications related to STN stimulation, however, are different from VIM or GPi stimulation. Involuntary movements can occur during the initial phase of STN stimulation but tend to be mild and generally respond to changes in the stimulation parameters [91, 92]. In a retrospective study, Volkmann noted that STN stimulation appeared to be associated with a higher incidence of adverse events compared with GPi stimulation at 1-year followup periods [93]. The STN DBS group had higher rates of

FIGURE 56-3. Postoperative axial and sagittal T1-weighted MRI scans showing the implanted STN DBS electrodes.

56.

movement disorders: deep-brain stimulation perspective

555

Table 56-1. Comparison of radiosurgery and deep-brain stimulation for movement disorders.

Experience Clinical results: tremor Clinical results: PD

Risk of hemorrhage Adjustable? Time to clinical effect? Reversibility? Repeatable? Targeting method Other complications

Radiosurgery

Radiofrequency lesioning

Deep-brain stimulation

40 years 24% to 70.6% tremor free 6% improvement in UPDRS III No signiﬁcant med change Dyskinesias improved 13% Very minimal No 1 to 12 months No No Imaging only

50 years Up to 85% tremor free 30% to 70% improvement in UPDRS III 30% decrease in meds Dyskinesias improved 70% 1% Lesion may be “sculpted” in OR Minutes to hours No Yes Imaging plus microelectrode recording and macrostimulation Lesion spread to adjacent structures Higher incidence of complications when performed bilaterally

20 years 55% to 95% tremor-free 60% improvement in UPDRS III 50% decrease in meds Dyskinesias improved 80% 1% to 3% Yes Minutes to hours Yes N/A Imaging plus microelectrode recording and macrostimulation Hardware infection/erosion Electrode breakage

Necrosis Lesion spread to adjacent structures

depression, anhedonia, hypophonia, dysarthria, and apraxia of eyelid opening. Nevertheless, the complication rates of STN DBS generally are mild and can be improved with changes in stimulator settings.

Parkinson’s Disease Disability Rating score had decreased from 28 to 11. She had no side effects. Although this is perhaps a promising case report, Gamma Knife subthalamotomy cannot be recommended as a safe option in patients with Parkinson disease at this time.

Radiosurgical Subthalamotomy Subthalamotomy in the form of ischemic strokes or hemorrhages have generally been known to cause hemiballism. However, despite this conventional teaching, there are case reports of parkinsonian patients with spontaneous STN hemorrhage who have experienced motor improvement without the occurrence of hemiballism [94, 95]. In addition, there are centers outside of the United States that have reported on the beneﬁts of therapeutic STN ablative surgery using traditional radiofrequency lesioning techniques. The Cuban group reported their results in 18 patients after bilateral dorsal subthalamotomy [96]. After a 16-month average follow-up period, there was a 58% improvement in UPDRS motor scores in the “off” period. Three patients developed severe generalized chorea, dysarthria, and balance instability that lasted 6 months from which they spontaneously recovered. The Bristol group reported their results on 50 subthalamotomies in 39 patients [97]. After unilateral STN lesions, the patients demonstrated a 46% improvement in the motor UPDRS score at 24 months. In addition, they were able to decrease their levodopa dose by 50%. With respect to complications, one patient developed an intracerebral hemorrhage, and one patient developed postoperative hemiballism on the side contralateral to the STN lesion that lasted for 3 weeks and then spontaneously disappeared. Hence, subthalamotomy via radiofrequency lesioning is a procedure in its infancy with approximately 50 patients in the reported literature. The results are promising, but safety and efﬁcacy remains to be proved. There is only a single case report in the literature with one patient treated with Gamma Knife subthalamotomy for Parkinson disease [98]. This one patient had a prior radiofrequency pallidotomy and then had a Gamma Knife procedure with a single 120-Gy, 4-mm shot to the subthalamic nucleus. Forty-two months after the procedure, her postoperative

Conclusion High-frequency electrical stimulation of basal ganglia targets has been proved to be effective and safe in the treatment of patients with Parkinson disease, essential tremor, and dystonia. There are numerous, well-conducted studies to this effect. In contrast, Gamma Knife radiosurgical lesioning procedures in the treatment of movement disorders have only limited documented beneﬁt and safety in the world literature (Table 56-1). At this time, the ﬁrst-line treatment for patients with medically refractory movement disorders is deep-brain stimulation, a therapy that can be titrated in the clinic to avoid side effects and to maximize clinical beneﬁt. Gamma Knife thalamotomy may be a treatment option in patients with essential tremor or tremor-predominant Parkinson disease who also have surgical contraindications such as coagulopathy or anatomic concerns such as multiple cavernous malformations. In addition, Gamma Knife radiosurgical lesioning for movement disorders should only be performed by neurosurgeons who are involved with deep-brain stimulation and thus are intimately familiar with the exquisite neuroanatomy of the basal ganglia. In contrast with Gamma Knife thalamotomy, Gamma Knife pallidotomy and subthalamotomy play only a limited role in the treatment of patients with movement disorders.

References 1. 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; 50:344–346. 2. Matsumoto K, Asano T, Baba T, Miyamoto T, Ohmoto T. Longterm follow-up results of bilateral thalamotomy for parkinsonism. Appl Neurophysiol 1976; 39:257–260.

556

j.y.k. lee et al.

3. Hirai T, Miyazaki M, Nakajima H, Shibazaki T, Ohye C. The correlation between tremor characteristics and the predicted volume of effective lesions in stereotaxic nucleus ventralis intermedius thalamotomy. Brain 1983; 106(Pt 4):1001–1018. 4. Benabid AL, Pollak P, Seigneuret E, Hoffmann D, Gay E, Perret J. Chronic VIM thalamic stimulation in Parkinson’s disease, essential tremor and extra-pyramidal dyskinesias. Acta Neurochir Suppl 1993; 58:39–44. 5. Benabid AL, Pollak P, Gao D, et al. Chronic electrical stimulation of the ventralis intermedius nucleus of the thalamus as a treatment of movement disorders.[comment]. J Neurosurg 1996; 84:203– 214. 6. Benabid AL, Pollak P, Gervason C, et al. Long-term suppression of tremor by chronic stimulation of the ventral intermediate thalamic nucleus. Lancet 1991; 337:403–406. 7. Pollak P, Benabid AL, Limousin P, Benazzouz A. Chronic intracerebral stimulation in Parkinson’s disease. Adv Neurol 1997; 74:213–220. 8. Koller W, Pahwa R, Busenbark K, et al. High-frequency unilateral thalamic stimulation in the treatment of essential and parkinsonian tremor. Ann Neurol 1997; 42:292–299. 9. Speelman JD, Bosch DA. [Continuous electric thalamus stimulation for the treatment of tremor resistant to pharmacotherapy]. Nederlands Tijdschrift voor Geneeskunde 1995; 139:926–930. 10. Alesch F, Pinter MM, Helscher RJ, Fertl L, Benabid AL, Koos WT. Stimulation of the ventral intermediate thalamic nucleus in tremor dominated Parkinson’s disease and essential tremor. Acta Neurochir 1995; 136:75–81. 11. Ondo W, Jankovic J, Schwartz K, Almaguer M, Simpson RK. Unilateral thalamic deep brain stimulation for refractory essential tremor and Parkinson’s disease tremor. Neurology 1998; 51: 1063–1069. 12. Kumar K, Kelly M, Toth C. Deep brain stimulation of the ventral intermediate nucleus of the thalamus for control of tremors in Parkinson’s disease and essential tremor. Stereotact Funct Neurosurg 1999; 72:47–61. 13. Hubble JP, Busenbark KL, Wilkinson S, et al. Effects of thalamic deep brain stimulation based on tremor type and diagnosis. Mov Disord 1997; 12:337–341. 14. Troster AI, Fields JA, Pahwa R, et al. Neuropsychological and quality of life outcome after thalamic stimulation for essential tremor. Neurology 1999; 53:1774–1780. 15. Limousin P, Speelman JD, Gielen F, Janssens M. Multicentre European study of thalamic stimulation in parkinsonian and essential tremor. J Neurol Neurosurg Psychiatry 1999; 66:289– 296. 16. Volkmann J, Herzog J, Kopper F, Deuschl G. Introduction to the programming of deep brain stimulators. Mov Disord 2002; 17: S181–187. 17. 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; 72:157–169. 18. Palur RS, Berk C, Schulzer M, Honey CR. A metaanalysis comparing the results of pallidotomy performed using microelectrode recording or macroelectrode stimulation.[comment]. J Neurosurg 2002; 96:1058–1062. 19. Deep-brain stimulation of the subthalamic nucleus or the pars interna of the globus pallidus in Parkinson’s disease. N Engl J Med 2001; 345:956–963. 20. Hariz M. Safety and risk of microelectrode recording in surgery for movement disorders. Stereotact Funct Neurosurg 2002; 78(3–4):146–157. 21. Binder DK, Rau GM, Starr PA. Risk factors for hemorrhage during microelectrode-guided deep brain stimulator implantation

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

for movement disorders. Neurosurgery 2005; 56:722–732; discussion 722–732. Oh MY, Abosch A, Kim SH, Lang AE, Lozano AM. Long-term hardware-related complications of deep brain stimulation. Neurosurgery 2002; 50:1268–1274; discussion 1274–1276. Kondziolka D, Whiting D, Germanwala A, Oh M. Hardwarerelated complications after placement of thalamic deep brain stimulator systems. Stereotact Funct Neurosurg 2002; 79:228–233. Benabid AL, Benazzouz A, Hoffmann D, Limousin P, Krack P, Pollak P. Long-term electrical inhibition of deep brain targets in movement disorders. Mov Disord 1998; 13:119–125. Tasker RR. Deep brain stimulation is preferable to thalamotomy for tremor suppression. Surg Neurol 1998; 49:145–153; discussion 153–154. Schuurman PR, Bosch DA, Bossuyt PM, et al. A comparison of continuous thalamic stimulation and thalamotomy for suppression of severe tremor.[comment]. N Engl J Med 2000; 342:461–468. Ohye C, Shibazaki T, Sato S. Gamma knife thalamotomy for movement disorders: evaluation of the thalamic lesion and clinical results. J Neurosurg 2005; 102(Suppl):234–240. Ohye C, Shibazaki T, Hirato M, Inoue H, Andou Y. Gamma thalamotomy for parkinsonian and other kinds of tremor. Stereotact Funct Neurosurg 1996; 66(Suppl 1):333–342. 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):120–127. Duma CM, Jacques D, Kopyov OV. The treatment of movement disorders using Gamma Knife stereotactic radiosurgery. Neurosurg Clin N Am 1999; 10:379–389. 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: 1044–1049. 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; 55:443–446. Young RF, Jacques S, Mark R, et al. Gamma knife thalamotomy for treatment of tremor: long-term results. J Neurosurg 2000; 93(Suppl 3):128–135. Kondziolka D, Lunsford LD, Claassen D, Maitz AH, Flickinger JC. Radiobiology of radiosurgery: part I. The normal rat brain model. Neurosurgery 1992; 31:271–279. Okun MS, Stover NP, Subramanian T, et al. Complications of gamma knife surgery for Parkinson disease. Arch Neurol 2001; 58:1995–2002. Siderowf A, Gollump SM, Stern MB, Baltuch GH, Riina HA. Emergence of complex, involuntary movements after gamma knife radiosurgery for essential tremor. Mov Disord 2001; 16:965–967. Andersson B, Larsson B, Leksell L, et al. Histopathology of late local radiolesions in the goat brain. Acta Radiol Ther Phys Biol 1970; 9:385–394. Larsson B, Leksell L, Rexed B, Sourander P, Mair W, Andersson B. The high-energy proton beam as a neurosurgical tool. Nature 1958; 182:1222–1223. Friehs GM, Noren G, Ohye C, et al. Lesion size following Gamma Knife treatment for functional disorders. Stereotact Funct Neurosurg 1996; 66(Suppl 1):320–328. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992; 76:53–61. Lang AE, Duff J, Saint-Cyr JA, et al. Posteroventral medial pallidotomy in Parkinson’s disease. J Neurol 1999; 246(Suppl 2): II28–41.

56.

movement disorders: deep-brain stimulation perspective

42. Siegfried J, Lippitz B. Bilateral chronic electrostimulation of ventroposterolateral pallidum: a new therapeutic approach for alleviating all parkinsonian symptoms. Neurosurgery 1994; 35:1126–1129; discussion 1129–1130. 43. Tronnier VM, Fogel W, Kronenbuerger M, Steinvorth S. Pallidal stimulation: an alternative to pallidotomy? J Neurosurg 1997; 87:700–705. 44. Durif F, Lemaire JJ, Debilly B, Dordain G. Long-term follow-up of globus pallidus chronic stimulation in advanced Parkinson’s disease. Mov Disord 2002; 17:803–807. 45. Krack P, Pollak P, Limousin P, et al. Opposite motor effects of pallidal stimulation in Parkinson’s disease. Ann Neurol 1998; 43:180–192. 46. Burchiel KJ, Anderson VC, Favre J, Hammerstad JP. Comparison of pallidal and subthalamic nucleus deep brain stimulation for advanced Parkinson’s disease: results of a randomized, blinded pilot study. Neurosurgery 1999; 45:1375–1382; discussion 1382– 1384. 47. Ghika J, Villemure JG, Fankhauser H, Favre J, Assal G, GhikaSchmid F. Efﬁciency and safety of bilateral contemporaneous pallidal stimulation (deep brain stimulation) in levodopa-responsive patients with Parkinson’s disease with severe motor ﬂuctuations: a 2-year follow-up review. J Neurosurg 1998; 89:713–718. 48. 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; 87:491–498. 49. 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. Neurology 1997; 49:249– 253. 50. 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: 73–82. 51. Merello M, Nouzeilles M, Kuzis G, et al. Unilateral radiofrequency lesion versus electrostimulation of posteroventral pallidum: a prospective randomized comparison. Mov Disord 1999; 14(1):50–56. 52. Volkmann J, Sturm V, Weiss P, et al. Bilateral high-frequency stimulation of the internal globus pallidus in advanced Parkinson’s disease. Ann Neurol 1998; 44:953–961. 53. Loher TJ, Burgunder JM, Pohle T, Weber S, Sommerhalder R, Krauss JK. Long-term pallidal deep brain stimulation in patients with advanced Parkinson disease: 1-year follow-up study. J Neurosurg 2002; 96:844–853. 54. Group DBSfPsDS. Deep-brain stimulation of the subthalamic nucleus or the pars interna of the globus pallidus in Parkinson’s disease. N Engl J Med 2001; 345:956–963. 55. Adler CH, Kumar R. Pharmacological and surgical options for the treatment of cervical dystonia. Neurology 2000; 55:S9–14. 56. Peppe A, Pierantozzi M, Altibrandi MG, et al. Bilateral GPi DBS is useful to reduce abnormal involuntary movements in advanced Parkinson’s disease patients, but its action is related to modality and site of stimulation. Eur J Neurol 2001; 8:579–586. 57. Galvez-Jimenez N, Lozano A, Tasker R, Duff J, Hutchison W, Lang AE. Pallidal stimulation in Parkinson’s disease patients with a prior unilateral pallidotomy. Can J Neurol Sci 1998; 25:300– 305. 58. Coubes P, Cif L, El Fertit H, et al. Electrical stimulation of the globus pallidus internus in patients with primary generalized dystonia: long-term results. J Neurosurg 2004; 101:189–194. 59. Coubes P, Roubertie A, Vayssiere N, Hemm S, Echenne B. Treatment of DYT1-generalised dystonia by stimulation of the internal globus pallidus. Lancet 2000; 355:2220–2221.

557

60. Kumar R, Dagher A, Hutchison WD, Lang AE, Lozano AM. Globus pallidus deep brain stimulation for generalized dystonia: clinical and PET investigation. Neurology 1999; 53:871–874. 61. Vingerhoets G, van der Linden C, Lannoo E, et al. Cognitive outcome after unilateral pallidal stimulation in Parkinson’s disease. J Neurol Neurosurg Psychiatry 1999; 66:297–304. 62. Troster AI, Fields JA, Wilkinson SB, et al. Unilateral pallidal stimulation for Parkinson’s disease: neurobehavioral functioning before and 3 months after electrode implantation. Neurology 1997; 49:1078–1083. 63. Johansson F, Malm J, Nordh E, Hariz M. Usefulness of pallidotomy in advanced Parkinson’s disease. J Neurol Neurosurg Psychiatry 1997; 62:125–132. 64. Kishore A, Turnbull IM, Snow BJ, et al. Efﬁcacy, stability and predictors of outcome of pallidotomy for Parkinson’s disease. Sixmonth follow-up with additional 1-year observations. Brain 1997; 120(Pt 5):729–737. 65. 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. 66. Masterman D, DeSalles A, Baloh RW, et al. Motor, cognitive, and behavioral performance following unilateral ventroposterior pallidotomy for Parkinson disease. Arch Neurol 1998; 55:1201– 1208. 67. Lang AE, Lozano AM, Montgomery E, Duff J, Tasker R, Hutchinson W. Posteroventral medial pallidotomy in advanced Parkinson’s disease. N Engl J Med 1997; 337:1036–1042. 68. Vitek JL, Bakay RA, DeLong MR. Microelectrode-guided pallidotomy for medically intractable Parkinson’s disease. Adv Neurol 1997; 74:183–198. 69. 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. Neurosurgery 1995; 36:1118–1125; discussion 1125–1127. 70. Baron MS, Vitek JL, Bakay RA, et al. Treatment of advanced Parkinson’s disease by posterior GPi pallidotomy: 1-year results of a pilot study. Ann Neurol 1996; 40:355–366. 71. Baron MS, Vitek JL, Bakay RA, et al. Treatment of advanced Parkinson’s disease by unilateral posterior GPi pallidotomy: 4year results of a pilot study. Mov Disord 2000; 15:230–237. 72. Young RF, Vermeulen S, Posewitz A, Shumway-Cook A. Pallidotomy with the gamma knife: a positive experience. Stereotact Funct Neurosurg 1998; 70(Suppl 1):218–228. 73. Friedman JH, Epstein M, Sanes JN, et al. Gamma knife pallidotomy in advanced Parkinson’s disease. Ann Neurol 1996; 39:535– 538. 74. 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; 212:143–150. 75. Friedman JH, Fernandez HH, Sikirica M, Stopa E, Friehs G. Stroke induced by gamma knife pallidotomy: autopsy result. Neurology 2002; 58:1695–1697. 76. Benabid AL, Pollak P, Gross C, et al. Acute and long-term effects of subthalamic nucleus stimulation in Parkinson’s disease. Stereotact Funct Neurosurg 1994; 62:76–84. 77. Limousin P, Krack P, Pollak P, et al. Electrical stimulation of the subthalamic nucleus in advanced Parkinson’s disease. N Engl J Med 1998; 339:1105–1111. 78. Krack P, Batir A, Van Blercom N, et al. Five-year follow-up of bilateral stimulation of the subthalamic nucleus in advanced Parkinson’s disease. N Engl J Med 2003; 349:1925–1934. 79. Kumar R, Lozano AM, Kim YJ, et al. Double-blind evaluation of subthalamic nucleus deep brain stimulation in advanced Parkinson’s disease. Neurology 1998; 51:850–855.

558

j.y.k. lee et al.

80. Welter ML, Houeto JL, Tezenas du Montcel S, et al. Clinical predictive factors of subthalamic stimulation in Parkinson’s disease. Brain 2002; 125:575–583. 81. Simuni T, Jaggi JL, Mulholland H, et al. Bilateral stimulation of the subthalamic nucleus in patients with Parkinson disease: a study of efﬁcacy and safety. J Neurosurg 2002; 96:666–672. 82. Charles PD, Van Blercom N, Krack P, et al. Predictors of effective bilateral subthalamic nucleus stimulation for PD. Neurology 2002; 59:932–934. 83. Kleiner-Fisman G, Fisman DN, Sime E, Saint-Cyr JA, Lozano AM, Lang AE. Long-term follow up of bilateral deep brain stimulation of the subthalamic nucleus in patients with advanced Parkinson disease. J Neurosurg 2003; 99:489–495. 84. Pahwa R, Wilkinson SB, Overman J, Lyons KE. Bilateral subthalamic stimulation in patients with Parkinson disease: long-term follow up. J Neurosurg 2003; 99:71–77. 85. Bejjani BP, Gervais D, Arnulf I, et al. Axial parkinsonian symptoms can be improved: the role of levodopa and bilateral subthalamic stimulation. J Neurol Neurosurg Psychiatry 2000; 68:595– 600. 86. Deuschl G, Fogel W, Hahne M, et al. Deep-brain stimulation for Parkinson’s disease. J Neurol 2002; 249:III/36–39. 87. Vingerhoets FJ, Villemure JG, Temperli P, Pollo C, Pralong E, Ghika J. Subthalamic DBS replaces levodopa in Parkinson’s disease: two-year follow-up.[comment]. Neurology 2002; 58:396– 401. 88. Krack P, Pollak P, Limousin P, et al. Subthalamic nucleus or internal pallidal stimulation in young onset Parkinson’s disease. Brain 1998; 121:451–457. 89. Krause M, Fogel W, Heck A, et al. Deep brain stimulation for the treatment of Parkinson’s disease: subthalamic nucleus versus

90.

91.

92.

93.

94.

95. 96.

97. 98.

globus pallidus internus. J Neurol Neurosurg Psychiatry 2001; 70: 464–470. Pinter MM, Alesch F, Murg M, Helscher RJ, Binder H. Apomorphine test: a predictor for motor responsiveness to deep brain stimulation of the subthalamic nucleus. J Neurol 1999; 246:907– 913. Limousin P, Pollak P, Hoffmann D, Benazzouz A, Perret JE, Benabid AL. Abnormal involuntary movements induced by subthalamic nucleus stimulation in parkinsonian patients. Mov Disord 1996; 11:231–235. Kumar R, Lozano AM, Sime E, Halket E, Lang AE. Comparative effects of unilateral and bilateral subthalamic nucleus deep brain stimulation. Neurology 1999; 53:561–566. Volkmann J, Allert N, Voges J, Weiss PH, Freund HJ, Sturm V. Safety and efﬁcacy of pallidal or subthalamic nucleus stimulation in advanced PD. [Erratum appears in Neurology 2001 Oct 9;57(7):1354.] Neurology 2001; 56:548–551. Sellal F, Hirsch E, Lisovoski F, Mutschler V, Collard M, Marescaux C. Contralateral disappearance of parkinsonian signs after subthalamic hematoma. Neurology 1992; 42:255–256. Vidakovic A, Dragasevic N, Kostic VS. Hemiballism: report of 25 cases. J Neurol Neurosurg Psychiatry 1994; 57:945–949. Alvarez L, Macias R, Lopez G, et al. Bilateral subthalamotomy in Parkinson’s disease: initial and long-term response. Brain 2005; 128:570–583. Gill SS, Heywood P. Bilateral dorsolateral subthalamotomy for advanced Parkinson’s disease. Lancet 1997; 350:1224. Keep MF, Mastrofrancesco L, Erdman D, Murphy B, Ashby LS. Gamma knife subthalamotomy for Parkinson disease: the subthalamic nucleus as a new radiosurgical target. Case report. J Neurosurg 2002; 97:592–599.

5 7

Movement Disorder: Medical Perspective Sangjin Oh and William J. Weiner

Introduction Movement disorders encompasses a wide range of neurologic diseases and syndromes that can be characterized by whether or not the motor manifestations are primarily hypokinetic or hyperkinetic. The hypokinetic disorders include such common syndromes as parkinsonism including Parkinson disease, multiple system atrophy (MSA), progressive supranuclear palsy (PSP), and various other syndromes that clinically manifest bradykinesia, muscle rigidity, and gait impairment. The hyperkinetic disorders characterize a wide range of syndromes and diseases including Huntington disease, dystonia, Tourette syndrome, essential tremor, myoclonus, and tardive dyskinesia. Each of the hyperkinetic syndromes clinically manifests either as tremor, chorea, tics, myoclonus, or dystonia. Clinical descriptions of these disorders can be found in a wide variety of references [1, 2]. This chapter will focus on the advantages and disadvantages of radiosurgery versus medical treatment of movement disorders. Radiosurgery has been used most extensively to treat Parkinson disease and essential tremor. The chapter will deﬁne the clinical syndromes of Parkinson disease and essential tremor, discuss the medical therapy that is available, and ﬁnally discuss the advantages and disadvantages of radiosurgery in these two disorders.

Parkinson Disease Parkinson disease is the most common identiﬁable syndrome that can be diagnosed in the context of parkinsonism. The most important clues to the diagnosis of Parkinson disease include, of course, elements of parkinsonism (usually two of three of the cardinal features: bradykinesia, resting tremor, cogwheel rigidity). These begin in an insidious, asymmetric fashion. To ﬁt the Parkinson disease diagnosis, the symptomatology has to be progressive. Many movement disorders specialists also require a history of sustained levodopa responsiveness. The presence of all these features will help ensure an accurate clinical diagnosis in most patients who present with parkinsonism [3]. On the other hand, none of these features, including unilateral onset and levodopa responsiveness, can deﬁnitely exclude other parkinsonian syndromes. Quite the contrary: unilateral

presentations are present in Parkinson disease in only about 70% of patients, and levodopa responsiveness can be seen in a wide variety of parkinsonian syndromes. Nonetheless, these clues to the diagnosis of Parkinson disease can be quite valuable clinically [4]. The insidious nature of the onset of Parkinson disease was ﬁrst recognized by James Parkinson in 1817. In An Essay on the Shaking Palsy [5], he wrote, “So slight and nearly imperceptible are the ﬁrst inroads of this malady, and so extremely slow is its progress, that it rarely happens, that the patient can form any recollection of the precise period of its commencement.” In the early evolution of the signs and symptoms of Parkinson disease, there is often very little functional impairment, and patients often choose to not take any symptomatic therapy until the evolving motor disorder begins to impinge on some aspect of their lives. Early treatment of Parkinson disease with either levodopa/carbidopa or one of the dopamine receptor agonists is very effective in alleviating motor symptomatology. This is referred to as “the honeymoon” period during which time the patient’s symptomatic medication works extremely well and there is very little in the way of adverse events. There are a wide range of therapeutic choices that are extremely effective in early Parkinson disease, and it would be highly unlikely that surgical intervention would ever be required in these patients. As Parkinson disease advances, the wide range of pharmacologic agents that are available to treat this disorder begin to become less effective. The most common problems that occur when patients have had the disorder for greater than 5 years are motor ﬂuctuations and dyskinesias. Motor ﬂuctuations consist of an erratic effect of drugs such as levodopa on the motor symptomatology. Patients begin to notice that 20 to 30 minutes after a dose of levodopa is taken, they feel stronger, their tremor lessens, and they walk better. They also begin to notice that 3½ hours after the ingestion of the last dose, the medication begins to wear off. Patients notice that their tremor begins to reemerge; they become slower and stiffer and have more difﬁculty carrying out the activities of daily living. Neurologists often refer to the period of time when the medication is producing its best effect as “on” and the time when the medication is working less effectively as “off” time. Patients began to experience more and more off time in the course of their day as the disease evolves. The other major complication that can be limiting in terms of therapeutic choices

559

560

s. oh and w.j. weiner

is the emergence of dyskinesias. Dyskinesia describes a wide range of involuntary movements (usually choreo-dystonic) other than tremor that begin to appear in patients with longstanding Parkinson disease who have been treated with dopaminergic agents. Dyskinesias are choreodystonic movements that occur in the limbs, axial regions, and can also cause lingual facial buccal movements or mouthing and chewing movements. Dyskinesias are precipitated by the total dopaminergic dose of medication the patient is receiving. Levodopa/carbidopa and the dopamine receptor agonists are the most potent drugs that induce dyskinesias. However, the MAO and COMT inhibitors can also induce dyskinesias. The therapeutic problem is that a patient’s increasing “off” time is an indication for the need for increased dopaminergic medication, but the appearance of increasingly severe dyskinesia is a sign that the dopaminergic medication needs to be decreased. It is in these patients that surgical options are often discussed. Surgical treatment options in patients with Parkinson disease are currently discussed when the major medical treatment parameters have been exhausted. This includes the proper dosing and use of levodopa/carbidopa in combination with dopamine receptor agonists, MAO and COMT inhibitors, and amantadine. Deep-brain stimulation (DBS) is the surgical modality that has become the treatment of choice for patients who have intractable medical problems related to their Parkinson disease. Deep-brain stimulation has been demonstrated to be an effective therapy for properly selected patients with Parkinson disease [6]. The timing of surgery and the selection of patients for surgery has become an intensely researched topic [7]. The timing of surgical procedures in Parkinson disease has traditionally been relegated to those patients who have essentially failed conventional medical therapy. There is growing interest in the concept that surgical treatment should be considered at an earlier stage in Parkinson disease before the patient’s disability and functional impairment becomes too advanced. This approach would still not include early patients who are doing extremely well on their medication. At the present time, patients with Parkinson disease who are suitable candidates for a surgical procedure include those patients who retain a good response to levodopa even for a short period of time, have no medical contraindications for surgery, have no signiﬁcant cognitive impairment, have no comorbid, serious psychiatric diagnoses, and have realistic expectations about the effects of surgery on their chronic progressive neurodegenerative disease [7, 8]. It is important for patients to understand that the surgical approach to Parkinson disease is palliative and will not result in a “cure.” In addition, patients must understand that the best response to surgery will result in more “on” time, but they will not be any better than their current best response to levodopa. This is a very important point, and patients must understand that their current best levodopa response in terms of the quality of their “on” time is the limit of what surgery can achieve. The purpose of the surgery with regard to motor ﬂuctuations is to allow the patient to have more good “on” time. However, the quality of the “on” time in terms of motor performance will be no better than the best levodopa-induced “on” time.

When the target is the subthalamic nucleus (STN), DBS often results in a reduction of the total amount of dopaminergic medication that is required to maintain a good motor response. It is probably this reduction in dopaminergic medication that results in the striking amelioration of dyskinesias in many patients with Parkinson disease who previously had moderate to severe dyskinetic movements. When the target is the globus pallidus interna in DBS surgery, dyskinesias are often not improved. DBS procedures have been performed with increasing frequency, and the literature and reports concerning the effects and complications related to DBS are far more documented and scrutinized than the results of radiosurgery in the same disorders [9–11]. Ablative surgery in Parkinson disease has a very long history [12–18]. Unfortunately, the role of lesion making and alleviation of the motor symptomatology of Parkinson disease has a very mixed history of enthusiasm for new lesion techniques followed by unrealistic expectations and the ultimate determination that the complication rate and beneﬁt from most of the previous lesion techniques was unwarranted. The enthusiasm for the surgical approach of DBS initially was related to its apparent effectiveness without the need to make an actual lesion in the central nervous system. The mechanism through which DBS alleviates motor symptoms in Parkinson disease is not understood. However, what has become apparent is that proper location of stimulating electrodes within a target whether it is the STN or the globus pallidus interna is extremely important. It has been repeatedly demonstrated that movement or misplacement of the electrodes by as little as 1 to 2 mm can make a tremendous difference in terms of the therapeutic beneﬁt the patient achieves. This is obviously very important because one of the problems with radiosurgery in terms of treatment of movement disorders is that the size and location of the lesion cannot be as exquisitely controlled. Patients with Parkinson disease should be treated with pharmacologic agents when signiﬁcant functional impairment begins. There is a wide variety of medical treatments available that provide many years of excellent symptomatic relief of the motor symptomatology of Parkinson disease. When the patient begins to develop motor ﬂuctuations and/or dyskinesias, surgical options can be considered. The most widely used surgical option at this time is DBS of the STN, which has been extensively studied. Radiosurgery lesioning in the treatment of parkinsonism must not only be compared with medical therapy but also with DBS. Radiosurgery lesioning in Parkinson disease was reviewed in the previous chapter, and because of the limited number of patients that have had this procedure, the nature of the studies, and serious adverse events that have been published, the difﬁculty with control of lesion size and location and the delay between treatment and therapeutic outcome currently indicate that if a surgical option is to be employed, DBS or a conventional ablative procedure is the better choice. Radiosurgery lesioning in Parkinson disease should be considered in patients who are too frail to undergo DBS procedures or who have severe complicating medical problems that would interfere with the placement of the electrodes.

57.

movement disorder: medical perspective

Essential Tremor Essential tremor (ET) is the second movement disorder that will be considered in this chapter. Essential tremor is probably the most common movement disorder. It is deﬁned by the presence of a postural and kinetic tremor primarily in the upper extremities. It usually has a symmetric onset, and often there is a family history of tremor. In 75% of patients with essential tremor, the tremor is alcohol responsive. This is an interesting feature that can be elicited by the history that a single glass of wine or cocktail will markedly alleviate the tremor for 30 to 45 minutes. The tremor is slowly progressive over decades but can eventually result in functional impairment related to holding glasses, eating soup, and buttoning clothes and results in a large scribbling handwriting [1, 2]. The medical treatment of essential tremor has been extensively reviewed [19]. Currently, propranolol (a nonselective beta-blocker) is the only medication approved by the U.S. Food and Drug Administration (FDA). However, prospective, randomized clinical trials have indicated that primidone, an anticonvulsant, is as effective in treating limb tremor as propranolol [19]. Both propranolol and primidone show a mean reduction in tremor magnitude by 50% and are considered ﬁrst-line therapy for ET. In addition, there are many other medications that can be used to treat tremor in patients not responsive to propranolol or primidone. These medications include alprazolam, atenolol, gabapentin, sotalol, and topiramate. In addition, limited studies have shown that clonazepam, clozapine, nadolol, and nimodipine may be effective in reducing tremor. In those patients who have severe tremor that results in disability and who do not respond to medical therapy, surgical options are available. The most common procedure performed today is DBS with the target being the ventralis intermedius (VIM) nucleus of the thalamus [20–22]. This procedure often results in marked and rather immediate control of the tremor. Thalamotomy has also been performed in a conventional method. However, bilateral thalamotomy often results in unacceptable side effects including speech dysfunction. DBS can be targeted to either one or both sides and does not result in as many adverse events as conventional ablative procedures in the thalamus. In addition, if the second stimulator is turned on and adverse events result, it can be turned off and those adverse events will dissipate. In the previous chapter, we have also reviewed the literature regarding radiosurgery lesioning in the thalamus and its effect on essential tremor. Unfortunately, the same criticism applies to these reports describing the results of radiosurgery lesioning in ET as mentioned previously for Parkinson disease. For the most part, the reports are uncontrolled with no rigorous studies performed. The same problems with lesion location, size, and time to effect after the radiosurgery lesion exist. At this time, radiosurgery lesioning for patients with essential tremor should be limited to those who are not only medically refractory but who are also too frail to undergo conventional ablative procedures or stimulating electrode placement in the thalamus.

561

References 1. Weiner WJ, Lang AE. Movement Disorders: A Comprehensive Survey. Mount Kisco, NY: Futura Publishing Company, 1989. 2. Watts RL, Koller WC. Movement Disorders: Neurologic Principles and Practice, 2nd ed. New York: McGraw-Hill, 2004. 3. Weiner WJ. A differential diagnosis of parkinsonism. Rev Neurol Dis 2005; 2(3):124–131. 4. Utti RJ, Baba Y, Whaley NR, et al. Parkinson disease: handedness predicts asymmetry. Neurology 2005; 64:1925–1930. 5. Parkinson J. An Essay on the Shaking Palsy. London: Sherwood, Neely and Jones, 1817. 6. Krack P, Batir A, Van Blercom N, et al. Five-year follow-up of bilateral stimulation of the subthalamic nucleus in advanced Parkinson’s disease. N Engl J Med 2003; 349:1925–1934. 7. Pahwa R, Factor SA, Lyons KE, et al. Practice Parameter: Treatment of Parkinson disease with motor ﬂuctuations and dyskinesia (an evidence-based review). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2006 2; [Epub ahead of print]. 8. Welter ML, Houeto JL, Tezenas du Montcel S, et al. Clinical predictive factors of subthalamic stimulation in Parkinson’s disease. Brain 2002; 125:575–583. 9. Beric A, Kelly PJ, Rezai A, et al. Complications of deep brain stimulation surgery. Stereotact Funct Neurosurg 2001; 77:73–78. 10. Oh MY, Abosch A, Kim SH, et al. Long-term hardware-related complications of deep brain stimulation. Neurosurgery 2002; 50:1268–1274. 11. Lyons KE, Wilkinson SB, Overman J, Pahwa R. Surgical and hardware complications of subthalamic stimulation: a series of 160 procedures. Neurology 2004; 63:612–616. 12. Putnam TJ. Treatment of unilateral paralysis agitans by section of the lateral pyramidal tract. Arch Neurol Psychiatr 1940; 44: 950. 13. Bucy JC. Cortical extirpation in the treatment of involuntary movement. Arch Neurol Psychiatr 1942; 21:551. 14. Meyers R. Surgical interruption of the pallidofugal ﬁbres: its effect on the syndrome paralysis agitans and technical considerations in its application. N Y State J Med 1942; 42:317–325. 15. Meyers R. The modiﬁcation of alternating tremors, rigidity and festination by surgery of the basal ganglia. Assoc Nerv Ment Dis 1942; 20:602–665. 16. Cooper IS. Litigation of the anterior choroidal artery for involuntary movements of parkinsonism. Arch Neurol 1956; 75:36– 48. 17. Svennilson E, Torvik A, Lowe R, Leksell L. Treatment of parkinsonism by stereotactic thermo lesions in the pallidal region. A clinical evaluation of 81 cases. Acta Psychiatr Neurol Scand 1960; 35:358–377. 18. Hassler R, Reichert T. Indikationen und lokalisations: methode de gezeilten hirnoperationen. Nervenarzt 1954; 25:441–447. 19. Zesiewicz TA, Elble R, Louis ED, et al. Practice parameter: therapies for essential tremor: report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2005; 64:2008–2020. 20. Pahwa R, Lyons K, Koller WC. Surgical treatment of essential tremor. Neurology 2000; 54(Suppl 4):S39–S44. 21. Benabid AL, Pollak P, Gao D, et al. Chronic electrical stimulation of the ventralis intermedius nucleus of the thalamus as a treatment of movement disorders. J Neurosurg 1996; 84:203–214. 22. Koller W, Pahwa R, Busenbark K, et al. High-frequency unilateral thalamic stimulation in the treatment of essential and parkinsonian tremor. Ann Neurol 1997; 42:292–299.

5 8

Psychiatric and Pain Disorders Jason Sheehan, Nader Pouratian, and Charles Sansur

Introduction Despite great advances in the medical management of pain and psychiatric disease, there remains a fraction of patients who are refractory to available medical therapies. For example, despite therapeutic progress in recent years, conventional treatment of anxiety disorders fails or has only a temporary effect in 20% of patients. Pain and psychiatric disorders are often severely disabling and are associated with rates of suicide comparable with those of depression. When medical therapy fails, options include open surgical procedures, consisting of lesioning procedures or implantation of deep-brain stimulators, and radiosurgery. Radiosurgical options are well suited for functional neurosurgery, in which the surgeon strives to create a focused lesion without injuring or affecting adjacent structures. The Gamma Knife was developed in the late 1960s as an alternative to open stereotactic lesioning for functional disorders. With the development of the Gamma Knife and a limited understanding of the neurophysiology of pain and psychiatric disorders, Dr. Leksell was obsessively interested in the treatment of patients with intractable pain and psychosurgery [1–4]. In his book Brain Fragments, Dr. Leksell wrote, “One can accept death, but one cannot accept the deep, devastating pain. Sharp, intractable pain is like hell ‘without escape, without hope and without Heliotrope when Venom burns.’ Standing at the bedside without ever having experienced pain, it is impossible to imagine the patient’s agony, and it is impossible to understand that a short time without pain can be extreme happiness” [2].

technique, Leksell would try it in the management of both intractable pain and tic douloureux. In fact, Leksell designed the radiation pattern of the ﬁrst Gamma Knife to be lens shaped so that its effect would be similar to that of a knife blade cutting pain ﬁbers. Leksell wanted to prove that the Gamma Knife could selectively ablate pain transmitted by C ﬁbers. In Brain Fragments, Leksell described the treatment of a 33-year-old accountant with facial dolorosa secondary to basal cell cancer of the tongue using the Gamma Knife. “The treatment lasts 1 hr, and the dose given is 15,000 rad. Meanwhile we eat sandwiches and drink Pilsner in the control room. When the radiation treatment is ﬁnished, the pain has already diminished” [2].

Radiosurgery and Trigeminal Neuralgia With the Gamma Knife unit installed in Stockholm, he asked Dr. Håkanson to ﬁnd two patients he previously treated with orthovoltage for trigeminal neuralgia. After the successful long-term outcome of these two patients, Dr. Leksell and Dr. Håkanson treated 48 patients between 1970 and 1978, using plain stereotactic skull X-rays and transoval cisternography using tantalum dust for stereotactic targeting [5].

Radiosurgery and Psychiatric Disorders Leksell also ﬁrst described psychoradiosurgery, in which he targeted the frontolimbic connections in both anterior internal capsules (capsulotomy) for selected cases of intractable anxiety and severe central pain [6]. Leksell ﬁrst employed the Gamma Knife to treat some psychiatric disorders.

Radiosurgery and Somatic Pain Leksell had a special interest in pain disorders likely in part due to the fact that a person close to him died a painful death. He focused upon acute pain associated with cancer and trigeminal neuralgia. Leksell did not really explore treatment for chronic pain. When Leksell started to treat acute pain, the understanding of the pathophysiology was scarce. Therefore, the results were generally discouraging. He was, at times, overly optimistic about the Gamma Knife results for pain syndromes. As soon as he developed a new

Disease Pathophysiology and Radiosurgical Targeting for Pain and Psychiatric Disorders The radiobiology of treating these disorders differs from other disease entities in that the targeted area is often a normal brain structure, such as the thalamus, pituitary gland, or internal capsule. The radiosurgical treatment of pain and psychiatric disease relies heavily on the concept that adjacent normal

563

564

j. sheehan et al.

structures receive a markedly lower dose (due to rapid radiation dose fall-off) and a lower dose rate (or rate of radiation administration). A higher dose rate (larger dose in an equivalent amount of time or same total dose applied over a shorter period of time) increases the lethality of the dose to the target due to greater interference with intrinsic cellular repair mechanisms during irradiation. The signiﬁcance of this effect is seen most clearly at a threshold dose rate of 1 Gy/min [7]. In order to understand the radiobiology of a single high dose of radiation on normal brain, the effect of Gamma Knife Surgery (GKS) on the normal parietal lobe of rats was studied. A dose of 50 Gy caused astrocytic swelling and ﬁbrin deposition in capillary walls without changes in neuronal morphology or breakdown of the blood-brain barrier at 12 months. At 75 Gy, more vigorous morphologic changes were seen in astrocytes within 4 months. In addition, necrosis, breakdown of the blood-brain barrier, and hemispheric swelling were noted. At 120 Gy, astrocytic swelling occurred within 1 week of irradiation, and necrosis was seen at 4 weeks, but it was not associated with hemispheric swelling [8]. Unlike treating a tumor or arteriovenous malformation, there is not a pathologic target for functional radiosurgery. Effective radiosurgical management of pain and psychiatric disease depends upon selecting appropriate targets. This, in part, relies on an understanding of the anatomic circuitry and pathophysiologic mechanisms underlying the functional disorder. The organization of the limbic system is the basis of

Coronal plane of anterior tip of lateral ventricle Ant. commissure Post. commissure Cingulate gyrus

Hippocampal gyrus Corpus callosum Orbital cortex

Uncus

CORONAL PLANE OF: BIMEDIAL LEUCOTOMY (FALCONER-SCHURR, JACKSON) INFERIOR-MEDIAL LEUCOTOMY (BAKER et al) Cingulotomy: (BALLANTINE et al) (MEYER et al) SITE OF LEUCOTOMY: (BAILEY et al) SITE OF 90Y LESION OF KNIGHT (IN SUBCOSTICAL WHITE MATTER) CONNECTIONS ORBITAL CORTEX TO UNCUS (VAN HOESEN et al) RECIPROCAL CONNECTIONS FRONTAL GRANULAR CORTEX

GYRUS

CINGULI (NAUTA) ANTERIOR CAPSULOTOMY (LEKSELL et al)

FIGURE 58-2. Neuroanatomic sites for various psychosurgeries are illustrated. (From Mindus P. Capsulotomy in Anxiety Disorders—A Multidisciplinary Study. Dissertation at the Karolinska Institute and Hospital, Stockholm, Sweden, 1991. Used with permission.) 4 6

CINGULUM BUNDLE

8

understanding for pain and psychoneurosurgery (Fig. 58-1) [9]. Neuroanatomic sites for various pain and psychosurgeries are depicted in Figure 58-2 [9].

9 23

24

CC ATR

ST

SCR 25 10

SE 12

P

FORNIX

AT

M TU

SM

HPT

13

MFB

DB

14

HYP UB OB

INS

OT AMYG

Pain

HAB

MTT

MB

IP

PIT HIPPOCAMPUS

CG LMA

PRESUBICULUM

G

HIPPOCAMPAL GYRUS

SOMATIC AND VISCERAL AFFERENTS

FIGURE 58-1. Limbic system: OB, olfactory bulb; LOT, lateral olfactory striae; INS, insula; UB, uncinate bundle; DB, diagonal band of Broca; AMYG, amygdala; SCR, subcallosal radiations; HYP, hypothalamus; AT, anterior thalamus; MB, mammillary body; MTT, mammillothalamic tract; ATR, anterior thalamic radiations; ST, stria terminalis; HAB, habenula; MFB, medial forebrain bundle; SM, stria medullaris; HPT, habenulointerpeduncular tract; IP, interpeduncular nucleus; LMA, limbic midbrain area; G, nucleus of Gudden; CG, central gray; and CC, corpus callosum. (From Mindus P. Capsulotomy in Anxiety Disorders—A Multidisciplinary Study. Dissertation at the Karolinska Institute and Hospital, Stockholm, Sweden, 1991. Used with permission.)

Radiosurgery has been used to treat different forms of pain including somatic pain (i.e., cancer related or non–cancer related), central pain, trigeminal neuralgia, and sphenopalatine neuralgia. All of these types of pain are a manifestation of different pathophysiologic processes. There is not a single anatomic circuit or target that is most appropriate. Treatment can target central anatomic structures that mediate the pain sensation (e.g., thalamus) or can target the nerve that carries the pain sensation (e.g., trigeminal nerve in trigeminal neuralgia). The lack of anatomic and pathophysiologic background knowledge of the mechanisms of pain makes management of pain by open or closed stereotactic techniques largely unsatisfactory. Chronic pain targeting with radiosurgery has been typically performed at the level of the medial thalamus. Young targeted the intralaminar, mediodorsal centromedian, and parafascicular nuclei in chronic pain patients [10–12]. PET or functional magnetic resonance imaging (MRI) guided cingulotomies have also been proposed for chronic pain treatment [13].

58.

psychiatric and pain disorders

565

Case Study 58-1 A 63-year-old male presented to the Lars Leksell Gamma Knife center with right V2 and V3 distribution trigeminal neuralgia for 12 years. The patient found it difﬁcult to shave and was unable to smile without excruciating pain. Initially, he had been managed on Tegretol, but his pain became refractory to this agent. Later, he had had a glycerol injection, but this approach offered him only about 2 months of relief. Using the model C Gamma Knife, the patient was treated with 80 Gy maximal dose to the trigeminal nerve root entry zone (Fig. 58-3). At 3 weeks after Gamma Knife surgery, the patient’s pain began to improve. By 6 months postoperatively, the patient was pain-free without medications. He has remained so for more than 2 years. He has had no development of sensory loss or dysesthesias to his face.

Trigeminal neuralgia (TGN) is a paroxysmal lancinating pain conﬁned to a distribution encompassing one or more of the branches of the trigeminal nerve on one side of the face. As early as 1941, Olivecrona understood and described that mechanical pressure along the root or at the level of the ganglion could be the cause of trigeminal neuralgia [14]. In pioneering work, Granit, Leksell, and Skoglund demonstrated that local pressure on nerve ﬁbers could result in painful afferent

FIGURE 58-4. An axial slice of the brain depicting the target areas for bilateral capsulotomies (asterisk). (From Mindus P. Capsulotomy in Anxiety Disorders—A Multidisciplinary Study. Dissertation at the Karolinska Institute and Hospital, Stockholm, Sweden, 1991. Used with permission.)

discharges from the injured neural segment [15]. More recently, Jannetta and others have suggested that vascular compression of the trigeminal nerve may be a causal agent in trigeminal neuralgia [16–18]. This fact led to the hypothesis of a causal relationship between vessel compression and trigeminal neuralgia and the devising of microvascular decompression surgery. Despite such hypotheses, the fact that balloon compression of the nerve can lead to symptomatic improvement in some patients underscores the true lack of understanding as to the underlying pathophysiology of trigeminal neuralgia [19, 20]. With radiosurgery, the usual target for treating trigeminal neuralgia is the trigeminal nerve root entry zone at the level of the pons. Much has been written about whether or not the trigeminal nerve should be targeted more proximally or distally [21–23] (Case Study 58-1). Rare cases of glossopharyngeal and sphenopalatine neuralgia treatment with radiosurgery have been described [24, 25]. Neither the radiosurgical targeting nor the outcome for these types of neuralgia is well deﬁned.

Obsessive-Compulsive Disorder and Depression

FIGURE 58-3. An axial T1-weighted MRI scan showing the placement of a 4-mm isocenter at the root entry zone of a patient with right-sided trigeminal neuralgia.

The most common radiosurgical target for obsessivecompulsive disorder (OCD) is the anterior limb of the internal capsule. A capsulotomy can be performed on the right hemisphere, left hemisphere, or both. The anatomic site for a capsulotomy is depicted in Figure 58-4 [9] and the appearance on MRI is shown in Figure 58-5 [9]. Lippitz et al. noted a correlation between clinical success and radiologic lesioning of the middle of the anterior limb of the internal capsule [26]. This was deﬁned as parallel to a plane encompassing the anterior commissure–posterior commissure line level at the level of the foramen of Monro and 4 mm rostral by a plane deﬁned by the internal cerebral vein [26]. Cingulotomy can also be performed for OCD and depression [27]. In addition, limbic leukotomy involving bilateral lesions of the cingulum with subcaudate tractotomy have been advocated for intractable depression [28–30].

566

j. sheehan et al. Somatic anxiety Musular temsion Psychic anxiety Psychasthenia

Inhibition of aggression HJ

Impulsivity Monotony avoidance

Post

Socialization Pre

Social desirability Detachment

FIGURE 58-5. Axial (left) and coronal (right) MRI scans illustrating the imaging appearance of bilateral capsulotomies (arrows) 7 years after surgery. (From Mindus P. Capsulotomy in Anxiety Disorders—A Multidisciplinary Study. Dissertation at the Karolinska Institute and Hospital, Stockholm, Sweden, 1991. Used with permission.)

Indirect aggression Verbal aggression Irritability Suspicion Guilt –1 SD

50

1 SD

Anxiety Leksell was impressed with the result yet disappointed with the prefrontal lobotomy technique made famous by Egas Moniz and Almeida Lima. He believed that the frontal lobes should have been surgically treated with more ﬁnesse and that psychosurgery in general should have been more simple yet elegant. Drawing from the work of Henry Wycis in Philadelphia and the anatomic studies of Professor Meyer in London, Leksell used thermal injury to perform a stereotactic capsulotomy on a successful industrialist who was completely incapacitated by anxiety [2]. The result from their early case was good and prompted more psychosurgery using electrodes and later the Gamma Knife. For intractable anxiety disorders, bilateral capsulotomies have been performed with a similar target to that for OCD [31–33].

Treatment Alternatives Chronic pain is initially treated with pharmacological treatment. Pain specialists utilize opioid agonists, nonsteroidal anti-inﬂammatories, local anesthetics, antidepressants, certain anticonvulsants, and muscle relaxants to optimize pain control. More recently, implantable drug delivery systems, transcutaneous electrical nerve stimulation, spinal cord stimulators, and vagal nerve stimulators have been implanted to assist with the treatment of chronic pain [34–36]. Pain rehabilitation programs can also be of use. For trigeminal neuralgia, medical management is the ﬁrst line of treatment for patients. However, many patients with this condition eventually fail medical therapy because of refractory pain or intolerable medication side effects. More invasive treatment options include microvascular decompression (MVD) and neuronal modulating procedures such as glycerol rhizolysis, radiofrequency rhizotomy, percutaneous balloon microcompression, and peripheral nerve blocks [16, 17, 19, 20, 37–39]. Obsessive-compulsive disorder, depression, and anxiety are initially managed with pharmacological agents and psychiatric intervention/counseling. Heterocyclic compounds, monoamine

oxidase inhibitors, selective serotonin reuptake inhibitors, bipolar agents, and anxiolytics/hypnotics have all been applied to these conditions with limited efﬁcacy. Many of the patients with intractable OCD, depression, or anxiety have a high rate of comorbid axis I diagnoses including personality disorders and other functional impairments further complicating the treatment and reducing the chances of a favorable therapeutic outcome [40]. Unfortunately, pharmacological therapies only beneﬁt between 50% and 70% of all OCD patients [41]. Fava and Davidson estimate that 29% to 46% of depressed patients fail to respond to pharmacological agents [42]. In addition to pharmacological treatment, deep-brain stimulation, vagal nerve stimulation, and transcranial magnetic stimulation have been utilized for OCD and depression [43]. The ability to reverse or modulate these neurologic interventions adds to the attractiveness of these interventions over a more static lesioning achieved with radiosurgery. Electroconvulsive therapy has been utilized for depression. It typically has a transient beneﬁt and must be readministered with some regularity.

The Case for Radiosurgery In most cases, radiosurgical treatment of pain and psychiatric disorders should be reserved for patients who fail medical management. However, as noted, this is not an inconsequential number of patients. In patients with psychiatric disease who require surgical intervention, functional radiosurgery offers several important clinical as well as scientiﬁc advantages over open techniques. The most important is patient tolerance. It is our experience that this psychologically vulnerable group of patients is much more willing to undergo a closed stereotactic procedure, which in contrast with open surgery leaves minimal, if any, external marks. Theoretically, the gradual development of the radiolesion may also allow the patient better psychological adjustment. The psychological rehabilitation phase is an important part of any psychosurgical procedure. Stereotactic radiosurgery also does not carry the same type and degree of risks that open surgery (e.g., microvascular

58.

567

psychiatric and pain disorders

decompression, deep-brain stimulation, etc) does. For instance, in a long-term series of microvascular decompression patients, there were the following risks from the series by Jannetta: 0.2% death; 0.1% brain-stem infarct; and 1% hearing loss [16]. None of these complications were observed in our series nor were they associated with radiosurgical treatment of trigeminal neuralgia in other major centers. Complications from hardware and the risk of hemorrhage for deep-brain and vagal nerve stimulators combine to yield a risk of 10% to 27% [44–46]. Ultimately, the patient must choose the type of intervention he or she is willing to undergo. Such a choice inevitably involves a weighing of the relative risks and beneﬁts of each intervention. Of all the indications, stereotactic radiosurgery has been most embraced for the treatment of medically refractory trigeminal neuralgia. The widespread application of radiosurgery for trigeminal neuralgia is in part a testament to its acceptable degree of pain relief and few side effects. Radiosurgery for chronic pain, OCD, intractable anxiety, and depression has been applied with far less frequency. This fact stems in part from the notion that such patients are usually managed by other types of physicians (e.g., psychiatrists, pain specialists, etc.) who have little familiarity with radiosurgery. Moreover, radiation oncologists and neurosurgeons are generally uncomfortable with treating these types of disorders without the close involvement of other specialists to evaluate and assist in the pre- and postoperative management.

Surgical Treatment Alternatives for Pain and Psychosurgeries Open Lesioning Although the results of stereotactic lesioning (e.g., thalamotomy, radiofrequency rhizotomy, glycerol injection) have been well studied and the beneﬁcial effects are immediate, there remains the potential for complications such as intracranial hemorrhage, stroke, and infection. In addition, these procedures carry with them an anesthetic risk, too.

Deep-Brain Stimulation and Vagal Nerve Stimulation The advantages of stimulation techniques include the reversible nature of the process and the ability to modulate the neural stimulation over time. However, patients are again exposed to the risks of open surgery including infection, hemorrhage, and anesthesia. In addition, rates of hardware complications/failure are not trivial, and the batteries need periodic replacement necessitating another surgical procedure [44–46].

Microvascular Decompression for Trigeminal Neuralgia This procedure remains the gold standard for surgical approaches to treat trigeminal neuralgia. Barker et al. reported excellent pain relief in 70% of patients and partial pain relief in another 4% 10 years postoperatively from a microvascular decompression [16]. However, there were the following risks from the series by Dr. Jannetta: 0.2% death; 0.1% brain-stem infarct; and 1% hearing loss [47]. Burchiel et al. in 1988 noted a 3.5% major

recurrence rate and 1.5% minor recurrence rate after invasive surgical approaches [48]. All surgical approaches including radiosurgery appear to have a “wearing off” effect in terms of pain relief over the years.

Radiosurgical Treatment Dosimetry The dose must be planned so that the steepest isodose gradient of the dose distribution (usually between the 50% and 70% isodose lines) coincides with the periphery of tissue being treated. This may require several overlapping ﬁelds of radiation, each using a different collimator size and a separate stereotactic focal point. Changing the relative time of radiation at each target may also change the isodose distribution. Finally, the radiation ﬁeld may be altered by blocking some of the radiation sources, also known as plugging. For radiosurgical capsulotomies and thalamotomies, focal lesions can be produced using a 4-mm collimator and three isocenters on each side for overlapping ﬁelds, creating a cylindrical lesion, with a maximum dose within the target volume of 200 Gy. The development of the lesions using such a plan has been followed by MRI and computed tomography (CT) scans every 3 months. On T2-weighted images, a high signal appears in the target area after approximately 3 months. This signal is most likely produced by local edema. The edema extends a maximal volume at around 9 months and then slowly subsides. The edema is directly related to the dose and to the volume radiated. It may be sufﬁcient to use only one isocenter and the 4-mm collimator. With these treatment parameters and a maximal dose of 180 Gy, a lesion measuring approximately 50 mm3 can be expected within several weeks with only minimal transient edema. Centers have utilized maximal target doses of 120 to 200 Gy for radiosurgical treatment of OCD, intractable anxiety disorders, depression, and chronic pain [10, 12, 31, 32, 40, 49, 50]. Kihlstrom et al. propose that a minimum dose of 110 Gy using the 4-mm collimator is required to create a permanent lesion [32]. Friehs et al. in 1996 reported lesion size after delivering a dose of 160 Gy using a 4-mm collimator [49] (Table 58-1). For trigeminal neuralgia, dose selection has been studied much more extensively. Most centers utilize maximal doses of 70 to 85 Gy [51–56]. Doses of 90 Gy or higher have been associated with increased risk of post-radiosurgical complications [53]. Case reports of glossopharyngeal and sphenopalatine neuralgia treatment with radiosurgery have been published [24, 25]. Pollock and Kondziolka performed Gamma Knife surgery twice on a patient with sphenopalatine neuralgia using a maximal dose of 90 Gy [24]. Doses are generally comparable with those used for trigeminal neuralgia. TABLE 58-1. Neuroanatomical structures and pathways of the limbic system. Time (months)

1 3 6 12

Diameter of necrosis (mm)

3 6 8 4

568

j. sheehan et al.

Outcomes Based on the Type of Radiosurgical Device

differences in target selection, or the type of pain (i.e., either atypical or typical trigeminal neuralgia) [55].

No trials have been performed to look at the results of psychosurgeries using different radiosurgical devices. In fact, very little psychosurgery has been done with radiosurgical devices other than the Gamma Knife. For trigeminal neuralgia, the results of linac-based and Gamma Knife radiosurgery appear comparable [22, 57, 58]. Outcomes appear generally favorable so long as the team is experienced and the device is dedicated to radiosurgical treatment.

Other Types of Pain

Radiosurgical Treatment Outcomes Pain: Trigeminal Neuralgia In 1953, radiosurgery with X-rays was ﬁrst utilized to treat TGN, and the long-term success (greater than 17 years) of two patients after radiosurgery was ﬁrst reported in 1971 [3]. Since then, radiosurgery has been employed to treat TGN. Radiosurgery has been viewed as a minimally invasive treatment option with few side effects. In various series, pain-free outcomes have ranged from 21.8% to 75%, and complication rates varied from 2.7% to 37% [3, 51, 52, 54–57, 59]. Most patients are treated with one isocenter, and the doses utilized from center to center vary very little. Nevertheless, there is a wide range in pain-free and pain-relief outcomes. Some of the differences between outcomes between radiosurgical centers may be a result of difference in patient populations. The other causes for the difference in reported outcomes presumably are the fashion in which data was gathered (e.g., neurosurgical evaluation, follow-up mailed questionnaire or phone survey, referring physician interview, etc.), length of follow-up, and the rigorous criteria utilized to meet either “pain-relief” or “pain-free” status. In our series, which included 136 patients with a median follow-up of 19 months, the median interval from the treatment to symptom improvement was 24 days (range, 1 to 180 days). Few patients experienced a beneﬁt more than 2 months after radiosurgery. At the last time of follow-up, 44% of patients were pain-free without medication, and 56% still had some degree of pain. The percentage of patients who were pain-free without medication at speciﬁc time points was as follows: 47% at 1 year; 45% at 2 years; and 34% at 3 years. Those who experienced some improvement in pain post-radiosurgery were as follows: 90% at 1 year; 77% at 2 years; and 70% at 3 years. The median time to recurrence of facial pain was 12 months [55]. When pain recurs, treatment options include microvascular decompression, glycerol injection, radiofrequency rhizotomy, and repeat radiosurgery. Factors associated with a better response to radiosurgical treatment of trigeminal pain include typical trigeminal neuralgia rather than atypical pain from multiple sclerosis or other causes, higher doses of radiation, a target closer to the brain stem, and no prior surgery [21, 59–62]. In our study, a univariate analysis revealed that a right-sided pain distribution and a previous neurectomy were related to a pain-free outcome. Multivariate analysis revealed that right-sided pain and age correlated with a pain-free outcome. In our analysis, it is notable that a pain-free outcome was not related to dose, sensory loss, slight

Early results using the Gamma Knife to produce thalamotomies for pain control were published by Steiner et al. [63] in 1980. All of the 52 patients treated suffered from terminal cancer and were treated prior to the advent of CT or MRI. Pneumoencephalography was used to target the thalamic centrum medianparafasciculus (CM-Pf complex). Good pain relief was obtained in 8 patients and moderate pain relief in 18. The patients had in general only temporary relief of pain. Of those with good pain relief, ﬁve died without recurrence of pain between 1 and 13 months after the procedure, and three had recurrence of pain at 3, 6, and 9 months. Doses between 100 and 250 Gy were tested. Observation of an actual lesion was only possible in 21 of 36 patients that had a postmortem examination. Not surprisingly, the presence of a lesion was associated with relief. Lesions were only reliably created with doses greater than 160 Gy. The collimators used were 3 by 5 mm and 3 by 7 mm. The most effective lesions were more medially located near the wall of the third ventricle, and the greatest relief was for face or arm pain [63]. These results were not particularly encouraging. However, with improvements in neuroimaging and alternate target selection, it is possible that more effective lesions can be produced. Recent reports seem to support this expectation. Hayashi et al. [64] in 2003 reported signiﬁcant pain reduction in patients with severe cancer pain and post-stroke thalamic pain after Gamma Knife lesioning of the hypophysis [64]. Using the 4-mm collimator and doses of 140 to 180 Gy, Young et al. have published effective pain relief in patients with chronic, intractable pain after medial thalamotomy with the Gamma Knife [10–12]. In a series of 15 patients followed for more than 3 months after a radiosurgical-induced medial thalamotomy, four (27%) were pain free and ﬁve others (33%) had greater than 50% pain relief [12]. Additional investigation must be conducted before the role of the Gamma Knife for pain treatment can be fully deﬁned.

Psychiatric Disease: Anxiety and OCD Mindus and colleagues at the Karolinska Institute reported the effects of bilateral anterior radiosurgical capsulotomies on the anxiety symptoms and personality characteristics. Using independent observers, the Comprehensive Psychopathological Rating Scale (CPRS), and the Karolinska Scales of Personality (KSP; designed to assess frontal lobe dysfunction and anxiety proneness), they compared patients who had been treated with conventional thermocoagulation and followed for 1 year with patients who had been treated with Gamma Knife surgery and followed for 7 years. The two groups were otherwise similar, each including patients who had a mean duration of psychiatric illness of 15 years and who had failed multiple other treatment interventions. The results of Gamma Knife capsulotomy were found to be comparable with those of capsulotomy performed by the thermocoagulation technique. In both groups, freedom from symptoms or considerable improvement was noted in 70% to 80% of patients, and none were worse after the operation. Negative effects on the personality were not noted. Representa-

58.

psychiatric and pain disorders

569

FIGURE 58-6. Karolinska Scales of Personality (KSP) scores reported by Mindus for a patient who underwent a capsulotomy for non– obsessive-compulsive disorder anxiety. SD, standard deviation. (From

Mindus P. Capsulotomy in Anxiety Disorders—A Multidisciplinary Study. Dissertation at the Karolinska Institute and Hospital, Stockholm, Sweden, 1991. Used with permission.)

tive KSP scores are illustrated in Figures 58-6 and 58-7 [9]. A number of patients who were preoperatively unable to work or function normally in society due to preoccupation with personal cleanliness and the inability to use public transportation resulting in domestic conﬁnement, aggravated psychological problems, deterioration of family relationships, and devastation of personal economy were able to return to their previous occupation and to a normal social function. In ﬁve of seven patients who underwent Gamma Knife capsulotomies, a lesion was demonstrated by MRI, and those were the patients who beneﬁted from the procedure. The lowest effective target dose was 160 Gy, whereas 100, 120, and 152 Gy failed to produce lesions [26, 32]. In preliminary results from an ongoing study of bilateral but two-session anterior capsulotomies using the Gamma

Knife, Greenberg et al. in 2003 reported 4 of 15 patients had at least a 35% decrease in the Yale-Brown Obsessive Compulsive scale and a minimum 15-point improvement in the Global assessment scale on 5-year follow-up. In another group of 16 patients who underwent two pairs of bilateral lesions during one session, 10 of 16 patients met the aforementioned criteria at the 3-year follow-up time point. Once achieved, this group noted generally stable beneﬁts [40]. We await further results from this group.

FIGURE 58-7. Karolinska Scales of Personality (KSP) scores reported by Mindus for a patient who underwent a capsulotomy for obsessivecompulsive disorder anxiety. SD, standard deviation. (From Mindus P.

Psychiatric Disease: Depression The long-term outcomes of radiosurgery for depression have not been well studied. Radiosurgery of patients with intractable depression requires a multidisciplinary approach and warrants

Capsulotomy in Anxiety Disorders—A Multidisciplinary Study. Dissertation at the Karolinska Institute and Hospital, Stockholm, Sweden, 1991. Used with permission.)

570

j. sheehan et al.

further investigation [40]. The work using thermocoagulation and cryoprobes to a perform limbic leukotomy may give direction to future work using radiosurgical devices to treat depression [28, 65].

Complications of Functional Radiosurgery Although stereotactic radiosurgery is not associated with some of the immediate risks of open surgical procedures, it nonetheless has risks. Except for trigeminal neuralgia, dose selection is not well studied. The wide range of doses may lead to an unpredictability of lesion size and subsequent outcome. Kihlstrom et al. noted that parenchymal reaction and time course to development of a lesion after radiosurgery for refractory anxiety disorders may be difﬁcult to predict [31]. Other potential complications associated with radiosurgery include radiation-induced parenchymal changes (some of which are reversible and others are irreversible), radiation-induced neoplasia, and vascular injury [66, 67]. More worrisome side effects of radiosurgical or thermocoagulation tractotomies, capsulotomies, cingulotomies, and limbic leukotomies include headache, confusion, somnolence, transient disinhibition, personality changes, seizures, incontinence, exacerbation of mania, apathy, and amotivation [26]. The exact incidence of these complications is difﬁcult to discern based on the existing reports. A noteworthy complication in the treatment of trigeminal pain is facial numbness or dysesthesias. The incidence of new trigeminal dysfunction varies from 6% to 66% [21, 52, 53, 55, 56]. Pollock et al. reported an association between higher radiation doses and the risk of trigeminal nerve dysfunction. In that study, 54% of patients treated with 90 Gy experienced facial numbness, whereas only 15% of patients treated with 70 Gy experienced a similar problem [53]. In our series, 12 of 136 (9%) patients developed new facial numbness after Gamma Knife surgery. In our study, only one patient received a dose of 90 Gy, and no facial numbness was noted in this case. We were unable to establish a signiﬁcant relationship between radiation dose and postoperative facial numbness. We did not observe cases of anesthesia dolorosa or absence of the corneal reﬂex in the 136 patients. Repeat Gamma Knife surgery was associated with an increased risk of facial numbness; whereas 7% of patients who had only one Gamma Knife surgery developed new facial numbness (p = 0.002, t-test) [55]. In a series of 18 patients who underwent repeat radiosurgery, Herman et al. noted an 11% incidence of new or worsened facial numbness that was not substantially elevated over the risk of facial numbness for those having only one Gamma Knife surgery [61]. However, Hasegawa et al. [60] and Shetter et al. [68] noted an increased risk of facial numbness associated with repeat Gamma Knife surgery.

Future Directions If it would be ethically acceptable, a control group of patients could be subjected to spending time in the collimator helmet without radiation. In a later stage, if this sham procedure is proved to give no result in comparison with the real procedure, the control group would receive the appropriate treatment. Such a controlled study is probably necessary before psychora-

diosurgery is generally accepted. Further efforts should also be made to study the biology of the developing lesions. Important and as yet unanswered questions include: 1. When does the functional effect of the radiation start? 2. What are the characteristics of the MRI and CT images at this time? 3. What are the long-term beneﬁts and risks of radiosurgery for psychiatric disorders and intractable pain? Even the issue of dose-volume relationships needs to be addressed further. PET or SPECT imaging may help to answer some of these questions, and pre- and posttreatment evaluation is planned for further series of patients. The experience from multiple centers suggests a mild degree of optimism for the use of radiosurgery in the treatment of intractable obsessivecompulsive disorder, and future research in psychiatric neurosurgery is proceeding in a cautious fashion. Any such work necessitates the coordination and effort of a multidisciplinary team. New neuroimaging modalities, fusion of such modalities with planning MRI, a better understanding of the radiobiology of radiosurgery, and a more thorough knowledge of the abnormal neuronal circuitry of psychiatric disorders all will be required to facilitate widespread application of radiosurgery for psychosurgery.

Conclusion Leksell’s personal battles with pain and his desire to perform more elegant psychosurgeries than prefrontal lobotomies led to the use of radiosurgical devices for the treatment of psychiatric disorders and pain. With the exception of trigeminal neuralgia, the application of radiosurgery to treat these problems has not been widely embraced. Obsessive-compulsive disorder has been the only psychiatric disorder for which radiosurgery has been reasonably well received. In his dissertation, Mindus wrote, “Considering the . . . proportion of cases intractable to conventional treatment, the use of psychosurgery is remarkably rare. It has, in fact, been described as ‘signiﬁcantly underutilised’ by some authors. Whether overutilised in the past, and underutilized in the present, it is the patients who pay the price. From an ethical viewpoint, we must therefore analyze factors which may lead to the inappropriate use of a treatment, be it overuse, abuse, or disuse” [9]. It is unfortunate that Leksell, who developed the Gamma Knife for functional disorders, never saw it embraced for such indications. Nonobliterative surgical approaches and pharmacological alternatives make the widespread embracement of radiosurgery for pain and psychosurgery unlikely. Initial experiences dating over the past 30 years have demonstrated limited efﬁcacy for radiosurgery to treat some of the most intractable of psychiatric disorders and chronic pain. Advances in neuroimaging, neuroanatomy, and psychiatry may lead to a resurgence of interest in this ﬁeld. The expansion of radiosurgical treatment of pain and psychosurgery depends heavily upon better elucidation of the neurophysiology of these disorders: such understanding will only be accomplished by works of excellence by many clinicians and scientists. Any future application of radiosurgery to treat these disorders should be done in the context of a multidisciplinary team.

58.

psychiatric and pain disorders

Acknowledgment. We thank Dr. Ladislau Steiner for assistance in writing this chapter. He also provided a copy of Dr. Mindus’ dissertation for us to review.

References 1. Helfant MH, Leksell L, Strang RR. Experiences with intractable pain treated by stereotaxic mesencephalotomy. Acta Chir Scand 1965; 129:573–580. 2. Leksell L. Brain fragments. In: Steiner L, Forster D, Backlund E-O, eds. Radiosurgery: Baseline and Trends. New York: Raven Press, 1992:263–292. 3. Leksell L. Sterotaxic radiosurgery in trigeminal neuralgia. Acta Chirurg Scand 1971; 137:311–314. 4. Leksell L. [Trigeminal neuralgia. Some neurophysiologic aspects and a new method of therapy]. Swedish Lakartidningen 1971; 68:5145–5148. 5. Håkanson S. Transoval trigeminal cisternography. Surg Neurol 1978; 10:137–144. 6. Bingley T, Leksell L, Meyerson BA, Rylander G. Long term results of stereotactic capsulotomy in chronic obsessivecompulsive neurosis. In: Sweet WH, Obrador S, MartinRodriguez JG, ed. Neurosurgical Treatment in Psychiatry, Pain, and Epilepsy. Baltimore: University Park Press, 1977:287–289. 7. Flickinger JC, Lunsford LD, Kondziolka D. Dose-volume considerations in radiosurgery. Stereotact Funct Neurosurg 1991; 57:99– 105. 8. Kamiryo T, Kassell NF, Thai QA, et al. Histological changes in the normal rat brain after gamma irradiation. Acta Neurochir (Wien) 1996; 138:451–459. 9. Mindus P. Capsulotomy in Anxiety Disorders—A Multidisciplinary Study. Dissertation at the Karolinska Institute and Hospital, Stockholm, Sweden, 1991. 10. Young RF, Jacques DS, Rand RW, et al. Technique of stereotactic medial thalamotomy with the Leksell Gamma Knife for treatment of chronic pain. Neurol Res 1995; 17:59–65. 11. Young RF, Jacques DS, Rand RW, et al. Medial thalamotomy with the Leksell Gamma Knife for treatment of chronic pain. Acta Neurochir 1994; 62(Suppl):105–110. 12. Young RF, Vermeulen SS, Grimm P, et al. Gamma Knife thalamotomy for the treatment of persistent pain. Stereotact Funct Neurosurg 1995; 64(Suppl 1):172–181. 13. Romanelli P, Esposito V, Adler J. Ablative procedures for chronic pain. Neurosurg Clin N Am 2004; 15:335–342. 14. Olivecrona H. Die Trigeminusneuralgie und ihre Behandlung. Nervenarzt 1941; 14:49–57. 15. Granit R, Leksell L, Skoglund CR. Fibre interaction in injured or compressed region of nerve. Brain 1944; 67:125–140. 16. Barker FG 2nd, Jannetta PJ, Bissonette DJ, et al. The long-term outcome of microvascular decompression for trigeminal neuralgia. N Engl J Med 1996; 334:1077–1083. 17. Jannetta PJ. Arterial compression of the trigeminal nerve at the pons in patients with trigeminal neuralgia. J Neurosurg 1967; 26(Suppl):159–162. 18. Lovely TJ, Jannetta PJ. Microvascular decompression for trigeminal neuralgia. Surgical technique and long-term results. Neurosurg Clin N Am 1997; 8:11–29. 19. Brown JA, Chittum CJ, Sabol D, et al. Percutaneous balloon compression of the trigeminal nerve for treatment of trigeminal neuralgia. Neurosurg Focus 1996; 1:e4. 20. Brown JA, Gouda JJ. Percutaneous balloon compression of the trigeminal nerve. Neurosurg Clin N Am 1997; 8:53–62. 21. Kondziolka D, Lunsford LD, Flickinger JC, et al. Stereotactic radiosurgery for trigeminal neuralgia: a multiinstitutional study using the gamma unit. J Neurosurg 1996; 84:940–945.

571

22. Massager N, Lorenzoni J, Devriendt D, et al. Gamma knife surgery for idiopathic trigeminal neuralgia performed using a faranterior cisternal target and a high dose of radiation. J Neurosurg 2004; 100:597–605. 23. Zhang P, Brisman R, Choi J, et al. Where to locate the isocenter? The treatment strategy for repeat trigeminal neuralgia radiosurgery. Int J Radiat Oncol Biol Phys 2005; 62:38–43. 24. Pollock BE, Kondziolka D. Stereotactic radiosurgical treatment of sphenopalatine neuralgia. Case report. J Neurosurg 1997; 87: 450–453. 25. Stieber VW, Bourland JD, Ellis TL. Glossopharyngeal neuralgia treated with gamma knife surgery: treatment outcome and failure analysis. Case report. J Neurosurg 2005; 102(Suppl):155–157. 26. Lippitz BE, Mindus P, Meyerson BA, et al. Lesion topography and outcome after thermocapsulotomy or gamma knife capsulotomy for obsessive-compulsive disorder: relevance of the right hemisphere. Neurosurgery 1999; 44:452–458; discussion 458–460. 27. Cosgrove GR, Rauch SL. Stereotactic cingulotomy. Neurosurg Clin N Am 2003; 14:225–235. 28. Kelly D, Mitchell-Heggs N. Stereotactic limbic leucotomy—a follow-up study of thirty patients. Postgrad Med J 1973; 49:865– 882. 29. Kelly D, Richardson A, Mitchell-Heggs N, et al. Stereotactic limbic leucotomy: a preliminary report on forty patients. Br J Psychiatry 1973; 123:141–148. 30. Richardson A. Stereotactic limbic leucotomy: surgical technique. Postgrad Med J 1973; 49:860–864. 31. Kihlstrom L, Guo WY, Lindquist C, et al. Radiobiology of radiosurgery for refractory anxiety disorders. Neurosurgery 1995; 36: 294–302. 32. Kihlstrom L, Hindmarsh T, Lax I, et al. Radiosurgical lesions in the normal human brain 17 years after gamma knife capsulotomy. Neurosurgery 1997; 41:396–401; discussion 401–392. 33. Leksell L, Herner T, Leksell D, et al. Visualisation of stereotactic radiolesions by nuclear magnetic resonance. J Neurol Neurosurg Psychiatry 1985; 48:19–20. 34. Khadilkar A, Milne S, Brosseau L, et al. Transcutaneous electrical nerve stimulation (TENS) for chronic low-back pain. Cochrane Database Syst Rev 2005; CD003008. 35. Multon S, Schoenen J. Pain control by vagus nerve stimulation: from animal to man . . . and back. Acta Neurol Belg 2005; 105: 62–67. 36. Turk DC. Clinical effectiveness and cost-effectiveness of treatments for patients with chronic pain. Clin J Pain 2002; 18:355– 365. 37. Hamid AI, Qureshi AA, Bhatti IH. Percutaneous radiofrequency retrogasserian rhizotomy for trigeminal neuralgia. J Pak Med Assoc 1993; 43:132–133. 38. Henson CF, Goldman HW, Rosenwasser RH, et al. Glycerol rhizotomy versus gamma knife radiosurgery for the treatment of trigeminal neuralgia: an analysis of patients treated at one institution. Int J Radiat Oncol Biol Phys 2005; 63:82–90. 39. Lunsford LD. Glycerol rhizotomy for trigeminal neuralgia. J Neurosurg 1987; 66:151–152. 40. Greenberg BD, Price LH, Rauch SL, et al. Neurosurgery for intractable obsessive-compulsive disorder and depression: critical issues. Neurosurg Clin N Am 2003; 14:199–212. 41. Rasmussen SA, Eisen JL. The epidemiology and clinical features of obsessive compulsive disorder. Psychiatr Clin North Am 1992; 15:743–758. 42. Fava M, Davidson KG. Deﬁnition and epidemiology of treatmentresistant depression. Psychiatr Clin North Am 1996; 19:179– 200. 43. Pallanti S, Hollander E, Goodman WK. A qualitative analysis of nonresponse: management of treatment-refractory obsessivecompulsive disorder. J Clin Psychiatry 2004; 65(Suppl 14):6–10.

572

j. sheehan et al.

44. Blomstedt P, Hariz MI. Hardware-related complications of deep brain stimulation: a ten year experience. Acta Neurochir (Wien) 2005; 147:1061–1064. 45. Kondziolka D, Whiting D, Germanwala A, et al. Hardwarerelated complications after placement of thalamic deep brain stimulator systems. Stereotact Funct Neurosurg 2002; 79:228– 233. 46. Smyth MD, Tubbs RS, Bebin EM, et al. Complications of chronic vagus nerve stimulation for epilepsy in children. J Neurosurg 2003; 99:500–503. 47. Jannetta PJ. Outcome after microvascular decompression for typical trigeminal neuralgia, hemifacial spasm, tinnitus, disabling positional vertigo, and glossopharyngeal neuralgia (honored guest lecture). Clin Neurosurg 1997; 44:331–383. 48. Burchiel KJ, Clarke H, Haglund M, et al. Long-term efﬁcacy of microvascular decompression in trigeminal neuralgia. J Neurosurg 1988; 69:35–38. 49. Friehs GM, Noren G, Ohye C, et al. Lesion size following Gamma Knife treatment for functional disorders. Stereotact Funct Neurosurg 1996; 66(Suppl 1):320–328. 50. Frighetto L, De Salles A, Wallace R, et al. Linear accelerator thalamotomy. Surg Neurol 2004; 62:106–113; discussion 113– 104. 51. Brisman R. Gamma knife surgery with a dose of 75 to 76.8 Gray for trigeminal neuralgia. J Neurosurg 2004; 100:848–854. 52. Petit JH, Herman JM, Nagda S, et al. Radiosurgical treatment of trigeminal neuralgia: evaluating quality of life and treatment outcomes. Int J Radiat Oncol Biol Phys 2003; 56:1147–1153. 53. Pollock BE, Phuong LK, Foote RL, et al. High-dose trigeminal neuralgia radiosurgery associated with increased risk of trigeminal nerve dysfunction. Neurosurgery 2001; 49:58–62; discussion 62–54. 54. Richards GM, Bradley KA, Tome WA, et al. Linear accelerator radiosurgery for trigeminal neuralgia. Neurosurgery 2005; 57:1193– 1200; discussion 1193–1200. 55. Sheehan J, Pan HC, Stroila M, et al. Gamma knife surgery for trigeminal neuralgia: outcomes and prognostic factors. J Neurosurg 2005; 102:434–441.

56. Tawk RG, Duffy-Fronckowiak M, Scott BE, et al. Stereotactic gamma knife surgery for trigeminal neuralgia: detailed analysis of treatment response. J Neurosurg 2005; 102:442–449. 57. Ma L, Kwok Y, Chin LS, et al. Comparative analyses of linac and Gamma Knife radiosurgery for trigeminal neuralgia treatments. Phys Med Biol 2005; 50:5217–5227. 58. Smith ZA, De Salles AA, Frighetto L, et al. Dedicated linear accelerator radiosurgery for the treatment of trigeminal neuralgia. J Neurosurg 2003; 99:511–516. 59. Maesawa S, Salame C, Flickinger JC, et al. Clinical outcomes after stereotactic radiosurgery for idiopathic trigeminal neuralgia. J Neurosurg 2001; 94:14–20. 60. Hasegawa T, Kondziolka D, Spiro R, et al. Repeat radiosurgery for refractory trigeminal neuralgia. Neurosurgery 2002; 50:494– 500; discussion 500–492. 61. Herman JM, Petit JH, Amin P, et al. Repeat gamma knife radiosurgery for refractory or recurrent trigeminal neuralgia: treatment outcomes and quality-of-life assessment. Int J Radiat Oncol Biol Phys 2004; 59:112–116. 62. Regis J, Metellus P, Dufour H, et al. Long-term outcome after gamma knife surgery for secondary trigeminal neuralgia. J Neurosurg 2001; 95:199–205. 63. Steiner L, Forster D, Leksell L, et al. Gammathalamotomy in intractable pain. Acta Neurochir (Wien) 1980; 52:173–184. 64. Hayashi M, Taira T, Chernov M, et al. Role of pituitary radiosurgery for the management of intractable pain and potential future applications. Stereotact Funct Neurosurg 2003; 81:75–83. 65. Price BH, Baral I, Cosgrove GR, et al. Improvement in severe self-mutilation following limbic leucotomy: a series of 5 consecutive cases. J Clin Psychiatry 2001; 62:925–932. 66. Ganz JC. Gamma knife radiosurgery and its possible relationship to malignancy: a review. J Neurosurg 2002; 97:644–652. 67. Lim YJ, Leem W, Park JT, et al. Cerebral infarction with ICA occlusion after Gamma Knife radiosurgery for pituitary adenoma: a case report. Stereotact Funct Neurosurg 1999; 72(Suppl 1): 132–139. 68. Shetter AG, Rogers CL, Ponce F, et al. Gamma knife radiosurgery for recurrent trigeminal neuralgia. J Neurosurg 2002; 97:536–538.

5 9

Intractable Epilepsies Jean Régis, Fabrice Bartolomei, and Patrick Chauvel

Introduction There are convincing arguments for investigating the potential role of radiosurgery in epilepsy surgery. We know that: • Radiosurgery (since its introduction in the 1950s) has been demonstrated to have advantages in terms of safety and efﬁcacy, for the treatment of numerous small, deeply seated intracerebral lesions. • Radiosurgical treatment of small cortico-subcortical lesions associated with epilepsy has been demonstrated to lead to seizure cessation in a high percentage (58% to 80% in arteriovenous malformation) of cases, long before the expected treatment of the lesion and sometimes even in spite of failing to cure the lesion itself. • Radiotherapeutic treatment of intractable epilepsies with or without space-occupying lesions can lead to a reduction in seizure frequency and/or severity. • Experimental models of epilepsies treated with radiation therapy have demonstrated a dose-dependent positive effect of radiation on the frequency and severity of the seizures and on the extent of discharge propagation. These clinical and experimental observations are leading to the concept of a speciﬁc antiepileptic effect. Lars Leksell conceived Gamma Knife (GK) radiosurgery as a tool for functional neurosurgery [1, 2]. Accordingly, he used GK in movement disorders, trigeminal neuralgia and other pain syndromes, but not for epilepsy surgery [3]. The ﬁrst radiosurgical treatments for epilepsy surgery were performed by Talairach in the 1950s [4]. Talairach was another pioneering expert in stereotaxis. Unlike Leksell, he had speciﬁc involvement in epilepsy surgery and led one of the ﬁrst large, comprehensive programs for epilepsy surgery. As early as 1974, he reported on the use of radioactive yttrium implants in patients with mesial temporal lobe epilepsy (MTLE) without space-occupying lesions and showed a high rate of seizure control in patients with epilepsies conﬁned to the mesial structures of the temporal lobe [5]. In 1980, Elooma [6], apparently ignoring the pioneer work of Talairach, promoted the idea of the use of focal irradiation for the treatment of temporal lobe epilepsy, based on the preliminary reports of Tracy, Von

Wieser, and Baudouin [7, 8]. Furthermore, clinical experience of the use of GK- and linac-based radiosurgery in arteriovenous malformations (AVMs) and cortico-subcortical tumors (mostly metastases and low-grade glial tumors) revealed an antiepileptic effect of radiosurgery in absence of necrotizing effect [9–11]. A series of experimental studies in small animals conﬁrmed this effect [12, 13] and has emphasized its relationship to the dose delivered [14–17]. Barcia-Salorio et al., and later Lindquist et al., reported small and heterogenous groups of patients treated with the aim of seizure cessation; results were however poor [18–21]. Unfortunately, these data were never published in peer-reviewed papers and precise information is unavailable [18–21]. The two major ﬁelds of expertise of the department of Stereotactic and Functional Surgery in Marseille are epilepsy surgery and radiosurgery. This context has therefore facilitated the investigation and development of a potential role for GK radiosurgery in the treatment of intractable epilepsy. Since March 1993, we have performed 106 cases of epilepsy surgery using GK radiosurgery. The majority of these patients presented with MTLE (73 patients) or hypothalamic hamartoma (HH; 38 patients). The rest of the patients suffered from severe epilepsy associated with small benign lesions (one ganglioglioma, three dysplasia, one periventricular heterotopia, one cavernous angioma), for which an epileptic zone was considered to be conﬁned to the surrounding cortex [22]. In HH, GK radiosurgery offers very low morbidity, with similar efﬁcacy when compared with microsurgical alternatives [23, 24]. This has led us to consider radiosurgery systematically as the ﬁrst-line treatment in patients with small HH of type I, II, III and possibly type IV [24]. In MTLE, on the other hand, in spite of a good short-term and middle-term safety-efﬁcacy ratio [25, 26], the use of GK is still regarded as being an experimental technique [26], given the well-established long-term safety and efﬁcacy of the microsurgical resection in the temporal lobe. Since 1994, we have promoted the idea that seizure cessation may be generated by a speciﬁc neuromodulatory effect of radiosurgery, without induction of a signiﬁcant amount of histologic necrosis [27–32]. The selection of the appropriate technical parameters (dose, volume target, etc.) allowing us to accurately obtain the desired functional effect without histologic damage remains an important challenge.

573

574

j. régis et al.

Hypothalamic Hamartomas In some patients, hypothalamic hamartomas (HHs) can be associated with catastrophic epilepsy. Usually, the seizures begin early in life and are often particularly drug-resistant from the outset. The evolution is unfavorable in the majority of the patients because of behavioral symptoms (particularly aggressive behavior) and mental decline, which occur as a direct effect of the seizures [33], due to an epileptic encephalopathy. Interestingly, in our experience, the reversal of this encephalopathy after radiosurgery seems to start even before complete cessation of the seizures and seems to be correlated with the improvement in background EEG activity. It is the authors’ speculation that these continuous discharges lead to the disorganization of several systems, including the limbic system, and that their disappearance accounts for the improvement seen in attention, memory, cognitive performance and impulsive behavior, and so forth. Here, the goal of radiosurgery is the reversal of the epileptic encephalopathy more than seizure cessation. Consequently, we consider that it is essential to operate on these young patients as early as possible, whatever the surgical approach considered (resection or radiosurgery). The intrinsic epileptogenicity of HH has been demonstrated [34, 35] even though the mechanisms of the epilepsy associated with HH are still debatable. The boundary of the target zone of treatment is that of the lesion visualized on magnetic resonance imaging (MRI). This contrasts greatly with cases of MTLE where there is no such clear delineation of an epileptogenic zone on the images used for planning radiosurgical intervention. The very good safety-efﬁcacy ratio observed on the occasion of a retrospective series [23] of 10 patients (all improved, 50% cured and no adverse effects except 1 case of poikilothermia) led us to organize a prospective multicenter trial. Sixty patients have been included prospectively since October 1999 including 37 with more far conﬁrmed than those of the more limited retrospective study. A minimum of 3 years follow-up after radiosurgical treatment is mandatory before considering a different treatment approach. Data for a more precise model of the relationship between the marginal dose and the psychiatric, cognitive, hormonal, memory, and seizure outcomes are also lacking. We pay special attention to the dose delivered to the mammillary body and to the fornix and we always try to tailor the dose plan for each patient, based on the use of a single run of shots with the 4-mm collimator. The number of patients operated by GK more than 3 years ago is 31. A satisfactory follow-up is available for 27 patients. Among those, 10 (37%) are seizure-free, 6 (22.2%) very much improved with a huge seizure reduction associated with a dramatic behavioral and cognitive improvement. Five (18.5%) patients with small hamartomas are only improved and under consideration for a new radiosurgery. Two have until now reported no signiﬁcant improvement. A microsurgical approach has been performed in four (14.8%) patients with quite large HH and poor efﬁcacy of radiosurgery. It is our policy, from the ﬁrst discussion with the patient and the family, to plan for a second radiosurgery in case of partial beneﬁt when the lesion is anatomically small and well deﬁned. The radiosurgical treatment has been repeated one time in nine patients. From the

radiology point of view, the majority of the patients have no magnetic resonance (MR) changes (28/31) and no clinical, even transient, side effects. Only four (14.8%) patients have experienced a transient worsening of the epilepsy, and three (11%) patients a transient poikilothermia. No patient until now has presented a permanent complication. Globally, a very good result has been obtained in 60% of the patients. The classiﬁcations of Valdueza and Arita [36, 37] are proving rather too simplistic and not sufﬁciently adapted to the pleomorphism of the lesions and the clinical and therapeutic consequences of these variations. As underlined by Palmini et al. [38], the exact location of the lesion in relation to the interpeduncular fossa and the walls of the third ventricle correlates with the extent of excision required, the seizure control, and the complication rate. This rationale led us to classify HH more precisely according to topology, relying on the pertinent features correlating with clinical semiology, prognosis, and surgical strategies (Fig. 59-1) [24]. In our experience, the indications for treatment are better reﬁned on the basis of such a topologic classiﬁcation. We have found that partial treatment (of the superior part of the lesion) or low-dose treatment always result in a negative outcome. We consider radiosurgery as a ﬁrst-line treatment for small type I, II, III or IV HH. In large type IV HH, a combined approach, with initial surgery for section or resection via a pterional approach, followed by radiosurgical treatment of the upper part, is recommended. In type II, an endoscopic approach or a transcallosal interforniceal approach may also be considered as an alternative to radiosurgery [39]. These upper approaches are restricted by the difﬁculty of identifying the boundary between the HH and the hypothalamus, mammillary body and fornix, yielding a signiﬁcant rate of post-

FIGURE 59-1. Hypothalamic hamartomas topological classiﬁcation. The exact location of the lesion in relation to the interpeduncular fossa and the walls of the third ventricle correlates with the extent of excision required, the seizure control, and the complication rate [24]. Although it may be an exceptional observation, type V (pedunculated) tends not to have neurologic symptoms (no epilepsy, no cognitive deterioration, and no behavioral disturbances). They may present with precocious puberty or be symptom-free [24]. Types I, II, III, and IV may cause seizures in many cases, as well as mental retardation, behavioral abnormalities, and precocious puberty. Type VI HHs are frequently found in patients with especially severe clinical presentations. We consider radiosurgery as a ﬁrst-line treatment for small type I, II, III, or IV HHs. (From Régis J, Hayashi M, Perez Eupierre L, et al. Gamma Knife surgery for epilepsy related to hypothalamic hamartomas. Acta Neurochir 2004; 91:33–50. With permission.)

59.

intractable epilepsies

treatment short-term memory deﬁcit and hormonal disturbances [24]. Two major questions remain. First, we know that complete treatment or resection of the lesion is not always mandatory [40, 41] but the required extent of resection for seizure cessation in an individual patient is still not predictable. Second, we know that these patients frequently present with an electro-clinical semiology suggesting involvement of the temporal or frontal lobe and that can mimic a secondary epileptogenesis phenomenon [35, 42]. In our experience, some of these patients can be completely cured by the isolated treatment of the HH, whereas in others, a partial result is obtained, with residual seizures despite a signiﬁcant overall psychiatric and cognitive improvement. In this second group, it is tempting to propose that such a secondary epileptogenic area accounts for the partial failure. Our initial results indicate that GKS is as effective as microsurgical resection and much safer for small type I to IV HHs. GKS also avoids the vascular risk related to radiofrequency lesioning or stimulation. The disadvantage of radiosurgery is its delayed action. Longer follow-up is mandatory for proper evaluation of the role of GKS. Results are faster and more complete in patients with smaller lesions inside the III ventricle (stage II). The early effect on subclinical EEG discharges appears to play a major role in the dramatic beneﬁt to sleep quality, behavior, and cognitive-developmental improvement. Gamma Knife surgery can safely lead to the reversal of the epileptic encephalopathy. Due to the very poor clinical prognosis of the majority of these patients with HH and the invasiveness of microsurgical resection, GK can be now be considered the ﬁrst-line intervention for small- to middle-size HHs associated with epilepsy, as it can lead to dramatic improvements to the future of these young patients. A second radiosurgical procedure is turning out to be very beneﬁcial and well tolerated in young patients with small hamartomas and a signiﬁcant improvement after the ﬁrst procedure. A major improvement of behavioral and cognitive abilities is frequently obtained in the young patients even, sometimes, in spite of remnant seizures. Radiosurgery must be proposed as soon as possible in order to offer to these young patients the highest probability to reverse the epileptic encephalopathy and its heavy comorbidity.

Mesial Temporal Lobe Epilepsy The ﬁrst Gamma Knife surgery operations for mesial temporal lobe epilepsy (MTLE) were performed in Marseille in March 1993. As far as no similar experience was available at this time in the literature, we were obliged to base our technical choices on hypothesis and experience of radiosurgery for other pathologic conditions. Four patients were treated with different technical strategies (dose, volume, target deﬁnition). The delayed huge radiologic changes observed some months after radiosurgery [32] led us to stop such treatment and follow these ﬁrst four patients. Because of the clinical safety of the procedure in these patients and the gradual disappearance of the acute MR changes after some months, we treated several new series of patients under strict prospective controlled trial conditions (with ethical committee approval). The treatment for the following 19 patients was based on that of the ﬁrst patient who

575

FIGURE 59-2. Dosimetry in a case of a typical right mesial temporal lobe epilepsy (24 Gy at the 50%).

had a successful outcome (as opposed to the three others who had partial or no effect). This “classic planning” was based on the use of two 18-mm shots, covering a volume of around 7 cm3 at the 50% isodose (24 Gy), and has turned out to produce a high rate of seizure cessation [25, 26]. For epileptologic reasons, as well as for safety reasons, the targeting was very much centered on the parahippocampal cortex and spared a signiﬁcant part of the amygdaloid complex and hippocampus. The reﬁnement of the Gamma Knife surgery technique, and the desire to ﬁnd a dose that would create less transient acute MR changes, led us to reduce the dose from 24 Gy to 20 and 18 Gy at the margin. However, this brought about a signiﬁcant decrease in the rate of seizure cessation. In the long-term, the rate of seizure freedom is 67% (12/18) for the patients treated with 24 Gy at the margin, 33% (6/18) for the one treated with 20 Gy, and 15% (1/6) for the one treated with 18 Gy (Fig. 59-2). Since 1993 [30], the systematic analysis of our clinical experience has led us to a series of observations: 1. Dramatic delayed MR changes are usually very well tolerated and progressively disappear with frequently no longterm MR evidence of any destructive effect [30, 32]. 2. The timetable of the clinical and radiologic events is quite standardized with a stable pattern and a variable delay [25]. Typically, the frequency of the seizures is not modiﬁed signiﬁcantly for the ﬁrst few months. Thereafter, there is a rapid and dramatic increase in auras for some days or weeks and then the seizures disappear. Usually, the peak in seizure cessation is observed around the 8th to 18th month with a

576

j. régis et al.

clear variability in the delay in onset. In one patient, this occurred 26 months after GK radiosurgery. 3. A verbal memory sparing especially on the dominant side is quite obvious by comparison with the microsurgical experience [43–45]. 4. Less severe depression is observed compared with the microsurgical experience [43, 44]. 5. A marginal dose between 20 and 24 Gy is required for a high rate of seizure cessation [29]. 6. Target selection must include the anterior parahippocampal cortex (entorhinal and perirhinal cortex), but complete inclusion of the amygdaloid nuclear complex and hippocampus is not mandatory [29, 46, 47]. 7. Patient selection must be restricted to pure MTLE (all the temporal lobe epilepsies treatable by a large corticectomy are not necessarily treatable by a very selective radiosurgery [29]. 8. A nondestructive effect of radiosurgery can be sufﬁcient for alleviate epilepsy [27]. 9. In the long-term, results are reasonably stable [48]. For more than 10 years, the Marseille Epilepsy Surgery group has been the only one reporting such a large clinical experience. The publication of some more limited and less favorable clinical experiences [49] have even led some authors to doubt the credibility of these results (see comment by Pollock [49]). Fortunately, the recent reports of the multicentric prospective trial [50] and the Spanish experience [51] are supporting very strongly our results and speciﬁcally the ﬁrst ﬁve points listed above. More data analysis and follow-up of these two groups are still mandatory for conﬁrmation of the four other points. The timetable of events after radiosurgery and the followup is quite standardized. Patients are informed that delayed efﬁcacy of radiosurgery is its main drawback. We usually consider a delay of 2 years as a minimum for post-radiosurgery follow-up. In the absence of initial radiologic changes or clinical beneﬁt, the recommendation is to wait for the onset of the MR changes and their subsequent disappearance. All our patients had the same pattern of MR changes regardless of marginal dose (18 to 24 Gy) or volume (5 to 8.5 cm3). However, the degree of these changes and their delay of onset varied according to the dose delivered to the margin, the volume treated, and the individual patient. In order to allow an optimal evaluation, we recommend that subsequent microsurgery not be considered before the third year after radiosurgery. Similarly, we believe that a patient who undergoes a corticectomy before the onset of the MR changes has occurred cannot be assumed to have failed radiosurgical treatment. Of course, before consideration of any further surgery, the question of the reason for the failure needs to be addressed. After reviewing ﬁles of patients treated for MTLE with radiosurgery, it was sometimes possible to identify likely causes of failure, such as: 1. Poor patient selection (e.g., patients with epilepsy involving more than the mesial temporal lobe structures). 2. Patients with the diagnosis of “treatment failure” (<3 years) who had been operated upon too early after radiosurgery [52].

3. Targeting of the amygdala and hippocampus (which is not in our opinion the optimal target in terms of safety and efﬁcacy) instead of parahippocampal cortex [53]. 4. Insufﬁcient dosage [52–54]. Our current strategy of treatment is based on our ﬁrst series of MTLE patients who were strictly selected and treated systematically with a very simple but very reproducible dose planning strategy [25, 30]. The identiﬁcation of putative improvements in the methodology requires a systematic analysis of the inﬂuence of the technical data from our experience and from the literature on the outcome of those patients.

The “Technical” Questions The Dose Issue The ﬁrst targets used in functional GK radiosurgery (capsulotomy, thalamotomy of VIM or the centromedianus, pallidotomy) were treated using high dose (300 to 150 Gy) delivered in very small volumes (3 to 5 mm in diameter) [3]. The goal was to destroy a predeﬁned very small anatomic structure with stereotactic precision. Quite a signiﬁcant variability in the delay and amplitude of the MR changes has been reported with ﬁxed regimen of doses [55, 56]. Barcia-Salorio et al. have presented several times a small and heterogeneous group of patients treated with different kinds of devices and dosage regimens [57]. Apparently, some of those patients had no expanding lesion and were treated with very large volumes and very low dosage (around 10 Gy). Based on this experience, several teams have made the assumption that very low doses, as low as 10 Gy to 20 Gy at the margin, should be as effective as the 24-Gy protocol (at the margin) that we used for our ﬁrst series of patients with MTLE [25, 26]. A cautious examination of the last proceeding of Barcia-Salorio et al. shows that the individual information concerning the dose at the margin, the volume, and the topography of the epileptogenic zone are not provided. Moreover, among the 11 patients reported, the real rate of seizure cessation is apparently only 36% (4/11), which is much lower than what we would expect with resection in MTLE [18]. In a heterogeneous group of 176 patients, Yang et al. conﬁrmed that only a very low rate of seizure control is achieved when low doses (from 9 to 13 Gy at the margin) are used [54]. The experience of the radiosurgical treatment of HH indicates that 18 Gy at the margin appears to be a threshold in terms of probability of seizure cessation [23]. In this group of patients (60 cases), only three showed MR changes. The majority of the AVM cases with worsening of the epilepsy were treated with a range of doses between 15 and 18 Gy. Similarly, poor results have been reported by Cmelak et al. in one case of MTLE treated with linac-based radiosurgery, with 15 Gy at the 60% isodose line, who underwent surgical resection 1 year later. In this case, the authors ﬁrst observed a slight improvement followed by an obvious worsening [52]. A recent de-escalation study has allowed us to demonstrate poorer results in patients receiving doses of 18 or 20 Gy at the margin compared with 24 Gy [46, 47]. Because of the rate of seizure cessation that is

59.

intractable epilepsies

achievable by conventional resection, a radiosurgical strategy associated with a much lower rate of seizure cessation appears unacceptable. Fractionated stereotactically guided radiotherapy has been demonstrated to fail systematically in controlling seizures. Among 12 patients treated by Grabenbauer et al., none have achieved seizure cessation [58, 59]; only seizure reduction was obtained in this series.

Is Radiosurgery a “Neuromodulation Therapy”? [27] The precise mechanism of action is not clearly identiﬁed. Experimental studies on small animals have demonstrated the antiepileptic effect of radiosurgery [12–14, 16], the dose dependence of this effect [14–16, 60] and the possibility of obtaining clear antiepileptic effect without macroscopic necrosis using certain doses [15]. Of course, the rat models of epilepsy are far from being good models of human MTLE. However, taking into account the huge difference in volume of the target, it is intriguing to notice that according to our clinical experience in humans, a similar maximum dose range of 40 to 50 Gy is currently the range of dose providing the optimal safety efﬁcacy ratio. More recently, Jenrow et al. [61] have demonstrated that the selective reduction of densities in the dentate granular cell layer and medial CA3 pyramidal cell layer is prevented or reversed by the irradiation at 25 but not at 18 Gy.

The Target Deﬁnition When the target is a lesion that is precisely deﬁned radiologically, the question of the selection of the marginal dose can be quite easily addressed by correlating safety-efﬁcacy individual outcome to the marginal dose. This can be reﬁned based on stratiﬁcation according to volume, location, age, and so forth. However, in patients presenting with MTLE, this process is invalid for two reasons. First, there is no consensus regarding the requirement for extent of mesial temporal lobe resection. Second, the concept of MTLE syndrome with a stable extent of the epileptogenic zone and surgical target is increasingly the topic of debate [62, 63]. The volume (in association with marginal dose) is wellknown to be a major determinant of the tissue effect, as shown in integrated risk/dose volume formulae [64]. In the ﬁrst series of patients that we treated, this marginal isodose volume (or prescription isodose volume) was approximately 7cm3 (range, 5 to 8.5). An attempt to correlate dose/volume and the effect on seizures and on the MR changes (as evaluated by volume of the contrast enhancement ring, extent of the high T2 signal, and the importance of the mass effect) has been published recently [46]. In this study, we found, not surprisingly, that the higher the dose and the volume, the higher the risk of having more severe MR changes, but also the higher the chance of achieving seizure cessation. However, these data have limited value. Hence, more precise identiﬁcation of those structures of the mesial temporal lobe that need to be “covered” by the radiosurgical treatment may allow more selective, but just as efﬁcacious, dose planning strategies, in spite of smaller prescription isodose volumes.

577

There is growing evidence to support the organization of the epileptogenic zone in networks, meaning that several different and possibly distant structures are discharging simultaneously at the onset of the electro-clinical seizure. This kind of organization explains why the risk of failure is so high when a simple topectomy (without preoperative investigations) is performed in severe drug-resistant epilepsies associated with a benign lesion [65]. This has been also reported in MTLE [62, 63] Certain nuclei of the amygdaloid complex, of the head, body, tail of the hippocampus, of the perirhinal, entorhinal (EC), and parahippocampal cortices may be associated with genesis of the seizures. The role of the EC cortex in epilepsy is supported by experimental studies in animals [66, 67]. The EC is considered to be the ampliﬁer of the “amygdalohippocampal epileptic system.” The pattern of the associated structures, including that of the structure playing the leader role, can vary signiﬁcantly from one patient to another [62, 63]. There is a subgroup of patients who have clonic discharges and the involvement of the EC, amygdala, and head of the hippocampus, with a clear leader role of the EC. Wieser et al. have analyzed the postoperative MR images of patients operated by Yasargil (amygdalohippocampectomy) and were able to correlate the quality of the resection of each substructure of the mesial temporal lobe area and the outcome with respect to seizures [68]. Only the quality of the removal of the anterior parahippocampal cortex was correlated strongly with a higher chance of seizure cessation [68]. We tried to perform a similar study in patients treated with GK radiosurgery [46]. We deﬁned and manually drew the limits of subregions on the stereotactic images of all these patients. The amygdala, the head, the body, and the tail of the hippocampus were ﬁrst delineated. The white matter, the parahippocampal cortex, and the cortex of the anterior wall of the collateral ﬁssure were then separately drawn and divided into four sectors in the rostro-caudal axis, corresponding with the amygdala, the head, the body, and the tail of the hippocampus [46].

Patient Selection Whang (without having ﬁrst performed speciﬁc preoperative epileptologic workup) treated patients with epilepsy associated with slowly growing lesions and observed seizure cessation in only 38% (12/31) of the patients [69]. This kind of observation emphasizes the importance of preoperative deﬁnition of the extent of the epileptic zone and of its relationship with the lesion [65, 70]. In our institution, the philosophy is to adapt the investigations for each individual case. In some patients, the electro-clinical data, the structural and functional imaging, and the neuropsychological examination are sufﬁciently concordant for surgery of the temporal lobe to be proposed without depth electrode recording. In other cases, the level of evidence for MTLE is judged insufﬁcient, and a stereoelectroencephalographic (SEEG) study is performed. The strategy of SEEG implantation is based on the primary hypothesis (mesial epileptogenic zone) and alternative hypotheses (early involvement of the temporal pole, lateral cortex, basal cortex, insular cortex, or other cortical areas). The goal of these studies is to record the patient’s habitual seizures, in

578

j. régis et al.

order to establish the temporo-spatial pattern of involvement of the cortical structures during these seizures. Clearly in these patients, the high resolution of depth electrode recording allows ﬁne tailoring of surgical resection, according to the precise temporo-spatial course of the seizures. The main limitation of radiosurgery is that of size of the target (prescription isodose volume). The radiosurgical treatment of MTLE is certainly the most selective surgical therapy for this group of patients. The requirements for precision and accuracy in the deﬁnition of the epileptogenic zone are consequently higher. Furthermore, if depth electrode investigation enables demonstration of a particular subtype of MTLE, this can lead to tailoring of the treatment volume and frequently allows this to be reduced.

The Potential Concerns The risk of long-term complications must always be cautiously scrutinized in functional neurosurgery. Radiotherapy is most frequently used in the brain for short-term life-threatening pathologies. The use of radiotherapy in young patients with benign disease such as pituitary adenomas or craniopharyngiomas has been associated with a signiﬁcant rate of cognitive decline [71, 72] and tumor genesis [73] including some carcinogenesis [74]. If the risk of radiation-induced tumor was similar with radiosurgery, we should have by now already observed numerous cases. However, such reported cases [75–77] are extremely rare and frequently fail to meet the classic criteria by which tumors are deemed to be “radiation-induced” [78] (for a review, see [79]). In fact, it is considered that, if this risk exists, it is likely to be far lower than the mortality risk associated with temporal lobectomy. Epilepsy is a life-threatening condition. The risk of sudden unexplained death in epileptic patients (SUDEP) is higher than in the general population [80, 81]. This risk is higher in patients treated with more than two antiepileptic drugs and IQ lower than 70 (as independent factors). Because seizure cessation after surgery reduces the mortality risk to that of the general population [81], microsurgical resection of the epileptogenic zone may confer a beneﬁt in terms of the possibility of immediate seizure cessation and therefore reduced mortality risk compared with the more delayed beneﬁts of radiosurgical treatment.

Our patients are systematically informed about this disadvantage of radiosurgery.

What Are the Current Indications? Hypothalamic hamartomas are already ﬁrst-line indication for radiosurgery when the lesion is small. We still consider the use of radiosurgery for MTLE to be experimental. Patients with clear MTLE on the dominant side, particularly when there is little atrophy and no or minimal verbal memory deﬁcit, are good candidates for radiosurgery. In our experience, the most important selection parameters are the demonstration of the purely mesial location of the epileptogenic zone, as well as clear understanding by the patient of the advantages, disadvantages, and limitations. Potential sparing of memory function is still a matter of debate and needs to be established using comparative studies. There is also a requirement for further demonstration of long-term efﬁcacy and safety of radiosurgery. Worldwide, microsurgical corticectomies for MTLE are proving to be very satisfactory due to the rarity of surgical complications and a high rate of seizure freedom. One other very good indication in our experience is that of patients with proven MTLE but previous failure of microsurgery, supposedly due to insufﬁcient posterior extent of the resection. Callosotomy can be performed by radiosurgery quite safely as previously reported by Shrottner et al. [82] (Fig. 59-3). When the epileptologic context makes realistic an anterior callosotomy and when the invasiveness of a microsurgical approach is a concern, radiosurgery can be a very appealing alternative. Finally, we started some years ago to treat epilepsies associated with benign lesions in highly functional areas (eight patients). We have treated small benign lesions like dysplasia or periventricular heterotopia (Fig. 59-4). The high functional risk of any epilepsy surgery in these patients has led us ﬁrst to perform a cautious analysis of the extent of the epileptogenic zone relying on SEEG investigation. When the SEEG conﬁrms the ictal onset in the lesion or its close vicinity and when the EZ is conﬁned to a small area, radiosurgery is systematically discussed.

FIGURE 59-3. Radiosurgical anterior corpus-callosotomy in a young patient with a bilateral mesial frontal lobe epilepsy (MR negative). Dose planning displayed on a sagittal view (160 Gy at max).

59.

intractable epilepsies

FIGURE 59-4. Periventricular heterotopia in a young female with a severe drug-resistant epilepsy (24 Gy at the 50%). SEEG investigation has allowed location of the epileptogenic zone as conﬁned to a tiny part

Conclusion The ﬁeld of epilepsy surgery is a new and promising one for radiosurgery. However, determination of the extent of the epileptogenic zone requires speciﬁc expertise, which is crucial in order to achieve a reasonable rate of seizure cessation. In addition, the huge impact of ﬁne technical detail on the efﬁcacy and eventual toxicity of the procedure means that, at present, its use for these indications remains under evaluation, and further prospective works are absolutely required. It is difﬁcult to know whether we really are at the dawning of a broader indication for the use of radiosurgery in forthcoming years; our ability to identify the correct technical strategies should determine whether this is so!

References 1. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chirurg Scand 1951; 102:316–319. 2. Leksell L. Sterotaxic radiosurgery in trigeminal neuralgia. Acta Chir Scand 1971; 137:311–314. 3. Lindquist C, Kihlstrom L, Hellstrand E. Functional neurosurgery—a future for the gamma knife? Stereotact Funct Neurosurg 1991; 57:72–81. 4. Talairach J, Bancaud J, Szikla G, et al. Approche nouvelle de la neurochirurgie de l’epilepsie. Méthodologie stéréotaxique et résultats thérapeutiques. Neurochirurgie 1974; 20:92–98.

579

of the periventricular heterotopia in front of the posterior part of the left temporal lobe. Seizure cessation has occurred more than 1 year after radiosurgery. The patient has been able to stop the medication.

5. Talairach J, Bancaud J, Szikla G, et al. Approche nouvelle de la neurochirurgie de l’epilepsie. Méthodologie stéréotaxique et résultats thérapeutiques. In: Congrés Annuel de la Société de Langue Française, eds. Neurochirurgie. Marseille: Masson, 1974: 205–213. 6. Elomaa E. Focal irradiation of the brain: an alternative to temporal lobe resection in intractable focal epilepsy? Med Hypotheses 1980; 6:501–503. 7. Baudouin M, Stuhl L, Perrard A. Un cas d’épilepsie focale traité par la radiothérapie. Rev Neurol 1951; 84:60–63. 8. Von Wieser W. Die Roentgentherapie der traumatischen Epilepsie. Mschr Psychiat Neurol 1939; 101:422–424. 9. Heikkinen ER, Konnov B, Melnikow L. Relief of epilepsy by radiosurgery of cerebral arteriovenous malformations. Stereotact Funct Neurosurg 1989; 53:157–166. 10. Rogers L, Morris H, Lupica K. Effect of cranial irradiation on seizure frequency in adults with low-grade astrocytoma and medically intractable epilepsy. Neurology 1993; 43:1599–1601. 11. Rossi G, Scerrati M, Roselli R. Epileptogenic cerebral low grade tumors: effect of interstital stereotactic irradiation on seizures. Appl Neurophysiol 1985; 48:127–132. 12. Barcia-Salorio JL, Vanaclocha V, Cerda M, Roldan P. Focus irradiation in epilepsy. Experimental study in the cat. Appl Neurophysiol 1985; 48:152. 13. Gaffey C, Monotoya V, Lyman J, Howard J. Restriction of the spread of epileptic discharges in cats by mean of Bragg Peak intracranial irradiation. Int J Appl Radiat Isotope 1981; 32:779–787. 14. Chen ZF, Kamiryo T, Henson SL, et al. Anticonvulsant effects of gamma surgery in a model of chronic spontaneous limbic epilepsy in rats. J Neurosurg 2001; 94:270–280.

580

j. régis et al.

15. Maesawa S, Kondziolka D, Dixon C, et al. Subnecrotic stereotactic radiosurgery controlling epilepsy produced by kainic acid injection in rats. J Neurosurg 2000; 93:1033–1040. 16. Mori Y, Kondziolka D, Balzer J, et al. Effects of stereotactic radiosurgery on an animal model of hippocampal epilepsy. Neurosurgery 2000; 46:157–165; discussion 165–158. 17. Ronne-Engström E, Kihlström L, Flink R, et al. Gamma Knife surgery in epilepsy: an experimental model in the rat. Presented at European Society for Stereotactic and Functional Neurosurgery, 1993. 18. Barcia-Salorio JL, Garcia JA, Hernandez G, Lopez-Gomez L. Radiosurgery of epilepsy: long-term results. Presented at European Society for Stereotactic and Functional Neurosurgery, 1993. 19. Lindquist C. Gamma knife surgery in focal epilepsy. 1 year followup in 4 cases. 1992. 20. Lindquist C, Hellstrand E, Kilström L, et al. Stereotactic localisation of epileptic foci by magnetoencephalography and MRI followed by gamma surgery. Presented at International Stereotactic Radiosurgery Symposium, 1991. 21. Lindquist C, Kihlström L, Hellstrand E, Knutsson E. Stereotactic radiosurgery instead of conventional epilepsy surgery. Presented at European Society for Stereotactic and Functional Neurosurgery, 1993. 22. Regis J, Bartolomei F, Hayashi M, et al. The role of gamma knife surgery in the treatment of severe epilepsies. Epileptic Disord 2000; 2:113–122. 23. Regis J, Bartolomei F, de Toffol B, et al. Gamma knife surgery for epilepsy related to hypothalamic hamartomas. Neurosurgery 2000; 47:1343–1351; discussion 1351–1342. 24. Régis J, Hayashi M, Perez Eupierre L, et al. Gamma Knife surgery for epilepsy related to hypothalamic hamartomas. Acta Neurochir 2004; 91:33–50. 25. Régis J, Bartolomei F, Rey M, et al. Gamma knife surgery for mesial temporal lobe epilepsy. Epilepsia 1999; 40:1551–1556. 26. Regis J, Bartolomei F, Rey M, et al. Gamma knife surgery for mesial temporal lobe epilepsy. J Neurosurg 2000; 93(Suppl 3): 141–146. 27. Regis J, Bartolomei F, Hayashi M, Chauvel P. Gamma Knife surgery, a neuromodulation therapy in epilepsy surgery! Acta Neurochir Suppl 2002; 84:37–47. 28. Régis J, Kerkerian-Legoff L, Rey M, et al. First biochemical evidence of differential functional effects following gamma knife surgery. Stereotact Funct Neurosurg 1996; 66:29–38. 29. Regis J, Levivier M. Radiosurgery for intractable epilepsy. Tech Neurosurg 2003; 9:191–203. 30. Régis J, Peragut JC, Rey M, et al. First selective amygdalohippocampic radiosurgery for mesial temporal lobe epilepsy. Stereotact Funct Neurosurg 1994; 64:191–201. 31. Regis J, Roberts D. Gamma Knife radiosurgery relative to microsurgery: epilepsy. Stereotact Funct Neurosurg 1999; 72(Suppl 1): 11–21. 32. Regis J, Semah F, Bryan R, et al. Early and delayed MR and PET changes after selective temporomesial radiosurgery in mesial temporal lobe epilepsy. AJNR Am J Neuroradiol 1999; 20:213– 216. 33. Deonna T, Ziegler AL. Hypothalamic hamartoma, precocious puberty and gelastic seizures: a special model of “epileptic” developmental disorder. Epileptic Disord 2000; 2:33–37. 34. Kuzniecky R, Guthrie B, Mountz J, et al. Intrinsic epileptogenesis of hypothalamic hamartomas in gelastic epilepsy. Ann Neurol 1997; 42:60–67. 35. Munari C, Kahane P, Francione S, et al. Role of the hypothalamic hamartoma in the genesis of gelastic ﬁts (a video-stereo-EEG study). Electroencephalogr Clin Neurophysiol 1995; 95:154– 160.

36. Arita K, Ikawa F, Kurisu K, et al. The relationship between magnetic resonance imaging ﬁndings and clinical manifestations of hypothalamic hamartoma. J Neurosurg 1999; 91:212–220. 37. Valdueza JM, Cristante L, Dammann O, et al. Hypothalamic hamartomas: with special reference to gelastic epilepsy and surgery. Neurosurgery 1994; 34:949–958; discussion 958. 38. Palmini et al. 39. Rosenfeld JV, Harvey AS, Wrennall J, et al. Transcallosal resection of hypothalamic hamartomas, with control of seizures, in children with gelastic epilepsy. Neurosurgery 2001; 48:108–118. 40. Pascual-Castroviejo I, Moneo JH, Viano J, et al. [Hypothalamic hamartomas: control of seizures after partial removal in one case]. Rev Neurol 2000; 31:119–122. 41. Watanabe T, Enomoto T, Uemura K, et al. [Gelastic seizures treated by partial resection of a hypothalamic hamartoma]. No Shinkei Geka 21998; 6:923–928. 42. Cascino GD, Andermann F, Berkovic SF, et al. Gelastic seizures and hypothalamic hamartomas: evaluation of patients undergoing chronic intracranial EEG monitoring and outcome of surgical treatment. Neurology 1993; 43:747–750. 43. Regis J, Bartolomei F, Hayashi M, Chauvel P. What role for radiosurgery in mesial temporal lobe epilepsy. Zentralbl Neurochir 2002; 63:101–105. 44. Regis J, Rey M, Bartolomei F, et al. Gamma knife surgery in mesial temporal lobe epilepsy: a prospective multicenter study. Epilepsia 2004; 45:504–515. 45. Regis Y, Roberts DW. Gamma Knife radiosurgery relative to microsurgery: epilepsy. Stereotact Funct Neurosurg 1999; 72:11–21. 46. Hayashi M, Bartolomei F, Rey M, et al. MR changes after Gamma knife radiosurgery for mesial temporal lobe epilepsy: an evidence for the efﬁcacy of subnecrotic doses. In: Kondziolka D, ed. Radiosurgery. Basel: Karger, 2002:192–202. 47. Hayashi M, Regis J, Hori T. [Current treatment strategy with gamma knife surgery for mesial temporal lobe epilepsy]. No Shinkei Geka 2003; 31:141–155. 48. Bartolomei et al. 49. Srikijvilaikul T, Najm I, Foldvary-Schaefer N, et al. Failure of gamma knife radiosurgery for mesial temporal lobe epilepsy: report of ﬁve cases. Neurosurgery 2004; 54:1395–1402; discussion 1402–1394. 50. Barbaro N, Larson D, Laxer K, McDermott M. Radiosurgical treatment of temporal lobe epilepsy. In: Society AE, ed. 2004. 51. Martinez R, Gil-Nagel A. Epilepsy surgery with Gamma Knife in MTLE. Presented at International Stereotactic Radiosurgery Symposium, Bruxelles, 2005. 52. Cmelak AJ, Abou-Khalil B, Konrad PE, et al. Low-dose stereotactic radiosurgery is inadequate for medically intractable mesial temporal lobe epilepsy: a case report. Seizure 2001; 10:442–446. 53. Kawai K, Suzuki I, Kurita H, et al. Failure of low-dose radiosurgery to control temporal lobe epilepsy. J Neurosurg 2001; 95: 883–887. 54. Yang KJ, Wang KW, Wu HP, Qi ST. Radiosurgical treatment of intractable epilepsy with low radiation dose. Di Yi Jun Yi Da Xue Xue Bao 2002; 22:645–647. 55. Kihlström L, Guo WY, Lindquist C, Mindus P. Radiobiology of radiosurgey for refractory anxiety disorders. Neurosurgery 1995; 36:294–302. 56. Kihlstrom L, Hindmarsh T, Lax I, et al. Radiosurgical lesions in the normal human brain 17 years after gamma knife capsulotomy. Neurosurgery 1997; 41:396–401; discussion 401–392. 57. Barcia-Salorio JL, Barcia JA, Hernandez G, Lopez-Gomez L. Radiosurgery of epilepsy. Long-term results. Acta Neurochir Suppl (Wien) 1994; 62:111–113. 58. Grabenbauer GG, Reinhold C, Kerling F, et al. Frationated stereotactically guided radiotherapy of pharmacoresistant temporal lobe epilepsy. Acta Neurochirurg 2002; 84:65–70.

59.

intractable epilepsies

59. Stefan H, Hummel C, Grabenbauer GG, et al. Successful treatment of focal epilepsy by fractionated stereotactic radiotherapy. Eur Neurol 1998; 39:248–250. 60. Sun B, deSalles AA, Medin PM, et al. Reduction of hippocampalkindled seizure activity in rats by stereotactic radiosurgery. Exp Neurol 1998; 154:691–695. 61. Jenrow KA, Ratkewicz AE, Lemke NW, et al. Effects of kindling and irradiation on neuronal density in the rat dentate gyrus. Neurosci Lett 2004; 371:45–50. 62. Bartolomei F, Wendling F, Bellanger J, et al. Neural networks involving the medial temporal structures in temporal lobe epilepsy. Clin Neurophysiol 2001; 112:1746–1760. 63. Spencer S, Spencer D. Entorhinal-hippocampal interactions in medial temporal lobe epilepsy. Epilepsia 1994; 35:721–727. 64. Flickinger JC. An integrated logistic formula for prediction of complications from radiosurgery. Int J Radiat Oncol Biol Phys 1989; 17:879–885. 65. Régis J, Bartolomei F, Kida Y, et al. Radiosurgery of epilepsy associated with Cavernous malformation: retrospective study in 49 patients. Neurosurgery 2000; 47:1091–1097. 66. Jones R, Heinemann U, Lambert J. The entorhinal cortex and generation of seizure activity: studies of normal synaptic transmission and epileptogenesis in vitro. In: Avanzini G, Engel J, Fariello R, Heinemann U, eds. Neurotransmitters in Epilepsy. Baltimore: Elsevier, 1992:173–180. 67. Wilson W, Swartzwelder H, Anderson W, Lewis D. Seizure activity in vitro: a dual focus model. Epilepsy Res 1988; 2:289– 293. 68. Wieser HG, Siegel AM, Yasargil GM. The Zurich amygdalohippocampectomy series: a short up-date. Acta Neurochir Suppl (Wien) 1990; 50:122–127. 69. Whang CJ, Kwon Y. Long-term follow-up of stereotactic Gamma Knife radiosurgery in epilepsy. Stereotact Funct Neurosurg 1996; 66(Suppl 1):349–356.

581

70. Kitchen N. Experimental and clinical studies on the putative therapeutic efﬁcacy of cerebral irradiation (radiotherapy) in epilepsy. Epilepsy Res 1995; 20:1–10. 71. Glosser G, McManus P, Munzenrider J, et al. Neuropsychological function in adults after high dose fractionated radiation therapy of skull base tumors. Int J Radiat Oncol Biol Phys 1997; 38:231–239. 72. McCord MW, Buatti JM, Fennell EM, et al. Radiotherapy for pituitary adenoma: long-term outcome and sequelae. Int J Radiat Oncol Biology Phys 1997; 39:437–444. 73. Strojan P, Popovic M, Jereb B. Secondary intracranial meningiomas after high-dose cranial irradiation: report of ﬁve cases and review of the literature. Int J Radiat Oncol Biol Phys 2000; 48:65–73. 74. Simmons NE, Laws E. Gliomas occurence after sellar irradiation: case report and review. Neurosurgery 1998; 42:172–178. 75. Kaido T, Hoshida T, Uranishi R, et al. Radiosurgery-induced brain tumor. Case report. J Neurosurg 2001; 95:710–713. 76. Shamisa A, Bance M, Nag S, et al. Glioblastoma multiforme occurring in a patient treated with gamma knife surgery. Case report and review of the literature. J Neurosurg 2001; 94:816–821. 77. Yu JS, Yong WH, Wilson D, Black KL. Glioblastoma induction after radiosurgery for meningioma. Lancet 2000; 356:1576–1577. 78. Cahan W, Woodard H, Higginbotham N, et al. Sarcoma arising in irradiated bone: report of eleven cases. Cancer 1948; 1:3–29. 79. Ganz JC. Gamma knife radiosurgery and its possible relationship to malignancy: a review. J Neurosurg 2002; 97:644–652. 80. Ficker D, So E, Shen W. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 1998; 51:1270–1274. 81. Sperling M, Feldman H, Kinman J, et al. Seizure control and mortality in epilepsy. Ann Neurol 1999; 46:45–50. 82. Shrottner et al.

6 0

Epilepsy: Surgery Perspective Keith G. Davies and Edward Ahn

Introduction A seizure is a sudden change in behavior characterized by changes in sensory perception or motor activity due to brain abnormal electrical discharge. A seizure may be provoked (for example, by trauma or fever) or unprovoked. Epilepsy is the occurrence of two or more unprovoked seizures. Epilepsy is a symptom, not a single disease entity, and has many causes. The classiﬁcation of the International League Against Epilepsy distinguishes two dichotomies in epilepsy syndromes: partial or focal onset (“localization-related”) versus generalized onset, and idiopathic (genetic) versus symptomatic. Speciﬁc syndromes can be identiﬁed in each group. Seizures of partial onset may secondarily generalize. Many epilepsy syndromes are effectively treated with antiepileptic medications but some are particularly resistant to medical therapy. It is the symptomatic and in particular the localization-related epilepsies that tend to be medication resistant. For example, for mesial temporal lobe epilepsy, antiepileptic drugs will yield a seizure-free rate of only 25% of patients [1]. Surgery has a place in the management of these patients. Surgically, the treatment of epilepsy may be approached in three different ways. These are not mutually exclusive: 1. Resection of a “focus” or epileptogenic area, if one can be identiﬁed. A basic principle of resection is that, while the epileptogenic area is removed, the integrity of functioning, eloquent cortex (e.g., language, memory, sensorimotor) should be preserved. Resection of a solitary focus offers the best results for surgery. The most common site for focal onset is the temporal lobe, particularly mesial temporal lobe epilepsy. 2. Prevention of propagation of a discharge, either interhemispheric as with corpus callosum section, or locally with multiple subpial transection. 3. Chronic intermittent electrical stimulation. Vagus nerve stimulation is a novel treatment modality that has antiepileptic effects and has become widely used since it was approved by the U.S. Food and Drug Administration (FDA) in 1997. Direct stimulation of brain, either deep structures or cortex, has been introduced in recent years [2]. In order to obtain the best possible result for seizure control and minimize complications, a challenge for selecting the appro-

priate surgical intervention is identifying the patient’s epilepsy syndrome and, where relevant, localization. This entails an evaluation of the seizure semiology, EEG (interictal and ictal), brain structural imaging studies magnetic resonance imaging (MRI), metabolic imaging (magnetic resonance spectroscopy, PET, SPECT), functional imaging (magnetoencephalography, functional MRI), and the results of neuropsychology testing. The intracarotid amobarbital (Wada) test is used to evaluate lateralization of language function as well as to evaluate the integrity of memory function on each side [3]. It can also provide information on lateralization and localization of pathology: intracarotid amobarbital test memory score asymmetry is associated with the presence of hippocampal sclerosis [4]. The criteria for patient selection for surgery vary among surgical centers, but many will employ ictal video EEG monitoring, which allows characterization of seizure semiology as well as simultaneous EEG recording. If localization cannot be obtained from scalp EEG recordings, then invasive monitoring can be performed with either intracerebral depth electrodes [5] or subdural strip electrodes inserted through burr holes [6]. This can provide important information as noninvasive EEG recording has a false localization or lateralization rate of about 10% and may also exclude patients from consideration for surgery who would beneﬁt from surgery [7]. The best surgical results are for resection for seizures of unifocal onset. Patients who are not candidates for a focal resection may be offered vagus nerve stimulation or corpus callosotomy.

Focal Resections Temporal Resection The most common site for focal onset is the temporal lobe, particularly the mesial temporal lobe, and the most common pathology in this case is hippocampal sclerosis, characterized by gliosis, dendritic abnormalities, and neuronal cell loss of varying severity in the cornu ammonis of the hippocampus with relative sparing of the CA2 sector [8]. The etiology of hippocampal sclerosis is not fully understood but it is associated with a distinct syndrome comprising early age of epilepsy onset, usually in childhood, early risk factors for epilepsy such as febrile

583

584

k.g. davies and e. ahn

convulsions in childhood, and MRI ﬁndings suggestive of hippocampal sclerosis such as unilateral atrophy, increased signal, and loss of the internal anatomy of the hippocampus [9]. The presence or absence of this syndrome has implications for the outcome for seizure control as well as for risks to cognitive function after surgery. Hippocampal sclerosis is the underlying pathology in about 60% of patients with mesial temporal lobe epilepsy (MTLE); low-grade intrinsic tumors, hamartomas or malformations, and neuronal heterotopia account for most of the remainder [8]. Some patients, localized by invasive EEG monitoring, have no identiﬁable pathology and the resected hippocampus is normal histopathologically [10]. Penﬁeld pioneered anterior temporal lobectomy (ATL) for intractable epilepsy in the 1940s and 1950s [11]. His technique was to operate on awake patients using local anesthesia, with intraoperative stimulation mapping that identiﬁed areas of eloquent cortex mediating somatosensory and motor function and, in the left hemisphere, language function, and intraoperative electrocorticography to identify areas of epileptogenesis. Penﬁeld’s practice was to “tailor” a resection according to the ﬁndings of intraoperative electrocorticography with the goal of maximizing the resection of epileptogenic tissue while preserving areas mediating function. With regard to language, Penﬁeld estimated that, in the left language-dominant hemisphere, the temporal (Wernicke) language area was 5 to 6 cm posterior to the temporal pole [11–13]. As it became clear that the ictal onset region was associated with mesial temporal pathology, particularly hippocampal sclerosis, in most cases of temporal lobe epilepsy, a “standard” ATL was introduced, such as the en bloc resection pioneered by Falconer [14, 15]. Falconer’s practice was to extend the resection 6 cm posterior to the temporal pole and, on the language-dominant side, to preserve the superior temporal gyrus, believing that language-eloquent cortex may reside within it [14]. Such standard ATL has become accepted surgical practice for epilepsy of temporal lobe onset. “Standard” in this context is something of a misnomer as there is variability in the details of the way in which the resection can be performed and in the extent of the lateral resection. The superior temporal gyrus may be either preserved or resected, and the resection may entail en bloc removal of lateral and mesial structures together, or separate dissection of the hippocampus after excision of the lateral temporal lobe, but resection of some lateral temporal neocortex, as well as mesial structures including hippocampus, is common to all of these procedures. Figure 60-1 illustrates standard ATL. For patients in whom the pathology and/or seizure onset has been identiﬁed to be mesial temporal, a modiﬁcation of the temporal resection is selective amygdalohippocampectomy, which entails no resection of lateral temporal cortex. A transcortical approach was ﬁrst described by Niemeyer in 1958 [16], and more recently a trans-sylvian approach has been popularized by Yasargil [17]. Figure 60-2 illustrates this procedure. A special case for temporal lobectomy is resection in the left, language-dominant hemisphere and whether language mapping should be performed before undertaking a resection of lateral temporal cortex. Ojemann has emphasized the variability of language sites in the left hemisphere that cannot be predicted by anatomical criteria alone [18]. Many workers therefore advocate language mapping, either intraoperatively

Sylvian fissure

3 3

1

4

Ventricle 2

Wing of sphenoid

Free edge of tentorium

5

Cut 1 6cm behind temporal pole to enter temporal horn of lateral vetricle Cut 2 Extension inferiorly to free edge of tentorium Cut 3 Along Sylvian fissure; on the dominant side the posterior 2/3 of the superior temporal gyrus is spread. This incision extends into the temporal horn by working from cut 1 Cut 4 Pia arachnoid cut over temporal pole Cut 5 Pia arachnoid cut just lateral to the free edge of tentorium

3

3

1

4 2

Temporal lobe 5

Free edge of tentorium

FIGURE 60-1. Diagram of the tissue typically removed and the technique for a standard anterior temporal lobectomy. (From Adams CBT. A Neurosurgeon’s Notebook. Oxford: Blackwell Science, 1998. Used with permission.)

in awake patients [18], or extraoperatively with implanted subdural grid electrodes [19]. This latter technique provides the additional advantage of allowing ictal EEG recordings for identifying the “epileptogenic zone” [19, 20]. Figure 60-3 illustrates a subdural grid electrode. For appropriately selected patients, temporal resection that includes the mesial structures (“limbic resection”) yields seizure-free rates of about 70%. In a multicenter survey of more than 3500 patients, the mean seizure-free (no seizures for at least 2 years) rate was 68% [21]. A further 24% were improved and 8% had no improvement. Selective amygdalohippocampectomy yields similar rates: 69% seizure-free, 26% improved, and 9% unchanged [21]. Seizure outcome results are better

60.

epilepsy: surgery perspective

585

for patients with hippocampal sclerosis compared with those without obvious pathology [22]. Complications of standard ATL include hemiparesis, usually due to vascular injury, either to the middle cerebral artery or to perforating branches from the internal carotid or posterior cerebral arteries. An upper quadrantic homonymous visual ﬁeld defect commonly occurs as a result of resection within Meyer’s loop of the optic radiation [23]. Decline in cognitive function, especially memory and language, can occur, particularly for resections on the language-dominant side. Right nondominant resections have relatively low risk for memory function [24]. The hippocampus is known to mediate episodic memory [25], and memory is particularly at risk in patients without hippocampal sclerosis [26]. The facet of language function that is especially at risk with left ATL is naming [27], and it is patients without hippocampal sclerosis and/or later epilepsy onset age (postadolescent) who are most vulnerable to declines

Sylvian fissure

FIGURE 60-3. Intraoperative photograph of implanted subdural grid electrodes on the left hemisphere. Rectangles outline electrodes where language responses were obtained. SF, sylvian ﬁssure. (From Jabbour RA et al. Atypical language cortex in the left temporal lobe. Relationship to bilateral language. Neurology 2004; 63:1833–1837. Used with permission.)

Sylvian vein Sylvian fissure incised to expose temporal horn

Middle cerebal artery and branches

Temporal horn Hippocampus anteriorly forming the pes hippocampus Fimbria Choroid plexus Amygdala

Dotted line indicates excision of amygdala and hippocampus

FIGURE 60-2. Diagram of the tissue removed and the approach for a trans-sylvian amygdalohippocampectomy. (From Adams CBT. A Neurosurgeon’s Notebook. Oxford: Blackwell Science. 1998. Used with permission.)

in naming ability [28, 29]. Patients with early (childhood) epilepsy onset tend to remain cognitively stable. Naming is classically associated with left temporal neocortical function [30], and it is not fully understood whether the naming decline may be due to resection of functioning hippocampus or whether the absence of hippocampal sclerosis is a marker for later epilepsy onset and the absence of reorganization of lateral temporal language that can occur with early epilepsy onset [28]. There is, however, also evidence that even with language mapping, naming is still at risk and, similar to the case for standard ATL, a risk factor is the absence of hippocampal sclerosis and/or later age of epilepsy onset [31]. Furthermore, there is also some evidence that a purely mesial temporal resection such as selective amygdalohippocampectomy results in naming decline, and the same risk factors apply [32]. A further complication of temporal resection that should be mentioned is the exacerbation, or the development de novo, of psychiatric disorders, particularly the interictal dysphoric disorder that is common with chronic intractable localizationrelated epilepsy. Blumer et al. reported that of 44 patients undergoing focal resection, 57% had interictal dysphoria preoperatively. After surgery, 39% developed either de novo psychiatric complications or exacerbation of preexisting interictal dysphoria [33].

Lesionectomy Versus Epilepsy Surgery Some mass lesions such as low-grade intrinsic tumors, cavernous malformations, or arteriovenous malformations are particularly epileptogenic. A controversy relating to the

586

k.g. davies and e. ahn

results depend upon the presence or absence of a mass lesion. In a multicenter survey, Engel found that in the presence of a lesion, 67% of patients became seizure-free, 21% were improved, and 12% were unchanged, whereas for nonlesional resections, 45% became seizure-free, 35% were improved, and 20% were unchanged [21].

Hypothalamic Hamartoma

FIGURE 60-4. Coronal MRI showing a right mesial temporal cavernous malformation.

resection of an epileptogenic mass lesion is whether to resect the lesion alone (“lesionectomy”) or whether to also resect adjacent epileptogenic cortex (“seizure surgery”). Many studies show that lesionectomy is effective in abolishing seizures, with seizure-free rates of at least 75% after lesionectomy for vascular malformations and low-grade intrinsic tumors [34–39]. Some studies have shown no difference in seizure outcome comparing lesionectomy with seizure surgery [39–41], but there is evidence that additional resection of epileptogenic cortex increases the likelihood of being seizure-free [42–44]. Furthermore, it is also known that completeness of resection of the lesion will result in better seizure control [38–39, 45], and the apparently superior results for seizure surgery may result from the resection of surrounding pathology in relation to the lesion rather than resection of areas of brain that have been “kindled” in response to repeated seizures [46]. For example, seizure outcome after resection of cavernous malformation (Fig. 60–4) is better when surrounding hemosiderin is resected as well as the lesion itself [47].

A distinct syndrome is the association of gelastic epilepsy (seizures with laughter as an ictal phenomenon) and hypothalamic hamartoma, ﬁrst described in 1957 [48]. These patients usually present in infancy or childhood, and the seizures tend to be medication resistant, progressively worsen over time, and other seizure types develop [49]. Behavioral disturbance is common, and precocious puberty may also be associated. Hypothalamic hamartomas may be sessile or pedunculated and contain nonneoplastic gray matter composed of hyperplastic neurons of varying size [50]. Figure 60-5 illustrates a sessile hypothalamic hamartoma. Initial surgical approaches to the treatment of epilepsy associated with hypothalamic hamartomas were based on the assumption that the seizures were arising in adjacent cortex, and cortical resections as well as corpus callosotomy were undertaken but with disappointing results [51, 52]. Subsequently, evidence accumulated that the epileptogenesis is in relation to the hamartoma itself [53, 54]. Considerations for direct surgery on the hypothalamus are its relative inaccessibility so that resection is technically demanding, and the risk of compromise of function of surrounding normal hypothalamus so that complications may be anticipated. Resections have often been subtotal, only partially effective in controlling seizures, and associated with signiﬁcant complications [49, 55]. There are three approaches: (1) transcallosal interforniceal approach, entering the third ventricle from superiorly; (2) pterional approach, exposing the inferior and lateral aspect of the hypothalamus; (3) anterior midline approach, entering the third ventricle through the lamina terminalis [56, 57]. Endoscopy may also be used with these approaches. In recent years, there have been two reports of surgical series conﬁrming that resection can result in good seizure relief but with signiﬁcant risk of complications. Rosenfeld et al.

Extratemporal Resection The most common site for extratemporal resection is frontal lobe, followed by occipital lobe. Likely underlying pathologies will be low-grade intrinsic tumors, hamartomas, malformations, and cortical dysplasia, but in many patients no obvious pathology can be identiﬁed, and localization of epileptogenesis is guided by corticography, especially from implanted subdural grid electrodes. The presence of possible functional areas may be a consideration depending on location, and intraoperative or extraoperative mapping with implanted subdural grid electrodes will be appropriate. Seizure outcome rates for extratemporal neocortical resections are not as good as for limbic resections. However, the

FIGURE 60-5. Sagittal (left) and coronal (right) MRI images showing a sessile hypothalamic hamartoma. (From Valdueza JM et al. Hypothalamic hamartomas: with special reference to gelastic epilepsy and surgery. Neurosurgery 1994; 34:949–958. Used with permission.)

60.

epilepsy: surgery perspective

reported ﬁve children operated on via the transcallosal interforniceal approach [58]. Complete or near-complete resection was achieved in all ﬁve patients. Three patients became seizurefree and two experienced occasional brief mild gelastic seizures. There was marked improvement in behavior in all ﬁve. There were no signiﬁcant neurologic complications, two patients had mild transient diabetes insipidus, and two increased appetite. Palmini et al. reported on the experience at four centers with 13 patients [59]. Two patients became seizure-free, and 11 had greater than 90% reduction or abolition of generalized seizure although they continued to have complex partial or absence seizures. Behavior improved in all patients. However, complications were signiﬁcant. Three patients had anterior thalamic infarcts, and one patient a capsular infarct. However, longlasting deﬁcits were minor.

587

Anterior partial (usually two thirds to three quarters) callosotomy does not usually have signiﬁcant cognitive consequences, but complete callosotomy often produces a hemisphere disconnection syndrome that can result in disabling symptoms of intermanual conﬂict [68]. The failure of interhemispheric transfer can be demonstrated in the conventionally organized patient (left hemisphere dominant for language) by failure of the patient to be able to name an object placed in the left hand (which after callosotomy has access only to the mute right hemisphere). Similarly, objects placed in the left visual ﬁeld cannot be named although the left hand (connected to the same hemisphere) can retrieve the shown object. It is because of the potential for the development of this syndrome that partial callosotomy is usually undertaken in the ﬁrst instance, and the section can be completed if there is an inadequate seizure control response.

Hemispherectomy Hemispherectomy, or more correctly, cerebral hemidecortication, for many years was the most effective procedure for eradicating seizures and resulted in seizure-free rates of about 80% [60]. However, the procedure is of limited application and is used only for patients with a unilaterally damaged hemisphere who already have a hemiparesis. Resection of epileptogenic hemisphere does not then result in an increase in the deﬁcit. Usually, the etiology is vascular damage from perinatal trauma, but infection, cortical dysplasia, and extensive vascular malformations have also been described. Krynauw ﬁrst described hemispherectomy for epilepsy in 1950 with a report of 12 patients with excellent outcome for seizure control. Interest in the procedure was stimulated, but subsequent reports of late complications such as hydrocephalus, cerebral hemosiderosis and intermittent bleeding into the resection cavity, and with a high mortality rate, resulted in a decline in enthusiasm [61]. Subsequent modiﬁcations by Rasmussen [62] and Adams [63] were able to reduce the complication rates while preserving the therapeutic beneﬁt. More recently, periinsular hemispherotomy has been described, a procedure that disconnects the entire hemisphere without the need for any removal of brain substance [64]. Hemispherectomy remains a valuable procedure for alleviation of seizures in appropriately selected patients.

Multiple Subpial Transection Local disconnection of epileptogenic cortex can be achieved by the technique of multiple subpial transection (MST) developed by Morrell [69]. The theoretical basis for this technique is that horizontal intragriseal ﬁbers are transected thereby preventing local synchronization of epileptogenic activity. Because the columnar organization is maintained, MST within eloquent cortex should not result in loss of function. This technique has application where resection is not feasible in view of the presence of function, and epileptogenicity has been identiﬁed with corticography. MST can be performed in combination with a resection, so it may be difﬁcult to ascertain how much beneﬁt and morbidity may be due to the transection rather than the resection, particularly as a single center will not be likely to have experience with large numbers of patients undergoing MST. A meta-analysis by Spencer et al. of 211 patients undergoing MST found that for those undergoing MST without resection, there was an excellent outcome (deﬁned as >95% reduction in seizures frequency) in 62% to 71% of patients [70]. This was only slightly less than patients undergoing MST with resection. New neurologic deﬁcits developed in 19% of those undergoing MST without resection.

Vagus Nerve Stimulation Corpus Callosotomy First described in 1940 [65], corpus callosotomy has been used for the treatment of generalized epilepsy or partial epilepsy with secondary generalization where the onset cannot be localized. Few patients become free of seizures after surgery. A multicenter survey found about 8% of patients became seizurefree, 70% were improved, and 31% were unchanged after callosotomy [21]. Wyler reported a series of 66 patients of whom 11% became seizure-free, and 68% had between 50% and 95% reduction in seizure frequency [66]. Corpus callosotomy is particularly effective for the drop seizures of Lennox-Gastaut syndrome [67], a symptomatic epilepsy syndrome comprising multiple seizure types, developmental delay, and abnormal interictal EEG. Drop seizures can result in signiﬁcant injuries.

Vagus nerve stimulation (VNS) is a relatively recently developed, novel treatment for intractable epilepsy. The desynchronization of electrocerebral activity by VNS was ﬁrst proposed by Zabara in 1985 [71]. VNS entails the subcutaneous implantation of a battery-driven pulse generator in the left upper chest connected to electrodes applied to the left vagus nerve (Fig. 60-5). The pulse generator may be programmed transcutaneously after implantation. The left vagus nerve is selected in view of the theoretical risk of cardiac dysrhythmia from innervation of the atrioventricular node by the right vagus nerve. The ﬁrst patient to be treated with VNS was implanted in 1988, and after multicenter, randomized, double-blind studies demonstrated a signiﬁcant beneﬁt. VNS therapy was approved by the FDA in 1997 for use in partial-onset epilepsy in patients over 12 years of age.

588

k.g. davies and e. ahn

The mechanism of action of VNS is unknown. Retrograde stimulation of the vagus nerve results in brain-stem stimulation via the dorsal nucleus of the vagus. The locus coeruleus is essential for the antiepileptic effect [72] implying that it is mediated by the release of norepinephrine. PET blood ﬂow studies have shown widespread changes in the brain, including increased blood ﬂow in cerebellar hemispheres, basal ganglia, insulae, hypothalamus, and right postcentral gyrus, and reduced blood ﬂow in limbic structures [73]. Changes in cerebrospinal ﬂuid neurotransmitter concentrations have been found, particularly increase in the inhibitory neurotransmitter GABA, as well as decreases in aspartate and glutamate [74]. VNS reduces seizure frequency and severity, but few patients become seizure-free. The four preapproval studies were prospective and randomized and compared high- with low-stimulation parameters. In the high-stimulation group, the median reduction in seizure frequency was 25% to 28%, 31% of patients had >50% reduction in seizure frequency, and 11% of patients >75% reduction in seizure frequency (for a review, see Ref. [75]). The beneﬁcial effect increases progressively with time; at 2 years, the mean reduction in seizure frequency was 41%. A recent long-term study found further progressive improvement, with, in this study, a mean seizure frequency reduction of 26% at 1 year, 30% after 5 years, and 52% after 12 years [76]. Seventy-nine percent of patients had >50% reduction in seizure frequency at last follow-up. Although approved by the FDA only for partial epilepsy, VNS has also been shown to be effective for generalized epilepsy [77, 78] and in children as well as adults [79, 80]. Labar et al. showed that in patients with generalized epilepsy, VNS produced a mean reduction in seizure frequency at 3 months of 46%, and 46% of patients had >50% reduction in seizure frequency [78]. Murphy et al. found that VNS in children gave a mean seizure frequency reduction of 31% at 6 months, and 42% at 18 months [79]. In patients with hypothalamic hamartoma, VNS has been shown to be very effective for control of seizures as well as behavior [81]. VNS is effective in children with Lennox-Gastaut syndrome [82]. A recent publication comparing VNS with corpus callosotomy found that VNS was not as effective as callosotomy for reducing generalized tonic-clonic seizures (50% vs. 79% seizure frequency reduction >50%), but the complication rate was higher for callosotomy, a more invasive procedure [83]. Stimulation may also be initiated by application of a magnet over the pulse generator at the ﬁrst indication of a seizure, and this results in reduced seizure severity in about 50% of patients [84]. Many patients ﬁnd this to be an important aspect of VNS therapy that gives some sense of control over an unpredictable condition. VNS is associated with side effects due to stimulation of the vagus nerve and its branches, including hoarseness, dysphonia, coughing, dysphagia and throat pain, which usually respond to modiﬁcation of the stimulation parameters and are well tolerated and resolve over time. VNS may be associated with the same psychiatric complications that follow signiﬁcant reduction in seizures by cortical resection, as well as antiepileptic medications [85].

Stereotactic Radiosurgery The antiepileptic effect of stereotactic radiosurgery (SRS) was ﬁrst demonstrated after its use to treat vascular malformations (Fig. 60-6) and tumors [86, 87]. It is in the treatment of partial symptomatic epilepsy that SRS is establishing a place as a therapeutic option, particularly for epileptogenic vascular malformations, hypothalamic hamartomas, and MTLE [52, 88–91]. SRS has been used for callosotomy; Pendl et al. described three patients who had ablation of the anterior third of the corpus callosum [92]. The severity and frequency of seizures were reduced in all three patients, and there were no complications apart from transient headache. Régis et al. reported a study of 49 patients with cavernous malformations and epilepsy treated with SRS at ﬁve centers [89]. They found that 53% of patients became seizure-free; 20% had signiﬁcant improvement and 26% were unchanged. The mean follow-up time was 24 months. Mesial temporal location was associated with poorer outcome. One patient had a hemorrhage and another developed radiation-induced edema. Schäuble et al. reported on 65 patients with cerebral arteriovenous malformations and found that, at a median of 48 months after treatment, 51% of patients were seizure-free, and 78% had signiﬁcant improvement in their seizures [91]. Smaller malformation size and a low seizure frequency score before treatment were associated with better outcome. Régis et al. have also reported on SRS treatment of hypothalamic hamartoma [88]. In a report of 10 patients treated at seven centers, they found that four patients became seizurefree, two had rare seizures, and two had improvement but continued to have generalized seizures. Two patients were considered to have insufﬁcient improvement, were treated again, and became seizure-free. Behavior improved in two patients. No complications were encountered apart from poikilothermia in one patient.

Stimulation site lead the VNS device

FIGURE 60-6. Diagram of the location of the implanted components for vagus nerve stimulation. (From Uthman BM et al. Treatment of epilepsy by stimulation of the vagus nerve. Neurology 1993; 43:1338– 1345. Used with permission.)

60.

epilepsy: surgery perspective

Selch et al. reported on three patients with hypothalamic hamartomas treated with SRS [93]. Two patients became seizure-free; one had a progressive decline in seizure frequency. There were no complications. It is in the treatment of MTLE, a more common syndrome than vascular malformations or hypothalamic hamartomas, that SRS may ﬁnd more widespread application for epilepsy. The ﬁrst treatment of MTLE by SRS was in 1993 in Marseille, and Régis et al. reported a series of seven patients in 1999 [90]. One patient remained seizure-free after treatment, the others had progressive decline in seizure frequency with cessation at about 10 months, and at follow-up at 24 to 61 months all except one were seizure-free. Subsequently, Régis et al. reported the results of a prospective study of 21 patients treated at three centers [94]. Thirteen patients had left MTLE. Twenty patients were available for follow-up, and the minimum follow-up was 2 years. The frequency of seizures progressively declined at each 6-month follow-up evaluation. The median seizure frequency was 6 prior to treatment and 0.3 at 2 years after treatment. At 2 years, 65% of patients were seizure-free. Five patients had transient side effects such as depression, headache, nausea, vomiting, and imbalance. Eight patients developed a quadrantic ﬁeld defect and one patient a hemianopia. At 2 years, another patient developed a quadrantic ﬁeld defect. No cognitive deﬁcits were observed. However, all these patients had hippocampal sclerosis on MRI and may therefore be expected to remain stable cognitively. One report of the failure of SRS to alleviate seizures came from Srikijvilaikul et al., who reported ﬁve patients with hippocampal sclerosis [95]. None became seizure-free. Two patients subsequently died of complications related to seizures, and three eventually had temporal resections that resulted in their becoming seizure-free. A prospective multicenter study was started in 2000 in the United States for patients with intractable epilepsy and MRI evidence of hippocampal sclerosis. In this study, patients who would be candidates for a temporal resection are offered SRS as an alternative treatment. A progress update in 2002 of 10 patients reported that none were seizure-free although four patients had >80% reduction in seizures [96]. The longest follow-up at that time was 18 months. The authors noted that after a period of 9 to 15 months, patients reported headaches, then an increase in auras, followed by a decrease in complex partial seizures. MRI demonstrated cerebral edema at this time. One patient developed papilledema as a result of the swelling and underwent an ATL.

Conclusion At present, there are relatively few patients treated for epilepsy with SRS to make reliable comparisons between surgery and SRS in the treatment of epilepsy, and follow-up periods are not sufﬁciently long to judge either long-term seizure outcomes or potential complications of SRS. Nevertheless, at present, SRS offers an attractive, noninvasive treatment option for lesions that are operatively technically challenging such as hypothalamic hamartomas; the seizure outcome results appear to be comparable and short-term complications less with SRS. For

589

vascular malformations, the results from studies suggest that SRS may not produce the excellent seizure-free rates that can result from resection, and, particularly for readily accessible lesions, resection may still be the treatment of choice. For deeply placed lesions, however, SRS would be an attractive alternative. Surgery would still be feasible if the results of SRS are unsatisfactory, as has been reported several times. For MTLE, the results for seizure control with SRS are comparable with temporal resection. One disadvantage of SRS is the delay in response of seizure control after treatment. SRS appears to be relatively safe, although the rate of visual ﬁeld defect is certainly not less than with resection. No cognitive sequelae have been reported so far, but patients with MTLE treated with SRS have all had hippocampal sclerosis, and it is difﬁcult to believe that the risks to memory and naming after treatment of left MTLE would be any less with SRS than they are with resection. Further studies should clarify the situation.

References 1. Glaser GH. Natural history of temporal lobe-limbic epilepsy. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. New York: Raven Press, 1987:13–30. 2. Morrell M. Brain stimulation for epilepsy: can scheduled or responsive neurostimulation stop seizures? Curr Opin Neurol 2006; 19:164–168. 3. Loring DW, Lee GP, Meador KJ. Intracarotid amobarbital (Wada) assessment. In: Wyler AR, Hermann BP, eds. The Surgical Management of Epilepsy. Boston: Butterworth-Heinemann, 1994: 97–110. 4. Davies KG, Bell BD, Dohan FC Jr, et al. Prediction of presence of hippocampal sclerosis from intracarotid amobarbital procedure memory asymmetry scores and epilepsy on set age. Epilepsy Res 1999; 33:117–123. 5. Spencer SS, Spencer DD, Williamson PD, et al. The localizing value of depth electroencephalography in 32 patients with refractory epilepsy. Ann Neurol 1982; 12:248–253. 6. Wyler AR, Ojemann GA, Lettich E, et al. Subdural strip electrodes for localizing epileptogenic foci. J Neurosurg 1984; 60:1195–1200. 7. Wyler AR, Ritchie ET, Hermann BP. Comparison of scalp to subdural recordings for localizing epilptogenic foci. J Epilepsy 1989; 2:91–96. 8. Babb TL, Brown WJ. Pathological ﬁndings in epilepsy. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. New York: Raven Press, 1987:511–540. 9. Wieser HG, Engel J Jr, Williamson PD, et al. Surgically remediable temporal lobe syndromes. In Engel J Jr, ed. Surgical Treatment of the Epilepsies. New York: Raven Press, 1993:49–63. 10. Wyler AR, Dohan FC Jr, Schweitzer JB, et al. A grading system for mesial temporal pathology (hippocampal sclerosis) from anterior temporal lobectomy. J Epilepsy 1992; 5:220–225. 11. Penﬁeld W, Baldwin M. Temporal lobe seizures and the technic of subtotal temporal lobectomy. Ann Surg 1952; 136:625–634. 12. Penﬁeld W, Jasper H. Epilepsy and the Functional Anatomy of the Human Brain. Boston: Little, Brown, 1954. 13. Penﬁeld W, Roberts L. Speech and Brain Mechanisms. Princeton: Princeton University Press, 1959. 14. Falconer MA. Anterior temporal lobectomy for epilepsy. In: Symon L, ed. Operative Surgery: Fundamental International Techniques. London: Butterworths, 1979:315–322.

590

k.g. davies and e. ahn

15. Falconer MA, Hill D, Meyer A, et al. Clinical, radiological, and EEG correlations with pathological changes in temporal lobe epilepsy and their signiﬁcance in surgical treatment. In: Baldwin M, Bailey P, eds. Temporal Lobe Epilepsy. Springﬁeld: Charles Thomas, 1958:396–410. 16. Niemeyer P. The transventricular amygdala-hippocampectomy in temporal lobe epilepsy. In: Baldwin M, Bailey P, eds. Temporal Lobe Epilepsy. Springﬁeld: Charles Thomas, 1958:461–482. 17. Yasargil MG, Teddy PJ, Roth P. Selective amygdalohippocampectomy. Operative anatomy and surgical technique. In: Symon L, ed. Advances and Technical Standards in Neurosurgery. Vienna: Springer-Verlag, 1985:93–123. 18. Ojemann G, Ojemann J, Lettich E, et al. Cortical language localization in left, dominant hemisphere. An electrical stimulation mapping investigation in 117 patients. J Neurosurg 1989; 71: 316–326. 19. Lüders HO, Awad I, Wyllie E, et al. Functional mapping of language abilities with subdural electrode grids. In: Wyler AR, Hermann BP, eds. The Surgical Management of Epilepsy. Boston: Butterworth-Heinemann, 1994:70–77. 20. Lesser RP, Gordon B, Fisher R, et al. Subdural grid electrodes in surgery of epilepsy. In: Lüders H, ed. Epilepsy Surgery. New York: Raven Press, 1991:399–408. 21. Engel J Jr, Van Ness PC, Rasmussen TB, et al. Outcome with respect to epileptic seizures. In: Engel J, Jr., ed. Surgical Treatment of the Epilepsies. New York: Raven Press, 1993:609–622. 22. Davies KG, Hermann BP, Dohan FC Jr, et al. Relationship of hippocampal sclerosis to duration and age of onset of epilepsy, and childhood febrile seizures in temporal lobectomy patients. Epilepsy Res 1996; 24:119–126. 23. Wyler AR. Focal cortical resections. In: Wyler AR, Hermann BP, eds. The Surgical Management of Epilepsy. Boston: ButterworthHeinemann, 1994:129–138. 24. Davies KG, Bell BD, Bush AJ, et al. Prediction of verbal memory loss in individuals after anterior temporal lobectomy. Epilepsia 1998; 39:820–828. 25. Squire LR, Zola-Morgan S. The medial temporal lobe memory system. Science 1991; 253:1380–1386. 26. Hermann BP, Seidenberg M, Haltiner A, et al. Relationship of age at onset, chronologic age, and adequacy of preoperative performance to verbal memory change after anterior temporal lobectomy. Epilepsia 1995; 36:137–145. 27. Hermann BP, Wyler AR, Somes G. Language function following anterior temporal lobectomy [see comments]. J Neurosurg 1991; 74:560–566. 28. Davies KG, Bell BD, Bush AJ, et al. Naming decline after left anterior temporal lobectomy correlates with pathological status of resected hippocampus. Epilepsia 1998; 39:407–419. 29. Hermann BP, Wyler AR, Somes G, et al. Dysnomia after left anterior temporal lobectomy without functional mapping: frequency and correlates. Neurosurgery 1994; 35:52–56; discussion 56–57. 30. McCarthy RA, Warrington EK. Cognitive Neuropsychology. New York: Academic Press, 1990. 31. Davies KG, Risse GL, Gates JR. Naming ability after tailored left temporal resection with extraoperative language mapping: increased risk of decline with later epilepsy onset age. Epilepsy Behav 2005; 7:273–278. 32. Bartha L, Trinka E, Ortler M, et al. Linguistic deﬁcits following left selective amygdalohippocampectomy: a prospective study. Epilepsy Behav 2004; 5:348–357. 33. Blumer D, Wakhlu S, Davies K, et al. Psychiatric outcome of temporal lobectomy for epilepsy: incidence and treatment of psychiatric complications. Epilepsia 1998; 39:478–486. 34. Casazza M, Avanzini G, Ciceri E, et al. Lesionectomy in epileptogenic temporal lobe lesions: preoperative seizure course and postoperative outcome. Acta Neurochir Suppl 1997; 68:64–69.

35. Casazza M, Broggi G, Franzini A, et al. Supratentorial cavernous angiomas and epileptic seizures: preoperative course and postoperative outcome. Neurosurgery 1996; 39:26–32; discussion 32– 24. 36. Cohen DS, Zubay GP, Goodman RR. Seizure outcome after lesionectomy for cavernous malformations. J Neurosurg 1995; 83: 237–242. 37. Dodick DW, Cascino GD, Meyer FB. Vascular malformations and intractable epilepsy: outcome after surgical treatment. Mayo Clin Proc 1994; 69:741–745. 38. Fried I. Management of low-grade gliomas: results of resections without electrocorticography. Clin Neurosurg 1995; 42:453–463. 39. Kim SK, Wang KC, Hwang YS, et al. Intractable epilepsy associated with brain tumors in children: surgical modality and outcome. Childs Nerv Syst 2001; 17:445–452. 40. Britton JW, Cascino GD, Sharbrough FW, et al. Low-grade glial neoplasms and intractable partial epilepsy: efﬁcacy of surgical treatment. Epilepsia 1994; 35:1130–1135. 41. Rossi GF, Pompucci A, Colicchio G, et al. Factors of surgical outcome in tumoural epilepsy. Acta Neurochir 1999; 141:819– 824. 42. Jooma R, Yeh HS, Privitera MD, et al. Lesionectomy versus electrophysiologically guided resection for temporal lobe tumors manifesting with complex partial seizures. J Neurosurg 1995; 83: 231–236. 43. Rassi-Neto A, Ferraz FP, Campos CR, et al. Patients with epileptic seizures and cerebral lesions who underwent lesionectomy restricted to or associated with the adjacent irritative area. Epilepsia 1999; 40:856–864. 44. Weber JP, Silbergeld DL, Winn HR. Surgical resection of epileptogenic cortex associated with structural lesions. Neurosurg Clin N Am 1993; 4:327–336. 45. Iannelli A, Guzzetta F, Battaglia D, et al. Surgical treatment of temporal tumors associated with epilepsy in children. Pediatr Neurosurg 2000; 32:248–254. 46. Morrell F, Wada J, Engel J Jr. Potential relevance of kindling and secondary epileptogenesis to the consideration surgical treatment for epilepsy. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. New York: Raven Press, 1987:701–707. 47. Baumann CR, Schuknecht B, Lo Russo G, et al. Seizure outcome after resection of cavernous malformations is better when surrounding hemosiderin-stained brain also is removed. Epilepsia 2006; 47:563–566. 48. Daly DD, Mulder DW. Gelastic epilepsy. Neurology 1957; 7:189– 192. 49. Berkovic SF, Andermann F, Melanson D, et al. Hypothalamic hamartomas and ictal laughter: evolution of a characteristic epileptic syndrome and diagnostic value of magnetic resonance imaging. Ann Neurol 1988; 23:429–439. 50. Valdueza JM, Cristante L, Dammann O, et al. Hypothalamic hamartomas: with special reference to gelastic epilepsy and surgery. Neurosurgery 1994; 34:949–958; discussion 958. 51. Cascino GD, Andermann F, Berkovic SF, et al. Gelastic seizures and hypothalamic hamartomas: evaluation of patients undergoing chronic intracranial EEG monitoring and outcome of surgical treatment. Neurology 1993; 43:747–750. 52. Parrent AG. Stereotactic radiofrequency ablation for the treatment of gelastic seizures associated with hypothalamic hamartoma. Case report. J Neurosurg 1999; 91:881–884. 53. Kuzniecky R, Guthrie B, Mountz J, et al. Intrinsic epileptogenesis of hypothalamic hamartomas in gelastic epilepsy. Ann Neurol 1997; 42:60–67. 54. Palmini A, Van Paesschen W, Dupont P, et al. Status gelasticus after temporal lobectomy: ictal FDG-PET ﬁndings and the question of dual pathology involving hypothalamic hamartomas. Epilepsia 2005; 46:1313–1316.

60.

epilepsy: surgery perspective

55. Paillas JE, Roger J, Toga M, et al. [Hamartoma of the hypothalamus. Clinical, radiological and histological study. Results of excision]. Rev Neurol (Paris) 1969; 120:177–194. 56. Berkovic SF, Arzimanoglou A, Kuzniecky R, et al. Hypothalamic hamartoma and seizures: a treatable epileptic encephalopathy. Epilepsia 2003; 44:969–973. 57. Polkey CE. Resective surgery for hypothalamic hamartoma. Epileptic Disord 2003; 5:281–286. 58. Rosenfeld JV, Harvey AS, Wrennall J, et al. Transcallosal resection of hypothalamic hamartomas, with control of seizures, in children with gelastic epilepsy. Neurosurgery 2001; 48:108–118. 59. Palmini A, Chandler C, Andermann F, et al. Resection of the lesion in patients with hypothalamic hamartomas and catastrophic epilepsy. Neurology 2002; 58:1338–1347. 60. Engel J Jr. Outcome with respect to epileptic seizures. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. New York: Raven Press, 1987:553–571. 61. Wilson PJ. Cerebral hemispherectomy for infantile hemiplegia. A report of 50 cases. Brain 1970; 93:147–180. 62. Rasmussen T. Hemispherectomy for seizures revisited. Can J Neurol Sci 1983; 10:71–78. 63. Adams CB. Hemispherectomy—a modiﬁcation. J Neurol Neurosurg Psychiatry 1983; 46:617–619. 64. Villemure JG, Mascott CR. Peri-insular hemispherotomy: surgical principles and anatomy. Neurosurgery 1995; 37:975–981. 65. Van Wagenen WP, Herren RY. Surgical division of commissural pathways in the corpus callosum: relation to spread of an epileptic attack. Arch Neurol Psychiatr 1940; 44:740–759. 66. Wyler AR. Corpus Callosotomy. In: Wyler AR, Hermann BP, eds. The Surgical Management of Epilepsy. Boston: ButterworthHeinemann, 1994:139–145. 67. Gates JR, Rosenfeld WE, Maxwell RE, et al. Response of multiple seizure types to corpus callosum section. Epilepsia 1987; 28: 28–34. 68. Reeves AG, Risse G. Neurological effects of callosotomy. In: Reeves AG, Roberts DW, eds. Epilepsy and the Corpus Callosum 2. New York: Plenum Press, 1995:241–251. 69. Morrell F, Whisler WW, Bleck TP. Multiple subpial transection: a new approach to the surgical treatment of focal epilepsy. J Neurosurg 1989; 70:231–239. 70. Spencer SS, Schramm J, Wyler A, et al. Multiple subpial transection for intractable partial epilepsy: an international metaanalysis. Epilepsia 2002; 43:141–145. 71. Zabara J. Inhibition of experimental seizures in canines by repetitive vagal stimulation. Epilepsia 1992; 33:1005–1012. 72. Krahl SE, Clark KB, Smith DC, et al. Locus coeruleus lesions suppress the seizure-attenuating effects of vagus nerve stimulation. Epilepsia 1998; 39:709–714. 73. Henry TR, Bakay RA, Votaw JR, et al. Brain blood ﬂow alterations induced by therapeutic vagus nerve stimulation in partial epilepsy: I. Acute effects at high and low levels of stimulation. Epilepsia 1998; 39:983–990. 74. Ben-Menachem E, Hamberger A, Hedner T, et al. Effects of vagus nerve stimulation on amino acids and other metabolites in the CSF of patients with partial seizures. Epilepsy Res 1995; 20:221–227. 75. Schachter SC, Saper CB. Vagus nerve stimulation. Epilepsia 1998; 39:677–686. 76. Uthman BM, Reichl AM, Dean JC, et al. Effectiveness of vagus nerve stimulation in epilepsy patients: a 12-year observation. Neurology 2004; 63:1124–1126.

591

77. Holmes MD, Silbergeld DL, Drouhard D, et al. Effect of vagus nerve stimulation on adults with pharmacoresistant generalized epilepsy syndromes. Seizure 2004; 13:340–345. 78. Labar D, Murphy J, Tecoma E. Vagus nerve stimulation for medication-resistant generalized epilepsy. E04 VNS Study Group. Neurology 1999; 52:1510–1512. 79. Murphy JV. Left vagal nerve stimulation in children with medically refractory epilepsy. The Pediatric VNS Study Group. J Pediatr 1999; 134:563–566. 80. Patwardhan RV, Stong B, Bebin EM, et al. Efﬁcacy of vagal nerve stimulation in children with medically refractory epilepsy. Neurosurgery 2000; 47:1353–1357; discussion 1357–1358. 81. Murphy JV, Wheless JW, Schmoll CM. Left vagal nerve stimulation in six patients with hypothalamic hamartomas. Pediatr Neurol 2000; 23:167–168. 82. Frost M, Gates J, Helmers SL, et al. Vagus nerve stimulation in children with refractory seizures associated with Lennox-Gastaut syndrome. Epilepsia 2001; 42:1148–1152. 83. Nei M, O’Connor M, Liporace J, et al. Refractory generalized seizures: response to corpus callosotomy and vagal nerve stimulation. Epilepsia 2006; 47:115–122. 84. Murphy JV, Torkelson R, Dowler I, et al. Vagal nerve stimulation in refractory epilepsy: the ﬁrst 100 patients receiving vagal nerve stimulation at a pediatric epilepsy center. Arch Pediatr Adolesc Med 2003; 157:560–564. 85. Blumer D, Davies K, Alexander A, et al. Major psychiatric disorders subsequent to treating epilepsy by vagus nerve stimulation. Epilepsy Behav 2001; 2:466–472. 86. Rogers LR, Morris HH, Lupica K. Effect of cranial irradiation on seizure frequency in adults with low-grade astrocytoma and medically intractable epilepsy. Neurology 1993; 43:1599–1601. 87. Steiner L, Lindquist C, Adler JR, et al. Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 1992; 77:1–8. 88. Regis J, Bartolomei F, de Toffol B, et al. Gamma knife surgery for epilepsy related to hypothalamic hamartomas. Neurosurgery 2000; 47:1343–1351; discussion 1351–1342. 89. Regis J, Bartolomei F, Kida Y, et al. Radiosurgery for epilepsy associated with cavernous malformation: retrospective study in 49 patients. Neurosurgery 2000; 47:1091–1097. 90. Regis J, Bartolomei F, Rey M, et al. Gamma knife surgery for mesial temporal lobe epilepsy. Epilepsia 1999; 40:1551–1556. 91. Schauble B, Cascino GD, Pollock BE, et al. Seizure outcomes after stereotactic radiosurgery for cerebral arteriovenous malformations. Neurology 2004; 63:683–687. 92. Pendl G, Eder HG, Schroettner O, et al. Corpus callosotomy with radiosurgery. Neurosurgery 1999; 45:303–307; discussion 307–308. 93. Selch MT, Gorgulho A, Mattozo C, et al. Linear accelerator stereotactic radiosurgery for the treatment of gelastic seizures due to hypothalamic hamartoma. Minim Invasive Neurosurg 2005; 48:310–314. 94. Regis J, Rey M, Bartolomei F, et al. Gamma knife surgery in mesial temporal lobe epilepsy: a prospective multicenter study. Epilepsia 2004; 45:504–515. 95. Srikijvilaikul T, Najm I, Foldvary-Schaefer N, et al. Failure of gamma knife radiosurgery for mesial temporal lobe epilepsy: report of ﬁve cases. Neurosurgery 2004; 54:1395–1402; discussion 1402–1394. 96. Barbaro NM, Laxer KD, Group RiES. Radiosurgery in the treatment of medial temporal lobe epilepsy [abstract]. Epilepsia 2002; 43:342.

6 1

Ocular and Orbital Lesions Gabriela Šimonová, Roman Liscˇák, and Josef Novotný Jr.

Introduction Technological advances in medical imaging, treatment planning, and radiation dose delivery have led to dramatic improvements in all treatment capabilities including radiosurgery. It is now possible to recognize intraocular tumors and other ophthalmologic diseases at an early stage because physicians are generally more aware and because of the availability and use of both direct and indirect ophthalmoscopy, ﬂuorescein angiography, ultrasonography, and magnetic resonance imaging. The basic advantage of radiosurgery is the possibility of applying relatively high doses in a single session to well-deﬁned intracranial targets. The ﬁrst experiences with ophthalmologic indications involved the treatment of patients suffering from uveal melanomas [1–6]. However, it would appear that the potential of radiosurgery to provide effective treatment for ophthalmologic indications is far greater and, besides uveal melanomas, the current spectrum of treated indications also includes eye metastases, advanced glaucoma, and age-related macular degeneration.

Technical and Radiophysical Aspects of Intraocular Stereotactic Radiosurgery or Radiotherapy Stereotactic radiosurgery (SRS) or stereotactic radiotherapy (SRT) was originally designed for the treatment of rigid centrally or almost centrally located intracranial targets [1, 2]. Stereotactic radiosurgery or radiotherapy of ophthalmic lesions is relatively far from these standard treatment conditions mainly due to very eccentric target location and motion of the eye. These aspects speciﬁc to the treatment of ophthalmic lesions should be taken into account. First of all, a proper eye ﬁxation or eye motion monitoring system is required because the eye can move during the treatment procedure. Very eccentric target volume location can also cause some inaccuracies in the treat-

ment planning calculations including relative dose distribution as well as absolute dose calculations (calculation of treatment time or monitor units). Finally, there may be also some technical inconveniences to treat patients due to limitations in the range of coordinate system in this very eccentric target location. Despite all these difﬁculties, SRS or SRT is a tempting alternative for ophthalmic lesions treatment mainly due to its very high accuracy and delivery of high conformal dose distribution to the small volume of the treatment target.

Eye Fixation For all kinds of SRS or SRT, it is imperative to employ a patient ﬁxation system that is highly reproducible and meets the accuracy requirements for high-precision treatments. Additionally, for treatments of ocular lesions, it is not sufﬁcient to ﬁxate the patient’s head but the eye itself has to be ﬁxated in the same position during imaging for treatment planning and during delivery of all treatment fractions. Several different eye-immobilization systems and techniques have been described in the literature [3–11]. In principle, eye-immobilization techniques can be divided into two groups: (1) passive immobilization techniques where the eye is immobilized by different mechanical means from outside and (2) active immobilization techniques where the patient controls the eye position (e.g., by ﬁxation on a light source). Tokuuye et al. described a mask ﬁxation technique with a plastic mold gently pressed down over the orbit to restrict ocular movements [3]. Zehetmayer et al. reported on a suction immobilization technique for SRS of intracranial malignancies at the Leksell Gamma Knife [4]. Langmann et al. used an invasive ﬁxation method of the globe by means of retrobulbar anesthetic blocking for eye treatments with the Leksell Gamma Knife [5]. Simonova et al. reported a series of patients with uveal melanomas treated with the Leksell Gamma Knife when an eye ﬁxation was achieved by applying two sutures in the rectus muscles and attaching these sutures to the Leksell stereotactic frame [6]. The same

593

594

g. šimonová et al.

technique of eye ﬁxation was also reported by Vladyka et al. when patients with glaucoma were treated with the Leksell Gamma Knife [7]. Active eye-immobilization techniques, where the patient controls the eye position (e.g., by ﬁxation on a light source) have been described in the literature for proton/helium ion beam therapy and recently also for linac-based SRT of malignant tumors [8, 9]. For proton/helium ion beam therapy, a surgical intervention is performed to position tantalum clips at the border of the visible tumor under diaphanoscopical control. These clips are used during treatment planning to deﬁne the target volume and to verify the position of the tumor during treatment delivery by using X-rays [10]. Recently, Petersch et al. described a new, noninvasive eye ﬁxation system for the application of linac-based image-guided/gated SRT of uveal melanoma [11]. The system is attached to a commercially available head-and-neck thermoplast mask system and is based on the patient’s ﬁxation on a light point. Computer-controlled eye monitoring provides quantitative information about the quality of the patient repositioning, both about the repositioning of the patient’s head within the mask and about the eye’s rotational state with respect to the reference position as determined during imaging for treatment planning prior the actual treatment. When eye position exceeds the preset geometric limits, the system is able automatically initiate interruption of patient irradiation. When invasive stereotactic frame ﬁxation to a patient’s head or invasive eye ﬁxation is used, then the treatment itself is limited to one single fraction or a very small number of fractions due to patient comfort. A prerequisite for fractionated SRT is the reproducible and reliable immobilization of the irradiated structures throughout the whole treatment. Whereas this is achieved in the region of the skull by invasive or noninvasive ﬁxation systems, treatments of intraocular lesions are more complicated due to the additional degrees of freedom caused by eye movements. Consequently, special eye motion monitoring devices as described above have to be employed during radiation delivery. An example of invasive ﬁxation by means of eye ﬁxation by applying two sutures in the rectus muscles and attaching these sutures to the Leksell stereotactic frame and an example of noninvasive eye ﬁxation used for image-guided/gated SRT are presented in Figure 61-1.

Treatment Planning A prerequisite for correct treatment planning is a good imaging of the target and surrounding structures, especially organs at risk. A typical imaging modality for ophthalmic lesions is magnetic resonance imaging (MRI), which can be supplemented by computed tomography (CT). Generally, MRI brings better contrast compared with CT when imaging intracranial and intraocular structures. However, whereas CT is supposed to be free of geometric distortions, MRI may produce some geometric image distortions (sometimes a few millimeters) that are not acceptable for SRS or SRT treatment planning. Generally, MRI geometric distortion is mainly a function of the MRI scanner employed, the MRI sequence used, image slice orientation, position in imaged object, and material and geometric parameters of the stereotactic frame [12–14]. The issue of

FIGURE 61-1. (A) Example of invasive eye ﬁxation by applying two sutures in the rectus muscles and attaching these sutures to the Leksell stereotactic frame. (B) Example of noninvasive eye ﬁxation used for image-guided/gated SRT. (From Petersch B, Bogner J, Dieckmann K, et al. Automatic real-time surveillance of eye position and gating for stereotactic radiotherapy of uveal melanoma. Med Phys 2004; 31:3521– 3527. Used with permission.)

image geometric distortion is even more crucial in the case of eye imaging when a target and other structures of interest are in a very frontal location. It is essential to check the extent of geometric distortions for each MRI scanner and sequence employed for eye imaging. It is beyond the scope of this chapter to give more details about MRI geometric distortion measurements. More details can be found in, for example, Refs. 12–14.

Leksell Gamma Knife Treatment planning for intraocular targets is very similar as for intracranial targets in the case of the Leksell Gamma Knife. Four different collimator helmets (4 mm, 8 mm, 14 mm, 18 mm) are available to create isocenters of essentially spherical dose distributions. To cover optimally the target volume, an arbitrary combination of all four collimator helmets can be used. Dose distribution can be tailored to the target volume by means of proper isocenter size selection, proper position in the stereotactic space, and proper weighting factor for each isocenter. To protect surrounding healthy tissue and mainly structures at risk, additional shaping of dose distribution can be done by blocking any of 201 Cobalt-60 beams. Shielding of properly selected number of beams can effectively reduce dose to critical structures such as optic nerve or eye lens. Target volume is usually covered by 50% isodose line. Treatment planning process is also a compromise between sufﬁcient conformity required for the treated target and treatment time (with increased number of isocenters, total treatment time is increased). Because of the very frontal location of eye volume, some mechanical difﬁculties can occur when treating a patient with the Leksell Gamma Knife. Depending on a patient’s head shape and volume, it might be difﬁcult to adjust properly the frame position to reach all Leksell stereotactic coordinates. Speciﬁcally, y (posterior-anterior) Leksell stereotactic coordinate can sometimes exceed limits engraved on the Leksell stereotactic frame. Because of this, a special y coordinate extender was developed at Na Homolce Hospital to be able to treat also y coordinates exceeding the limits of the commercially available Leksell stereotactic frame (Fig. 61-2). Treatment of eye lesions with the Leksell Gamma Knife requires typically an extremely large neck rake (when supine

61.

ocular and orbital lesions

FIGURE 61-2. (A) Special y coordinate extender developed at Na Homolce Hospital used for the treatment of eye lesion when y coordinate exceeds the limits of the commercially available Leksell stereotactic frame. (B) Typical treatment of eye lesion on the Leksell Gamma Knife with patient in the prone position.

position is used) that can be hardly tolerated or even impossible to perform for some patients. Consequently, for a patient’s better comfort, the treatment is typically done in the prone patient position (Fig. 61-2). It was determined at the Na Homolce Hospital that there is no effect on target positioning geometric accuracy when imaging for treatment planning is done in supine and treatment in prone patient positions for eye invasively ﬁxed by applying two sutures in the rectus muscles and attaching these sutures to the Leksell stereotactic frame.

Linear Accelerator In contrast with the Leksell Gamma Knife, there are many different techniques available when performing treatment planning for linear accelerator. Non-coplanar arcs represent the standard treatment technique. Non-coplanar arcs treatments use circular collimators with ﬁxed diameters. The resulting isodose distribution is essentially spherical and very similar to the one created in the case of the Leksell Gamma Knife. There is usually a wide range (typically 5 to 40 mm) of circular collimators depending on the manufacturer of the system. Individual dose distribution can be further optimized by combination of proper number and size of circular collimators. However, total number of isocenters is much smaller compared with the Leksell Gamma Knife (usually does not exceed 5). Dose distribution of individual isocenters can be optimized by diameter of collimator, number of arcs, arc length, gantry angles, arc weighting factor, and table angles. Another option is non-coplanar conformal static ﬁelds. Non-coplanar conformal static ﬁelds can be applied using customized blocks or miniature-multileaf collimator with leaf width in the millimeter range. The use of miniature-multileaf collimator has several advantages compared with customized blocks: (1) generating miniature-multileaf collimator shapes takes less time than molding blocks, (2) there is no need to reenter the treatment room during multiple ﬁeld treatments, and (3) any modiﬁcation in the shape can be performed easily. Single-isocenter technique with selected number of conformal ﬁelds (typically 5 to 15) is used. Individual ﬁelds are shaped using beam’s-eye view technique. Weighting factors of single

595

individual beams, gantry angles, and table angles are used for ﬁnal dose distribution optimization. Dynamic non-coplanar arcs are another treatment technique. A selected number of conformal isocentric arcs are used. Beam’s-eye view technique is used for individual beam shaping. Finally, intensity-modulated SRS or SRT is an option for the treatment planning. A selected number of ﬁelds optimized by proper weighting factor, gantry and table angles are used as in the non-coplanar conformal static ﬁelds technique. Inverse treatment planning is used to generate optimal intensity through individual beams. Georg et al. compared different dosimetric characteristics of treatment plans when using the above-mentioned techniques for the treatment planning of uveal melanomas [15]. Conclusions of their study were that static conformal and dynamic arc treatment techniques present dosimetric advantages over conventional non-coplanar arcs using circular collimators while being simultaneously highly efﬁcient in treatment planning and delivery. Dynamic arc SRS or SRT combines the homogeneous dose distribution and the high degree of static conformal beams with the steep dose gradients and smeared-out low-dose volumes of arc beam therapy. Intensity-modulated SRS or SRT does not show clear impact on further reduction of doses to organs at risk and depends probably on the target volume location within the eye.

Dosimetry Because of very eccentric target volume location, some inaccuracies in the treatment planning calculations including relative dose distributions as well as absolute dose calculations (calculation of treatment time or monitor units) can be expected. This issue is even more important for surface structures of eye (e.g., eye lens, eye lid, cornea, etc.) than for target volumes itself as these structures are located in depths smaller than build up of employed photon energies. Petersch et al. reported excellent agreement between measurements and treatment planning system calculations for uveal melanomas at depth 15 to 20 mm [11]. However, at small depths (<10 mm), the treatment planning system cannot model the inﬂuence of absorption in eye motion monitoring device accurately if beam passed perpendicularly through it. Consequently, Petersch et al. suggested avoidance of frontal beams when doing treatment planning as a solution to reduce signiﬁcantly dosimetric errors. Detailed dosimetric measurements were also carried out at Na Homolce Hospital where treatment planning system calculations of target volume relative dose distributions and absolute dose calculations in different depths were compared with experimental measurements. Various treatment plans for different ophthalmic indications (uveal melanoma, glaucoma, retinoblastoma) calculated for the Leksell Gamma Knife were evaluated. There was observed good agreement between treatment planning system relative dose distribution calculations for target volume and performed measurements (Fig. 61-3). For the calculation of absolute doses, there was observed very good agreement at depths 15 to 20 mm. Typical deviations between treatment planning system calculations and actual measurements were within 3%. However, measurements performed at

596

g. šimonová et al.

Clinical Aspects Uveal Melanomas Pathophysiology

FIGURE 61-3. (A, B, C) Example of relative dose distribution treatment planning system calculations veriﬁcation with polymer gel dosimeter. (A) Three-dimensional treatment plan in LGP. (B) Polymerization that occurred in the polymer gel. (C) Relative dose proﬁles (calculated by LGP—solid line, measured with polymer gel—dotted line). (D) Tissue-equivalent built-up material that is attached to the treated eye to improve inaccuracies in absolute dose calculations in surface structures. (E) Corresponding MR localization scan visualized in LGP together with head contour.

depth less than 10 mm showed large deviations of about 15% to 20% compared with the treatment planning system calculations. The dose delivered to patient was always smaller than that calculated by the treatment planning system. To improve inaccuracies in absolute dose calculations in surface structures, it was suggested to use proper tissue equivalent built-up material that is attached on the treated eye (Fig. 61-3) and improves treatment planning system calculations.

Uveal melanoma represents the most common primary malignant tumor of the eye in adults with a peak incidence between 55 and 70 years, in 6 to 7 cases per million people per year, and approximately 50% of patients with diagnosed melanomas of the chorioid or ciliary body will die from this tumor within 15 years [16]. The incidence is rare in patients younger than 20 years [17]. From a larger cohort, we can see that eyes from black patients represent less than 1% of the total [18]. The incidence of cutaneous melanomas has been increasing in frequency over the past several decades and the dependence on latitude has been observed, reﬂecting levels of exposure to ultraviolet light. This trend has not been evident for patients with uveal melanomas [18]. Uveal melanomas are neoplasms arising from the uveal tract, which includes the iris, ciliary body, and chorioid. Two types of tumor growth were recognized, nodular and inﬁltrative. Histologically, the tumor consists of spindle cells, epithelioid cells, and intermediate cells similar to those of cutaneous melanoma, but the prognosis is better than that associated with cutaneous melanomas. Several staging systems have been used, one of the most commonly reported was based on the 1988 American Joint Staging System (Table 61-1) [19, 20]. The uveal tract is a vascular structure with no lymphatic drainage. Lymphatic node involvement (auricular, submandibular, and cervical nodes) can be diagnosed on rare occasions when subconjunctival extension of primary tumor has been observed. The major routes of spread are local growth and by bloodstream, with predilection for hematologic dissemination to the liver, lung, and brain. Liver metastases are diagnosed in two thirds of patients with disseminated diseases. Systemic examination is recommended for all patients and includes CT of the chest and ultrasonography, CT or MRI of the liver, and CT or MRI of the brain to detect other organ dissemination. Wide variations in the grading of malignancy can be observed, from relatively benign types with several years survival without dissemination, to other tumors with fast multiorgan dissemination, which lead to the patient’s death within a few months.

TABLE 61-1. Staging of uveal melanoma (T, N, M) (American Joint Staging System, 1998). T1 T2 T3 T4 N0 N1 M0 M1

10 mm or less diameter and 3 mm thickness More than 10 mm to 15 mm diameter and more than 3 mm to 5 mm thickness More than 15 mm in diameter and more than 5 mm thickness Tumor with extrascleral extension No regional lymph node metastasis Regional lymph node metastasis No distant metastasis Distant metastasis

61.

ocular and orbital lesions

597

Tumor Location

Surgical Treatment

The iris melanomas (incidence about 10%) are single or multiple elevated lesions with changes of iris color. These types of melanomas are much more benign, smaller in size, readily visible, and early diagnosed. It can be difﬁcult to assess the potential of the malignant cells, even in those with documented growth, because iris nevi may also grow. For this reason, local resection is usually effective whenever it seems to be necessary. Chorioidal melanomas typically (but with some variation) grow in nodular form, and in cases of continuing growth many tumors break through the Bruch membrane. The Bruch membrane is a base membrane–connective tissue complex lying between the retina pigment epithelium and the choriocapillaris. When the melanomas break through the Bruch membrane, they extend into the subsensory retina space, which gives them a mushroom or collar-button shape. These characteristics, as well as the detachment of the retina, are typical for uveal melanomas. This type of tumor growth is sometimes accompanied by vitreous hemorrhage with symptoms of sudden visual loss. Melanomas involving the ciliary body are rare, carry a poor prognosis, and are often diagnosed late. Melanomas can also have diffuse, inﬁltrative growth to the ciliary body and a pattern called ring melanoma. These tumors are diagnosed with difﬁculty and late, because they are accompanied by relatively little visible tumor mass effect and they occur in the diagnostically “silent” eye area. Any extra scleral extension of the tumor usually confers a poor prognosis. The diffuse growing melanomas are usually more aggressive, with more malignant cell types. Tumor growth is also inﬂuenced on the potential doubling time (T pot) which varies widely between 60 and 350 days (the median of T pot being about 70 days) [21]. A longer T pot for a signiﬁcant proportion of tumors can explain the relatively long survival of these patients and also the relatively slow tumor response after irradiation.

Surgical methods include resection of the eye wall (for very small visible melanomas of the iris), enucleation, and orbital exenteration [20, 22, 23]. The role of enucleation in the management of posterior to equator localized melanomas is controversial, but enucleation remains the standard option for most large chorioidal melanomas that cause severe secondary glaucoma or are invading the optic nerve.

Primary Tumor Diagnosis Melanoma of the ciliary body and chorioid have typical clinical features, and the correct diagnosis for a majority of patients can be made by taking a history and performing a complete ocular examination. The ophthalmologic examination includes external ocular examination, indirect ophthalmoscopy, ﬂuorescein angiography, and ultrasonography A, B. Magnetic resonance imaging plays a fundamental role in the treatment planning of radiosurgery or stereotactic radiotherapy. Ultrasonography plays a basic role, not only at the time of diagnosis but during follow-up and is used to obtain precise measurement of the tumor base and height. The height of tumor in particular represents one of the most important factors inﬂuencing the treatment decision, and it is used during followup to evaluate the tumor response.

Treatment Methods Treatment modalities strongly depend on the site of the primary tumor and its volume and metastatic extent. Historically, the ﬁrst known treatment modality was enucleation [20, 22].

Radiotherapy Preoperative or postoperative irradiation is a promising approach under evaluation that could also decrease the chances of malignant dissemination, however recent study observed no beneﬁt [19, 24]. Postoperative external beam radiotherapy has been used in patients with large tumors with extrascleral or extraocular extension. The recommended total applied dose is 60 Gy in daily fractionations.

Brachytherapy The most frequently clinically used therapeutic system for delivering radiation is brachytherapy (plaque radiotherapy). A number of radionuclides, including cobalt-60, gold-198, iodine125, ruthenium-106, iridium-192, and palladium-103, have been used for plaque therapy with varying results according to the retrospective studies [25–34]. A metal shield containing small radioactive seeds is sutured to the outside of the sclera overlying the tumor with radiation doses principally delivered to the base of the tumor. The radioactive plaque is left in place for approximately 5 to 7 days. Cobalt-60 was one of the ﬁrst isotopes to be used in brachytherapy, but it is a high-energy gamma emitter and thus there exists a risk of damaging healthy ocular structures several millimeters from the source. Other isotopes have been used: iodine-125 is also a gamma emitter but has lower energy than cobalt-60 and, because of this, it is associated with a lower rate of complications. Brachytherapy leads to long-term local tumor control in 85% to 90% of patients and can be used to treat small melanomas [25–34]. The best results can be achieved in tumors less than 10 mm in height and with a base diameter not exceeding 16 mm. Patients with extrascleral tumor extension, ring melanoma, and tumor involvement of more than half of the eye are not suitable for plaque therapy.

Proton Beam Therapy Patients with tumors too large for brachytherapy or tumors considered too close to the optic nerve have been recommended for proton beam therapy (PBT). Early studies with PBT recognized that severe damage was caused to the eyelid and the anterior parts of the eye. Patients with tumors localized near the eye surface had a higher incidence of severe toxicity because the sparing effect of Bragg peak was absent. Many of these patients had chronically painful and cosmetically unsatisfactory eyes. Local tumor control in 90% to 98% of patients was achieved by this method in a large series [35–41]. The technique was subsequently improved and the complications rate dropped accordingly.

598

g. šimonová et al.

Heavy Charged Particles Therapy Helium-ion irradiation is uniquely suited to the precise localization of irradiation because of the sharp fall-off of the dose at the distal end of the Bragg peak, the sharp lateral edges of the beam, and the ability to tailor the depth penetration and spread by a modulated Bragg peak. Uveal melanomas are indicated for this treatment. When the charged particles are used to treat the uveal melanoma, the normal tissue of the anterior part of the eye can receive a radiation dose comparable with the tumor dose. This fact causes radiation toxicity in the anterior parts of the eye leading to side effects such as eyelash loss, lacrimal damage, conjunctival changes, dry eye syndrome, keratitis, and neovascular glaucoma [42–45]. Local tumor control is achieved in 96% of patients and the enucleation rate is 19% (3% for local failure and 16% because of complications) [42].

Stereotactic Radiotherapy The advantage of stereotactic irradiation using a linear accelerator is mainly due to the possibility of fractionation treatment compared with the Leksell Gamma Knife. Fractionation may decrease late-radiation–related toxicity, and this fact was the subject of intensive study [4, 8, 9, 11, 15, 46–48]. The reported mean applied minimal dose ranged from 25 to 45 Gy in two fractions and 15 Gy in three equal fractions. Local tumor control, deﬁned as freedom of local progression, was reported in 98% of patients [48]. Further clinical studies using fractionated stereotactic irradiation are justiﬁed to optimize the effective dose and fractionation pattern.

Radiosurgery The ﬁrst experiments with radiosurgery for the treatment of uveal melanomas were conducted using the rabbit eye melanoma model. The reported tumor lethal dose in this experiment range from 60 to 90 Gy applied in a single session. A preliminary study of 11 treated patients with uveal melanomas was published in 1992 and suggested that local tumor control could be achieved with a minimal dose between 60 and 90 Gy in a single dose using the Leksell Gamma Knife. Some observations

were made on patients after radiosurgery who had their eyes secondarily enucleated. All eyes with secondary enucleations presented an initial height exceeding 9 mm and the second reason for the high overall enucleation rate was the high number of ciliary body tumors [49]. Several authors reported the results of radiosurgery in uveal melanomas and some of them observed relatively high radiation-related toxicity [5, 6, 49–53]. The minimum effective dose for long-lasting local tumor control as well as the tolerance doses for critical ocular structures is not yet well deﬁned.

Radiosurgery (Personal Experience) We studied a group of 126 patients with uveal melanoma treated using the Leksell Gamma Knife over a period of 8 years with all survivors undergoing minimal follow-up 24 months after the treatment. The late effect of normal tissue (LENT) subjective objective management analysis (SOMA) scoring system is acceptable for measurement of radiation side effects (Table 61-2).

Imaging and Treatment Planning MRI plays a fundamental role in the imaging of ocular tumors for treatment planning, and the following sequences have been used: the native T1- and T2-weighted sequences, then postcontrast T1-weighted sequences and three-dimensional with slice thickness 1 to 3 mm. The melanotic uveal melanomas are well deﬁned on native T1 (Fig. 61-4A) and T2 (Fig. 61-4B) sequences and have typical postcontrast enhancement on T1-weighted MRI (Fig. 61-4C). This postcontrast enhancement of melanomas differentiates the boundary between the tumor volume and the accompanying retinal detachment (Fig. 61-4C). The rare amelanotic melanomas are difﬁcult to differentiate on native T1 (Fig. 61-5A) but are well deﬁned on T2-weighted magnetic resonance sequence (Fig. 61-5B). The postcontrast enhancement is not so intensive as in the melanotic forms (Fig. 61-5C). The treatment planning volume (PTV) includes the growth

TABLE 61-2. Toxicity scoring system (RTOG/EORTC LENT SOMA). Grade 1

Grade 2

Grade 3

Grade 4

Cornea

Increased tearing on exam

Noninfectious keratitis

Infectious keratitis, corneal ulcer

Iris

Rubeosis only

Rubeosis, increased intraocular pressure

Neovascular glaucoma, ability to count ﬁngers at 1 m

Optic nerve

Afferent pupillary defect with normal-appearing nerve

Lens

Asymmetric lenticular opacities, no visual loss

Retina

Microaneurysms, nonfoveal exudates, minor vessel attenuation, extrafoveal pigment changes Loss of episcleral vessels

≤¼ pallor with asymptomatic visual ﬁeld defect Moderate lenticular changes with mild-moderate visual loss Cotton wool spots

>¼ pallor or central scotoma Moderate lenticular changes with several visual loss Massive macular exudation, focal retinal detachment

Panophthalmitis, corneal scar, ulceration leading to perforation of globe/loss of globe Neovascular glaucoma without ability to count ﬁngers at 1 m; complete blindness Profound optic atrophy, complete blindness

≤50% scleral thinning

>50% scleral thinning

Sclera

Severe lenticular changes

Opaque vitreous hemorrhage, complete retinal detachment, blindness Scleral or periosteal graft required due to perforation

61.

ocular and orbital lesions

599

FIGURE 61-4. Melanotic type of uveal melanoma. (A) Native T1-weighted MRI. (B) T2-weighted MRI. (C) Postcontrast T1-weighted MRI.

tumor volume and 1- to 2-mm safety margins in all three diameters. The minimal dose for radiosurgery ranged from 35 to 45 Gy with a median dose of 41 Gy in our series. The majority of melanomas are irradiated using 4-mm and 8-mm collimator helmets (Fig. 61-6) and the plugging pattern aim to spare critical structures such as cornea (Fig. 61-7) or optic nerve (Fig. 61-8).

Follow-up Patients should be examined by an ophthalmologist at regular intervals (24 hours after SRS, 1 week, 3 months, and then every 6 months after SRS; magnetic resonance can be repeated every year for the ﬁrst 2 years after radiosurgery and then every 2 to 3 years after the treatment). The following ophthalmologic methods were employed for diagnostic investigation and patient follow-up at regular intervals 6 months after radiosurgery: ultrasonography, indocyanine and ﬂuorescein angiography, optical coherence tomography, ultrabiomicroscopy (for tumors located anterior to the equator), visual function, and biomicroscopy.

Local Tumor Response Tumor regression is deﬁned as a decrease in tumor height registered by A and B ultrasonography scans and by control MRI. In our personal experience, the tumor regression can be achieved in 70% of patients. The maximum local effect has been recorded during the interval of 20 to 30 months after radiosurgery. Most tumor response has been achieved within

24 months after SRS (Fig. 61-9). Patients with tumor growth progression or without any response 30 months after treatment were candidates for retreatment by radiosurgery, and patients with a tumor height of more than 10 mm or extra scleral growth were indicated for enucleation (5% in our series).

Survival Patients younger than 50 years have the best prognosis, with a pre-equatorial location of the tumor, when tumor height does not exceed 5 mm, gross tumor volume is not larger than 500 mm3, and there is no other organ dissemination (Figs. 61-10 to 61-14).

Complications The tolerance of normal tissues is in general the limiting factor for the dose that can be given to target volume. Radiation therapy is associated with a broad spectrum of normal-tissue reactions, and no reporting of outcome of radiotherapy or radiosurgery is satisfactory without a thorough description of the treatment-related toxicity. Radiation toxicity is relatively well-known from external beam fractionated radiotherapy, and the most common complication is primarily due to radiation vasculopathy. Neovascular glaucoma has been described as the most serious problem after external beam radiotherapy and radiosurgery [56]. The development of severe radiation-induced toxicity for the various eye critical structures is complicated by other factors (such as hypertension, diabetes), the local effects of pathologic process, the segment of the eye involved, the size and location of the tumor, the type of radiation and the

FIGURE 61-5. Amelanotic type of uveal melanoma. (A) Native T1-weighted MRI. (B) T2-weighted MRI. (C) Postcontrast three-dimensional MRI.

600

g. šimonová et al.

FIGURE 61-6. (A) Axial, (B) coronal plane, and (C) sagital plane. Uveal melanoma. Characteristics: The tumor base is 9 mm and the tumor height is 3 mm. Gross tumor volume is 84 mm3, the planned treat-

ment volume is 183.3 mm3, minimum dose is 44 Gy on a 60% isodose, maximum dose is 73.3 Gy. The maximum doses to critical structures: lens 4 Gy, cornea 2 Gy, eyelid 2 Gy.

left

FIGURE 61-7. Demonstration of the effect of plugged sources for protection of cornea for patient with uveal melanoma treated by 8-mm and 14-mm collimators. (A) Treatment plan without plugging.

right

(B) Treatment plan with plugging. The dose to cornea in this case did not exceed 14 Gy (given on displayed 20% isodose). (C) Plug pattern used for 14-mm collimator. Plugged sources are indicated in black.

left

right

FIGURE 61-8. Demonstration of the effect of plugged sources for protection of optic nerve for patient with uveal melanoma treated by two 4-mm collimators. (A) Treatment plan without plugging. (B) Treat-

ment plan with plugging. (C) Plug pattern used for plugging both 4-mm collimators. Plugged sources are indicated in black.

FIGURE 61-9. Kaplan-Meier cumulative decrease of tumor height curve after Leksell Gamma Knife.

FIGURE 61-10. Kaplan-Meier cumulative survival curves after Leksell Gamma Knife for patients divided according to age.

61.

1.0

1.0 Anterior to equator

Cumulative survival after LGK

Cumulative survival after LGK

601

ocular and orbital lesions

0.8 Posterior to equator

0.6

Anterior and posterior to equator or the whole bulbus

0.4

0.2 p = 0.046 (Log rank) 0.0

Tumor volume <500 mm3

0.8

0.6

Tumor volume 500–1500 mm3

Tumor volume >1500 mm3

0.4

0.2

p = 0.003 (Log rank) p = 0.019 (Cox)

0.0 0

20

40

60

80

100

0

120

20

40

60

80

100

120

Time [months] FIGURE 61-11. Kaplan-Meier cumulative survival curves after Leksell Gamma Knife for patients divided according to tumor location.

Time [months] FIGURE 61-13. Kaplan-Meier cumulative survival curves after Leksell Gamma Knife for patients divided according to tumor volume.

technique used. Two types of complications are deﬁned: any radiation-related morbidity that occurs within the ﬁrst 90 days after the end of treatment is usually regarded as early and all other complications as late. The clinical course of the two types of reactions is different, early reactions tending to be transient whereas late reactions are often irreversible. Biologically, it has been established that higher grades of reactions are on average seen at later times than lower grades and that increased treatment toxicity may considerably shorten the latent period [54, 55]. Many schemas have been devised for recording the late effects of radiotherapy treatments including the Radiation Therapy Oncology Group (RTOG) and LENT SOMA scales widely used in clinical practice. The tables for scoring ocular toxicity are documented in Table 61-2 [54]. The most common late toxicity for all types of irradiation is retinopathy, cataracts, secondary glaucoma, and optic neuropathy [6, 20, 47, 49, 56–58]. The presence of iris rubeosis can be chosen as a marker of severe ocular damage, and many authors have reported that secondary neovascular glaucoma is

the most common cause of the toxicity-related enucleation [37, 56]. Every new treatment method needs detailed and qualiﬁed radiation morbidity scoring. The initial steps identify the tolerance doses for critical structures and the effective doses for the target volume. In an analysis of late toxicity (126 patients), we recorded the following results: signiﬁcantly lower toxicity in the optic nerve was observed when the maximum dose was less than 10 Gy (incidence of grades 3 or 4 only in 2.4%), in the cornea when maximum dose did not exceed 10 Gy (incidence of toxicity grades 3 or 4 in 3%), in the lens when the maximum dose did not exceed 7 Gy (incidence of toxicity grades 3 or 4 in 7.7%), and in the iris when the maximum dose did not exceed 15 Gy (incidence of grades 3 or 4 late toxicity in 4.6%). Detection of risk factors can improve treatment results and decrease the adverse effects, and an appropriate length of follow-up is needed for patients after radiosurgery (a minimum of 2 years, preferably 5 years) because the hazard of late toxicity is increased with the follow-up time. An acceptably low incidence of late radiation–related complications for cornea, iris,

1.0

Without dissemination

1.0 Cumulative survival after LGK

Cumulative survival after LGK

Tumor height <5 mm

0.8 Tumor height 5–10 mm

0.6 Tumor height >10 mm

0.4

0.2 p = 0.017 (Log rank) 0.0

0

20

40

0.8

0.6

0.4

0.2

With dissemination

p < 0.001 (Log rank) p < 0.004 (Cox)

0.0 60

80

100

120

Time [months] FIGURE 61-12. Kaplan-Meier cumulative survival curves after Leksell Gamma Knife for patients divided according to tumor height.

0

20

40

60

80

100

120

Time [months] FIGURE 61-14. Kaplan-Meier cumulative survival curves after Leksell Gamma Knife for patients divided according to disease dissemination.

602

g. šimonová et al.

and optic nerve is observed when the maximum dose to these structures does not exceed 10 Gy. The most frequent and severe radiation-related side effect was secondary radiation-related retinopathy, secondary neovascular glaucoma, and optic neuropathy. Severe grades 3 or 4 retinopathy was noticed in 15% of treated patients, radiation-related late toxicity grades 3 or 4 in lens was observed in 26%, optic neuropathy in 9%, and secondary glaucoma in 18% of cases, and 11% patients were enucleated for this reason during the ﬁrst 2 years after radiosurgery. The median time to occurrence of secondary neovascular glaucoma was 18 months and so a higher incidence of enucleation during a longer follow-up could be expected. In our series, we did not observe any signiﬁcant inﬂuence of the minimum dose and tumor location, but a signiﬁcantly lower incidence of secondary glaucoma was noticed when the volume planning target volume (PTV) of the peripheral (prescribed) isodose was less than 1000 mm3 with an incidence of 6.9%. The worsening of useful vision caused by radiation-related toxicity during 3 years after radiosurgery was observed in 35% of patients and was due to tumor progression in 3%, optic nerve neuropathy in 8%, retinopathy in 9%, and neovascular glaucoma in 14%. There exists a difference in the area of severe complications and it is based on the radiation damage of different eye tissues. The most acceptable are complications concerning the lens because ophthalmologists can treat them more easily. The main problems in the eye bulb are complications of the iris, cornea, and optic nerve. The optic nerve tolerance is well deﬁned from other radiosurgery procedures especially from the irradiation of pituitary adenomas, meningiomas, and craniopharyngiomas. The tolerance dose of the optic nerve ranged from 8 to 10 Gy and the same doses are well tolerated by other critical ocular structures with an incidence of late complications not higher than 5% [6]. Late radiation responses occur after latent periods of 3 months and many years [5, 6, 20, 56–58]. An extended follow-up period is therefore required in order to obtain a reliable estimate of the incidence of late complications. Care must be taken when evaluating late effects in a group of patients with short life expectancy or a short follow-up, who may not be at risk long enough to develop any late complications of interest. The majority of side effects are recorded 9 to 36 months after radiosurgery and radiotherapy and, for example, the latent period between irradiation and development of cataract is about 4 years, depending on the dose to this structure, and the shortest observed latency is 6 months.

Prognostic Factors The main aim of radiosurgery is not only to preserve the bulb and visual function but also to achieve long-term local tumor control and to protect against hematologic dissemination of malignant melanoma. The prognosis of patients with uveal melanoma is determined by tumor size, cell type, tumor location (poor results for ciliary body tumors), and the extent of the disease [6, 10, 19–24, 25–36, 40–45]. Five-year survival rates of patients with small tumors (heights less than 8 mm) and with no detected metastatic disease outside the eye are above 80%, and for lager tumors (over 10 mm in height) with metas-

tases outside the eye, the 5-year survival period ranged from 10% to 30%. The most important prognostic factor is the presence or absence of the organ dissemination at the time of uveal melanoma diagnosis. The potential dissemination has been discussed several times. Some authors maintain that dissemination from uveal melanomas starts when the tumor is larger than 7 mm in diameter, and growth from a 7- to 10-mm diameter increases the risk of metastatic disease to approximately 16%. This fact was supported by clinical observations [21]. There still exists a serious gap in our knowledge of growth of melanomas in situ and the documentation of the time intervals between the diagnosis of primary tumor and the appearance of the other organ dissemination. Tumor growth strongly depends on tumor doubling time. The numerous accounts of tumor doubling times in uveal melanomas have been reported, and in the majority of them the estimated doubling time was longer than 69 days. On theoretical grounds, it has been suggested that metastatic death (after the primary tumor has released malignant cells into the systemic circulation) will occur at a time interval derived from multiplication of the tumor doubling time by 35 to 40 [21]. Taking these two assumptions together, it may be accepted for practical prognostic consideration in individual patients that death from dissemination is unlikely to occur before 6 years after dissemination of the ﬁrst viable embolus of malignant tumor cells. The minimum latent interval of 6 years implies that metastatic death within 7 to 8 years after local treatment on the uveal melanoma is nearly always due to pretreatment dissemination [48]. The chance of metastasis-caused death within 5 years totally depends on the presence or absence of hematologically disseminated tumor cells or diagnosed metastases at the time of local treatment. Some data from the analyzed clinical studies strongly supports the assumption that the rate of death by metastatic disease is inﬂuenced by local tumor control: improvement in local tumor control rate results in a longer survival. One of the fundamental experiments and analysis of results was reported from Switzerland; a study was based on 15 years of treatment of uveal melanomas with proton beam therapy [42]. The minimal latent interval of 5 years implies that metastatic death 5 years after local treatment is nearly always due to pretreatment dissemination. This is the one reason that statistically signiﬁcant prognostic factors for survival in tumors less than 8 mm in diameter should never be established 5 years after local treatment, and the recommended interval of follow-up and analysis seems to be 8 to 10 years after the treatment. The minimal interval for the evaluation of tumor response and treatment related toxicity is about 3 to 5 years [59].

Other Aspects of Eye Preservation Methods The psychosocial aspect of conservative treatment can also be intensively studied. The psychological problems of some conservatively treated patients can be caused by frequent follow-up tests, treatment of any reoccurrence, treatment of complications, and the possibility of persistent malignant cells in the eye involved. On the other hand, the impact of enucleation on vision-dependent daily activities must also be discussed.

61.

603

ocular and orbital lesions

TABLE 61-3. Comparison of different treatment methods.

Treatment method

Brachytherapy [25–30] Proton beam therapy [35–38] Helium-ions therapy [42–45] Stereotactic radiotherapy [47, 48] Radiosurgery [5, 6, 49–53]

Median follow-up (months)

Minimal dose (Gy)

Tumor response (%)

5-year local tumor control (%)

Radiationrelated retinopathy (%)

Neovascular glaucoma (%)

Enucleation rate (%)

39–96 40 96 36 36

70–250 40–70 F 50–80 F 45–60 F 35–90

69–96 91–100 87–96 94 70–90

82–92 87–100 80 NA NA

22–63 NA NA 26 23–70

3–11 8 35 9–20 18–23

10–21 9 20 NA 18

Note: Fractionated treatment is indicated in the table by F; NA indicates data not available.

Trials and Treatment Methods Comparison The management of uveal melanomas still remains controversial. The available data have not yet established whether enucleation or conservative treatment including radiosurgery is more effective in prolonging patients’ survival as the treatment results in terms of survival are similar. The multicenter Collaborative Ocular Melanoma Study (COMS) was initiated at 32 clinical centers in the United States and in Canada in 1986 and this study included (1) a trial for patients with medium chorioidal melanomas who are assigned randomly to treatment by primary enucleation or iodine-125 plaque therapy; (2) a trial for patients with larger melanomas, assigned randomly for treatment by primary enucleation with or without preceding radiotherapy; (3) a trial of small uveal melanomas (a prospective observational study) [22]. Proton beam therapy was analyzed in Switzerland, and the treatment results over 15-year periods were reported. Effective local tumor control was achieved in 90% and later in 96% of treated patients [36–39]. Ten years survival of patients with local controlled tumors was 72.6% and 47.5% in patients with local recurrence of melanoma. The helium-ions therapy trial documented local tumor control of 85% to 90% [42–45]. The radiosurgery series reported the local tumor control of about 70% to 75% [6, 49–51].

Treatment Methods Comparison Brachytherapy represents two surgical interventions: the ﬁrst is plaque application and the second replacement. Proton beam therapy is accompanied by pretreatment application of target markers to the exposed sclera using tantalum clips. Stereotactic radiotherapy uses the fractionation regimes. Radiosurgery is combined with the stereotactic frame and eye ﬁxation. The treatment results in terms of tumor response and local control are similar to those of brachytherapy, stereotactic irradiation, helium-ions therapy, proton beam therapy, or enucleation and range between 85% and 97% [36–39, 42–45, 58–60]. Long-term local tumor control of more than 90% and an incidence of rubeosis of iridis in 34% of patients has been achieved using proton beam therapy after a median follow-up of 36 months [45]. A major predictive factor for the development of rubeosis iris and later neovascular glaucoma was a large tumor size unsuitable for brachytherapy. Stereotactic irradiation using a linear accelerator with applied doses of 45 to 70 Gy delivered in one or three fractions is an effective method with the following results: tumor height reduction was achieved in 97%, sec-

ondary enucleation were performed in 13%, and the incidence of secondary glaucoma was 20% also including iris or chamber angle neovascularization [48]. In a prospective study of heliumion therapy, the incidence of secondary glaucoma was 29% after a mean observation period of 53 months [45]. Radiosurgery data was reported by Rennie, and in this study, after delivering a single fraction of 70 Gy, the majority of patients developed signiﬁcant complications after a median follow-up of 24 months [53]. Retinopathy was found in 78% and secondary neovascular glaucoma in 43% of patients. Our radiosurgery data were mentioned above. Comparison of different treatment methods is given in Table 61-3.

Treatment Decision Rationale criteria for selecting a treatment modality should be dependent upon the interrelations between the clinical characteristics of tumor and the morbidity associated with the available treatment modalities. Small tumors (less than 10 mm in height) and localized pre-equatorially can be successfully treated using brachytherapy. Tumors localized equatorially or posterior to the equator and not higher than 10 mm can be irradiated using the Leksell Gamma Knife. Small melanomas (height 5 to 10 mm) can be treated effectively using Leksell Gamma Knife and in cases with PTV not exceeding 1000 mm3 with less than 7% risk of neovascular glaucoma. Tumors more than 10 mm in height, especially melanomas of the ciliary body, progression of growth after irradiation, tumors combined with neovascular glaucoma, or vitreous hemorrhage are candidates for enucleation (Case Studies 61-1 and 61-2).

Case Study 61-1 Uveal melanoma in intimate contact with the disk of optic nerve. Characteristics: The gross tumor volume (GTV) was 951 mm3, the treatment planning volume (50% isodose curve) was 2700 mm3, the tumor height was 10 mm, and the minimal dose applied on 50% isodose curve was 40 Gy, using 1 isocenter (14-mm collimator) (Fig. 61-15). The maximum doses to the critical structures: optic nerve 40 Gy, cornea 6 Gy, lens 10 Gy, eyelid 6 Gy. The ultrasonography and control MRI 12 months after radiosurgery documented tumor regression and the development of postirradiation retinopathy grade 2. No acute toxicity was observed. Progression-free survival is 24 months after radiosurgery (Fig. 61-16).

604

g. šimonová et al.

FIGURE 61-15. Post–equatorial uveal melanoma. (A) Pretreatment funduscopy and treatment planning MRI. (B) Native T1-weighted MRI. (C) T2-weighted MRI.

FIGURE 61-16. (A) The tumor response 12 months after radiosurgery and the development of postradiation retinopathy (funduscopy control). (B, C) Control MRI 12 months after radiosurgery.

Case Study 61-2 Patient with useful vision and diagnosis of uveal melanoma. The gross tumor volume was 480 mm3, the tumor height was 6 mm, the planning treatment volume (57% isodose curve) was 890 mm3, and the minimal applied dose to 57% isodose curve was 40 Gy.

FIGURE 61-17. (A) Pretreatment, postcontrast T1-weighted MRI. (B) T2-weighted MRI.

The maximal dose to critical structures: optic nerve 10 Gy, lens 7 Gy, cornea 4 Gy, eyelid 3 Gy. Pretreatment MRI scan is shown in Figure 61-17. Control MRI was performed 18 months after radiosurgery and documented the tumor regression (Fig. 61-18). No acute or late toxicity was recorded and no change of vision. Progression-free survival is 30 months.

FIGURE 61-18. Control MRI 18 months after radiosurgery. (A) Postcontrast T1-weighted MRI. (B) T2-weighted MRI.

61.

ocular and orbital lesions

Retinoblastoma Pathophysiology Retinoblastoma represents the most common primary malignant tumor of the eye in children. Although this tumor accounts for only 1% of all malignant tumors in children, it has been studied intensively because of its genetic features.

605

ultrasonography, and MRI postcontrast T1-weighted sequence. When the patient has no prior history of malignant tumors, a systemic examination should be performed to identify the primary tumor and metastatic extent outside the eye. Chorioidal metastases generally appear as a creamy yellow subretinal mass, often accompanied by retinal detachment. Ultrasonographic ﬁndings can usually clearly differentiate chorioidal metastases from chorioidal melanomas.

Diagnosis and Treatment The majority of cases are diagnosed within the ﬁrst 2 years of life. The treatment strategy depends on clinical staging (disease conﬁned to the retina, disease conﬁned to the globe, extraocular extension, and presentation of distant metastases) and includes the following methods: tumor resection, enucleation, chemotherapy, and external beam radiotherapy.

Radiosurgery (Personal Experience) Over 3 years, four patients were admitted for stereotactic irradiation using the Leksell Gamma Knife and they had vitreous seeding of malignant cells. All patients were pretreated according to SIOP (Society International of Pediatric Oncology) protocol and relapsed after standard treatment. The only other possibility was enucleation. The whole vitreous body was irradiated using the Leksell Gamma Knife in a single session, collimator diameter 14 mm, the median PTV 2900 mm3 was and the median of minimum dose was 15 Gy. To avoid growth retardation of the irradiated eye and bony structures, the minimal dose did not exceed 15 Gy in all patients, and for this as well as the risk of complications from the previous treatment, the applied doses were relatively low and three of the four patients were enucleated with a median of 2 months after radiosurgery for local progression. The role of stereotactic irradiation and radiosurgery in the treatment of primary tumors still remains unclear.

Orbital and Uveal Metastases of Carcinomas Pathophysiology Metastatic lesions in every subsite of the eye and the orbita have been described several times but predominately affect the posterior uveal tract [61]. The autopsy studies estimate the incidence of these metastases to be 4% to 12% in adult patients with cancer of all histologic types but more than one third in patients with breast cancer. The overall incidence of ocular metastases is difﬁcult to quantify, as many patients remain asymptomatic. However, with improvements in systemic treatment, patients may be living longer and may, therefore, manifest intraocular and intraorbital lesions more frequently than in the past.

Diagnosis A careful history and ophthalmologic examination is generally sufﬁcient to diagnose ocular metastases, and the complete examination should include slit-lamp examination, funduscopy,

Treatment Brachytherapy Plaque therapy is a standard treatment method using the same radionuclides as for uveal melanomas. Candidates for this type of irradiation are patients with a solitary lesion and a mean diameter of the pathologic lesion not exceeding 8 to 10 mm.

Radiotherapy External beam fractionated radiotherapy represents the local effective treatment with a low rate of late complications. The recommended total dose is 30 to 40 Gy with 2 to 3 Gy per fraction per day [62].

Stereotactic Radiotherapy This treatment modality applies higher radiation doses using a fractionation scheme but, at present, there is still a lack of information about its use in clinical practice.

Radiosurgery Radiosurgery may extend treatment options for uveal metastases because of its speciﬁc advantages: (1) the possibility of applying a high radiation dose and (2) evidence of its sparing effect on the surrounding critical structures. The recommended minimal doses are similar to those for brain metastases, ranging between 20 and 30 Gy, and depend on the tumor volume, location, and histology. Additional study of the efﬁcacy and complications of this technique compared with external beam radiotherapy is indicated as the role of this method still remains open.

Prognosis The diagnosis of ocular metastases generally carries a poor prognosis with a median survival period of 9 months, with the literature reporting a range of 4 to 13 months (Case Study 61-3).

Case Study 61-3 A patient with spinocellular cell carcinoma. The minimum applied dose was 25 Gy on a 50% isodose curve (Fig. 61-19). Regression of tumor on control MRI 24 months after radiosurgery is shown in Figure 61-20.

606

g. šimonová et al.

FIGURE 61-19. (A, B) Pretreatment MRI of spinal cell carcinoma (patient treated by Leksell Gamma Knife, minimum applied dose was 25 Gy on a 50% isodose curve) (A) in axial and (B) in coronal plane.

Ocular and Orbital Malignant Lymphomas and Leukemia Inﬁltrates This special group is represented by orbital malignant lymphomas and leukemia pathologic inﬁltrates. These types of malignancies are typically radiosensitive and chemosensitive. The tumor lethal doses ranged from 30 to 40 Gy, with 2 Gy per fraction per day, and external beam radiotherapy is highly effective with a low incidence of radiation-related toxicity caused by the fact that the effective tumor dose is lower than the tolerance doses for critical structures [63]. At present, there is no real indication for using radiosurgery in these types of radiosensitive malignancies, and the question of its use in persistent or recurring pathologic lesions after previous radiotherapy and chemotherapy is still open.

Age-Related Macular Degeneration Pathophysiology Age-related macular degeneration is characterized by degeneration in the retinal layers accompanied by formation of subfoveal neovascular membrane that compromises vision. The growth factors plays a fundamental role in age-related macular degeneration, and these growth factors are produced by inﬂammatory cells and hypoxic or ischemic retinal tissues, including the neurosensory layer and the retinal pigmented epithelium. This pathologic process induces scarring and destroys the macular retina, causing severe visual changes and visual loss in people over 50 years of age [64–66]. The scheme of neovascular membrane development is 1. Tissue degeneration; 2. Tissue hypoxia and ischemia; 3. Production of vascular endothelial growth factor– speciﬁc endothelial cell mitogen released from hypoxic cells; 4. Vascular endothelial growth factor attaches to receptor veins; 5. Mitogenic stimuli for endothelial cells; 6. Mitosis of endothelial cells and endothelial cells migration into tissue when they join with other endothelial cells and to form new vessels;

FIGURE 61-20. Control MRI 2 years after radiosurgery documents tumor regression.

7. Tissue matrix consists of macrophages, retinal pigmented epithelium, ﬁbroblasts, and microvessels; 8. Membrane formation.

Diagnosis A complete ophthalmologic examination, including ﬂuorescein angiography, is needed to diagnose age-related macular degeneration. The membrane mass can be diagnosed using postcontrast T1-weighted MRI.

Treatment Methods Radiotherapy The aim of radiotherapy (as for radiosurgery) seems to be (1) to control growth-factor formation and (2) to alter the degenerative process. The persistence of hypoxic stimuli that upregulate growth factors may explain why the membrane regrows and disease recurs. Age-related macular degeneration is a degenerative process with proliferative component, and irradiation can play an antiproliferative role. The experience from radiotherapy showed that neovascular membranes are composed of endothelial cells with proliferation more rapidly than the endothelial cells of the retina and may be more sensitive than the retinal vasculature [64]. Radiotherapy uses the fractionation scheme and recommended doses ranged from 14 to 20 Gy in 8 to 10 daily fractions. The experience of radiotherapy shows that this modality can affect active subretinal neovascularization but is unlikely to prevent new neovascularization lesions being produced by this chronic disease.

Proton Beam Therapy The proton beam therapy was reported as local effective treatment, and the recommended minimal dose is 14 Gy in a single session [65].

Photocoagulation The management of chorioidal subuveal neovascularization is still controversial and difﬁcult. Laser photocoagulation, which is indicated for patients with uveal membrane that adjoins but does not involve the macula, can control the evolving membrane. The strict criteria of the Macular Photocoagulation Study for patient selection shows that only 10% of patients are eligible for this type of treatment [66].

61.

ocular and orbital lesions

Other Methods The pilot studies of surgical excision of membrane and the use of anti-angiogenic agents to prevent new vessel formations have not been successful.

Radiosurgery Initial results from patients treated by Leksell Gamma Knife for age-related macular degeneration (ARMD) showed that radiosurgery is able to affect tissues beneath the retina (neovascular membrane in ARMD) without damaging the overlying retinal structures. Haas in his pilot study had investigated the effect of single-fraction (10 Gy at the 90% isodose) Gamma Knife surgery in patients with classic subfoveal neovascular membrane due to ARMD [67]. The regression was observed in l of 10 patients.

Radiosurgery (Personal Experience) In our study, we treated 11 patients with neovascular membrane (NM) in ARMD with a dose of 15 Gy at a 50% isodose in a single fraction. Both ultrasonography and ﬂuorescein angiography demonstrated a regression of the neovascular complex or its stabilization in 90% of patients. An enlargement of the NM was found in one patient [68]. The patients have to undergo the same procedure (the frame ﬁxation, the eye immobilization, and build-up application) as in treatment of uveal melanomas, and the procedure has been described in detail above. The planning treatment volume includes the postcontrast enhanced lesion (neovascularization membrane) on a three-dimensional Tl-weighted sequence. The effective minimal dose based on personal experience is 15 Gy, generally applied on a 50% isodose curve, and the use of 8- or 4-mm collimators is recommended to decrease the doses to the critical structures, especially to the disk of the optic nerve. The adequate dose to the optic nerve in patients with useful visual function has not to exceed 10 Gy. Eye structures other than the retina received doses lower than 10 Gy. Figure 61-21 documents the neovascular membrane on postcontrast T1-weighted MRI and 15 Gy and 10 Gy isodose curves. Regression of neovascular membrane after radiosurgery (typically late effect) is observed after median time ranging from 9 to 24 months after radiosurgery (Fig. 61-22, pretreatment angiography and ultrasonography; and Fig. 61-23 documents regression 12 months after radiosurgery). The application of a

607

stereotactically guided single-dose irradiation is another promising treatment modality.

Follow-up Careful follow-up is necessary to integrate into the treatment, and the intervals at which this takes place depends on estimates as to when local effect—responses and recurrences of diseases or radiation-related toxicity—might occur. A reasonably careful follow-up schedule might be every 3 months for the ﬁrst year after treatment and then every 6 or 12 months for up to 5 years. Radiosurgery could play a signiﬁcant role in very advanced forms of neovascular membranes, which are not treatable using other procedures. On this basis, we have constructed a new prospective, nonrandomized study, which is now under way. The deﬁnitive analysis of results and complications in agerelated macular degeneration after radiosurgery still has not been completed, and further investigation is warranted. Additional study of the efﬁcacy and complications of this technique in comparison with external beam radiotherapy is indicated, and the role of this method still remains open.

Glaucoma Pathophysiology Glaucoma is a chronic, slowly progressive, usually bilateral neuropathy of the optic nerve. Untreated glaucoma eventually leads to complete loss of vision. About 10% of patients with glaucoma become uni- or bilaterally blind [7, 68]. We currently understand the pathophysiology of glaucoma to be a progressive loss of ganglion cells resulting in visual ﬁeld damage related to the intraocular pressure. Although many clinicians now feel that there are several factors involved in the pathogenesis of glaucoma, the only rigorously proven treatment method is the lowering of intraocular pressure.

Treatment Methods Conventional Treatment The conventional antiglaucomatous treatment—local and systemic pharmacotherapy, laser and cryotherapy, and incisional surgery—sometimes fails, and progressive optic nerve head

FIGURE 61-21. (A, B, C) The treatment planning of age-related macular degeneration with 15 Gy and 10 Gy isodose curves.

608

g. šimonová et al.

FIGURE 61-22. (A, B) Pretreatment angiography and (C) ultrasonography of age-related macular degeneration.

FIGURE 61-23. (A, B) Control angiography of ARMD and (C) ultrasonography 12 months after radiosurgery.

neuropathy is manifested. Then we face a difﬁcult situation of advanced and sometimes painful glaucoma with major defects in visual functions and, in the ﬁnal stages of the disease, enucleation of the painful blind eye must be considered. Therefore, a new effective method of treatment has to be sought.

Radiosurgery During radiosurgical treatment of patients with uveal melanomas localized near the ciliary body, in the past years at our center we observed that the radiation exerted a positive inﬂuence on painful secondary glaucoma with decrease of intraocular pressure [7]. The inﬂuence on pain is well-known from radiotherapy of degenerative and proliferative benign processes. The mechanism of this effect has still not been fully explained. The ciliary body, as the source of intraocular aqueous production, had been simultaneously irradiated using four 8mm collimators with median of minimal dose 15 Gy on a 50% isodose curve (Fig. 61-24). The dose to cornea should not exceed 7 to 8 Gy, and in some cases the use of plugging can help to decrease the dose to this structure.

The ciliary body as a target volume is located near the anterior to equator, and the doses for cornea, lens, and anterior chamber are higher than to the optic nerve and retina. A typical acute, transient reaction is lacrimation with incidence of 30%. The transient erythema on eyelid was observed after a single dose of 7 Gy and higher. Late-toxicity grade changes 1 or 2 of cornea was observed after Dmax 10 Gy. The worsening of lens opacity (grades 3 or 4) was recorded after the maximum dose to this structure higher than 10 Gy. The disappearance of pain after radiosurgery was observed in 77% with secondary glaucoma, and in all these patients no medication (analgesics) was required. The median period of the analgesic effect was 6 weeks (range, 2 to 32 weeks). In secondary glaucoma, the median of intraocular pressure (IOP) was 51.3 mm Hg (range, 23 to 68 mm Hg) but the median of IOP fell to 27 mm Hg (range, 8 to 48 mm Hg) after radiosurgery. In primary open angle glaucoma, the less raised IOP also fell from a median of 25.3 mm Hg (range, 17 to 35 mm Hg) to 16.1 mm Hg (range, 12 to 28 mm Hg) after radiosurgery [7]. Patients with advanced glaucoma (after standard treatment) have been included into a prospective, nonrandomized

FIGURE 61-24. The whole ciliary body irradiated with minimal dose 15 Gy on a 50% isodose curve shown on (A) axial, (B) coronal, and (C) sagittal images.

61.

ocular and orbital lesions

trial, and future investigation and analysis of results and complications is warranted. The risk of radiation in benign diseases must be balanced against the risk of alternative treatments, which are not negligible.

Conclusion The treatment of ocular tumors and other eye or orbital lesions can extend conservative therapeutic options for these types of diseases with vision or eye preservation. Further clinical studies using radiosurgery in ocular disease are necessary to optimize effective doses and to analyze factors inﬂuencing the length of survival with a careful evaluation of the side effects. The main aim of further clinical studies is to improve local treatment effects while decreasing severe radiation-related complications.

References 1. Lunsford LD, Kondziolka D, Flickinger JC. Gamma Knife Brain Surgery. Progress in Neurological Surgery, Vol. 14. Basel: Karger AG, 1998. 2. Gildenberg PL, Tasker RR. Textbook of Stereotactic and Functional Neurosurgery. New York: McGraw-Hill, 1998. 3. Tokuuye K, Akine Y, Sumi M, et al. Fractionated stereotactic radiotherapy for choroidal melanoma. Radiother Oncol 1997; 43:87–91. 4. Zehetmayer M, Menapace R, Kitz K, Ertl A. Suction ﬁxation system for stereotactic radiosurgery of intraocular malignancies. Ophthalmologica 1994; 208:119–121. 5. Langmann G, Pendl G, Schröttner O. Gamma knife of uveal melanoma radiosurgery for intraocular melanomas. Preliminary report. Spectrum Augenheilkd 1995; 9(Suppl 1):16–21. 6. Šimonová G, Novotný J Jr, Lišcˇák R, Pilbauer J. Leksell gamma knife treatment of uveal melanoma. J Neurosurg 2002; 97(Suppl 5):635–639. 7. Vladyka V, Lišcˇák R, Šimonová G, et al. Progress in glaucoma tretament research: a nonrandomized prospective study of 102 patients with advanced refractory glaucoma treated by Leksell gamma knife radiation. J Neurosurg 2005; 102(Suppl):214–219. 8. Bellmann C, Fuss M, Holz FG. Stereotactic radiation therapy for malignant choroidal tumors, preliminary, short-term results. Ophthalmology 2000; 107:358–365. 9. Debus J, Fuss M, Engenhart-Cabillic R. Stereotaktische Konformierende Bestrahlung von Aderhautmetastasen. Ophthalmologe 1998; 95:163–167. 10. Courdi A, Caujolle JP, Grange JD. Results of proton therapy of uveal melanomas treated in Nice. Int J Radiat Oncol Biol Phys 1999; 45:5–11. 11. Petersch B, Bogner J, Dieckmann K, et al. Automatic real-time surveillance of eye position and gateing for stereotactic radiotherapy of uveal melanoma. Med Phys 2004; 31:3521–3527. 12. Novotny J Jr, Vymazal J, Novotny J, et al. Does new magnetic resonance imaging technology provide better geometrical accurac during stereotactic imaging? J Neurosurg 2005; 102(Suppl):8–13. 13. Fransson A, Andreo P, Pötter R. Aspects of MRI image distortions in radiotherapy treatment planning. Strahlenther Onkol 2001; 177:59–73. 14. Walton L, Hampshire A, Forster D. Accuracy of stereotactic localization using magnetic resonance imaging. A comparison between two-and three-dimensional studies. Stereotact Funct Neurosurg 1996; 66(Suppl):49–56.

609

15. Georg D, Dieckmann K, Bogner J, et al. Impact of micromultileaf collimator on stereotactic radiotherapy of uveal melanoma. Int J Radiat Oncol Biol Phys 2003; 55:881–891. 16. Sing AD, Topham A. Incidence of uveal melanoma in the United States: 1973–1997. Ophthalmology 2003; 110(5):956–961. 17. Egan KM, Seddon JM, Glynn RJ, et al. Epidemiologic aspects of uveal melanoma. Surv Ophthalmol 1998; 116:366–370. 18. Margo CE, Mc Lean IW. Malignant melanoma of the chorioid and ciliary body in black patients. Arch Ophthalmol 1984; 102:77–79. 19. Fine SL, Straastma BR, Earle JD, et al. Failure of preenucleation radiation to decrease uveal melanoma mortality: the Collaborative Ocular Melanoma Study Steering Committee. Am J Opthalmol 1989; 107(4):440–442. 20. Collaborative Ocular Melanoma Study Group. Histopathologic characteristics of uveal melanomas in eyes enucleated from Collaborative Ocular Melanoma Study: COMS report No. 6. Am J Ophthalmol 1998; 125:745–766. 21. Manschot WA, van Strik R. Uveal melanoma: therapeutic consequences of doubling times and irradiation results. Opthalmology 1992; 16:91–99. 22. Straasma BR, Fine SL, Early JD, et al. Enucleation versus plaque irradiation for chorioidal melanoma. Ophthalmology 1988; 95: 1000–1004. 23. Damato BE, Paul J, Foulds WS. Risk factors for residual and recurrent uveal melanoma after trans-scleral local resection. Br J Ophthalmol 1996; 80:102–1088. 24. Finger PT. Radiation therapy for chorioidal melanoma (therapeutic review). Surv Ophthalmol 1997; 42:215–232. 25. Nag S, Quivey JM, Earle JD. The American Brachytherapy Society recommendations for brachytherapy of uveal melanomas. Int J Radiat Oncol Biol Phys 2003; 56(2):544–555. 26. Karvat A, Duzenli C, Ma R, et al. The treatment of chorioidal melanoma with 198 Au plaque brachytherapy. Radiother Oncol 2001; 59:153–156. 27. The Collaborative Ocular Melanoma Study (COMS). Randomized trial of I-125 brachytherapy for medium chorioideal melanoma. III. Initial mortality ﬁndings. COMS report, 2001. Arch Ophthalmol 119:969–982. 28. Potter CL, Shields CL, Shields JA, et al. Plaque radiotherapy for juxtapapillary chorioidal melanoma. Arch Ophthalmol 1996; 114:1357–1365. 29. Fontanesi J, Meyer D, Shizhao X, et al. Treatment of chorioidal melanoma with I-125 plaque. Int J Radiat Oncol Biol Phys 1993; 26:619–623. 30. Packer S, Rotman M, Slaniotro P. Iodine-125 irradiation of chorioidal melanoma: clinical experience. Ophthalmology 1984; 91:1700–1708. 31. Char DH, Quivey JM, Astro JR, et al. Helium ions versus iodine 125 brachytherapy in the management of uveal melanoma. A prospective, randomized, dynamically balanced trial. Ophthalmology 1993; 100:1547–1554. 32. Lommatzch PK. Results after beta-irradiation (106 Ru/106 Rh) melanomas: 20 years’ experience. Br J Ophthalmol 1986; 70: 844–885. 33. Potter R, Jansen K, Prott FJ. Ruthenium-106 eye plaque brachytherapy in the conservative treatment of uveal melanoma: evaluation of 175 patients treated with 150 Gy from 1981 to 1989. Front Radiat Oncol 1997; 30:143–149. 34. Finger PT, Berson A, Syechter A. Palladium-103 plaque radiotherapy for chorioidal melanoma: an 11-year study. Int J Radiat Oncol Biol Phys 2002; 54(5):1438–1445. 35. Egger E, Zografos L, Giotein G. Proton beam irradiation of chorioidal melanoma: techniques and results. In: Sagerman KH, Alberti WE, eds. Radiotherapy of Intraocular and Orbital Tumors, 2nd rev. ed. Berlin: Springer-Verlag, 2003:67–80.

610

g. šimonová et al.

36. Seddon JM, Gragoudas ES, Albert DM, et al. Comparison of survival for patients with uveal melanoma after treatment with proton beam irradiation. Am J Ophthalmol 1985; 99:282– 290. 37. Egan KM, Gragoudas ES, Seddon JM. The risk of enucleation after proton beam irradiation of uveal melanoma.Ophthalmology 1989; 96:1377–1382. 38. Gragoudas ES, Egan KM, Seddon JM, et al. Intraocular recurrence of uveal melanoma after proton beam irradiation. Ophthalmology 1992; 99:760–766. 39. Seddon JM, Gragoudas ES, Albert DM, et al. Visual outcome after proton beam irradiation of uveal melanoma. Ophthalmology 1986; 93:666–674. 40. Seddon JM, Gragoudas ES, Albert DM. Comparison of survival rates for patients with uveal melanoma after treatment with proton beam irradiation or enucleation. Am J Ophthalmol 1985; 99:282–290. 41. Gragoudas ES, Seddon JM, Egan K. A randomized controlled trial of varying radiation doses in the treatment of chorioidal melanoma. Arch Ophthalmol 2001; 18:773–778. 42. Castro JR, Char DH, Petti PL, et al. 15 years experience with helium ion therapy for uveal melanoma. Int J Radiat Oncol Biol Phys 1997; 39:989–996. 43. Char DH, Castro JR, Quivey JM, et al. Helium charged particle therapy for chorioidal melanoma. Ophthalmology 1980; 87: 565–570. 44. Char DH, Saunders W, Castro JR. Helium ion therapy for chorioidal melanoma. Ophthalmology 1983; 90:1219–1225. 45. Char DH, Kroll SM, Castro J. Ten-year follow up of helium ion therapy for uveal melanoma. Am J Ophthalmol 1998; 125:81–89. 46. Bogner J, Petersch B, Goerg D, et al. A noninvasive eye ﬁxation and computer-aided eye minitoring system for linear acceleratorbased stereotactic radiotherapy of uveal melanoma. Int J Radiat Oncol Biol Phys 2003; 56(4):1128–1136. 47. Dieckmann K, Georg D, Zehetmayer M, et al. Linac based stereotactic radiotherapy of uveal melanoma: 4 years clinical experince. Radiother Oncol 2003; 67:199–206. 48. Zehetmayer M, Kitz K, Menapace R, et al. Local tumor control and morbidity after one to three fractions of stereotactic external beam irradiation for uveal melanoma. Radiother Oncol 2000; 55:135–144. 49. Zehetmayer M, Menapace R, Kitz K, et al. Stereotactic irrradiation of uveal melanoma with Leksell gamma unit. Front Radiat Ther Oncol 1997; 30:47–55. 50. Zehetmayer M, Menapace R, Kitz K, et al. Stereotactic irradiation of uveal melanoma with Leksell gamma unit. In: Weigel T, Bornfeld N, Foester MH, Hinkelbein W, eds. Radiotherapy of Ocular Disease. Basel: Karger, 1997:165–177. 51. Marchini G, Gerosa M, Piovan E, et al. Gamma Knife stereotactic radiosurgery for uveal melanoma: clinical results after 2 years. Stereotact Funct Neurosurg 1996; 66(Suppl 1):208–213.

52. Rand RW, Khonsary A, Brown WJ, et al. Leksell stereotactic radiosurgery in the treatment of eye melanoma. Neurol Res 1987; 9:142–146. 53. Rennie I, Forster D, Kemeny A, et al. The use of single fraction Leksell stereotactic radiosurgery in the treatment of uveal melanoma. Acta Ophthalmol Scand 1996; 74:558–562. 54. LENT SOMA tables. Radiother Oncol 1995; 35:17–60. 55. Leer JWH, Van Houtte P, Davelaar J. Indications and treatment schedules for irradiation of benign diseases: a survey. Radiother Oncol 1998; 48(3):249–257. 56. Kim MK, Char DH, Castro JL, et al. Neovascular glaucoma after helium ion irradiation for uveal melanoma. Ophthalmology 1986; 93:189–193. 57. Park SS, Walsh SM, Gragoudas ES. Visual ﬁeld deﬁcits associated with proton beam irradiation for parapapillary chorioidal melanoma. Ophthalmology 1996; 103:110–116. 58. Saornil MA, Egan KM, Gragoudas ES, et al. Histopathology of proton beam- irradiated vs enucleated uveal melanomas. Arch Ophthalmol 1992; 110:1112–1118. 59. Augsburger JJ, Gonder JR, Amsel J, et al. Growth rates and doubling times of posterior uveal melanomas. Arch Ophthalmol 1984; 91:1709–1715. 60. Sahel JA, Pesavento R, Frederick AR, et al. Melanoma arising de novo over a 16-months period. Arch Ophtalmol 1988; 106: 381–385. 61. Bloch RS, Gartner S. The incidence of ocular metastatic carcinoma. Arch Ophthalmol 1971; 85:673–675. 62. Rudoler SB, Corn BW, Shileds C, et al. External beam irradiation for chorioid metastases: identiﬁcation of factors predisposing to long-term sequelae. Int J Radiat Oncol Biol Phys 1997; 38(2): 2551–256. 63. Bhatia S, Paulino A, Buatii JM. Curative radiotherapy for primary orbital lymphoma. Int J Radiat Oncol Biol Phys 2002; 54(3): 818–823. 64. Berson AM, Finger PT, Sherr DL. Radiotherapy for age-related macular degeneration: preliminary results of a potentially new treatment. Int J Radiat Oncol Biol Phys 1996; 36(4):861–865. 65. Slater JM, Archambeau JO, Miller D, et al. The proton treatment center at Loma Linda University Medical Center: rationale for the description of its development. Int J Radiat Oncol Biol Phys 1991; 22:383–389. 66. Macular Ophthalmolocoagulation Study Group. Laser photocoagulation of subuveal neovascular lesions in age-related macular degeneration: results of a randomized clinical trial. Arch Ophthalmol 1991; 109:1220–1231. 67. Haas A, Papaefthymiou G, Langmann G. Gamma knife treatment of subfoveal, classic neovascularization in age-related macular degeneration: a pilot study. J Neurosurg 2000; 93(Suppl 3): 172–176. 68. Fuchs HJ, Nissen KR, Goldschmidt E. Glaucoma blindness in Denmark. Acta Ophtalmol 1992; 70:73–78.

6 2

Stereotactic Body Radiation Therapy Laura A. Dawson

Introduction Technological advancements in imaging and radiation planning and delivery have made it possible for cranial stereotactic radiosurgery techniques to be applied to tumors outside of the brain. Although high-dose radiation therapy may be delivered in a single fraction, referred to as extracranial stereotactic radiosurgery (SRS), more often, high-precision radiation is delivered in more than one fraction, leading to the ﬁeld of stereotactic body radiation therapy, or SBRT. SBRT refers to the use of a limited number of high-dose fractions delivered very conformally to targets with high accuracy, using biologic doses of radiation far higher than those used in standard fractionation. Stereotactic refers to the use of a reference system to aid in localizing the tumor. An external reference system, such as a stereotactic head frame for cranial SRS or internal ﬁducial markers including the tumor itself may be used for localization and guidance. For tumors ﬁxed to boney anatomy, such as base of skull cancers, image guidance using boney anatomy may be appropriate; however, often the position of the tumor is not highly correlated with the bones or an external reference system, and imaging and targeting of the tumor itself may be required. SBRT is becoming more popular, as evidenced by increasing recognition at international meetings, a textbook devoted to the topic [1], and an American Society for Therapeutic Radiology and Oncology (ASTRO) consensus document on SBRT [2]. The ASTRO consensus document deﬁnes SBRT as highprecision radiotherapy delivered in three or more fractions to very potent doses of highly conformal radiation with steep dose gradients around the target. SBRT is “an evolving technology,” and deﬁnitions may change with time. The philosophies and techniques of SBRT can be applied to longer fractionation regimens (when the target in intimately associated with a critical serial functioning normal tissue, for example) or shorter fractionations, including SRS. Similar to cranial SRS, multiple static or dynamic beams in a variety of beam arrangements, with or without segments or intensity modulation, can be used to produce a dose distribution in which isodose lines tightly conform to the target volume. Although most reports on SBRT refer to megavoltage photon irradiation, protons are not excluded from the SBRT deﬁnition. Inhomogeneity within the target volume is permitted and hot

spots are encouraged within the planning target volume (PTV) to increase the chance of tumor ablation. Dose is generally prescribed to an isodose line covering the PTV with a very steep dose gradient outside the PTV. SBRT is a noninvasive, outpatient intervention, generally complete within a week or two, not substantially delaying systemic therapy. The short radiation treatment time and high dose per fraction in SBRT have potential radiobiological therapeutic advantages compared with conventional fractionation, due to less tumor repopulation and repair. Furthermore, the shorter SBRT treatment times are more convenient for patients and have resource utilization advantages. Alternative local treatment options to SBRT are available for some tumor sites. These options, including surgery and radiofrequency ablation, are most often invasive, with associated risks such as tumor seeding. The nonsurgical ablative treatments are not well suited for tumors greater than 5 cm in diameter. Although tumors of this size have not been the focus of most SBRT research, larger tumors that are unresectable have few therapeutic options, and SBRT should be investigated in this setting. The rationale for SBRT is that there is a need for improved local therapies for many primary cancers and also in the situation when there are “oligo” (i.e., isolated) metastases, speciﬁcally in sites where surgery has been shown previously to be able to cure patients with “oligo” metastases (e.g., colorectal cancer, renal cell cancer, and sarcoma). SBRT is less invasive than surgery, and once the safety of SBRT has been established, SBRT has the potential to be used in place of surgery in situations when surgery may be associated with high risk. Even in incurable situations, SBRT may also be used to control local disease to improve symptoms and quality of life when the role of conventional radiation therapy is limited. Advancements outside of radiation oncology also provide rationale for SBRT. Functional imaging such as positron emission tomography (PET) allows better patient selection for SBRT. Furthermore, improvements in systemic therapy more likely to control micrometastases provide rationale for improving local therapies, such as SBRT, to reduce large foci of tumor burden. In this chapter, a historical perspective, rationale, technical details, and literature on SBRT outcomes and complications will be reviewed. Speciﬁc applications of SBRT to be reviewed include lung cancer, hepatocellular carcinoma, paraspinal

611

612

l.a. dawson

malignancies, renal cell carcinoma, pancreas cancer, and SBRT for oligo-metastases in general.

Historical Perspective The ﬁrst extracranial site to be investigated with SRS techniques was the spine in 1993. A rigid skeletal ﬁxation device that immobilized the spine above and below the target was used as a stereotactic reference frame to guide radiation [3]. Around the same time, Blomgren and Lax from Sweden began applying stereotactic techniques to body targets such as lung and liver tumors. They developed a body stereotactic reference frame to aid in guidance of radiation [4]. Blomgren et al. reported their initial experience in 1995, in which 31 patients with primarily solitary lung and liver tumors were treated from doses of 7.7 to 45 Gy in one to four fractions. Reproducibility of this approach was found to be within 5 to 8 mm for 90% of setups with diaphragm motion reduced to 5 to 10 mm with abdominal pressure. At the Karolinska Institute in Stockholm, Sweden, since the introduction of SBRT in 1991, close to 2000 patients have been treated with SBRT [1]. Clinical outcome and toxicity data after SBRT have now been published by numerous international groups, primarily from Europe and Asia, in which a variety of primary and metastatic tumors have been treated using a variety of fractionation regimens, ranging from 7.7 to 60 Gy in 1 to 10 fractions. In Germany, single fractions of 24 Gy and three fractions of 12.5 Gy have been reported to be safe and effective at eradicating small primary and metastatic liver cancers [5, 6]. In Asia, there is substantial experience using 3- to 10-fraction SBRT for primary lung and liver tumors [7–9], with encouraging outcomes. Experience in SBRT in North America is growing. However, mature clinical outcomes have only recently been reported. The ﬁrst North American study of SBRT was published in 2003 by Timmerman et al. [10]. This was a dose escalation trial for primary lung cancer, in which 60 Gy in three fractions was found to be tolerable for peripheral small primary lung cancers. Based on this experience, the Radiation Oncology Therapy Group (RTOG) initiated a multi-institutional phase II trial of SBRT in lung cancer in 2004. Other North American groups have also recently reported on their experience with SBRT for liver metastases [11–13], and an RTOG study of 10-fraction SBRT for liver metastases is planned to open in 2006. Paraspinal cancer SBRT is also an area of active research [14, 15]. SBRT and body SRS have also been investigated for head and neck tumors [16], as well as for tumors of the retroperitoneum and pelvis.

Radiobiology Radiobiologic Models Traditional radiobiologic modeling may not be entirely valid for very short fractionation schemes. The linear quadratic model may overpredict tumor cell kill and underpredict toxicity at increased dose per fraction. However, the linear quadratic model can be useful in guiding development of most suitable SBRT fractionations [17]. Fowler et al. have predicted that local

control should be excellent for non–small cell lung cancer with biological doses of the order 100 Gy in 2-Gy equivalent dose or higher (for an α/β of 10). Fractionation schemes with an estimated tumor control probability of 95% or greater include 45 Gy to 60 Gy in three fractions and 60 Gy in ﬁve fractions. It has been suggested that the overall treatment time should be kept to within 3 weeks to minimize the adverse effects of repopulation. Other radiobiologic models that have been used in the setting of hypofractionation include the repair-misrepair model [18] and the modiﬁed linear quadratic (MLQ) model [19].

Tumor Response to SBRT For most solid malignancies, a dose response has been observed; thus, there is rationale for dose escalation. Given that dose escalation with conventionally fractionated radiation therapy requires substantial prolongation of treatment time, which is associated with repopulation and a reduced chance of tumor control, there is rationale for hypofractionation. The disadvantages of hypofractionation include reduced reassortment and reoxygenation of tumor and reduced repair and repopulation of normal tissues, leading to a reduction in therapeutic ratio. Conversely, the advantages of hypofractionation include less repair and repopulation of tumors and less reassortment and reoxygenation of normal tissues. One substantial difference in SBRT compared with conventional radiation therapy or conformal radiation therapy is that the very high biological doses used in SBRT are much more likely to completely ablate the tumor. The most common SBRT fractionation schemes utilizes one to ﬁve fractions of greater than 6 Gy per fraction, and as high as 20 Gy per fraction, usually prescribed to the periphery of the PTV. Although these highly biologically potent prescription doses are associated with a very high tumor control probability, there is also a higher risk of normal tissue toxicity if normal tissues tolerances are not respected and/or if a systematic error occurs in treatment delivery when a target is adjacent to critical normal tissue. One potential disadvantage of SBRT is that there is less time for the beneﬁts of chemosensitization if concurrent chemotherapy is to be used. Conversely, if enough dose can be delivered, the need for chemosensitization is reduced. In contrast, there is strong rationale for use of concurrent hypoxic targeted agents with SBRT, as less tumor reoxygenation occurs with fewer fractions. Concurrent chemotherapy or other sensitizing agents can be delivered with SBRT, although the therapeutic ratio is optimized primarily using physical methods (highly conformal radiotherapy) rather than biological methods (fractionation). Another potential disadvantage of SBRT is that there is a risk that the treatment delivery is so complex that the actual delivery time may become long (e.g., more than 30 minutes). This may lead to increased clonogenic survival and less chance of tumor control [20]. Furthermore, when the treatment time becomes too long, there is an increased chance that the patient, tumor, and normal tissues may change position, size, and shape during treatment, leading to alterations in the delivered dose compared with the prescribed dose. Finally, although modeling suggests an increased tumor control probability with SBRT rather than escalated dose conformal radiation, and initial clinical results after SBRT are

62.

stereotactic body radiation therapy

excellent, there is no level 1 evidence that the same outcomes could not be obtained in similar patients treated with escalated dose conformal radiation therapy. Hopefully, clinical trials will address this once SBRT quality assurance guidelines become available and the safety of hypofractionated SBRT has been established.

Normal Tissue Response to SBRT Normal tissues that function primarily as parallel normal tissues are made up of predominately structurally deﬁned functional subunits (FSUs) (e.g., peripheral lung, liver, kidney). Serial functioning normal tissues, characterized by a chain of function, are composed of primarily undeﬁned FSUs (e.g., gastrointestinal mucosa, trachea, spinal cord, bronchus). Parallel organ toxicity is mostly related to volume irradiated (e.g., mean dose, volume treated to a threshold dose such as V20), whereas serial organ toxicity is mostly related to the maximum dose delivered to that tissue. Although useful, this distinction between serial and parallel functioning organs is likely too simplistic, and most normal tissues likely have both parallel and serial functionality. This has been demonstrated in elegant rat model experiments of spinal cord tolerance to radiation therapy by van der Kogel et al. [21]. The rat spinal cord tolerance to radiation was found to be dependent on the volume irradiated and the spatial distribution of dose. The gray matter of the cord was found to be most resistant, whereas lateral white matter was most sensitive. As the volume of normal tissues irradiated to high doses with SBRT is generally less than the volume of normal tissue irradiated with conventional fractionation, for normal tissues that are primarily parallel functioning, if the volume irradiated is low enough, the risk of toxicity may be extremely low despite delivery of very high doses to a small volume. However, for serial organs that are in close proximity to tumors, even a small volume irradiated to a high dose may lead to irreversible toxicity. Thus, SBRT should be used cautiously for tumors adjacent to serial functioning organs such as the esophagus or spinal cord. An increase in the number of fractions may improve the therapeutic ratio and should be considered in place of hypofractionated SBRT in this situation. With a dramatic increase in the fraction size and total dose with SBRT, the repair mechanisms may not be initiated as they would at a lower dose per fraction, potentially leading to permanent damage to the normal tissue within the high dose volumes and/or unpredictable effects outside the high dose volume. Given the potential for normal tissue injury and that the volume tolerance of normal tissues to such high doses per fraction is unknown, most SBRT has been applied to small tumors (<7 cm) in which the volume of normal tissue around the tumor is small. SBRT is being investigated for larger tumor volumes [12], and it is hopeful that clinical data will eventually be obtained to provide guidance to what the dose-volume toxicity relationship is for organs irradiated with inhomogeneous doses from hypofractionated fractionation schemes. As very steep dose gradients are used with SBRT with rapid fall-off of dose in surrounding normal tissues, not only is there substantial variability of dose throughout normal tissues, but also there is substantial variability in the dose per fraction. Thus, when dose-volume characteristics of a normal tissue asso-

613

ciated with toxicity risk are investigated, correction for the variability in dose per fraction should be considered [22]. Unfortunately, accurate knowledge of the α/β ratio for normal tissues is lacking. Clinical trials of novel fractionation regimens are required to conﬁrm outcomes and toxicities associated with the many different fractionation schemes used in SBRT.

Rationale Natural History General SBRT has been investigated for primary and metastatic lung cancer, primary and metastatic liver cancer, primary and metastatic paraspinal malignancies, isolated nodal metastases, and cancers of the retroperitoneal space and the pelvis including renal cell cancer and pancreatic cancer. Many of these primary tumors have been reported to have a dose response with increased chance at tumor control with escalated doses. For lung cancer, hepatocellular cancer, and retroperitoneal and paraspinal sarcoma, local recurrence remains the predominant pattern of disease recurrence after conventional radiation therapy. Local recurrences are often associated with substantial morbidity and can lead to death. Escalated dose with standard fractionation is associated with very long treatment times, and thus SBRT has the potential to deliver very high biological equivalent doses in a far shorter time.

Lung Cancer Lung cancer is the ﬁrst most common cause of cancer death, with the majority of patients being unresectable. Surgery is associated with the most mature 5-year survival rates for non– small cell lung cancer: 60% to 80% for stage I, 30% to 40% for stage II, and less for more advanced disease. Thus, surgery is the preferred treatment for early lung cancer. Most patients have underlying lung disease and, in many patients, surgery is associated with high risks of morbidity or is not possible at all. Radiation therapy with or without chemotherapy has played a key role in the treatment of lung cancers. However, some patients cannot be treated with conventional fractionation, due to very poor lung function. Patients with unresectable early disease (T1–3N0), especially those with peripherally located tumors, are the most appropriate patients for SBRT. Clinical trials have demonstrated improved outcomes with shortening of the overall treatment time for advanced stage lung cancer, providing rationale for reducing time of treatment for early lung cancer.

Hepatocellular Carcinoma Hepatocellular carcinoma is the sixth most common cancer in the world (626,000 cases/year) and the third most common cause of cancer-related death (598,000 deaths/year). Although predominately a problem in developing countries, the incidence is expected to rise over the next decade in North America. Unfortunately, the overall 5-year survival rate for all patients

614

l.a. dawson

with hepatocellular carcinoma has remained steady at 3% to 5%, and there is a need for improved therapies and improved integration of therapies. Although surgery and transplant offer patients a chance at cure, with 5-year survival rates ranging from 30% to 70%, these options are only feasible in approximately 15% of all patients with hepatocellular carcinoma. Technical advancements in imaging, planning, and delivery of radiation have made it possible for radiation to be used in the setting of hepatocellular carcinoma, as dose can be spared from the surrounding uninvolved liver while high doses can be delivered to the focal tumors [23–25]. With increased dose conformality, less liver volume is required to be irradiated, with a reduced risk of liver toxicity and the potential for hypofractionated dose escalation (i.e., SBRT).

Pancreatic Cancer Pancreatic cancer is the fourth cause of cancer death in North America, with approximately 3000 new cases being reported per year in Canada alone. Pancreatic cancer has a particularly poor prognosis with a median survival of 6 months, and a 5-year survival rate of less than 5%. The majority of patients are not candidates for surgical resection. Radiation therapy has a role in locally advanced therapy, but conventional fractionated radiation has not led to large improvements in local control or survival. Pancreas cancer has a high propensity to metastasize distantly, and systemic therapy should be considered for the majority of pancreatic cancer patients.

Renal Cell Cancer More than 30,000 cases of renal cell cancer are diagnosed annually in the United States, with approximately 40% of those cases fatal. After curative resection for renal cell cancer, the most common recurrence pattern is distant, with local recurrences being rare. Interleukin-2 is a treatment option for systemic disease, but the response rate is less than 20%. For isolated sites of metastases, surgical resection is sometimes offered to patients. Although challenging to interpret the outcomes of these patients due to selection bias, the 5-year survival rates of 35% to 44% are often better than systemic therapy alone. In one clinical trial that randomized 240 patients with metastatic renal cell cancer to nephrectomy and immunotherapy versus immunotherapy alone, survival was signiﬁcantly improved with nephrectomy, suggesting that reduction of tumor burden may be worthwhile [26]. Although historically, renal cell cancer has been thought to be “radioresistant,” several clinical studies have disputed this. Thus, in place of surgery, SBRT could be used to improve tumor control and survival.

Paraspinal Malignancies Tumors of the spine and paraspinal region are most often extradural metastases, with approximately 18,000 new cases of spinal metastases developing in patients annually in North America. Most often, spinal metastases are symptomatic, with pain a common presenting symptom. The proximity of the spinal cord to paraspinal tumors and the morbidity associated with spinal cord myelitis make the treatment of primary and metastatic spinal and paraspinal malignancies challenging. This is a site

where the ability to tightly conform high-dose isodoses around the tumor with a steep gradient to limit dose to the spinal cord have large potential advantages. Setup accuracy must be good enough to allow a reduction in PTV margins. For paraspinal metastases, many patients have incurable disease, and reducing overall treatment time with SBRT is more convenient than longer fractionations.

Oligo Metastases Oligo, or isolated, metastases can be cured in the liver and lung in 25% to 50% of patients. However, such aggressive local therapy has traditionally been limited to few select clinical scenarios, including liver or thoracic resection of colorectal metastases, and metastectomy of isolated metastases from renal cell carcinoma or sarcoma. In the era of improved systemic therapies that have a better chance to treat occult disease, as well as improved imaging that may detect patients with oligo metastases who would otherwise not be diagnosed, there is more rationale to attempt to expand the guidelines for aggressive local therapy for isolated metastases. In the future, molecular markers may also help to better deﬁne in which patients such aggressive local therapy is beneﬁcial. Clinical trials are warranted to test the hypothesis that aggressive local therapies should lead to improved survival.

Clinical Trials No randomized trials of SBRT have been completed to date, although one was initiated in Germany to compare one-fraction SRS to three-fraction SBRT for liver metastases. There is little mature data from prospective multiinstitutional trials of SBRT, as few multi-institutional studies have been completed and published. However, there are several ongoing multi-institutional studies in North America and Europe; thus, local control, survival, and toxicity outcomes are expected to become available in the future.

Expected Outcomes SBRT is noninvasive, delivered as an outpatient within a short time period, not signiﬁcantly delaying systemic therapies. The short radiation treatment time and high dose per fraction in SBRT have potential radiobiological therapeutic advantages, with less tumor repopulation and repair. SBRT is convenient for patients and has a large potential beneﬁt in resource utilization compared with conventional fractionation. SBRT delivered with appropriate PTV margins, organ immobilization or tracking, image guidance, and enough dose should result in excellent local control for most clinical situations that it has been investigated in. For small tumors, even without organ immobilization and image guidance, local control is likely with a variety of fractionation regimens, assuming PTV margins are large enough. Improved local control may result in improved survival, but clinical trials are required to conﬁrm in which patient populations this is most likely. Toxicities after SBRT include toxicities after conventional therapy as well as some serious toxicities unique to SBRT. Furthermore, the late sequelae from SBRT may not manifest until many years after therapy.

62.

stereotactic body radiation therapy

Despite encouraging single-institutional experience and multi-institutional experience from Asia and Europe, there is a need for multi-institutional North American trials. Ultimately, randomized controlled trials in SBRT to determine in which patient populations these techniques have the most beneﬁt are required.

615

concurrent hepatic arterial ﬂoxuridine for chemosensitization. This treatment strategy has been used at the University of Michigan, with mature local control and survival data for primary and metastatic liver cancers [29]. The median survival of hepatocellular carcinoma, cholangiocarcinoma, and metastatic colorectal cancer patients was 15.2 months, 13.3 months, and 17.2 months, respectively, all higher than historical controls. The actuarial 3-year survival for all patients was 17%.

Alternative Treatments Metastases

Alternative Radiation Fractionations General Most often, the sites chosen for SBRT require that high precision radiation therapy be used to ensure that the dose gradients are steep and the volume of normal tissue irradiated is small. Thus, the technological innovations that allow highly conformal radiation therapy to be delivered are also applicable to SBRT. Dose escalation using conventional fractionation high-precision radiotherapy is an alternative to SBRT, at the expense of longer treatment times. The technological advancements that make very high precision radiation therapy possible require increased quality assurance and increased time and effort for each fraction delivered. Thus, hypofractionated SBRT has resource utilization and patient convenience implications, in addition to theoretical radiobiologic beneﬁts. Alternative, more conventional fractionation schedules include 35 fractions, hyperfractionation, and accelerated fractionation. When the PTV overlaps with serial functioning normal tissues (such as the spinal cord or stomach), there are beneﬁts to reducing the dose per fraction and increasing the number of fractions, as this may reduce the risk of normal tissue late toxicity.

If the patient’s overall performance status is poor, low-dose palliative radiation therapy (e.g., 8 Gy in one fraction) is most often a more reasonable treatment option than SBRT. Single, 5, and 10 fractions of simple palliative radiation therapy are also more commonly more suitable for metastases treatment.

Alternative Local Therapies General Surgery has a well established role in many cancers, and until mature outcome and safety data are available after SBRT, surgery should remain the standard of care in resectable patients. Nonsurgical ablative treatments are becoming more widely available and are used in more widespread indications. The long-term outcomes and safety of these procedures is not as clearly established as for surgery. However, for some sites, outcome and safety data is mature. For the most part, these ablative techniques are most appropriate for smaller tumors, less than 5 cm in maximum diameter, away from large vessels.

Lung Cancer Lung Cancer For inoperable lung cancer, outcomes after conventional radiation therapy are worse then those in surgical series. This is partially due to selection bias. Nonetheless, many inoperable patients progress locally despite conventional radiation therapy, and there is the need for improved local therapies for these patients. Dose escalation to the primary tumor with sparing of dose to regional nodes has been used using conformal radiation safely, with improved outcomes [27]. For patients with small T1/T2N0 unresectable primary lung cancers, “mildly” hypofractionated radiotherapy schemes targeting the tumor without any regional nodal radiation have been used [28]. After 48 Gy in 12 fractions for T12N0 non–small cell carcinomas, 1- and 2-year survival rates were 80% and 46%, respectively, with a 2-year failure from local recurrence rate of 71% [28]. Without stereotactic or conformal radiation techniques, subcutaneous ﬁbrosis and skin toxicity was observed using this fractionation. Increased dose conformality using techniques such as SBRT should allow more dose to be delivered to the tumor while sparing dose to the peripheral subcutaneous tissue, which should facilitate safer dose escalation and improved local control.

Liver Cancer An example of an alternative fractionation for liver cancer is hyperfractionated radiation, 1.5 Gy twice daily, delivered with

Surgery is associated with the most mature 5-year survival results for non–small cell lung cancer, with 5-year survival of 60% to 80% for patients with stage I non–small cell lung cancer. Thus, surgery is the standard of care. Unfortunately, many lung cancer patients have comorbidities that increase surgical morbidity. Other local therapies such as radiofrequency ablation are being investigated for small primary and metastatic lung cancers.

Liver Cancer Surgery is the optimum local therapy for hepatocellular carcinoma primary and liver metastases from colorectal cancer, if the tumor can be removed with negative margins and minimal toxicity. For oligo-metastases from colorectal cancer, surgery is associated with 5-year survival rates of 40% to 70%, with a serious morbidity and mortality rate of approximately 5%. Liver transplant is an option for some patients with a single hepatocellular carcinoma less than 5 cm in maximum diameter or three lesions less than 3 cm in maximum diameter, with no macrovascular invasion or extrahepatic disease. Five-year survival rates after transplant range from 50% to 71%. Unfortunately, prolonged wait times for donor organs are associated with tumor progression and some patients becoming unsuitable for transplant or dying while on the wait list.

616

l.a. dawson

As only 15% of patients with hepatocellular carcinoma or liver metastases are candidates for resection or transplant, other local therapies have been investigated. Radiofrequency ablation (RFA) is the most common ablative technique, associated with excellent local control for small liver cancers. Other ablative techniques, including percutaneous ethanol injection, microwave coagulation therapy, laser-induced thermotherapy, and high-intensity focused ultrasound (HIFU) are being investigated in this setting. The most suitable tumors for these interventions are less than 5 cm in maximum diameter, as there is an increased risk of local recurrence in larger tumors and in tumors adjacent to major blood vessels. Furthermore, there are increased risks in tumors adjacent to the diaphragm and the stomach. A ﬁnal disadvantage to the majority of ablative approaches is that they are invasive and associated with a risk of tumor seeding.

Other Alternative Therapies Alternative treatments of chemotherapy, supportive care, or palliative care may be more appropriate than SBRT in many patients. Supportive and palliative care includes the use of percutaneous and internal stents and surgical bypass procedures that can often relieve symptoms faster than any radiation therapy option. Systemic chemotherapy is becoming more effective in many sites of cancers, including colorectal cancer and breast cancer. Thus, systemic therapy should be considered early once a diagnosis of metastases has been made. First- and possibly secondline chemotherapy should be used prior to SBRT. This may lead to tumor shrinkage and safer SBRT, while screening out the patients who develop widespread metastases quickly. In tumor sites with a propensity to metastasize distantly (e.g., pancreas cancer), systemic therapy should always be used if possible. For tumor sites without effective systemic therapy (e.g., hepatocellular carcinoma), SBRT prior to chemotherapy is reasonable. For hepatocellular carcinoma, transarterial chemoembolization (TACE) has been shown to have a modest survival beneﬁt compared with supportive care in patients with unresectable hepatocellular carcinoma in two randomized trials and a meta-analysis [30]. The patients most likely to beneﬁt from TACE are those without macrovascular tumor invasion. Systemic chemotherapy has had limited impact in hepatocellular carcinoma.

Immobilization Overview Given the sensitivity of highly conformal SBRT plans to setup uncertainty and organ motion, reduction of geometric uncertainties and organ motion is important. The choice of patient position and immobilization may impact setup error, organ motion, as well as intrafraction motion secondary to patient discomfort. For example, prostate motion due to breathing is reduced in the supine position compared with the prone position [31, 32]. Detailed comparisons of different immobilization devices and patient positions have not been investigated in the

setting in SBRT. The different immobilization devices and strategies that have been used for SBRT are described below.

Paraspinal Tumor Immobilization Similar to intracranial SRS with a cranial halo secured to the skull with transcutaneous pins, rigid ﬁxation systems have been used for paraspinal and spinal SBRT. Hamilton in 1995 immobilized the spine for SBRT with rigid skeletal ﬁxation above and below the involved segments. With this system, excellent accuracy, with less than 2-mm offsets, was observed [33]. However, such an approach is invasive, and avoidance of invasive ﬁxation systems is desirable to minimize risks. Noninvasive stereotactic systems using a frame, with or without implanted ﬁducial markers, have been used for paraspinal tumor SBRT with accuracy within 2 mm [34]. Opticalguided three-dimensional ultrasound has also been used for spinal SBRT to ensure that the patient dose not move during radiation [35].

Body Tumor Immobilization Nonrigid ﬁxation has been performed for SBRT with specialized body frames that have ﬁducial systems attached to the frame and a device to control respiration using abdominal compression [36]. Abdominal compression using such frames has been found to reduce diaphragm caudal cranial motion to less than 5 to 10 mm in most patients. Reproducibility of target positioning using these frames has been reported to be better than conventional immobilization systems, with positional deviations of lung cancer and liver cancer position less than 10 mm in 98% of patients [36–38]. An option to these specialized body frames that facilitates guidance is to not use the frames for guidance, but to image internal anatomic references, such as bones near the target or the soft tissue tumor itself, to deﬁne the treatment coordinates. Repositioning and repeat veriﬁcation imaging is required to ensure the patient was moved to the correct position. A variety of immobilization devices can been used with this strategy, as long as they keep the patient immobilized and preferably are (relatively) comfortable to minimize intratreatment patient motion.

Correction for Breathing Motion Organ motion due to physiologic functions during a radiation treatment fraction can be substantial. For example, the liver can move up to 5 cm in the caudal-cranial direction during free breathing [39], causing motion of the upper abdominal and lower thoracic cavity. As this motion can result in alternations in target and normal organ volume deﬁnitions, PTV margins and the entire dose distribution, interventions to reduce the impact of intratreatment organ motion are required for many SBRT patients, to facilitate dose escalation and reduce the volume of normal tissue irradiated. Strategies to compensate for breathing motion include the use of abdominal pressure, voluntary shallow breathing, voluntary deep inspiration, voluntary breath holds at variable phases of the respiratory cycle, active breathing control

62.

stereotactic body radiation therapy

(ABC), gated radiotherapy, and real-time tumor tracking. Although voluntary breath holds may be beneﬁcial for some patients, there is potential for leaking air and patient error, particularly for patients with lung disease. ABC refers to organ immobilization with breath holds that are controlled, triggered, and monitored by a caregiver. In approximately 60% of patients with liver cancer, ABC was able to be used successfully, with excellent reproducibility of the liver relative to the vertebral bodies within the time period of one radiation fraction (intrafraction reproducibility, σ, of the liver relative to the vertebral bodies: 1.5 to 2.5 mm) [40, 41]. However, with ABC, from day to day the position of the immobilized liver varies relative to the bones (interfraction reproducibility, σ, 3.4 to 4.4 mm), providing rationale for daily image guidance when ABC is used for liver SBRT (Figs. 62-1 and 62-2). Gated radiotherapy, with the beam triggered to be on only during a predetermined phase of the respiratory cycle, usually refers to the use of an external surrogate for tumor position (as opposed to direct tumor imaging) to gate the radiation. This can be used to reduce the volume of normal tissue irradiated. Changes in baseline organ position can occur from day to day [42], and thus image guidance is important to avoid geographic misses, especially in the setting of SBRT. Tumor tracking is another approach to reduce adverse effects of organ motion. An elegant real-time tumor tracking system consisting of ﬂuoroscopic X-ray tubes in the treatment room allowing visualization of radiopaque markers in tumors was ﬁrst described by Shirato et al. The linear accelerator is triggered to irradiate only when the marker is located within the planned treatment region [43]. As an alternative to turning the radiation beam off when the tumor moves outside treatment region, multileaf collimators, the couch position, or the entire accelerator on a robotic arm may move with the tumor to ensure adequate tumor coverage at all times (e.g., CyberKnife image-guided radiosurgery system; Accuray, Sunnyvale, CA). The latter system uses dual orthogonal ﬂuoroscopy tubes to track radiopaque markers in or near the tumor at a preset frequent.

617

FIGURE 62-2. Patient at CT simulator with active breathing control to stabilize the liver during imaging.

There are advantages to gating, breath hold, and tracking in exhalation phase of the respiratory breathing cycle versus inhalation. These include the fact that exhale tends to be more reproducible and is longer than inhalation, so that treatment during exhale reduces duty time. However, in certain situations, there may be rationale for breath hold, gating, or tracking during inhalation. For example, for lung tumors and/or tumors adjacent to the heart, inhalation will reduce the density of the lungs and/or may move critical structures away from the target volume. Other approaches to minimize respiratory and nonrespiratory organ motion include maintaining the same preparative regimen prior to each treatment and ensuring comfortable immobilization and short overall treatment time to reduce patient movement due to discomfort and physiologic change in organs such as stomach ﬁlling. Another general intervention is patient feedback, either auditory or visual. Ideally, feedback would be from direct tumor imaging; however, feedback from imaging of adjacent organs, spirometry, nasal ﬂow monitoring, external marker position, or optical monitoring of body contour are often more practical options. If indirect measures of organ position are used, conﬁrmation for an individual patient that the indirect measure is indeed directly related to organ position is mandatory.

Treatment Planning Overview

FIGURE 62-1. Example of patient immobilized with active breathing control to stabilize the liver during SBRT.

As dose gradients are steeper and doses are higher with SBRT than with conventional radiation therapy, the consequences of error in tumor delineation, errors introduced by dosimetry and geometric uncertainties may be more deleterious. Thus, all aspects of treatment planning that are important in conformal radiation planning are even more crucial in SBRT, especially for tumors in close proximity to critical normal tissues, where a systematic error could lead to permanent serious toxicity if the normal tissue planned to be spared from radiation is irradiated to the high doses planned for the tumor.

618

l.a. dawson

Imaging at Simulation At the time of simulation, patient positioning and the imaging modality (CT, MRI), resolution (e.g., CT thickness), phase of contrast (e.g., arterial IV contrast for hepatocellular carcinoma) must be chosen carefully. Motion must be considered at this time, as breathing introduces artifacts in the tumor deﬁnition, normal tissue deﬁnition, and resultant errors in tumor control probability (TCP) and normal tissue complication probability (NTCP). Furthermore, if motion is not considered, there is potential for a systematic error from the time of simulation to the time of treatment to occur. Motion due to breathing is largest for tumors near the diaphragm (i.e., the upper abdomen and the lower thorax). One method to account for motion is to eliminate it, for example with a breath-hold scan. Diagnostic breath-hold scans are often obtained in the inhale position, which may not correspond with the treatment position (e.g., exhale breath hold). An effort to conduct all imaging to be used for planning with the patient in the same position, with the same phase of breath hold is required. If breath hold is not possible, reduction of breathing motion may help reduce the negative impact of breathing motion. However, even with a small range of breathing motion, errors in tumor and normal tissue volumes may occur, resulting in a geographic miss or excessive toxicity. An option to breath-hold imaging is to obtain a fourdimensional imaging data set. From this, any position could be used for planning and image guidance [42]. Planning on the exhale data set with asymmetric PTV margins is an option [44], as is planning using the mean tumor position. The phase of the breathing cycle in which the patient is planned should correspond with the phase of breathing cycle used for image guidance and treatment.

Target Volumes A decision has to be made regarding whether a margin is required for microscopic disease risk or the clinical target volume (CTV) margin. Although in many SBRT series, no extra margin for CTV has been used, in some situations there may be a risk of microscopic disease in adjacent normal tissues, especially for larger tumors. For most clinical situations, there is little clinical radiographic-pathologic data to provide guidance regarding the most appropriate CTV margins required. This issue is complicated in SBRT series, as there is always a dose gradient from the prescription dose to a “microscopic dose,” treating some volume of normal tissue around the gross tumor volume (GTV) to a microscopic dose. It is possible that the extent of subclinical microscopic dose around the GTV may be larger than the steep dose gradient, especially as the tumor size increases. If a CTV margin is used, the dose required to go to the CTV should be reduced compared with that to the GTV, given the much lower burden of disease within the CTV. At Princess Margaret Hospital in Toronto, for our present SBRT liver cancer studies, a 5-mm margin around the GTV within the liver is used to deﬁne the CTV. The planned minimum dose to the CTV PTV is 27 Gy in six fractions, while the dose to the GTV PTV may be as high as 60 Gy in six fractions at the periphery. Careful radiologic-pathologic studies accounting for organ deformation and shrinkage are required to determine whether the use of a CTV margin is required or not.

Finally, appropriate PTV margins must be used to ensure that the actual planned doses are delivered to the tumor. The PTV margins must consider setup uncertainty and internal organ motion. Individual institution setup uncertainty data should be used if available. Individual patient internal organ motion (e.g., breathing motion) may be used in this information is known. The use of a population PTV margin to ensure coverage of 95% of the population 95% of the time may require very large PTV margins to be used, especially in the situation of SBS or SBRT delivered in few fractions. Of note, the classic PTV margin papers on PTV margin determination, such as that published by Van Herk et al. [45], do not apply to very tightly conformal plans delivered in a few fractions, such as those used in most SBRT cases. Thus, the margin recipes that are used for conventional radiation planning may be inappropriate for SBRT plans. Modeling has demonstrated that when PTV margins are too small, higher doses must be delivered for the same tumor control probability [46].

Planning Often, many beams of low weight are used to develop a highly conformal SBRT dose distribution. Sometimes, non-coplanar beams or arcs are used if required to reduce the dose to normal tissues. For example, 8 to 12 beams may be used for a typical lung SBRT plan (Fig. 62-3). One strategy to obtain highly steep dose gradients at the edge of the PTV is to close the aperture of the beams to coincide or be within the PTV outline, not leaving a gap for penumbra as usually done for conformal radiation therapy. When enough beams are summed together, the PTV may be covered by a lower isodose such as 60% that is often at the steepest part of the dose gradient. Resultant high doses/hot spots occur in the center of the PTV, perhaps giving the highest dose to the center hypoxic volume (although the potential beneﬁt of this is unproven) (Fig. 62-4). Of note, the use of many

FIGURE 62-3. SBRT beam arrangement plan for early lung cancer. Patient was treated with 60 Gy delivered in three fractions.

62.

stereotactic body radiation therapy

beams for SBRT is not mandatory. For a peripheral liver tumor, three to ﬁve beams may be used to reduce the overall radiation path length through the liver. The use of segments within the beams can adjust the dose distribution to ensure that the hot spot is within the GTV while reducing the overall integral dose to the liver and other normal tissues (Fig. 62-5). As the adverse effects of dosimetric errors are most pronounced with SBRT, appropriate methods to consider corrections for heterogeneities should also be used. Typical prescription doses for SBRT range from 5 Gy in 10 fractions to 20 Gy in 3 fractions. One to ﬁve fractions are most often used, with a dose per fraction usually greater than 6 Gy. The common feature to most of the SBRT fractionation schemes is that they are biologically potent. Multiple-fraction regimens have some radiobiologic advantages to single-fraction SBRT. Clinical data is not available to provide guidance for the most appropriate fractionation for each clinical scenario. However, when the PTV is in very close proximity to normal tissues that function serially, it is reasonable to consider prolonging fractionation to minimize the risk of toxicity to the serial function normal tissue. It is a challenge to relate the prescribed dose to tumor control probability, as inhomogeneous doses are generally used and various delivery and veriﬁcation techniques are utilized. For a moving target not treated with image guidance, the actual delivered minimal dose to the tumor may be lower than the prescribed dose. Furthermore, there is heterogeneity in patterns of prescribing dose. One method to account for inhomogeneous dose distributions is to use equivalent uniform dose (EUD) to tumor for reporting [47]. Of course, accounting for individual patient motion in the dose distribution or eliminating motion would help ensure that the reported doses better reﬂect delivered doses as well. Unfortunately, it is not possible to account for motion in current commercially available planning systems.

FIGURE 62-4. SBRT plan for early lung cancer. Patient was treated with 60 Gy delivered in three fractions.

619

FIGURE 62-5. SBRT plan for an isolated liver metastases. Patient was treated with 54 Gy in six fractions.

Image Guidance Overview Image guidance at the time of treatment can improve setup accuracy, reduce PTV margins and the volume of normal tissue irradiated, facilitating safe dose escalation. Traditionally, surrogates for the target have been used in guiding the placement of treatment ﬁelds. For example, skin marks for patient position are routinely used for initial patient setup in practice. For some clinical situations, the use of skin marks to align the patient can be done with high precision. However, in most body tumors, the internal structures cannot be accurately localized with the use of skin marks. The use of boney anatomy with electronic portal imaging is another standard practice in radiation therapy. However, for many clinical situations, the boney anatomy is not well correlated with the internal tumor position. Options for locating internal anatomy include the use of implanted radiopaque ﬁducial markers as surrogates for the target, tissues adjacent to the tumor, or the tumor itself. Fiducial markers may also be used to measure organ motion and or track/gate the beam. Many of the published reports of SBRT have been on patients treated with conventional linear accelerators. However, more specialized treatment units are now available with the potential to allow soft tissue image guidance and reduced PTV margins. Examples of such systems include a lightweight and robotic linear accelerator (CyberKnife; Accuray, Sunnyvale, CA) and modiﬁed linear accelerators to allow image guidance including accelerators such as Novalis (BrianLAB, Inc., Westchester, IL), Synergy (Elekta Oncology, Stockholm, Sweden), Trilogy (Varian Medical Systems, Palo Alto, CA), Artiste (Siemens, Concord, CA), and Tomotherapy (Madison, WI). Proton and heavy particles can also be used to produce very

620

l.a. dawson

conformal dose distributions, with steep gradients and less integral dose, reducing the low doses delivered to adjacent normal tissues. Special quality assurance procedures need to be conducted prior to SBRT as tolerance for error is reduced with such high-precision RT plans. For SBRT, it has been recommended that radiation delivery equipment should have mechanical tolerances within ± 2 mm [2]. Quality assurance measures need to be speciﬁed for all aspects of SBRT planning and delivery, including image acquisition and correction of potential systematic errors based on the virtual patient model upon which the plan is designed. Equipment-speciﬁc quality assurance procedures need to also be completed. With the availability of volumetric imaging at treatment, there is a need for image registration methods that can use that information. Often for SBRT, geometric distortions between imaging data sets of the same organ exist, and deformable image registration methods are required to register such imaging data sets accurately [48].

Image-Guidance Strategies The two primary correction strategies that may be used to reduce setup error are an on-line approach and an off-line approach. The on-line approach refers to the use of daily imaging prior to every fraction with correction for offsets in position greater than a predeﬁned threshold prior to every radiation fraction. An off-line approach refers to the collection of imaging data with high frequency at the beginning of therapy (e.g., ﬁrst 5 fractions), followed by an ofﬂine analysis to determine the patients systematic (mean offset) and random (standard deviation, σ) setup errors. A correction in position is then made to consider the systematic error, with possible replanning to individualize the PTV margins based on the patients’ random setup error as sampled at the beginning of therapy. On-line correction strategies reduce both systematic and random setup errors, with a greater reduction in error compared with the off-line approach, at the expense of more time and cost. On-line correction strategies are most appropriate for hypofractionated SBRT, as there are generally few fractions to collect setup data and there is rationale for maximizing improvements in setup error. Imaging at the time of treatment can be used for localization for guidance, veriﬁcation, and also as a quality assurance tool. Veriﬁcation that the appropriate dose is actually delivered is also important for SBRT. Soft tissue imaging at the time of treatment can allow a measurement of the impact of geometric uncertainties (such as organ deformation) at the time of treatment. When image guidance and repositioning are used, imaging after repositioning should be used to ensure the positioning moves were made in the correct direction. When repositioning moves are required due to changes in internal organ position, replanning is not routinely conducted. However, when substantial changes in organ position and or tumor size or breathing pattern occur, the dose delivered may be altered due to changes in radiation path lengths and position of organs with difference heterogeneities. The magnitude of dosimetric differences due to such changes should be studied to better deﬁne in which situations replanning may be of beneﬁt.

Two-Dimensional Image Guidance Equipment Orthogonal Imaging Orthogonal megavoltage (MV) portal ﬁlms and more recently images from electronic portal imaging devices (EPIDs) have traditionally been used for image guidance and may be appropriate for targets adherent to the bones. MV images of large ﬁelds of view can be obtained with 2 to 8 monitor unit (MU) per image. These images not only can guide therapy but also can verify the shape and orientations of the treatment ﬁelds. If radiopaque ﬁducial markers are inserted in or near the tumor, the ﬁducial markers themselves may be used for guidance. Other alternatives for guidance include using surrogates that are in close proximity to the tumor, for example the diaphragm as a surrogate for liver tumors [49, 50]. An example of an anterior-posterior (AP) MV image for liver cancer guidance, in which the diaphragm is used for cranial-caudal positioning, is shown in Figure 62-6. Due to the low contrast of MV radiographs and the doses delivered with repeat MV imaging, orthogonal kV radiographs and kilovoltage (kV) ﬂuoroscopy have also been used for image guidance of tumors and/or ﬁducial markers, either immediately prior to each radiation fraction [49, 50] or throughout radiation delivery [43]. kV X-ray tubes may be ceiling or wall mounted or attached to the linear accelerator. With both MV and kV orthogonal imaging, alignment tools registering the images to digitally reconstructed radiographs (DRRs) can improve the accuracy and efﬁciency of image matching to determine the offsets in position. Such alignments tools include template matching tools based on therapists’ visualization of anatomy and/or automated image registration of the region of interest (e.g., mutual information). Decision rules including tolerance levels for repositioning must be integrated with the overall system.

Real-Time Tumor Tracking Real-time tumor tracking while the radiation beam is on is another approach to reduce adverse effects of organ motion. An elegant highly integrated tracking system consisting of four ceiling-mounted ﬂuoroscopic X-ray tubes and four ﬂoormounted ﬂat panel imagers in the treatment room allowing visualization of radiopaque markers in tumors was ﬁrst described

FIGURE 62-6. MV AP image and kV cone beam CT coronal reconstruction, as example of image-guidance tools for liver cancer SBRT.

62.

stereotactic body radiation therapy

by Shirato et al. [43]. This system has a temporal resolution of 30 frames per second and a precision of 1.5 mm. The linear accelerator is triggered to irradiate only when the ﬁducial marker is located within a predeﬁned volume. This system has been used for image-guided radiation therapy of lung, liver, and paraspinal malignancies. As an alternative to turning the radiation beam off when the tumor moves outside the treatment region, multileaf collimators, the couch position, or the entire accelerator on a robotic arm may move with the tumor to ensure adequate tumor coverage (e.g., CyberKnife image-guided radiosurgery system). The latter lightweight (330 lb) linear accelerator mounted on a robotic arm uses 6 MV, 5- to 60-mm collimators, and a dose rate of 300 to 400 MU per minute, combined with dual orthogonal ﬂuoroscopy tubes to track radiopaque markers in or near the tumor at a preset frequent. When the beam is on, infrared external surrogates are continuously monitored, while the internal anatomy is monitored every few seconds with kV imaging. The external surrogates are used for determining the breathing model, and the model is updated based on the X-ray data obtained every few seconds. The robotic linear accelerator responds to motion by moving to an appropriate position, within a range of ±10 mm x, y, z and ±1° pitch and roll, ±3° yaw. This system was found to have an 0.3-mm accuracy when tested in phantom studies. Disadvantages of this system include the need for ﬁducial markers, long potential delivery times (up to 90 minutes), lack of suitability for large tumors with motion more than 10 mm, and highly inhomogeneous dose distributions. Another system (Novalis; BrainLAB, Heimstetten, Germany) also acquires kV orthogonal images and matches the images to DRRs obtained from the planning CT. The imaging axes are not coincident with isocenter, and a translation of patient position is required between imaging and treatment. Accuracy of this system has been reported to be within 3% and 3-mm distance to agreement. Recently, wireless transponders and infrared cameras to track tumors have also been proposed as an imageless localization system (Calypso Medical Technologies, Seattle, WA).

621

error were observed using ultrasound guidance, with a residual mean three-dimensional residual setup error vector of 4.6 mm (±3.4 mm) after image guidance [52]. Advantages of ultrasound are that it is widely available and relatively inexpensive. Disadvantages of ultrasound image guidance are that the accuracy of the system for image guidance requires specialized training and is user dependent.

In-Room CT The placement of a diagnostic CT scanner in the treatment room with a known geometric relationship to the linear accelerator is one approach for volumetric imaging immediately prior to treatment with the patient in their immobilization device. Uematsu et al. have used this approach to treat liver and lung cancers with SBRT . Uematsu et al. reported that with repeat CT for frameless head and neck cancer radiotherapy, there was a geometric vector error ranging from 0.1 to 0.9 mm [53]. Multiple manufacturers have developed products of this type, including Siemen’s Primatrom, Mitsubishi’s accelerator in combination with a General Electric CT scanner, and Varian’s ExaCT targeting system [54–56]. All systems place the CT scanner in close proximity to the linear accelerator, allowing the couch to be moved from the imaging position to the treatment position. The CT scanner gantry is often translated during acquisition to minimize couch motion. Accuracy has been reported to be under 0.5 mm [55], and it has been reported to be improved with ﬁducial markers from 0.7 mm to 0.4 mm [54]. Advantages of in-room CT include that state of the art, diagnostic-quality CT can be used for optimal image quality and robustness. Disadvantages of this system are that the imaging and treatment isocenters are not coincident. Accuracy of motion from the CT scanner gantry, the accelerator couch, and the coincidence of the CT and linear accelerator isocenters needs to be veriﬁed. Limiting patient movement between imaging and treatment (e.g., couch retraction <1 m) should improve setup accuracy, however, organ motion between imaging and delivery may occur.

Three-Dimensional Volumetric Image Guidance

kV Cone Beam CT

Technological advances allowing volumetric imaging allow image guidance immediately prior to treatment using the tumor or a soft tissue organ in close proximity to the tumor for guidance, rather than the boney anatomy. Advantages of volumetric imaging systems includes that adjacent normal organs can also be visualized for more accurate avoidance of critical structures. Some of these volumetric imaging techniques can also measure tumor motion due to breathing.

Jaffray et al. ﬁrst described the concept of cone beam CT for image-guided radiation therapy in 1997 [57]. Cone beam CT refers to combined kV X-ray imaging and MV radiation delivery in one integrated gantry-mounted system. Advancements in large-area ﬂat panel detector technology facilitated volumetric imaging to be acquired in a single rotation of the linear accelerator gantry. Planar kV images projections are obtained as the gantry rotates about the patient on the linear accelerator table, over 30 seconds to 4 minutes. Cone beam CT threedimensional volume reconstruction images may then be obtained for position veriﬁcation or for image guidance (Fig. 62-2). Geometric calibration methodologies for cone beam CT systems [58] and quality assurance recommendations [59] have recently been proposed. Doses delivered to obtain cone beam CT scans typically range from 0.5 to 2 Gy, which is substantially less than the dose from MV orthogonal images. Three vendors have developed kV cone beam CT systems: Elekta Synergy (Fig. 62-7), Varian Trilogy, and Siemens Artiste. Examples of liver and lung cancer

Ultrasound Buatti et al. have used a three-dimensional ultrasound reconstruction combined with optical tracking for pelvic and retroperitoneal tumor localization for SBRT. This device, under the trade name of Radio Camera (Zed), uses a position sensor unit to track optically the position of infrared markers arranged in an array to form a rigid body [51]. Ultrasound has also been used for image guidance of liver and pancreas cancers. Signiﬁcant reductions in residual setup

622

l.a. dawson

images acquired with the Elekta cone beam CT system are shown in Figure 62-8. In addition to providing volumetric imaging for veriﬁcation and guidance, these systems have the ability to be used for real-time kV tracking; the latter application has not been used clinically. Similar to in-room CT, artifacts may occur with kV cone beam CT reconstructions due to high Z structures such as surgical clips, hip prostheses, and dental ﬁllings. Methods to reduce these artifacts have been developed.

MV Cone Beam CT CT imaging using MV beams has also been explored for more than 20 years [60–63] and has also been made possible with advances in portal imaging technology. Advantages of MV cone beam CT are that the treatment MV beam is used to obtain the imaging, requiring less modiﬁcation to the linear accelerator; the electron density estimates for treatment planning are accurate; and there is no high Z artifact that is associated with kV imaging. MV cone beam CT has been used to aid in lung cancer SBRT, as described by Nakagawa et al. in 2000. MVCT aided lung SBRT was used for treatment of 22 lung tumors [64]. Pouliot et al. recently reported the feasibility of acquiring low-exposure megavoltage cone beam CT. Phantom and pig cadaver head and neck images were acquired using a linear accelerator dose rate of 0.01 to 0.08 MU per image, for a set of 90 to 180 projections, acquired in 1° to 2° increments over 45 to 60 seconds. MV cone beam CT scans were obtained with doses of approximately 5 cGy. Despite the low efﬁciency of this system, visibility of high-contrast structures, such as air and bone, was reasonable [65].

Planning CT

Cone Beam CT #1

Cone Beam CT #2

Cone Beam CT #3

FIGURE 62-8. Planning CT and kV cone beam CT from each lung cancer SBRT fraction. GTV from the planning CT scan is shown in orange and PTV in green.

MV Tomotherapy MV tomotherapy combines tomographic scanning capabilities, from a conventional CT detector, with a linear accelerator mounted on a rotating gantry. Simpson described the initial development of an MVCT scanner for radiation therapy in 1982 [60]. More recently, the TomoTherapy treatment platform has become available for image guidance and veriﬁcation. The MV treatment beam is used to obtain imaging, with a lower energy, 3.5 MV instead of 6 MV. Computer-controlled multileaf collimators, also on the rotating gantry, have two sets of leaves that open and close to modulate the radiation beam while the couch advances the patient through the gantry bore, for helical intensity modulated radiation therapy (IMRT). Similar to MV cone beam CT, there are no high Z artifacts with MV tomotherapy.

Clinical Outcomes Overview

FIGURE 62-7. kV cone beam CT unit (Elekta Oncology, Stockholm, Sweden).

A difﬁculty in comparing clinical results after SBRT is that patients may vary tremendously in their performance status, comorbidities, their prognostic factors, treatment delivered, and treatment goals. There is selection bias in all singleinstitution series and even multi-institutional studies. Thus, detailed comparisons cannot be made with reliability. Local control is an important end point as it gives information regarding the efﬁcacy of SBRT, a potent local therapy designed to eradicate or ablate disease. Local control is not so easy to deﬁne. Occasionally, there may be complete resolution

62.

stereotactic body radiation therapy

of disease for small tumors (e.g., in the liver). However, most often there is a residual scar that may remain forever. In the lung, changes near the irradiated volume are common, and it is challenging to know when the disease recurs. The test of time and prolonged progression-free survival are ultimately helpful in this regard. Even PET may have false positives after SBRT. Overall survival, progression-free survival, and distant metastases should also be reported as these reﬂect SBRT efﬁcacy and appropriate patient selection for SBRT. For clinical situations in which the goal of SBRT is symptom relief (e.g., paraspinal recurrence), pain relief and quality of life may be the most relevant end points. However, local control and survival should also be reported, as they again reﬂect SBRT efﬁcacy and appropriateness of patient selection. Ideally, the optimal method to determine the efﬁcacy of SBRT is to compare SBRT to alternative local therapies or conventional radiation therapy in randomized controlled trials. This would control for selection bias and inconsistencies with reporting local control. To date, SBRT has not been tested in randomized trials.

Lung Cancer (Case Study 62-1) SBRT has perhaps been most widely studied in primary and metastatic lung cancer [4, 7, 10, 56, 64, 66–72]. A summary of

Case Study 62-1 Lung Cancer SBRT (Courtesy of Dr. Andrea Bezjak, PMH, Toronto, Canada) Patient History Mrs. P is a 77-year-old woman with a recently diagnosed T1N0 squamous cell lung cancer of the right upper lobe. She has signiﬁcant chronic obstructive lung disease, which has resulted in pulmonary hypertension and a degree of cardiac insufﬁciency. She was deemed to be at a higher risk for surgical complications if she were to undergo resection of her lung cancer. Staging, which included a PET scan, revealed no evidence of metastases. On exam, Mrs. P was in no respiratory distress at rest but mildly dyspneic on talking. She appeared her stated age. Her vital signs were stable. There was no palpable lymphadenopathy. Air entry was mildly reduced bilaterally. Cardiac exam was unremarkable; there was no hepatosplenomegaly, and neurologic exam was normal.

Treatment Mrs. P was approached for participation in the RTOG phase II study of SBRT for patients with early-stage resectable peripheral tumors who are deemed inoperable for medical reasons. She was interested in participating, especially in view of fewer visits for radiation therapy compared with the alternative, more standard protracted fractionation schedule. Informed consent was obtained, and she was registered in the study. She was planned in the supine position, using a vac-lock immobilization board and stereotactic frame. Fluoroscopy was used to assess tumor motion, which was less than 5 mm; thus, abdominal compression or breathing control was not needed. In addition to a standard helical planning CT, Mrs. P

623

clinical outcomes after SBRT for primary and metastatic lung cancer is shown in Tables 62-1 and 62-2. Initial reports from the Karolinska Institute included patients who were treated for lung metastases. Since then, patients with primary non–small cell lung cancer have also been treated. Eight to 20 Gy per fraction delivered in two to ﬁve fractions were the common SBRT fractionations used. The overall 5-year survival of 35 patients with stage I non–small cell lung cancer was 35%. Only two stage II patients were treated. The 2-year survival of 25 patients with stage III disease was 40%, 11 of whom developed local-regional disease as the site of ﬁrst recurrence [73]. In 2001, Uematsu et al. reported on 50 patients with T1/2N0 disease who were medically inoperable (21) or refused surgery (29), treated with 60 Gy in 5- to 10-fraction SBRT. Some patients received conventional radiation therapy prior to SBRT in an attempt to shrink the tumor. In-room CT was used for image guidance. The 3-year overall survival was 66% in these patients [7]. Onishi et al. summarized the results of 241 patients treated at 13 institutions in Japan using a variety of techniques, all with non-coplanar arcs of multiple static beams and interventions to reduce respiratory motion [74]. Fractionation schemes were heterogeneous, but included SBRT, ranging from 18 to 75 Gy in 1 to 22 fractions. With a median follow-up of 18

underwent a four-dimensional CT to assess three-dimensional tumor motion during free breathing. The four-dimensional CT was not used for treatment planning (as it was not permitted on the RTOG protocol) but was used to verify that the margins applied included the tumor in all phases of respiration. The GTV consisted of the tumor as visualized on CT. The PTV around the GTV consisted of three-dimensional expansions of 5 mm in all directions other than the cranial caudal direction in which expansions were 10 mm. Normal tissues that were contoured included both lungs, heart, trachea and proximal bronchi, distal bronchial tree, esophagus, spinal canal, and brachial plexus. The protocol includes strict dose criteria to these normal tissues at risk that cannot be exceeded. The prescription isodose covering the PTV was 60 Gy in three fractions. Seven non-coplanar beams were used to conform the dose to the PTV, with careful attention paid to the dose spillage region that should not be larger than 2 cm around the PTV (Figs. 62-9 and 62-10). The three fractions of RT were delivered 3 to 5 days apart, with pretreatment steroids as per the RTOG protocol. Prior to each treatment, kV cone beam CT was used for image guidance. Mrs. P was repositioned for every offset of 3 mm or more. Veriﬁcation repeat imaging was conducted with MV portal images after each repositioning. The average time of veriﬁcation imaging and treatment was 45 minutes. Mrs. P tolerated treatment well with only mild fatigue and an episode of chest discomfort not related to treatment; she had no skin erythema and no worsening of her respiratory symptoms. A follow-up CT scan showed shrinkage of the tumor at 3 months; a residual nodule remains and has not changed in follow-up to date. She remains well clinically and free of disease elsewhere.

624

l.a. dawson

FIGURE 62-9. Baseline CT of isolated liver metastases and follow-up imaging at 3 months and 18 months revealing no evidence of progressive disease.

months, pulmonary complications grade 2 or higher occurred in 2.1% of patients, with local recurrence in only 10.4% of patients. A higher local recurrence was observed in patients treated with doses less than an equivalent dose of 100 Gy in 2 Gy per fraction. The 3-year survival was 56%. Nagata et al. found that four fractions of 12 Gy was associated with 100% local control in 16 patients with T1N0 non–small cell lung cancer, whereas a local recurrence was observed after 10 Gy × 4 [67]. Wulf et al. reported on 61 patients with primary (20) and metastatic (41) lung cancer treated with 26 to 37.5 Gy in one to three fractions [66]. Two patients (3%) developed transient pneumonitis, with no other serious toxicity. Actuarial local control was 92% and 80% for primary and metastatic lung cancer, respectively. One-year and 2-year survival rates were

FIGURE 62-10. SBRT dose distribution for lung cancer.

52% and 32% for primary lung cancer and 83% and 33% for metastases. Twelve-month progression-free survival was 60% and 35% for primary and metastatic lung cancer, respectively [66].

TABLE 62-1. Selected results of primary non–small cell lung cancer treated with SBRT. Local control

Survival

Ref.

Year

No. patients

Total dose (Gy)

No. of fractions

Percent (%)

Time (months)

Percent (%)

Time (months)

3 57 53 51 54 6 56 52 55 50

2001 2002 2002 2002 2003 2003 2003 2003 2003 2004

43 22 5 16 241 37 15 10 9 20

50 48–60 20–30 15–66 18–75 24–60 15 19–26 30–40 26–27.5

5 8 1 1–5 1–22 3 1 1 3–4 1–3

94 94 100 94 90 83 74 80 90 92

20 2–6 20 16 18 15 7 NR 18 7

66 NR NR 79 56 54 NR 64 100 32

36 NR NR 24 36 15 NR 24 18 24

Time (months)

Percent (%)

Time (months)

20 12 16 7 18 9

66 NR NR NR NR 33

36 NR NR NR NR 24

NR, not reported.

TABLE 62-2. Selected results of lung metastases treated with SBRT. Local control Ref.

Year

No. targets

58

1995

14

3 53 51 56 55 50

2001 2002 2002 2003 2003 2004

23 18 9 8 19 51

NR, not reported.

Total dose (Gy)

No. of fractions

Percent (%)

20–45

1–3

75–95

50 20–30 15–66 15 30–40 26–27.5

5 1 1–5 1 3–4 1–3

100 78 66 74 88 80

Survival

62.

625

stereotactic body radiation therapy

Hara et al. reported on single-fraction SRS for primary and metastatic lung cancers less than 4 cm in maximum diameter. Local recurrences were more common if the dose was less than 30 Gy in one fraction versus 30 Gy or higher (3 of 10 recurrences versus 1 of 13 recurrences) [69]. More recently, Timmerman et al. completed a threefraction dose escalation SBRT study for non–small cell lung cancer [10]. The starting fraction size was 8 Gy, with the highest fraction size investigated of 20 Gy, delivered in three fractions for a total dose of 24 to 60 Gy. Thirty-seven patients with medically inoperable stage I non–small cell lung cancer were treated on this study. No maximally tolerated dose was found, and 20 Gy in three fractions was concluded to be tolerable in well-selected patients. One patient developed grade 3 pneumonitis and another developed grade 3 hypoxia. One asymptomatic cardiac effusion was also observed. Both these patients were treated with less than 18 Gy per fraction. Radiographic responses were seen in 87% of patients. Based on these results, a RTOG study of SBRT using three fractions of 20 Gy has been

Case Study 62-2 Patient History Mr. D is a 72-year-old man who developed an isolated liver metastases from rectal cancer treated with low anterior resection and postoperative radiation 3 years earlier. Initial staging revealed a T3N1 adenocarcinoma of the rectum, with no evidence of liver metastases on CT. Mr. D received 6 months of adjuvant 5-ﬂuorouracil containing chemotherapy. On a routine follow-up, he was found to have an elevated serum carcinoembryonic antigen (CEA), suggesting recurrent rectal cancer. An abdominal CT revealed a new 4-cm enhancing tumor in the central liver, consistent with metastases. Further staging investigations revealed no other metastases. Mr. D was seen by surgical and medical oncology; however, due to underlying severe cardiac and pulmonary comorbidities, surgery was not feasible. The risks of chemotherapy were estimated to be higher than potential beneﬁts, and chemotherapy was not offered. Mr. D had a history of stable angina and a grade 3 ventricle, controlled on multiple medications; he also had severe chronic obstructive pulmonary disease, requiring regular treatment with oral steroids and hospitalizations once or twice a year. Despite these comorbidities, cardiac and lung function appeared to be stable over the past 2 years. Mr. D was interested in curative treatment for his isolated liver metastases from rectal cancer. On exam, Mr. D was a mildly obese, barrel-chested man who appeared his stated age. His vital signs were stable. He had occasional wheezes on auscultation. There was no hepatosplenomegaly or stigmata of chronic liver disease.

Treatment Mr. D was treated on an in-house protocol of escalated dose SBRT for liver metastases. He was positioned supine in a vac-lock immobilization board. Simulation consisted of kV ﬂuoroscopy to assess diaphragm motion during free breathing and the need for immobilization with breath hold. The range of caudal cranial free breathing diaphragm motion was 5 mm,

TABLE 62-3. RTOG SBRT lung study normal tissue permitted tolerances, in three fractions. Spinal cord Esophagus Brachial plexus Heart Trachea Lung

Maximum dose Maximum dose Maximum dose Maximum dose Maximum dose V13Gy Mean dose

19 Gy in 6 Gy per fraction 27 Gy in 9 Gy per fraction 24 Gy in 8 Gy per fraction 30 Gy in 10 Gy per fraction 30 Gy in 10 Gy per fraction <10% <7 Gy

initiated for patients with peripheral lung cancers. Normal tissue constraints used in this study are shown in Table 62-3.

Liver Cancer (Case Study 62-2) The liver cancer SBRT experience is summarized in Table 62-4. Blomgren et al. from the Karolinska Institute in Sweden ﬁrst reported outcomes after 20 to 45 Gy in 1 to 4 fractions to treat

and thus Mr. D was recommended to be imaged and treated without breath hold. He underwent a contrast planning CT scan with a 3-mm slice thickness. The GTV consisted of the enhancing liver metastases as visualized on venous phase CT. A CTV was included and consisted of an 8-mm three-dimensional expansion within the liver parenchyma. PTVs around the GTV and CTV consisted of three-dimensional expansions of 5 mm in all directions other than the cranial caudal direction, which was 7 mm. The prescription isodose covering the GTV PTV was 45.6 Gy, delivered in six fractions of 7.6 Gy per fraction. Normal tissue constraints used in this in-house six-fraction phase I study are shown in Table 62-5. The prescription isodose covering the CTV PTV was 27 Gy in six fractions. The mean dose to the liver minus the GTV was 12.7 Gy in six fractions, with an associated risk of radiation-induced liver disease (RILD) of less than 1%. The plan is shown in Figure 62-11. Treatment was delivered three times a week over 2 weeks. Prior to each treatment, orthogonal MV imaging was used to localize the diaphragm for caudal cranial positioning and the vertebral bodies for anterior-posterior and medial lateral positioning. Mr. D was repositioned for every offset of 3 mm or more. Veriﬁcation repeat imaging was conducted after each repositioning. The average time of veriﬁcation imaging and treatment was 25 minutes. Mr. D tolerated treatment well with no grade 3 toxicity. His only reported toxicity was mild transient fatigue.

Outcome At 3 months after SBRT for the liver metastases, Mr. D had a near-complete response to his therapy. At 18 months after SBRT, he continues to have no evidence of progressive disease within the liver and outside of the liver. His follow-up liver CT scans reveal contraction within the portion of liver treated to high dose and hypertrophy in the lateral lobes of the liver (see Fig. 62-9).

626

l.a. dawson

TABLE 62-4. Results of liver cancer treated with SBRT. Local control Ref.

Year

No. patients

58 58 62 74 9 8

1995 1995 2001 2001 2005 2005

23 23 23 37 18 51

Tumor type Metastases

HCC

Cholangiocarcinoma

17 23 56 16 22

11

1

4 2 8

21

Survival

Total dose (Gy)

No. of fractions

Percent (%)

Time (months)

Percent (%)

Time (months)

7.7–45 15–45 30 14–26 36–60 24–60

1–4 1–3 3 1 3 6

29 43 61 67 NR NR

6–23 1.5–38 24 18 NR NR

NR NR 71 72 NR NR

6–26 2 days to 39 12 12 NR NR

NR, not reported; HCC, hepatocellular carcinoma.

TABLE 62-5. SBRT liver study normal tissue permitted tolerances, in six fractions [12]. Liver

Iso-NTCP model, based on effective liver volume irradiated (For <10% risk, suggest mean liver dose <20 Gy)

Esophagus

Maximum dose <30 Gy

Stomach

Maximum dose <30 Gy

Duodenum

Maximum dose <30 Gy

Large bowel

Maximum dose <30 Gy

Kidney

D67% <15 Gy

Cord

Maximum dose <25 Gy

with one patient alive with no evidence of disease at 101 months. Herfarth et al. from Heidelberg University used single-fraction SRS to treat 60 liver tumors in 37 patients (4 primary and 56 metastases). The median tumor size was 10 cm3 (1 to 132 cm3), with an upper maximum diameter of 6 cm. The single dose was safely escalated from 14 Gy to 26 Gy, and a maximum tolerated dose was not found [5]. With a median follow-up of 5.7 months, there was no major toxicity. Ninety-eight percent of all tumors were locally controlled after 6 weeks, with complete and partial responses seen in 4 and 28 patients, respectively. The actuarial local control was 81% 18 months after therapy. Normal tissue constraints included 30% and 50% of the liver less than 12 Gy and 7 Gy, respectively, and maximum dose to esophagus and stomach was 14 Gy and 12 Gy, respectively [77]. Wulf et al. from the University of Wurzburg used SBRT delivered in three fractions (30 Gy total) to treat 23 patients with solitary liver tumors. No grade 3 toxicity was observed. Crude local control was 76% and 61% at 1 and 2 years. Overall survival was 71% and 41% at 1 and 2 years [6]. Schefter et al. recently published a multi-institutional trial of three-fraction SBRT for liver cancer [13]. Patients with one to three liver metastases with a maximum tumor diameter of 6 cm were eligible. Twenty-ﬁve tumors in 18 patients with a variety of diagnoses (16 metastases, 2 hepatocellular carcinoma) were treated with 36 Gy to 60 Gy, in three fractions delivered within 14 days. Dose escalation by 6 Gy occurred only after a minimum of three patients were followed without toxicity 30 days after SBRT. The dose was generally prescribed to the 80% to 90% isodose from highly conformal plans

Norm. Volume

29 liver tumors in 23 patients (8 patients with hepatocellular carcinoma, 1 with intrahepatic cholangiocarcinoma, and 14 with metastases) [4]. Radiographic responses were observed in 29% and 43% of evaluable patients with primary and metastatic liver cancer, respectively. Two of the patients with hepatocellular carcinoma were alive with no progressive disease 12 and 39 months after treatment. Complete responses occurred quickly for small tumors, but the time to maximal response was prolonged for larger tumors. For example, maximal response was seen 16 months after SBRT for a liver metastases 13 cm in maximum dimension. Using this approach, the mean liver dose generally ranged from 1 to 8 Gy, with a maximum of 18 Gy delivered in three fractions for one patient. Several serious toxicities were seen after hepatocellular carcinoma SBRT, including one sudden death 2 days after 30 Gy in one fraction to a large tumor, two cases of radiation-induced liver disease (RILD) 1.5 and 2.5 months after SBRT (45 and 30 Gy in three fractions, respectively), and one subcapsular bleed 2 weeks after SBRT. Hemorrhagic gastritis was seen in one patient with liver metastases in which the gastric wall received 7 Gy × 2 [4]. The clinical outcomes after SBRT for liver cancers were updated in 1998 by Blomgren et al. [75]. With a median followup of 9.6 months after SBRT for 17 patients with 21 liver metastases, stable disease was seen in 10 tumors, partial response in 4, and complete response in 4. The crude local control rate was 95% with a mean survival of 17.8 months. Gunven et al. reported on the feasibility of using SBRT for recurrent liver metastases after hepatic resection for colorectal cancer metastases [76]. SBRT fractionations of 20 Gy × 2 or 15 Gy × 3 were used to treat four liver-conﬁned recurrences judged to be appropriate for SBRT (of 5 liver recurrences occurring in 18 consecutively treated patients). No serious toxicity was observed, and all tumors were locally controlled 13 to 101 months after surgery,

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

PTU

Bronchus Spinal Canal

0

Rt & Lt Lung

1000

2000

3000 4000

5000

6000

7000

Dose (cGy) FIGURE 62-11. Dose-volume histograms, demonstrating little dose delivered to the majority of the lung, bronchi, and cord. Dose to lung cancer PTV 60 Gy × 3.

62.

stereotactic body radiation therapy

obtained from dynamic arc therapy or non-coplanar beams. At least 700 cm3 of uninvolved liver had to receive less than 15 Gy in three fractions. Most patients were given prophylactic antiemetic therapy with or without one dose of dexamethasone. The tumor volume was 17.8 cm3 (3 to 98 cm3). The median percentage of PTV covered by the prescription dose was 98% (91% to 100%). No dose-limiting toxicity was seen. Twelve patients were alive with a median survival of 7.1 months after enrollment (3.8 to 12.3 months). Six patients died 3.1 to 18.9 months after enrollment, from progressive liver metastases (1), extrahepatic disease (4), and a preexisting medical condition (1). A tumor equivalent uniform dose (EUD) of more than 54 Gy in three fractions was associated with improved local control compared with an EUD of less than 54 Gy in three fractions [47]. At the Princess Margaret Hospital in Toronto, a phase I study of SBRT for patients with unresectable primary or metastatic liver cancer is ongoing [12]. This protocol uses an individualized treatment approach similar to that described by McGinn et al. from the University of Michigan, in which the prescribed tumor dose is dependent on volume of liver irradiated [78]. There is no upper limit on tumor size in this protocol. To date, 55 patients (23 with hepatocellular carcinoma, 9 with intrahepatic cholangiocarcinoma, and 22 with liver metastases) have been treated with doses ranging from 24 Gy to 54 Gy, delivered in six fractions over 2 weeks. No dose-limiting toxicity has been observed, although two patients with cirrhosis developed worsening of their liver function from Child score A to Child score B 3 months after treatment. Radiographic responses have been common (Fig. 62-12). The largest experience with radiation therapy for hepatocellular carcinoma is from Asia, where excellent long-term local control and survival has been reported in patients treated with a variety of radiation fractionation schemes. Wu et al. reported on 94 patients with hepatocellular carcinoma treated with TACE followed by SBRT [79]. Forty-eight to 60 Gy was delivered in 4 to 8 Gy per fraction over 17 to 26 days using conformal radiation therapy. The response rate was 91%, with excellent 1- and 3-year survival rates of 93% and 26%, respectively.

627

In summary, many series suggest that high-dose SBRT can be delivered safely to both liver metastases and hepatocellular carcinoma. European and Asian series demonstrate very encouraging response rates with some patients having longterm progression-free survival. Despite this experience, the dose-volume tolerances of critical normal tissues such as the liver are not well understood.

Paraspinal Malignancies Hamilton et al. reported on nine patients treated in the prone position to a median dose of 10 Gy in a single fraction, prescribed to the 80% isodose line. The cord doses ranged from 8 to 15 Gy. All patients had previously undergone radiotherapy. At a median follow-up of 6 months, all patients were controlled in irradiated sites [33]. Ryu et al. described outcomes after 11 Gy to 25 Gy in one to ﬁve fractions to treat spinal tumors using the CyberKnife system. No toxicity was seen at a minimum of 6-month followup in patients treated with a maximum dose to the cord of 8 Gy. All tumors had no progression at last follow-up [80]. One hundred twenty-ﬁve lesions in 115 patients were treated in a similar fashion with 12 to 20 Gy in one fraction prescribed to the 80% isodose line. Seventy-four of 79 patients presenting with pain had reported a reduction in pain after SBRT, with no toxicity reported [81]. More recently, Yamada et al. reported on the use of IMRT SBRT for patients with gross tumor adjacent to the spinal cord where greater than 54 Gy in conventional fractionation would have required irradiation of the spinal cord. Thirty-ﬁve patients were irradiated, including 14 with primary spinal tumors, 11 with recurrent sarcoma, and 3 chordoma patients. The previously irradiated patients had been treated with a median dose of 30 Gy in 10 fractions. In 23 patients, pain was the predominant presenting symptom. The median pretreatment pain score (6.6 of 10) was reduced to 1.1 of 10 after radiation therapy. The median dose delivered previously was 30 Gy in 10 fractions. The median reirradiation dose was 20 Gy in ﬁve fractions. The median dose delivered to primary tumors was 70 Gy in 2 Gy per fraction [14]. Chang et al. also have used SBRT to treat metastatic spinal tumors, using 30 Gy in ﬁve fractions, with a limit to the spinal cord of 10 Gy maximum dose. Twenty-seven Gy in nine fractions was used with a spinal cord limit of 9 Gy per fraction. Fourteen of 15 patients were free of local progression at last follow-up [15].

Pancreas Cancer

FIGURE 62-12. SBRT plan for patient with isolated liver metastases treated with 45.6 Gy in 7.6 Gy per fraction ×6 fractions.

Two prospective studies have recently been published reporting outcomes after SBRT for pancreatic cancer. The results of these studies differ. The ﬁrst study by Koong et al. from Stanford University reported outcomes of a phase I study of singlefraction SBRT in unresectable pancreatic cancer [82]. Fifteen patients were treated, three of whom had received prior chemotherapy and three of whom had had prior gastrojejunostomy. Three patients received 15 Gy; 5 patients received 20 Gy; and 7 patients received 25 Gy, prescribed to a peripheral covering isodose. All patients had three to ﬁve ﬁducial markers implanted in the tumor to facilitate image guidance during SBRT. Patients were treated with CyberKnife, and orthogonal

628

l.a. dawson

X-rays were used to ensure appropriate positioning of the ﬁducial markers, with the pancreas immobilized in voluntary breath hold. The treatment time ranged from 3 to 6 hours. A small volume of duodenum was treated to 22.5 Gy in some patients, without development of grade 3 or higher gastrointestinal toxicity. Local control was reported in 6 of 6 patients with evaluable disease treated with 25 Gy. Distant recurrences were common, and the median survival of all patients was 11 months. The second study from Denmark reported on 22 patients with unresectable pancreatic cancer who were treated with 15 Gy × 3 (with 95% prescribed to the CTV and 67% prescribed to the PTV) [83]. Patients were treated using linear accelerator–based SBRT, immobilized in stereotactic body frames using abdominal pressure to reduce organ motion due to breathing. Fifteen patients had repeat CT scans with less than 5 mm offsets in pancreas position reported. Portal imaging was done prior to each fraction, with repositioning if necessary. SBRT was delivered within 5 to 10 days, without concurrent chemotherapy. The toxicity was severe, and local control and survival were very poor. Only 9% of patients had a partial response. The median survival was 5.7 months, with a 1-year survival rate of 5%. Despite the use of prophylactic ondansetron and pantoprazole for 4 weeks, acute toxicity at 14 days after therapy was pronounced, with worsening of nausea and pain that resolved in 8 or 12 patients 3 months after SBRT. Four patients developed gastrointestinal toxicity consisting of gastritis, ulceration, and/ or perforation. Reasons for the difference in these studies are not clear. Both studies reported a high rate of distant recurrences, supporting the role of chemotherapy in this setting. The delivered doses were similar, with slightly higher biologic doses in the Danish study (15 Gy × 3 vs. 25 Gy). However, the Danish study observed very poor local control. The study from Stanford used real-time image guidance and breath hold to immobilize the pancreas, which may have reduced the chance of a marginal miss. The Danish study treated patients with abdominal compression and did not use real-time imaging to verify pancreas position during SBRT. It is possible that lower doses than intended may have been delivered to the tumor periphery due to organ motion from breathing. The differences in gastrointestinal toxicities should be regarded with some caution, as the total number of patients is small with a short combined followup time. Although it is possible that a small volume of duodenum may be able to receive 22.5 Gy safely as suggested by Koong et al., more patients need to be treated at this dose level to have more conﬁdence in the tolerance of the duodenum to single-fraction SRS. The Danish study highlights the caution that needs to be used when SBRT is used in tumors adjacent to serial functioning organs. Based on the results of these studies, pancreatic cancer SBRT in which PTV margins overlap with the luminal gastrointestinal organs should be done cautiously and only on protocol. Off protocol, combined chemotherapy and more conventionally fractionated radiation therapy should reduce the risk of gastrointestinal toxicity. Given the encouraging results reported by Koong et al., it is likely that they will obtain more mature toxicity and outcome data in a larger number of patients; such data will be a valuable contribution to the SBRT literature and may perhaps highlight how technical aspects of SBRT may be related to clinical outcomes.

Renal Cell Cancer Qian et al. observed 31% response rate after 48 Gy in ﬁve fractions for primary and metastatic renal cell carcinoma [84]. At the Karolinska Institute, 58 patients with 162 lesions have been treated with 30 to 40 Gy in two to four fractions, primarily for lung metastases from renal cell carcinoma. Lack of in-ﬁeld progression was observed in 90% of cases at 37 months median follow-up. The response rate was 55%, with 30% complete responses and 22% partial responses. SBRT was also used in patients who developed second renal cell carcinomas in their remaining kidney. The ability of SBRT to control such tumors in sensitive critical normal tissues is exciting, and this experience should help in determination of what the hypofractionated dose-volume tolerance of the kidney is [85].

Prognostic Factors Prognostic factors after SBRT are generally those that predict good outcome and survival to any therapy (i.e., validated prognostic factors). These include patient, tumor, and treatment factors, including performance stats and TNM stage. Sometimes organ function is related, for example in the treatment of hepatocellular carcinoma. Prognostic factors speciﬁc to SBRT will need to be deﬁned in future clinical trials. For liver metastases, prognostic factors are best described by Fong et al. for patients treated with surgical resection [86]. The risk factors for recurrences included positive margin, extrahepatic disease, node-positive disease, disease-free interval from primary to metastases <12 months and largest tumor >5 cm, and carcinoembryonic antigen (CEA) >200 ng/mL. No patient with a score of 5 was a long-term survivor. Otherwise, 5- and 10-year survival was 37% and 22%. Such prognostic factors likely would apply to liver metastases patients treated with SBRT. Similarly, prognostic factors for other tumors treated by other therapeutic modalities may be appropriate to apply to SBRT, at least until prognostic factors from SBRT series are established.

Complications Recognition Toxicities that may occur after fractionated radiation therapy may also occur after SBRT. There are also some toxicities that are unique to SBRT. Some of these toxicities are highlighted below.

Liver Radiation has historically played a minor role in the management of patients with unresectable liver cancer, primarily due to the low tolerance of the whole liver to radiation. There is a 5% risk of radiation-induced liver toxicity after uniform wholeliver radiation of 28 Gy and 32 Gy in 2 Gy per fraction for liver metastases and primary liver cancer, respectively [87]. The most common liver toxicity observed in North America is radiationinduced liver disease, or RILD, a clinical syndrome of anicteric hepatomegaly, ascites, and elevated liver enzymes (particularly serum alkaline phosphatase) occurring between 2 weeks to 3

62.

629

stereotactic body radiation therapy

months after external beam radiation. Treatment for RILD consists of supportive measures. Spironolactone diuretics and steroids are often used, although there is no evidence that they change the natural history of RILD. Most cases resolve with conservative treatment, but some cases lead to irreversible liver failure and occasionally death. RILD has been observed 1.5 and 2.5 months after SBRT (45 and 30 Gy in three fractions, respectively) [4]. Reactivation of viral hepatitis and precipitation of underlying liver disease can also occur after radiation therapy for hepatocellular carcinoma [88]. Radiographic changes after liver SBRT is well documented by Herfarth et al. [89]. They include a type 1 reaction (hypodensity in portal venous contrast phase, isodensity in late contrast phase), type 2 (hypodensity in portal venous phase and hyperdensity in late contrast phase); type 3 is isodensity and hyperdensity in postvenous contrast and hyperdensity in late venous phase. It is unknown whether these geographic changes are related to ﬁbrosis or clinically meaningful liver toxicity. However, most often, liver contracture occurs after the changes described above. Conversely, regeneration occurs in the portions of the liver spared from high-dose radiation. After highly potent SBRT, there is the potential for different hepatobiliary toxicities to occur, including biliary sclerosis and hepatic subcapsular injury. A subcapsular bleed was observed 2 weeks after SBRT in a patient with two anterior tumors, both treated with high-dose SBRT [4].

Luminal Gastrointestinal The esophagus, stomach, duodenum, and large bowel are at risk of injury from radiation if dose cannot be spared from these organs. Gastrointestinal bleeding is the most common luminal toxicity after high-dose irradiation, although small bowel obstruction, gastric outlet obstruction, and ﬁstula formation are also possible late sequelae. Esophageal ulcers and fatal bleeds have been observed 3 to 9 months after SBRT for lung cancer. Hemorrhagic gastritis has also been reported in a patient with liver metastases in which the gastric wall received 7 Gy × 2 [4].

Lung Pneumonitis is the dose-limiting complication after conventional radiation therapy for lung cancer. It is a clinical diagnosis of exclusion, consisting of cough, fever, pleuritic pain, dyspnea, with occasional respiratory failure and general inﬂammation. Characteristic radiographic ﬁndings are often present, most often in the distribution of the radiation ﬁelds. Steroids are often used to treat pneumonitis.

Pneumonitis is unusual in patients with small tumors generally selected for SBRT. However, other pulmonary toxicities that may be permanent may be seen after SBRT. Bronchial damage from SBRT may be severe and associated with total obliteration of the airway lumen and downstream atelectasis. This often appears as a wedge-like collapse on follow-up imaging. The volume of atelectasis is related to the proximity of the bronchus injured. Tracheal stricture with subsequent lung collapse or injury is another toxicity that is possible after highly potent SBRT to medial lung cancers.

Other Rib and skin toxicity may be more pronounced with SBRT than after conventional fractionated radiation therapy. Rib fractures and pleural or hepatic capsular pain may also occur after SBRT of peripherally located lung or liver tumors. The pain is usually transient but may require long- term analgesics. Severe acute and subacute skin reactions are more likely if there is overlap of entry and exit beams. Barrier creams, wet compresses, and occasionally antibiotic creams may be used for acute skin reactions. For nonhealing skin ulcers, hyperbaric oxygen and/or surgery may be required. Asymptomatic pericardial effusions and pleural effusions can also occur after SBRT [10]. It is not clear if these are clinically signiﬁcant. A sudden death was observed 2 days after 30 Gy delivered in one fraction to a large hepatocellular carcinoma [4]. Although this may have been unrelated to SBRT, it is possible that a tumor lysis syndrome or sudden release of cytokines may have contributed to the sudden event. Hopefully, future studies will report such toxicity even if it is not obviously related to treatment.

Avoidance/Partial Volume Effects The tolerance of organs to hypofractionated regimens remains largely unknown. However, based on the limited literature, estimates of partial volume tolerance to normal tissues from others’ experiences are included here. Tables 62-4 through 62-6 summarize some of the normal tissue constraints used in previous or ongoing clinical trials.

Liver Tolerance The Lyman NTCP model has been used to describe the partial volume effect of the liver to hyperfractionated radiation, and as expected it is a highly volume dependent organ with risk of toxicity highly correlated with mean liver dose [87]. There is a 5% risk of radiation-induced liver toxicity after uniform whole-

TABLE 62-6. SBRT liver study normal tissue permitted tolerances, in three and one fraction. Schefter [13], 3 fx

Herfarth [77], 3 fx

Herfarth [77], 1 fx

Liver

>700 cm liver

<15 Gy

D50% < 15 Gy D30% < 21 Gy

Stomach

Maximum

<30 Gy

Kidney

D67%

<15 Gy

Cord

Maximum

<18 Gy

fx, fraction.

3

D50% < 7 Gy D30% < 12 Gy

630

l.a. dawson

liver radiation of 28 Gy and 32 Gy in 2 Gy per fraction for liver metastases and primary liver cancer, respectively [87]. Extrapolation of this modeling to the hypofractionated setting and/or to different types of radiation plans in different patient populations needs to be done cautiously. Based on the model and on the rare occurrence of RILD after SBRT, it is highly likely that for a low effective liver volume irradiated, very high biological doses can be delivered in a variety of SBRT fractionations with little risk of RILD. Based on Schefter’s multi-institutional study of 18 patients with liver metastases, a recommendation was made that at least 700 cm3 of uninvolved liver receive less than 15 Gy in three fractions [13]. The median mean liver dose (to uninvolved liver) in the North American study was 15.3 Gy (3.3 to 23.9 Gy). Most patients were given prophylactic antiemetic therapy with or without one dose of dexamethasone. Herfarth recommends that no more than 50% of the liver receive more than 15 Gy in three fractions (or 7 Gy in one fraction), and 30% of the liver not receive more than 21 Gy in three fractions (or 12 Gy in one fraction), as more than 45 patients with liver metastases were treated with SBRT with these tolerance criteria without development of RILD (Herfarth KK and Wulf J, personal communication, 2005). There is no data regarding dose tolerances for biliary sclerosis.

Lung Tolerance The dose-volume constraints for lung pneumonitis, atelectasis, and other pulmonary toxicity after SBRT have not been well described. However, 60 Gy in three fractions is likely to lead to some local permanent injury of the bronchial tree and lung parenchyma in the high dose volume. Thus, local atelectasis, usually consisting of radiographic abnormalities in peripherally selected tumors for SBRT, is likely. More clinically substantial atelectasis may occur if centrally located tumors are treated with SBRT.

Luminal Gastrointestinal Tolerance Esophageal ulcers and fatal bleeds have been observed 3 to 9 months after SBRT for lung cancer [6]. Hemorrhagic gastritis has also been reported in a patient with liver metastases in which the gastric wall received 7 Gy × 2 [4]. We have observed one serious gastric bleed 6 months after SBRT for liver metastases, where a maximum dose of 33 Gy in six fractions was delivered to less than 0.5 cm3 of the gastric wall. At the time of the bleed, tumor recurrence was eroding though the stomach, which likely contributed to the toxicity [12]. Acute and late gastrointestinal toxicity has been observed after pancreatic cancer SBRT. The majority of patients treated in Denmark with 15 Gy × 3 for unresectable pancreatic cancer developed pronounced acute toxicity, despite the use of prophylactic ondansetron and pantoprazole. Four of 22 (18%) patients developed late gastrointestinal toxicity consisting of gastritis, ulceration, and/or perforation. Although actual delivered doses to the duodenum are not known, this highlights the caution that needs to be used when SBRT is used in tumors adjacent to serial functioning organs such as the duodenum [83]. Interestingly, another study of SBRT for pancreas cancer reported on 15 patients treated with 15 to 25 Gy single-fraction

SRS, with a small volume of duodenum receiving up to 22.5 Gy without development of grade 3 or higher gastrointestinal toxicity [82]. More patients are required to be treated with SBRT to have better conﬁdence in these dose tolerances of the duodenum to SBRT irradiation.

Spinal Cord Tolerance The doses we generally accept to the spinal cord (45 Gy in 1.8 Gy per fraction) are associated with a risk of myelitis of less than 1%. In some clinical situations, higher doses need to be delivered to the spinal cord in order to increase the tumor control probability. The spinal cord dose that has been estimated to be associated with a risk of myelitis of 5% risk is 57 Gy in 2 Gy per fraction [90]. The data on spinal cord tolerance after hypofractionation is more difﬁcult to interpret, as the patients treated with hypofractionation previously were most often treated with palliative intent and may not have lived to see late toxicity. For this reason, the following reports likely underestimate the true risk of toxicity. Spinal cord myelitis has been reported after 4 Gy × 10 (12/430, 3%) and 5.8 Gy × 6 (8/71, 11%) and was not reported after 3 Gy × 15, as summarized by Schultheiss et al. in 1995 [90]. Myelitis has not been reported in more recent published literature of SBRT. Results from some elegant experiments of spinal cord tolerance of the rat spinal cord to radiation therapy by van der Kogel et al. have recently been reported [21]. The rat spinal cord tolerance to radiation was found to be dependent on the volume irradiated and the spatial distribution of dose. The gray matter of the cord was found to be most resistant, whereas lateral white matter was most sensitive. Furthermore, the presence of a “bath” of low dose around the high dose maximum dose dramatically altered the cord tolerance to radiation. These ﬁndings should be considered when designing novel SBRT treatment protocols for paraspinal tumors.

Dose-Volume Tolerance Summary Other organs, some previously not considered dose-limiting normal tissues, need to be considered during SBRT planning. The dose-volume tolerance to SBRT is not well understood for most normal tissues, including, but not limited to, the esophagus, ribs, heart parenchyma, pericardium, coronary arteries, brachial plexus, peripheral nerves, subcutaneous tissue, and bone. Data is required to validate proposed normal tissue tolerance constraints used in previous papers and protocols. Some dose limits used in trials are shown in Tables 62-3, 62-5, and 62-6.

Conclusion SBRT is an exciting new ﬁeld of radiation oncology that brings together many of the technological advancements that have occurred recently in radiation oncology. SBRT is not speciﬁc to one planning system or delivery method but requires utilization of high-quality imaging for target deﬁnition, immobilization, high precision planning, and delivery and image guidance. Using these technological advancements, potent radiobiologi-

62.

stereotactic body radiation therapy

cal doses can be delivered in convenient fractionation schemes, generally ranging from one to ﬁve fractions. Preliminary data suggest that SBRT can be delivered safely with a high likelihood of local control and an acceptable safety proﬁle for most sites. Clinical outcomes are expected to be improved even further if SBRT can be combined with other treatments such as surgery and chemotherapy, and future research efforts should include determining optimal combinations of SBRT with other therapeutic modalities. The high doses and highly conformal SBRT plans are more prone to error introduced by geometric or dosimetry uncertainties. Such errors could lead to permanent and serious late normal tissue toxicity, and thus caution is required when SBRT is used clinically. Serious late toxicities have been reported after SBRT, including radiation-induced liver injury and gastrointestinal bleeding. As we gain increased experience with SBRT, the partial volume tolerances to hypofractionated heterogeneous radiation delivered to normal tissues should be able to be better deﬁned. Optimal fractionations, methods of guidance, and clinical applications of SBRT are not well established, and clinical trials continue to be required in this ﬁeld, especially in North America.

References 1. Kavanagh BD, Timmerman R, eds. Stereotactic Body Radiation Therapy. Philadelphia: Lippincott Williams & Wilkins, 2005. 2. Potters L, Steinberg M, Rose C, et al. American Society for Therapeutic Radiology and Oncology and American College of Radiology practice guideline for the performance of stereotactic body radiation therapy. Int J Radiat Oncol Biol Phys 2004; 60(4):1026–1032. 3. Hamilton AJ, Lulu BA. A prototype device for linear acceleratorbased extracranial radiosurgery. Acta Neurochir Suppl 1995; 63: 40–43. 4. Blomgren H, Lax I, Naslund I, Svanstrom R. Stereotactic high dose fraction radiation therapy of extracranial tumors using an accelerator. Clinical experience of the ﬁrst thirty-one patients. Acta Oncol 1995; 34(6):861–870. 5. Herfarth KK, Debus J, Lohr F, et al. Stereotactic single-dose radiation therapy of liver tumors: results of a phase I/II trial. J Clin Oncol 2001; 19(1):164–170. 6. Wulf J, Hadinger U, Oppitz U, Thiele W, Ness-Dourdoumas R, Flentje M. Stereotactic radiotherapy of targets in the lung and liver. Strahlenther Onkol 2001; 177(12):645–655. 7. Uematsu M, Shioda A, Suda A, et al. Computed tomographyguided frameless stereotactic radiotherapy for stage I non-small cell lung cancer: a 5-year experience. Int J Radiat Oncol Biol Phys 2001; 51(3):666–670. 8. Onishi H, Kuriyama K, Komiyama T, et al. Clinical outcomes of stereotactic radiotherapy for stage I non-small cell lung cancer using a novel irradiation technique: patient self-controlled breathhold and beam switching using a combination of linear accelerator and CT scanner. Lung Cancer 2004; 45(1):45–55. 9. Hiraoka M, Nagata Y. Stereotactic body radiation therapy for early-stage non-small-cell lung cancer: the Japanese experience. Int J Clin Oncol 2004; 9(5):352–355. 10. Timmerman R, Papiez L, McGarry R, et al. Extracranial stereotactic radioablation: results of a phase I study in medically inoperable stage I non-small cell lung cancer. Chest 2003; 124(5):1946–1955. 11. Fuss M, Thomas CR Jr. Stereotactic body radiation therapy: an ablative treatment option for primary and secondary liver tumors. Ann Surg Oncol 2004; 11(2):130–138.

631

12. Dawson LA, Eccles C, Craig T. Individualized image guided iso-NTCP based liver cancer SBRT. Acta Oncol 2006; 45:856– 864. 13. Schefter TE, Kavanagh BD, Timmerman RD, et al. A phase I trial of stereotactic body radiation therapy (SBRT) for liver metastases. Int J Radiat Oncol Biol Phys 2005; 62(5):1371– 1378. 14. Yamada Y, Lovelock DM, Yenice KM, et al. Multifractionated image-guided and stereotactic intensity-modulated radiotherapy of paraspinal tumors: a preliminary report. Int J Radiat Oncol Biol Phys 2005; 62(1):53–61. 15. Chang EL, Shiu AS, Lii MF, et al. Phase I clinical evaluation of near-simultaneous computed tomographic image-guided stereotactic body radiotherapy for spinal metastases. Int J Radiat Oncol Biol Phys 2004; 59(5):1288–1294. 16. Le QT, Tate D, Koong A, et al. Improved local control with stereotactic radiosurgical boost in patients with nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 2003; 56(4):1046– 1054. 17. Fowler JF, Tome WA, Fenwick JD, Mehta MP. A challenge to traditional radiation oncology [erratum appears in Int J Radiat Oncol Biol Phys. 2005; 61(3):962]. Int J Radiat Oncol Biol Phys 2004; 60(4):1241–1256. 18. Tobias CA. The repair-misrepair model in radiobiology: comparison to other models. Radiat Res Suppl 1985; 8:S77–95. 19. Guerrero M, Li XA. Extending the linear-quadratic model for large fraction doses pertinent to stereotactic radiotherapy. Phys Med Biol 2004; 49(20):4825–4835. 20. Zheng XK, Chen LH, Yan X, Wang HM. Impact of prolonged fraction dose-delivery time modeling intensity-modulated radiation therapy on hepatocellular carcinoma cell killing. World J Gastroenterol 2005; 11(10):1452–1456. 21. van Luijk P, Bijl HP, Konings AW, et al. Data on dose-volume effects in the rat spinal cord do not support existing NTCP models. Int J Radiat Oncol Biol Phys 2005; 61(3):892–900. 22. Craig T, Eccles C, Bissonnette J, Dawson LA. Physical versus biological dose for normal tissue complication probability in hypofractionated radiation therapy of unresectable liver cancer. Int J Radiat Oncol Biol Phys 2004; 60(1):S583–S584. 23. Dawson LA, McGinn CJ, Normolle D, et al. Escalated focal liver radiation and concurrent hepatic artery ﬂuorodeoxyuridine for unresectable intrahepatic malignancies. J Clin Oncol 2000; 18(11):2210–2218. 24. Seong J, Park HC, Han KH, Chon CY. Clinical results and prognostic factors in radiotherapy for unresectable hepatocellular carcinoma: a retrospective study of 158 patients. Int J Radiat Oncol Biol Phys 2003; 55(2):329–336. 25. Seong J, Park HC, Han KH, et al. Local radiotherapy for unresectable hepatocellular carcinoma patients who failed with transcatheter arterial chemoembolization. Int J Radiat Oncol Biol Phys 2000; 47(5):1331–1335. 26. Flanigan RC, Salmon SE, Blumenstein BA, et al. Nephrectomy followed by interferon alfa-2b compared with interferon alfa-2b alone for metastatic renal-cell cancer. N Engl J Med 2001; 345(23):1655–1659. 27. Narayan S, Henning GT, Ten Haken RK, et al. Results following treatment to doses of 92.4 or 102.9 Gy on a phase I dose escalation study for non-small cell lung cancer. Lung Cancer 2004; 44(1): 79–88. 28. Cheung PC, Yeung LT, Basrur V, et al. Accelerated hypofractionation for early-stage non-small-cell lung cancer. [see comment]. Int J Radiat Oncol Biol Phys 2002; 54(4):1014–1023. 29. Ben-Josef E, Normolle D, Pan C, et al. A phase II trial of highdose conformal radiation therapy with concurrent hepatic artery ﬂuorodeoxyuridine for unresectable intrahepatic malignancies. J Clin Oncol 2005; 23(24):8739–8747.

632

l.a. dawson

30. Llovet JM, Real MI, Montana X, et al. Arterial embolisation or chemoembolisation versus symptomatic treatment in patients with unresectable hepatocellular carcinoma: a randomised controlled trial.[see comment]. Lancet 2002; 359(9319):1734–1739. 31. Dawson LA, Litzenberg DW, Brock KK, et al. A comparison of ventilatory prostate movement in four treatment positions. Int J Radiat Oncol Biol Phys 2000; 48(2):319–323. 32. Malone S, Crook JM, Kendal WS, Szanto J. Respiratory-induced prostate motion: quantiﬁcation and characterization. Int J Radiat Oncol Biol Phys 2000; 48(1):105–109. 33. Hamilton AJ, Lulu BA, Fosmire H, et al. Preliminary clinical experience with linear accelerator-based spinal stereotactic radiosurgery. Neurosurgery 1995; 36(2):311–319. 34. Yenice KM, Lovelock DM, Hunt MA, et al. CT image-guided intensity-modulated therapy for paraspinal tumors using stereotactic immobilization. Int J Radiat Oncol Biol Phys 2003; 55(3): 583–593. 35. Meeks SL, Buatti JM, Bouchet LG, et al. Ultrasound-guided extracranial radiosurgery: technique and application. Int J Radiat Oncol Biol Phys 2003; 55(4):1092–1101. 36. Lax I, Blomgren H, Naslund I, Svanstrom R. Stereotactic radiotherapy of malignancies in the abdomen. Methodological aspects. Acta Oncol 1994; 33(6):677–683. 37. Wulf J, Hadinger U, Oppitz U, et al. Stereotactic radiotherapy of extracranial targets: CT-simulation and accuracy of treatment in the stereotactic body frame. Radiother Oncol 2000; 57(2): 225–236. 38. Lohr F, Debus J, Frank C, et al. Noninvasive patient ﬁxation for extracranial stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 1999; 45(2):521–527. 39. Booth JT, Zavgorodni SF. Set-up error & organ motion uncertainty: a review. Australas Phys Eng Sci Med 1999; 22(2):29–47. 40. Dawson LA, Brock KK, Kazanjian S, et al. The reproducibility of organ position using active breathing control (ABC) during liver radiotherapy. Int J Radiat Oncol Biol Phys 2001; 51(5): 1410–1421. 41. Eccles C, Brock KK, Hawkins MA, et al. Reproducibility of liver using active breathing control. Int J Radiat Oncol Biol Phys 2006; 64(3):751–759. 42. Sonke JJ, Zijp L, Remeijer P, van Herk M. Respiratory correlated cone beam CT. Med Phys 2005; 32(4):1176–1186. 43. Shirato H, Shimizu S, Kitamura K, et al. Four-dimensional treatment planning and ﬂuoroscopic real-time tumor tracking radiotherapy for moving tumor. Int J Radiat Oncol Biol Phys 2000; 48(2):435–442. 44. Balter JM, Lam KL, McGinn CJ, et al. Improvement of CT-based treatment-planning models of abdominal targets using static exhale imaging. Int J Radiat Oncol Biol Phys 1998; 41(4):939–943. 45. van Herk M, Remeijer P, Rasch C, Lebesque JV. The probability of correct target dosage: dose-population histograms for deriving treatment margins in radiotherapy. Int J Radiat Oncol Biol Phys 2000; 47(4):1121–1135. 46. van Herk M, Remeijer P, Lebesque JV. Inclusion of geometric uncertainties in treatment plan evaluation. Int J Radiat Oncol Biol Phys 2002; 52(5):1407–1422. 47. Kavanagh BD, Timmerman RD, Benedict SH, et al. How should we describe the radiobiologic effect of extracranial stereotactic radiosurgery: equivalent uniform dose or tumor control probability? Med Phys 2003; 30(3):321–324. 48. Brock KK, Sharpe MB, Dawson LA, et al. Accuracy of ﬁnite element model-based multi-organ deformable image registration. Med Phys 2005; 32(6):1647–1659. 49. Dawson LA, Eccles C, Bissonnette JP, Brock KK. Accuracy of daily image guidance for hypofractionated liver radiotherapy with

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

active breathing control. Int J Radiat Oncol Biol Phys 2005; 62(4):1247–1252. Balter JM, Brock KK, Litzenberg DW, et al. Daily targeting of intrahepatic tumors for radiotherapy. Int J Radiat Oncol Biol Phys 2002; 52(1):266–271. Buatti JM, Meeks SL, Howes TL. Stereotactic body radiation therapy for retroperitoneal and pelvic tumors. In: Kavanagh BD, Timmerman R, eds. Stereotactic Body Radiation Therapy. Philadelphia: Lippincott Williams & Wilkins, 2005:137. Fuss M, Salter BJ, Cavanaugh SX, et al. Daily ultrasound-based image-guided targeting for radiotherapy of upper abdominal malignancies. Int J Radiat Oncol Biol Phys 2004; 59(4):1245– 1256. Uematsu M, Sonderegger M, Shioda A, et al. Daily positioning accuracy of frameless stereotactic radiation therapy with a fusion of computer tomography and linear accelerator (focal) unit: evaluation of z-axis with a z-marker. Radiother Oncol 1999; 50: 337–339. Court L, Rosen I, Mohan R, Dong L. Evaluation of mechanical precision and alignment uncertainties for an integrated CT/LINAC system. Med Phys 2003; 30:1198–1210. Kuriyama K, Onishi H, Sano N, et al. A new irradiation unit constructed of self-moving gantry-CT and linac. Int J Radiat Oncol Biol Phys 2003; 55:428–435. Onishi H, Kuriyama K, Komiyama T, et al. A new irradiation system for lung cancer combining linear accelerator, computed tomography, patient self-breath-holding, and patient-directed beam-control without respiratory monitoring devices. Int J Radiat Oncol Biol Phys 2003; 56(1):14–20. Jaffray DA, Drake DG, Moreau M, et al. A radiographic and tomographic imaging system integrated into a medical linear accelerator for localization of bone and soft-tissue targets. Int J Radiat Oncol Biol Phys 1999; 45:773–789. Cho Y, Moseley DJ, Siewerdsen JH, Jaffray DA. Accurate technique for complete geometric calibration of cone-beam computed tomography systems. Med Phys 2005; 32(4):968–983. Jaffray DA, Bissonnette J, Craig T. X-ray imaging for veriﬁcation and localization in radiation therapy. In: Van Dyk J, ed. The Modern Technology of Radiation Oncology. Madison, WI: Medical Physics Publishing; 2005:278–279. Simpson RG, Chen CT, Grubbs EA, Swindell W. A 4-MV CT scanner for radiation therapy: the prototype system. Med Phys 1982; 9(4):574–579. Swindell W, Simpson RG, Oleson JR, et al. Computed tomography with a linear accelerator with radiotherapy applications. Med Phys 1983; 10(4):416–420. Mackie TR, Holmes T, Swerdloff S, et al. Tomotherapy: a new concept for the delivery of dynamic conformal radiotherapy. Med Phys 1993; 20:1709–1719. Mackie TR, Kapatoes J, Ruchala K, et al. Image guidance for precise conformal radiotherapy. Int J Radiat Oncol Biol Phys 2003; 56(1):89–105. Nakagawa K, Aoki Y, Tago M, et al. Megavoltage CT-assisted stereotactic radiosurgery for thoracic tumors: original research in the treatment of thoracic neoplasms. Int J Radiat Oncol Biol Phys 2000; 48(2):449–457. Pouliot J, Bani-Hashemi A, Chen J, et al. Low-dose megavoltage cone-beam CT for radiation therapy. Int J Radiat Oncol Biol Phys 2005; 61(2):552–560. Wulf J, Haedinger U, Oppitz U, et al. Stereotactic radiotherapy for primary lung cancer and pulmonary metastases: a noninvasive treatment approach in medically inoperable patients. Int J Radiat Oncol Biol Phys 2004; 60(1):186–196. Nagata Y, Negoro Y, Aoki T, et al. Clinical outcomes of 3D conformal hypofractionated single high-dose radiotherapy for one or

62.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

stereotactic body radiation therapy

two lung tumors using a stereotactic body frame. Int J Radiat Oncol Biol Phys 2002; 52(4):1041–1046. Hof H, Herfarth KK, Munter M, et al. Stereotactic single-dose radiotherapy of stage I non-small-cell lung cancer (NSCLC). Int J Radiat Oncol Biol Phys 2003; 56(2):335–341. Hara R, Itami J, Kondo T, et al. Stereotactic single high dose irradiation of lung tumors under respiratory gating. Radiother Oncol 2002; 63(2):159–163. Lee SW, Choi EK, Park HJ, et al. Stereotactic body frame based fractionated radiosurgery on consecutive days for primary or metastatic tumors in the lung. Lung Cancer 2003; 40(3):309– 315. Whyte RI, Crownover R, Murphy MJ, et al. Stereotactic radiosurgery for lung tumors: preliminary report of a phase I trial. Ann Thoracic Surg 2003; 75(4):1097–1101. Fukumoto S, Shirato H, Shimzu S, et al. Small-volume imageguided radiotherapy using hypofractionated, coplanar, and noncoplanar multiple ﬁelds for patients with inoperable Stage I nonsmall cell lung carcinomas. Cancer 2002; 95(7):1546–1553. Song DY, Blomgren H. Stereotactic body radiation therapy for lung tumors. In: Kavanagh BD, Timmerman R, eds. Stereotactic Body Radiation Therapy. Philadelphia: Lippincott Williams & Wilkins, 2005:101–102. Onishi H, Nagata Y, Shirato H, et al. Stereotactic hypofractionated high-dose irradiation for patients with stage I non-small cell lung carcinoma: clinical outcomes in 241 cases of a Japanese multi-institutional study. Int J Radiat Oncol Biol Phys 2003; 57(Suppl 1):142. Blomgren J, Lax I, Goranson H, et al. Radiosurgery for tumors in the body: clinical experience using a new method. J Radiosurg 1998; 1:63–74. Gunven P, Blomgren H, Lax I. Radiosurgery for recurring liver metastases after hepatectomy. Hepat Gastroenterol 2003; 50(53): 1201–1204. Wulf J, Herfarth K. Normal tissue dose constraints in stereotactic body radiation therapy for liver tumors. In: Kavanagh BD, Timmerman R, eds. Stereotactic Body Radiation Therapy. Philidelphia: Lippincott Williams & Wilkins, 2005:44. McGinn CJ, Ten Haken RK, Ensminger WD, et al. Treatment of intrahepatic cancers with radiation doses based on a normal tissue complication probability model. J Clin Oncol 1998; 16(6):2246– 2252.

633

79. Wu DH, Liu L, Chen LH. Therapeutic effects and prognostic factors in three-dimensional conformal radiotherapy combined with transcatheter arterial chemoembolization for hepatocellular carcinoma. World J Gastroenterol 2004; 10(15):2184–2189. 80. Ryu SI, Chang SD, Kim DH, et al. Image-guided hypo-fractionated stereotactic radiosurgery to spinal lesions. Neurosurgery 2001; 49(4):838–846. 81. Gerszten PC, Ozhasoglu C, Burton SA, et al. Evaluation of CyberKnife frameless real-time image-guided stereotactic radiosurgery for spinal lesions. Stereotact Funct Neurosurg 2003; 81(1–4):84–89. 82. Koong AC, Le QT, Ho A, et al. Phase I study of stereotactic radiosurgery in patients with locally advanced pancreatic cancer. Int J Radiat Oncol Biol Phys 2004; 58(4):1017–1021. 83. Hoyer M, Roed H, Sengelov L, et al. Phase-II study on stereotactic radiotherapy of locally advanced pancreatic carcinoma. Radiother Oncol 2005; 76(1):48–53. 84. Qian G, Lowry J, Silverman P, et al. Stereotactic extra-cranial radiosurgery for renal cell carcinoma. Int J Radiat Oncol Biol Phys 2003; 57(Suppl 2):S283. 85. Wersall JP, Kavanagh BD. Stereotactic body radiation therapy for renal cell carcinoma. In: Kavanagh BD, Timmerman R, eds. Stereotactic Body Radiation Therapy. Philadelphia: Lippincott Williams & Wilkins, 2005:129–133. 86. Fong Y, Fortner J, Sun RL, et al. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 1999; 230(3):309–318; discussion 18–21. 87. Dawson LA, Normolle D, Balter JM, et al. Analysis of radiationinduced liver disease using the Lyman NTCP model. Int J Radiat Oncol Biol Phys 2002; 53(4):810–821. 88. Cheng JC, Wu JK, Lee PC, et al. Biologic susceptibility of hepatocellular carcinoma patients treated with radiotherapy to radiation-induced liver disease. Int J Radiat Oncol Biol Phys 2004; 60(5):1502–1509. 89. Herfarth KK, Hof H, Bahner ML, et al. Assessment of focal liver reaction by multiphasic CT after stereotactic single-dose radiotherapy of liver tumors. Int J Radiat Oncol Biol Phys 2003; 57(2): 444–451. 90. Schultheiss TE, Kun LE, Ang KK, Stephens LC. Radiation response of the central nervous system. Int J Radiat Oncol Biol Phys 1995; 31(5):1093–1112.

6 3

Stereotactic Body Radiation Therapy: Fractionated Radiation Therapy Perspective Gordon W. Wong, Rafael R. Mañon, Wolfgang Tomé, and Minesh Mehta

Introduction The introduction of intracranial stereotactic radiosurgery in 1951 is widely attributed to Lars Leksell [1]. As initially developed, the method delivered a single dose of radiation to a target lesion, which was localized using the principles of stereotaxy. Current treatment strategies in stereotactic body radiation therapy (SBRT) are an extension of these principles applied to extracranial sites. Techniques that allow precise targeting and treatment are used to deliver one or usually more large fractions of radiation to extracranial sites. SBRT incorporates two speciﬁc innovations: a variety of systems (ultrasound, computed tomography, beacons) that allow adequate visualization of targets, primarily to ensure accurate repositioning for repeat fractions, often referred to as image-guided radiotherapy, or IGRT; and systems that eliminate, minimize, or compensate for tumor motion, the latter often referred to as 4D radiotherapy. These critical advances facilitate single-fraction treatment or hypofractionation, with increased daily doses and decreased overall treatment times, and avoid the projected concomitant increase in toxicity by minimizing the volumes irradiated. SBRT is attractive because it allows the delivery of an ablative dose of radiation to tumor sites, with the potential to improve local tumor control. SBRT also shortens the length of the treatment course, making therapy more convenient for the patient and potentially further improving tumor control by overcoming accelerated repopulation. In this context, we will examine the beneﬁts and disadvantages of SBRT relative to conventionally fractionated radiotherapy. Stereotactic body radiation therapy has been applied to a number of sites outside the CNS, the most common being lung and liver. The bulk of our experience at the University of Wisconsin is in the application of this modality to the treatment of early-stage (T1/T2) non–small cell lung cancer (NSCLC). The discussion of general principles presented here will primarily

focus on this topic. The radiobiologic principles utilized can be applied to SBRT of other parallel organs such as the liver and kidney, given that their respective tolerance doses are considered.

Standard Dose Fractionated Radiation in Unresectable T1/T2 Lung Cancer Surgical resection in patients with early-stage NSCLC offers 5year survival rates in the range of 60% to 70% [2, 3]. This remains the treatment of choice for early-stage lung cancer; however, many patients with early-stage disease are unable to undergo surgery because of physiologic limitations, such as inadequate pulmonary reserve, cardiac disease, or other comorbidities. Deﬁnitive radiation therapy has historically been considered a reasonable alternative to surgery for these patients, with survival rates ranging from 10% to 30%, with “standard” schedules of 45 to 66 Gy in 1.8- to 2.0-Gy fractions [4–8]. A review from Duke University found a 30% incidence of death after local failure and a 30% incidence of distant failure with median doses of 60 to 66 Gy [9]. An analysis by Krol et al. of patients with T1/T2 NSCLC treated with 60-Gy split course or 65-Gy continuous radiation demonstrated a 15% and 31% 5year overall and cause-speciﬁc survival, respectively [10]. Cheung et al. reported on patients with T1–T4 NSCLC treated with 52.6-Gy continuous radiation and found a 16% overall survival and 13.9% relapse-free survival at 5 years. There was a local component of failure in 68.9% of the patients [11]. All of these experiences found little beneﬁt from prophylactic nodal irradiation. Given that early-stage lung cancer is not very commonly metastatic, these results underscore the relatively poor results of standard radiation approaches to control this disease and naturally lead to the question: “Would more dose-intense approaches to radiation delivery yield superior results?”

635

636

g.w. wong et al.

NSCLC If no prolif.

1.

80

94

100

=

60

0

=

1.

γ5

with proliferation, as published by Martel et al 1999

40 20 Tp = 3 days Tk = 28 days γ = 0.66 Gy/d

5

0

An analysis of some of the dose regimens used in the stereotactic treatment of lung lesions is shown in Table 63-1.What is evident at ﬁrst glance is that the biologically equivalent doses (BEDs) for acute effects (assuming an α/β = 10 Gy) are 1.3 to 1.7 times higher with SBRT compared with standard fractionation treatments. Why would such high doses be necessary, and what underscores the failure of standard radiation therapy? To answer this question, we must examine several lines of data. A postulated major radiobiologic limitation to conventionally fractionated schedules is accelerated repopulation, which tends to limit gains in tumor control probability achieved by increasing total dose [14]. Martel et al. described the results of a dose escalation study wherein they showed a clear dependence of local progression-free survival on dose, with 85 Gy deemed necessary for a 30-month local-tumor control probability of 50% [12]. Their results are summarized in Fig. 63-1(dashed curve), which illustrates the progression-free survival and D50 as a function of total radiation dose in 2-Gy fractions with and without repopulation. D50 is deﬁned as the dose necessary to achieve a tumor control probability (TCP) of 50%. The relative slope at D50 is γ5, which measures the increase in the TCP curve as a percentage change in the total dose and is dimensionless. This graph helps us understand the poor results with standard radiation therapy of 60 to 70 Gy with the predicted 30-month local progression-free survival of 15% and 24%, respectively. Doses up to 100 Gy, given at 2-Gy fractions, ﬁve times a week, would be required for 90% control probability. However, this would require an overall treatment time of approximately 10 weeks, which is unduly extensive. SBRT overcomes accelerated repopulation by completing therapy before accelerated repopulation becomes a factor. The no-repopulation (solid) curve in Figure 63-1 is applicable for modeling SBRT results. Note that the no-repopulation slope is steeper at a γ5 of 1.94, instead of 1.5 as in the original Martel curve [13], and its D50 is 70 instead of 84.5 Gy. The resulting curve with increased γ5 and lower D50 implies greater innate effectiveness of an accelerated schedule in terms of tumor control. A second consideration comes from post hoc analysis of survival as a function of actual treatment duration in clinical trials where the planned duration was ﬁxed, but subsets of

% Progression -free Survival of patients at 30 months (Martel et al. 1999 )

γ5

Would Dose-Intense Radiation Therapy Yield Superior Results?

0 40

50

60

70

80

90

100

Total dose in 2 Gy fractions (= NTD) FIGURE 63-1. Calculated progression-free survival for patients with NSCLC 30 months from treatment as a function of total dose in 2-Gy fractions with and without repopulation [13]. (Reprinted from Fowler JF, Tome WA, Fenwick JD, et al. A challenge to traditional radiation oncology. Int J Radiat Oncol Biol Phys 2004; 60(4):1241–1256. With permission from Elsevier.)

patients required treatment interruptions, thereby lengthening the delivery duration. One such analysis demonstrated a survival loss of 1.6% per day of treatment prolongation beyond the planned 6.5 weeks [15]. Level I evidence supporting this hypothesis comes from the European continuous hyperfractionated accelerated radiation therapy (CHART) study. This study tested the concept of “dose-dense radiotherapy” using an accelerated hyperfractionated regimen, delivering all treatment within 2 weeks. It demonstrated a 24% reduction in the relative risk of death and a 23% reduction in the relative risk of local progression with the accelerated hyperfractionation arm [16]. The Eastern Cooperative Oncology Group (ECOG) hyperfractionated accelerated radiation therapy (HART) study also demonstrated a large beneﬁt in survival with “dose-dense radiotherapy” [17]. These two trials provide clinical proof for the concept of accelerated repopulation and suggest that schedule shortening may be an effective strategy to minimize the impact of this phenomenon. A similar trial in head and neck cancer found that patients overexpressing EGFR beneﬁted from schedule shortening, putatively implicating EGFR upregulation as a molecular mechanism or marker for accelerated repopulation [18].

TABLE 63-1. Analysis of selected dose regimens. Total dose

Conventional fractionation 60 Gy, 30 fractions 70 Gy, 35 fractions SBRT 48 Gy, 4 fractions 45 Gy, 3 fractions 48 Gy, 3 fractions 60 Gy, 5 fractions 60 Gy, 3 fractions 69 Gy, 3 fractions

RE 1 + d/10

1.2 1.2 2.2 2.5 2.6 2.2 3.0 3.3

NTD Gy (2-Gy fractions)

Estimated progression-free survival at 30 months (assuming no hypoxia) (%)

72 84

60 70

16 26

106 113 125 132 180 228

88 94 104 110 150 190

84 95 99 >99 >99 >99

BED, Gy10

Note: BED (biologically effective dose) and NTD (normalized total dose) were calculated using the linear quadratic formulation assuming α/β = 10 Gy. Using Martel et al. [12] and proliferation subtracted at PFS at 30 months Tp = 3 days and Tk = 28 days, PFS for patients with NSCLC was estimated at 30 months. Repopulation was only accounted for in the conventional fractionation schedules. Source: Adapted from Fowler JF, Tome WA, Fenwick JD, et al. A challenge to traditional radiation oncology. Int J Radiat Oncol Biol Phys 2004; 60(4):1241–1256.

63.

OER 3.0

–2

IC OX

–6

LL CE

–8

S

Log 10 TUMOR CELLS SURVIVING

HYPOX 20%

0

–4

20% HY IF N PO OT XIC RE CE OX LLS YG EN AT ED

–10 3F x 23Gy_24Gy

–12 0

637

stereotactic body radiation therapy: fractionated radiation therapy perspective

20

40 60 80 100 120 140 160 180 200 Total Dose in 2Gy fractions = NTD (Gy) FIGURE 63-2. Cell survival as a function of total dose for oxic versus 20% hypoxic cells without reoxygenation [13]. (Reprinted from Fowler JF, Tome WA, Fenwick JD, et al. A challenge to traditional radiation oncology. Int J Radiat Oncol Biol Phys 2004; 60(4):1241–1256. With permission from Elsevier.)

Hypoxia is another major factor limiting the success of radiotherapy and necessitating very high doses for long-term control. Figure 63-2 illustrates cell killing with multifraction radiation therapy [13]. The graph with closed circles represents cell killing in a tumor model that is well-oxygenated. In this scenario, biologically equivalent doses in the range used in conventional fractionated radiation therapy (60 to 70 Gy in 30 to 35 fractions) would have a good chance of controlling a 1- to 10-g mass. However, if reoxygenation is incomplete so that a proportion of cells are hypoxic, two- to threefold higher doses may be necessary to have a reasonable chance of tumor control by reducing the tumor burden by a factor of 10−10 to 10−11. The question of which tumors will beneﬁt from radiation dose intensiﬁcation remains unanswered until reliable methods of measuring hypoxia are available [13].

Single Dose Versus Fractionated Radiation Therapy

the cell cycle to shift into more radiosensitive phases [13]. Therefore, single-fraction SBRT would be expected to require signiﬁcantly higher doses than fractionated SBRT. This issue is addressed further in the next section.

What Have We Learned from Mouse Models? In the 1970s, Fowler et al. evaluated the control of mammary tumors in C3H mice with radiation therapy [22]. Several regimens were tested with and without misonidazole. Results applicable to the current discussion are presented in Figure 63-3. The ﬁgure depicts tumor control probability (TCP) as a function of overall treatment time. The top curve illustrates TCP of mouse tumors treated with misonidazole prior to irradiation, and the lower curve illustrates TCP of control mice breathing room air (a tumor system that is approximately 10% hypoxic) [22]. It is evident that TCP is signiﬁcantly lower for hypoxic tumors treated with one- or two-fraction courses. This difference in TCP decreases as the number of fractions increases to three and beyond, which demonstrates the importance of reoxygenation even in treatment schedules using only a few fractions. Although single-fraction SBRT may be adequate for small tumors in anatomically favorable locations, we do not yet have the experience to discern which tumors are amenable to single-fraction treatment by reliably measuring the degree of hypoxia in a clinical setting. However, it is evident that much higher doses may be needed to achieve tumor control with single-fraction versus hypofractionated SBRT.

Clinical Experience The initial SBRT experiences in the treatment of primary liver tumors, NSCLC, and metastatic lesions to the lung, liver, abdomen, and pelvis have been published and are extensively reviewed in several chapters in this book. A summary of the published results is presented in Table 63-2. Many of these studies have treated a small number of patients, and no studies have directly compared SBRT with conventional schedules. C3H transplanted mouse mammary tumors 100

% TCP at 150 d

The parameters of high dose per fraction, small treatment volume, and rapid dose fall-off, which are intrinsic to radiosurgery, make precision in treatment delivery during SBRT of great importance [19–21]. The complexity of set-up can make the delivery of each fraction both labor intensive and taxing on the patient. Single-dose SBRT is attractive because it has the advantages of patient convenience, decreased staff involvement, and lower consumption of resources. Despite its convenience, single-fraction SBRT is not the best option for several reasons. First, multifraction treatment affords the luxury of compensating for small errors in set-up, delivery, or patient motion after each fraction. In single-fraction SBRT, a geometric miss or a tumor subvolume that is underdosed cannot be compensated. In addition, single-fraction SBRT is not the ideal radiobiologic option. As previously stated, reoxygenation of tumors is important in reducing their radioresistance. Single-fraction treatment gives no chance for hypoxic cells to reoxygenate or for cells in a resistant phase of

With Misonidazole 35 min before each Fr Based on Tpots of 1.6 vs 5 d

80 60 40

Controls breathing air.

20 1F 3F 5F 0

0

9F

15F

5 10 15 20 Overall Treatment Time (days) for mice Probable overall treatment times of x 3 for people FIGURE 63-3. Tumor control probability (TCP) as a function of time with or without misonidazole in the treatment of mammary tumors in mice.

638

g.w. wong et al.

TABLE 63-2. Published data for SBRT to various targets.

Author and institution

N

Location

Blomgren et al. [23] Karolinska Hospital, Sweden

31

Lung, liver, and retroperitoneal tumors Primary hepatic lesions

7 14 5 3

Blomgren et al. update [24] Karolinska Hospital, Sweden

Wulf et al. [25] University of Würzburg, Germany Herfath et al. [26] German Cancer Research Center Hof et al. [27] German Cancer Research Center Uematsu et al. [28, 29] National Defense Medical College, Japan Nagata et al. [30] Kyoto University, Japan Nagata et al. update [31] Kyoto University, Japan Fukumoto et al. [32] Hokkaido University, Japan Timmerman et al. [33, 34] Indiana University Onishi et al. [35] 13 Japanese institutions Wulf et al. [36] University of Würzburg, Germany Wada et al. [37] Yamagata City Hospital, Japan; Tohoku University School of Medicine, Japan

Total dose (Gy)

Fractions

8–66

Median follow-up (months)

Local failures

Local control (%)

1–4

7

77

15–45

1–3

0

100

8–64 20–60 30–48

1–4 1–3 2–3

5 ?1* 1

64

4

92

Toxicities ≥ grade 3

2 patients with unretractable ascites resulting in deaths, 1 subcapsular hemorrhage 1 hemorrhagic gastritis

15–45

1–5

11

Metastatic hepatic lesion Pulmonary metastases Retroperitoneal and skeletal metastases Lung, liver, and retroperitoneal tumors Primary hepatic lesions

15–45

1–3

0

100

17

Metastatic hepatic lesion

20–45

1–4

1

94

13 15

Pulmonary metastases Extrahepatic abdominal lesions

15–45 18–30

1–3 2–5

1 2

92 87

27

Primary/recurrent NSCLC, pulmonary metastases T1–T3 NSCLC subset Liver lesions Primary and metastatic liver

14–30

2–4

8

4

85

14–30 28–30 14–26

2–3 3–4 1

8 9 9.5

1 4 12

88 83 66

None None None

14.9

2

80

None

>60

0

100

None

19

1 3 4

97 67 90

None None None

50

8 24 35

11

67

1 ventricular bleed requiring transfusion

2 patients with unretractable ascites resulting in deaths, 1 subcapsular hemorrhage 1 hemorrhagic gastritis, 1 duodenal ulcer 1 chronic cough 1 ventricular bleed requiring transfusion, 1 hemorrhagic duodenal ulcer One grade 3 pneumonitis, 1 pulmonary hemorrhage

10

T1–T2 NSCLC

19–26

1

50

T1–T2 NSCLC

50–60

5–10

31 9 40

40–48 48 40–48

4 4 4

44 12 13

T1–T3 NSCLC Pulmonary metastases Primary and metastatic lung T1–T3 NSCLC Pulmonary metastases Pulmonary metastases

48 48 60

4 4 5

19

0 2 0

100 83 100

None None None

17

T1–T2 NSCLC

48–60

8

24–44

1

94

None

44

T1–T2 NSCLC

24–72

3

18

6

86

One grade 3 pneumonitis, one grade 3 hypoxia

2

T1–T2 NSCLC

18–75

1–25

24

33

87

6 patients: pneumonitis, esophagitis, dermatitis

Abdominal and pelvic lesions Abdominal subset Pelvic subset

10–15

2–3

20

5

76

None

10–15 11–15

2–3 3–3

0 5

100 67

45

3

6

82

45

3

2

85

21 6 15 34 13

Primary NSCLC, lung and liver metastases Primary NSCLC subset

18

None

63.

stereotactic body radiation therapy: fractionated radiation therapy perspective

Early clinical data from the Karolinska Hospital were reported by Blomgren et al. [23]. Of the ﬁrst 31 patients treated with SBRT using a stereotactic body frame for immobilization, 9 were treated for primary intrahepatic lesions to 15 to 45 Gy in 1 to 3 fractions. Of the 7 evaluable patients, no local failures occurred. Fourteen patients were treated for metastatic liver lesions to 8 to 64 Gy in 1 to 4 fractions, with 5 subsequent local failures. Five patients were treated for metastatic pulmonary lesions to 20 to 60 Gy in 1 to 3 fractions with radiographic regression in 4 of the patients. Fever and nausea were common in patients treated for liver lesions. Major complications were only seen in the patients treated for primary hepatic lesions, including two patients who developed ascites and one with subcapsular hemorrhage. Overall, 43 lesions were treated in 31 patients, with 7 local failures. In a subsequent update of 75 lesions in 50 patients, 4 local failures were reported (95% local control); of 20 primary hepatic lesions in 11 patients treated to 15 to 45 Gy in 1 to 3 fractions, there were no local failures; of the 21 hepatic metastases in 17 patients treated to 20 to 45 Gy in 2 to 4 fractions, there was 1 local failure; and of the 17 metastatic thoracic lesions in 13 patients treated to 15 to 45 Gy in 1 to 3 fractions, there was 1 local failure [24]. Wulf et al. reported on the experience from the University of Würzburg, Germany, on SBRT to lung and liver lesions [25]. All patients were treated using a stereotactic body frame for immobilization. Twenty-seven thoracic lesions were treated, including 12 primary T1 to T3 NSCLC, 4 local recurrences of lung carcinoma, and 11 pulmonary metastases. Twenty-four of the 27 patients received 10 Gy × 3, one patient received 7 Gy × 2 as a boost to conventional treatment, and two patients received 7 Gy × 4 due to proximity of critical structures. Four of the treated pulmonary lesions failed, including one failure in the eight patients treated with curative intent for primary NSCLC. Twenty-four liver lesions were treated with 30 Gy in three fractions, with the exception of one patient receiving 28 Gy in four fractions due to proximity to the esophagus. Four local failures occurred in this group. For all patients treated, there were only two major toxicities, including a grade 3 esophagitis and a death secondary to a pulmonary hemorrhage 9 months after SBRT from a local recurrence in a previously irradiated region. A phase I/II study from the German Cancer Research Center that included 35 patients was reported by Herfarth et al. [26]. Fifty-ﬁve unresectable hepatic lesions, including 4 primary hepatic and 51 metastatic tumors, were treated with single-dose SBRT. Dose escalation over the course of the trial ranged from 14 to 26 Gy, with no major complications. With a median follow-up of 5.7 months, 13 patients died of progressive disease, and there were 12 local failures, 8 in patients receiving lower doses. The same group also treated 10 patients with T1/ T2 NSCLC with single-dose stereotactic radiation of 19 to 26 Gy. After a median follow-up of 14.9 months, there were two local failures and ﬁve systemic metastases [27]. Wada et al. reported the results of treating 42 lesions in 34 patients [37]. There were 14 primary T1/T2 NSCLC, 6 hepatocellular carcinomas, and 23 lung or liver metastases treated to 45 Gy in 3 fractions using a body immobilization device with an abdominal belt to minimize respiratory movement. With a median follow-up of 19.3 months, there were six local failures. Only grade 1 or 2 toxicities were observed, including pneumo-

639

nitis and mild chest or abdominal discomfort after procedure. Univariate analysis showed a statistically signiﬁcant higher TCP for tumors <3 cm versus ≥3 cm of 95.0% and 58.3%, respectively. Uematsu et al. reported the 5-year results in 50 patients with stage I NSCLC treated at the National Defense Medical College in Japan [28]. Patients received 50 to 60 Gy in 5 to 10 fractions over 1 to 2 weeks. The FOCAL system was utilized for daily patient positioning and imaging localization, without any mechanical immobilization by employing a fusion of computed tomography (CT) and linear accelerator with a shared table. There were no major toxicities, and local failure occurred in three patients over a median follow up of 36 months. There were ﬁve isolated nodal failures and two nodal and distant failures. In an 8-year update, no additional local failures were documented [29]. Nagata et al. reported on the experience at Kyoto University in Japan. Forty patients with lung lesions <40 mm, primary or metastatic, were treated with SBRT using a stereotactic frame, vacuum pillow, and diaphragm control device [30]. The initial three patients received 40 Gy in four fractions, with the remainder receiving 48 Gy in four fractions. Only grade 1 toxicities were seen. Of the 31 patients treated with primary NSCLC, there were three distant failures and one local failure in a patient who received 40 Gy. There were three local failures in nine patients treated for metastatic pulmonary lesions. Nagata et al. more recently updated their experience in 77 patients, 55 with primary lung cancer (mostly stage I). With a median follow-up of 19 months in 44 patients with T1/T2 NSCLC receiving 48 Gy in four fractions over 1 to 2 weeks, no local failures were observed. There were two local failures in 10 patients with metastatic lesions that received 48 Gy in four fractions, but no local failures have been reported to date in metastatic lesions in 12 patients that received 60 Gy in ﬁve fractions [31]. Hypofractionated image-guided radiotherapy for 22 patients with stage I NSCLC was reported by Fukumoto et al. from Hokkaido University in Japan [32]. Treatment volumes were based on the visualized lesions during baseline, peak expiration, and peak inspiration on CT scan without breathing or mechanical restrictions. Eleven patients with peripheral tumors were given 60 Gy in eight fractions, and 11 patients with central tumors received 48 Gy in eight fractions over 2 weeks. Seventeen patients were eligible for evaluation over a follow-up period of 24 to 44 months, with no toxicities and only one local failure. However, four patients progressed regionally or distally. Timmerman et al. reported the phase I dose escalation results from Indiana University using SBRT for early-stage lung cancer [33, 34]. Forty-four patients with medically inoperable T1/T2 NSCLC ≤ 7 cm were treated to deﬁne a maximum tolerated dose. Doses were initiated at 8 Gy × 3 fractions, with escalating doses of an additional 2 Gy per fraction in each cohort. Patients were immobilized with a stereotactic body frame. Grade 3 toxicities were seen in two patients, one with pneumonitis and another with hypoxia. After a median follow-up of 15.2 months, local failure occurred in six patients, all of whom received less than 18 Gy × 3, and distant failure occurred in nine patients. The maximum tolerated dose was not reached, but patients with 5- to 7-cm primary lesions experienced bother-

640

g.w. wong et al.

FIGURE 63-4. Screen shot from TomoTherapy treatment of a patient with an 11-mm recurrent stage I NSCLC who received 60 Gy in ﬁve fractions. The upper left image is the MVCT (megavoltage CT) obtained by TomoTherapy, the lower left image is the treatment planning CT,

and the larger image on the right is the MVCT image with overlaid contours from the treatment planning CT to ensure proper patient positioning and target localization.

some grade 2 toxicities, leading to a dose reduction in this group to 22 Gy × 3. Onishi et al. recently reported multi-institutional retrospective data in 245 Japanese patients with T1/T2 NSCLC <6 cm treated at 13 institutions [35]. Various schedules with doses ranging from 18 to 75 Gy in 1 to 25 fractions were employed. Immobilization techniques also varied among the institutions. Six patients experienced greater than grade 2 toxicities, including pneumonitis, esophagitis, and dermatitis. With a median follow-up of 24 months, 33 (13.5%) patients experienced local failure. However, the rate was 8.1% for those who received a BED ≥ 100 Gy compared with 26.4% for BED < 100 Gy. Patients who received a BED ≥ 100 Gy also had a statistically signiﬁcant improvement in overall survival. There were 20 regional and 36 distant failures. At the University of Wisconsin, we have treated 14 patients with primary or recurrent NSCLC with SBRT. One patient was treated with 45 Gy in ﬁve fractions, whereas the remainder received 60 Gy in ﬁve fractions. With a median follow-up of 12 months, there have not been any local or regional failures or any major complications of treatment, other than one episode of subcutaneous ﬁbrosis resulting in neuropathic pain, controlled with Tegretol. Four of these patients have been treated with TomoTherapy. These patients were treated while immobilized in a double vacuum-based immobilization system and aligned using image-guided positioning. The TomoTherapy treatment machine incorporates megavoltage CT (MVCT), which can be fused and compared with the treatment planning CT, prior to each treatment. Representative images of MVCT and treatment planning CT can be seen in Figure 63-4. Wulf et al. also reported on SBRT as a boost treatment for 21 patients with abdominal or pelvic lesions [36]. These patients received 10 to 15 Gy in two to three fractions of 5 Gy while immobilized in a stereotactic body frame. After a median follow-up of 20 months, there were no local failures in six abdominal lesions and ﬁve failures in 15 pelvic lesions treated. They theorized that insufﬁcient dose may have contributed to the failures. There were no acute toxicities greater than grade

2. However, late effects included one patient with rectal bleeding and two patients experienced intestinal-vaginal ﬁstulas.

Cost-Effectiveness and Quality of Life Cost-effectiveness and quality of life for SBRT has not been adequately addressed, as the technology is in its infancy, with rapid changes. Of the institutions that perform SBRT, the approach and methods can vary widely, thus affecting cost. There are greater requirements in set-up and treatment in comparison with conventional treatment. More precise technologies incorporate one or more of the following: gating, PET/CT, and daily image veriﬁcation, and the frequency of the use of these inﬂuences cost. Immobilization is more involved with most institutions using various types of stereotactic body frames, vacuum pillows, abdominal compression devices, and ﬁducial markers. A greater involvement of staff is required with intimate involvement of physics, dosimetry, and therapists. However, the total number of fractions and treatment duration is much shorter than conventional schedules. Local control for early-stage NSCLC ranges from approximately 80% to 90%. However, these are early results in a small number of patients, without direct comparison with the current standard of care. The true effectiveness cannot be ascertained without wellcontrolled clinical trials. Quality of life issues have not been addressed. However, the incidence of major toxicities is fairly low as described in the previous section.

Conclusion Stereotactic body radiation therapy is a viable option when surgical resection as the current standard of care cannot be offered or is refused by the patient. However, SBRT is still in an early evolutionary phase. The optimum dose-fractionation schedules have not been established, but many institutions have published encouraging data with local control rates of approxi-

63.

stereotactic body radiation therapy: fractionated radiation therapy perspective

641

TABLE 63-3. Advantages and disadvantages of single-fraction versus multifraction SBRT versus conventional radiotherapy.

Addresses hypoxia

Overcomes repopulation Efﬁcacy

Cost-effectiveness

Convenience

Single-fraction SBRT

Multifraction SBRT

Conventional radiotherapy

No, single fraction does not allow for reoxygenation or cellcycle phase shift Treatments completed before repopulation occurs May be more effective, but no strong clinical evidence and technique has less room for error over a single-fraction course Requires more resources and staff, but overall decreased with single fraction Most convenient with a singlefraction treatment

Yes

Yes

Treatments completed before repopulation occurs May be more effective, but no strong clinical evidence at this time

Requires high doses to compensate for repopulation Standard option for unresectable patients

Requires more resources and staff, but fewer fractions

Standard option for unresectable patients

Fewer fractions, more convenient for patients, but more time required per fraction

Can initiate treatment quickly with simpler planning and setup, but requires 6 to 7 weeks of treatment

mately 80% to 90%, especially with BED > 100 Gy. Radiobiologic modeling-based schedules have been proposed and are in the early phases of being tested clinically [13]. Some of the advantages and disadvantages of single-fraction SBRT, multifraction SBRT, and conventional radiation schedules are summarized in Table 63-3. Well-controlled clinical trials are necessary to address the unanswered questions of proper technique, dose fractionation, quality of life, cost-effectiveness, and comparison with surgical resection and conventional radiotherapy.

References 1. Levin VA. Cancer in the Nervous System, 2nd ed. Oxford: Oxford University Press, 2002. 2. Mountain CF. A new international staging system for lung cancer. Chest 1986; 89(4 Suppl):225S–33S. 3. Naruke T, Goya T, Tsuchiya R, et al. Prognosis and survival in resected lung carcinoma based on the new international staging system. J Thorac Cardiovasc Surg 1988; 96(3):440–447. 4. Armstrong JG, Minsky BD. Radiation therapy for medically inoperable stage I and II non-small cell lung cancer. Cancer Treat Rev 1989; 16(4):247–255. 5. Coy P, Kennelly GM. The role of curative radiotherapy in the treatment of lung cancer. Cancer 1980; 45(4):698–702. 6. Dosoretz DE, Katin MJ, Blitzer PH, et al. Radiation therapy in the management of medically inoperable carcinoma of the lung: results and implications for future treatment strategies. Int J Radiat Oncol Biol Phys 1992; 24(1):3–9. 7. Haffty BG, Goldberg NB, Gerstley J, et al. Results of radical radiation therapy in clinical stage I, technically operable nonsmall cell lung cancer. Int J Radiat Oncol Biol Phys 1988; 15(1): 69–73. 8. Kaskowitz L, Graham MV, Emami B, et al. Radiation therapy alone for stage I non-small cell lung cancer. Int J Radiat Oncol Biol Phys 1993; 27(3):517–523. 9. Sibley GS. Radiotherapy for patients with medically inoperable Stage I nonsmall cell lung carcinoma: smaller volumes and higher doses—a review. Cancer 1998; 82(3):433–438. 10. Krol AD, Aussems P, Noordijk EM, et al. Local irradiation alone for peripheral stage I lung cancer: could we omit the elective regional nodal irradiation? Int J Radiat Oncol Biol Phys 1996; 34(2):297–302.

11. Cheung PC, Mackillop WJ, Dixon P, et al. Involved-ﬁeld radiotherapy alone for early-stage non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 2000; 48(3):703–710. 12. Martel MK, Ten Haken RK, Hazuka MB, et al. Estimation of tumor control probability model parameters from 3-D dose distributions of non-small cell lung cancer patients. Lung Cancer 1999; 24(1):31–37. 13. Fowler JF, Tome WA, Fenwick JD, et al. A challenge to traditional radiation oncology. Int J Radiat Oncol Biol Phys 2004; 60(4):1241–1256. 14. Mehta M, Scrimger R, Mackie R, et al. A new approach to dose escalation in non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 2001; 49(1):23–33. 15. Fowler JF, Chappell R. Non-small cell lung tumors repopulate rapidly during radiation therapy. Int J Radiat Oncol Biol Phys 2000; 46(2):516–517. 16. Saunders M, Dische S, Barrett A, et al. Continuous hyperfractionated accelerated radiotherapy (CHART) versus conventional radiotherapy in non-small-cell lung cancer: a randomised multicentre trial. CHART Steering Committee. Lancet 1997; 350(9072):161–165. 17. Belani CP, Wang W, Johnson DH, et al. Phase III study of the Eastern Cooperative Oncology Group (ECOG 2597): induction chemotherapy followed by either standard thoracic radiotherapy or hyperfractionated accelerated radiotherapy for patients with unresectable stage IIIA and B non-small-cell lung cancer. J Clin Oncol 2005; 23(16):3760–3767. 18. Bentzen SM, Atasoy BM, Daley FM, et al. Epidermal growth factor receptor expression in pretreatment biopsies from head and neck squamous cell carcinoma as a predictive factor for a beneﬁt from accelerated radiation therapy in a randomized controlled trial. J Clin Oncol 2005; 23(24):5560–5567. 19. Kavanagh BD, Timmerman RD. Stereotactic Body Radiation Therapy. Philadelphia: Lippincott Williams & Wilkins, 2005. 20. Wulf J, Hadinger U, Oppitz U, et al. Stereotactic radiotherapy of extracranial targets: CT-simulation and accuracy of treatment in the stereotactic body frame. Radiother Oncol 2000; 57(2):225–236. 21. Wulf J, Hadinger U, Oppitz U, et al. Impact of target reproducibility on tumor dose in stereotactic radiotherapy of targets in the lung and liver. Radiother Oncol 2003; 66(2):141–150. 22. Fowler JF, Sheldon PW, Denekamp J, et al. Optimum fractionation of the C3H mouse mammary carcinoma using x-rays, the hypoxic-cell radiosensitizer Ro-07-0582, or fast neutrons. Int J Radiat Oncol Biol Phys 1976; 1(7–8):579–592.

642

g.w. wong et al.

23. 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(6):861–870. 24. Blomgren H, Lax I, Goranson H, et al. Radiosurgery for tumors in the body: clinical experience using a new method. J Radiosurg 1998; 1(1):63. 25. Wulf J, Hadinger U, Oppitz U, et al. Stereotactic radiotherapy of targets in the lung and liver. Strahlenther Onkol 2001; 177(12):645–655. 26. Herfarth KK, Debus J, Lohr F, et al. Stereotactic single-dose radiation therapy of liver tumors: results of a phase I/II trial. J Clin Oncol 2001; 19(1):164–170. 27. Hof H, Herfarth KK, Munter M, et al. Stereotactic single-dose radiotherapy of stage I non-small-cell lung cancer (NSCLC). Int J Radiat Oncol Biol Phys 2003; 56(2):335–341. 28. Uematsu M, Shioda A, Suda A, et al. Computed tomographyguided frameless stereotactic radiotherapy for stage I non-small cell lung cancer: a 5-year experience. Int J Radiat Oncol Biol Phys 2001; 51(3):666–670. 29. Uematsu M, Shioda A, Taira H, et al. Computed tomography (CT)-guided stereotactic radiation therapy (SRT) for stage I nonsmall cell lung cancer (NSCLC): 8-year results of 50 initial patients. Int J Radiat Oncol Biol Phys 2003; 57(2 Suppl 1):S281. 30. Nagata Y, Negoro Y, Aoki T, et al. Clinical outcomes of 3D conformal hypofractionated single high-dose radiotherapy for one or two lung tumors using a stereotactic body frame. Int J Radiat Oncol Biol Phys 2002; 52(4):1041–1046.

31. Nagata Y, Takayama K, Aoki T, et al. Clinical outcome of 3-D conformal hypofractionated high-dose radiotherapy for primary and secondary lung cancer using a stereotactic technique. Int J Radiat Oncol Biol Phys 2003; 57(2 Suppl 1):S280. 32. Fukumoto S, Shirato H, Shimzu S, et al. Small-volume imageguided radiotherapy using hypofractionated, coplanar, and noncoplanar multiple ﬁelds for patients with inoperable stage I nonsmall cell lung carcinomas. Cancer 2002; 95(7):1546–1553. 33. Timmerman R, Papiez L, McGarry R, et al. Extracranial stereotactic radioablation: results of a phase I study in medically inoperable stage I non-small cell lung cancer. Chest 2003; 124(5):1946–1955. 34. Timmerman RD, Papiez L, McGarry R, et al. Extracranial stereotactic radioablation: results of a phase I study in stage I non-small cell lung cancer. Int J Radiat Oncol Biol Phys 2003; 57(2 Suppl 1): S280. 35. Onishi H, Araki T, Shirato H, et al. Stereotactic hypofractionated high-dose irradiation for stage I nonsmall cell lung carcinoma: clinical outcomes in 245 subjects in a Japanese multiinstitutional study. Cancer 2004; 101(7):1623–1631. 36. Wulf J, Hadinger U, Oppitz U, et al. Stereotactic boost irradiation for targets in the abdomen and pelvis. Radiother Oncol 2004; 70(1):31–36. 37. Wada H, Takai Y, Nemoto K, et al. Univariate analysis of factors correlated with tumor control probability of three-dimensional conformal hypofractionated high-dose radiotherapy for small pulmonary or hepatic tumors. Int J Radiat Oncol Biol Phys 2004; 58(4):1114–1120.

6 4

Stereotactic Body Radiation Therapy: Brachytherapy Perspective Caroline L. Holloway, Desmond O’Farrell, and Phillip M. Devlin

Introduction Brachytherapy was conceptualized in the early 1900s when Pierre Curie suggested that a small radium tube could be inserted directly into cancer. Experiments carried out at that time showed tumor shrinkage in response to the directed radiation. The ﬁrst therapeutic brachytherapy treatments used surface molds to deliver low dose rate (LDR) radiation to skin cancers, and by 1903 the ﬁrst intracavitary radium treatment for cervical cancer was delivered. The use of brachytherapy waned over the 20th century with the advent of external beam radiation therapy and the development of the linear accelerator. However, brachytherapy continued to be integral in the treatment of many Gynecologic cancers. Developments in image guidance, high dose rate (HDR) and pulsed dose rate (PDR) after-loading machines, and the ability to computer-optimize treatments have again made brachytherapy a prominent component of radiation oncology departments. Brachytherapy is used for curative-intent treatments in sarcomas, Gynecologic, prostate, and head and neck malignancies, partial breast irradiation, and can be used with surface applicators for skin cancer. It is also part of the management of cancer in the lung in both curative and palliative settings. Brachytherapy can also be used in the palliative setting in rectal, esophageal, and hepatobiliary cancers. The similarities between stereotactic body radiation therapy (SBRT) and brachytherapy are surprising. Both modalities use image guidance to deliver conformal dose to the target with steep dose gradients. SBRT consists of a limited number of high-dose fractions with the most common SBRT fractionation schemes between one and ﬁve fractions at doses of 6 to 20 Gy per fraction. This is similar to many HDR brachytherapy fractionation schemes. Both SBRT and fractionated HDR brachytherapy can have similar treatment times and, therefore, similar dose rates and radiobiologic considerations. Both modalities present characteristic inhomogeneities and resultant hot spots within the planning target volume (PTV). SBRT can take advantage of intensity modulation of treatment beams to deliver a conformal dose and sparing of normal structures. Optimized

HDR and PDR brachytherapy allows manipulation of the dwell times of the radiation isotope to deliver conformal treatments, and LDR can take advantage of differing emitted energies of therapeutic isotopes to treat the target and spare organs at risk (OAR). The two modalities differ in their method of delivery, as SBRT is a relatively noninvasive technique, whereas most brachytherapy procedures are invasive. Within this chapter, we would like to compare the disease sites currently under study with SBRT (as described in the preceding chapter) to brachytherapy and highlight some novel therapeutic brachytherapy options. As well, we will brieﬂy discuss the issues of patient selection, cost-effectiveness, and quality of life for both brachytherapy and SBRT.

Disease Sites Lung Cancer The described role for SBRT in lung cancers is for early-stage, small-volume peripheral tumors in nonsurgical candidates. This curative intent treatment involves signiﬁcant treatment planning time and relies on the accurate deﬁnition of the target and the ability to avoid geographic misses. Recent advances in targeting treatment volumes through 4D planning, breath hold techniques, and tumor tracking technology aids in this endeavor. The fundamental principle in SBRT is to provide a conformal prescription dose to the target with rapid fall-off of dose. To achieve this, multiple beams are generally used. It is therefore important to understand the integral dose delivered to the surrounding normal tissues. Currently, radiation pneumonitis in patients treated with SBRT is rare because of the small size of the tumors targeted, but as studies look to include larger-sized tumors, the integral dose to the lung must be monitored closely. Studies using standard fractionation of external beam radiation therapy (EBRT) have shown the risk of pneumonitis to relate to the mean lung dose (MLD) as well as the volume of lung receiving 30 Gy (V30) and V20 [1]. Recent studies suggest that V05 to V13 should also be evaluated in the ipsilateral lung [2].

643

644

c.l. holloway et al.

Brachytherapy can be used in either the curative or palliative setting for lung cancer. Permanent radioactive implantation for locally advanced disease with close or microscopically positive margins is practiced in some centers. A few centers have reported high feasibility and excellent local in-ﬁeld control rates with this technique [3–6]. A ﬂexible mesh-based suture seed is made to conform to irregular targets and can deliver an additional radiation dose not feasible with conventional EBRT. The technique employed at this institution is depicted in the accompanying ﬁgures. The area at risk is measured intraoperatively at the time of surgery and 1 cm is added in all directions for the margin. A custom Vicryl mesh is cut generously to this measurement. Parallel lines at 1 cm are drawn on the mesh to guide the placement of the suture seeds. Vicryl suture with I125 seeds is sewn into the mesh and ﬁxed with small surgical clips (Fig. 64-1). This mesh is taken into the operative bed and sutured to the area at risk (Fig. 64-2). There are many variables that need to be considered in using this technique including the previous external beam dose and the radiation tolerances of the normal structures in the vicinity of the implant. The isodose geometry will vary as a function of the source activity, the size of the implant, and the spacing of the seeds. Irregular surfaces, curvatures, and absorbing potential of tissue surrounding the implant will also affect the dose gradients (Fig. 64-3). The toxicities associated with this procedure relate to the normal structures that are in the area of the implant. Stewart et al. [7] reported two cases of esophageal ﬁstula after partial esophageal wall resection and brachytherapy in the setting of recurrent carcinoid tumors in previously irradiated ﬁelds. One of the disadvantages of this procedure is that the radiation oncologist, physicist and radiation safety ofﬁcer need to be on-call during the course of the open thoracotomy. Radiation safety education for the hospital staff, family, and patient is essential. In short, there is scant literature with some caution as to the limits of this therapy. Further investigation is required. Endoluminal brachytherapy is a simple palliative technique that has been shown to successfully relieve hemoptysis, dyspnea, and cough in patients with endobronchial disease. This relief has, in the palliative setting, correlated with an overall survival beneﬁt in some series [8]. This therapy is most often performed

FIGURE 64-1. Vicryl mesh with I125 suture seeds.

FIGURE 64-2. Placement of I125 seed mesh implant.

with high-dose rate after-loading fractionation [6, 9]. Although the placement of the catheter requires anesthesia or conscious sedation, it is most often well tolerated. In conjunction with direct visualization of the tumor at the time of catheter placement, the treatment planning is often performed with plain ﬁlm images characterizing the treatment length. The dose is prescribed to a reference radius, usually 10 mm (Fig. 64-4). For low performance status patients, a single fraction (8 to 10 Gy) can be delivered the same day and the catheter removed. However, for higher performance patients, multiple fractions or serial implants, 1 or 2 weeks apart, can be prescribed. The most common fractionation pattern for fractionated therapy is 5 Gy × 3, given twice daily over 1½ days. Disadvantages of this therapy are the limit to the depth that the dose can be thrown. If one attempts to exceed 1 cm, the endoluminal mucosal dose becomes intolerable and can lead to ulceration and hemoptysis. The catheter is not usually centered within the lumen, and the eccentric dose in relationship to the lumen can allow for a very high endobronchial mucosal dose. There are case reports of

FIGURE 64-3. Isodose distribution of I125 mesh placed on the superior vena cava.

64.

stereotactic body radiation therapy: brachytherapy perspective

FIGURE 64-4. Endobronchial isodose distribution; 100% isodose at 1-cm reference radius.

endoluminal direct freehand seed placement for bronchial stump tumors. However, homogeneity of dose is difﬁcult to achieve with this technique and is not recommended.

Liver Tumors

645

This approach selectively targets the ionizing radiation to the tumors within the liver by microembolization of exit vessels from the tumor site, followed by infusion of the sphere slurry into the hepatic artery [11–13]. Average tumor doses in excess of 200 Gy are delivered [12]. A phase III trial of chemotherapy plus SIR-Spheres showed signiﬁcant partial and complete response rates in the experimental group. The median time to disease progression in the liver was signiﬁcantly longer in the microsphere arm: 15.9 versus 9.7 months (p = 0.001). There was no increase in grades III to IV treatment-related toxicity and no loss of quality of life for patients [12]. 99mTc-MAA scintigraphy is required pretreatment in order to exclude patients with signiﬁcant lung shunting as radiation pneumonitis is a serious potential complication of this therapy [14]. This novel approach relies heavily on the skills of the interventional radiologist and in the United States also needs to involve the radiation oncologist, as these spheres have been characterized as brachytherapy devices by regulatory bodies. The role of the radiation oncologist as the authorized user of this isotope, similar to permanent radioactive seed implantation, requires the brachytherapy intraoperative team to be on-call and ready for procedures. This methodology involves the risk of a radioactive spill. In this regard, the isotope is much more like iodine-131 colloid and strontium.

Bile Duct Tumors Inoperable bile duct cancer remains a therapeutic challenge for radiation therapy. These tumors, though small, quickly lead to obstructive symptoms and patient decompensation. Palliative bile duct brachytherapy has relieved obstruction temporarily. Recent reports of more aggressive approaches combine conventional external beam radiotherapy, brachytherapy, and concurrent chemotherapy. The goal is to deliver a deﬁnitive dose while respecting the relatively low radiation tolerance of the liver. A study from Italy reviewed 22 patients with either unresectable or residual extrahepatic bile duct tumors [10]. The patients received between 39.6 and 50.4 Gy with EBRT in conjunction with 5-ﬂuorouracil infusional chemotherapy. Twelve of the patients received a boost using intraluminal LDR I192 wires. Local control was 44.5 months, and overall survival was 23 months. Two of the 12 patients who received brachytherapy developed late toxicity in the form of duodenal ulcerations. Although this trial shows protraction of the median survival, the speciﬁc role of brachytherapy is questioned. Disadvantages of biliary brachytherapy are similar to disadvantages of endobronchial brachytherapy. Biliary scarring or necrosis could be the result of overextending the prescription distance from the isotope, either in high- or low-dose-rate brachytherapy. One must also always take into account the radiation tolerance of nearby normal structures.

Liver Parenchyma SBRT for hepatocellular carcinoma and liver metastasis shows encouraging response rates and some long-term survivors. Dose-volume tolerance of critical normal tissues though has not yet been well characterized. A novel brachytherapy approach for solid tumors is the use of selective internal radiation therapy (SIRT). This technique in primary and metastatic solid tumors in the liver has been studied with interarterial radioactive yttrium-90 microspheres.

Pancreas The literature for SBRT in localized pancreatic cancer has been presented in the preceding chapter. There is need for further studies to understand the role for SBRT in this disease site and the toxicities associated with high dose per fraction treatment. The same is true for the use of brachytherapy. Local control reported in case series and retrospective reviews is ∼70% when brachytherapy either in the form of intraoperative radiation (IORT), seed implantation, or HDR endoluminal catheters are used in conjunction with EBRT [15, 16]. Research on intratumoral injection of a phosphorus-32 silicon particle is being studied on human pancreatic carcinoma xenographs in nude mice [17] and may provide another form of SIRT in the future.

Patient Selection, Quality of Life, and Cost-Effectiveness In choosing patients appropriate for brachytherapy or SBRT tumor, patient and treatment factors must all be considered. The performance status of the patient including the ability to tolerate operating room (OR) procedures and anesthesia or conscious sedation is important when considering brachytherapy. In SBRT, the patients must be compliant and remain in the treatment position for variable lengths of time. The size and location of the tumor in question must also be considered. Brachytherapy in most sites is limited to small volumes of disease that can be easily accessed by percutaneous, endoluminal, or open techniques. SBRT has the ability to treat larger tumors and tumors in areas that brachytherapy may not be able to easily access without major surgical procedures. Both brachytherapy and SBRT offer short treatment courses that are favored by many patients.

646

c.l. holloway et al.

The effect on quality of life for SBRT treatments is not well-known. This technique is novel and even large studies, such as RTOG 0236 in lung cancer, are still looking at toxicity and effectiveness rather than quality of life measurements. The cost of radiation programs involving SBRT or brachytherapy is difﬁcult to deﬁne. Both SBRT and brachytherapy involve a team of radiation oncologists, therapists, and physicists. Both also rely on imaging modalities such as computed tomography (CT) and magnetic resonance imaging (MRI) for deﬁnition of volumes and treatment planning. Brachytherapy is given in conjunction with EBRT or surgery in many cases, therefore, one needs to include surgical OR time and EBRT planning and treatment machines in a cost estimate. Unique to brachytherapy are the different applicators required for treatment delivery and the purchase of isotopes. In addition, software to plan both HDR and LDR brachytherapy differs from EBRT planning software. SBRT involves a short treatment course, however, SBRT requires many hours of physician, therapist, and physicist time for contouring, planning, quality assurance, and calibration. Whether the shorter treatment course will be cost-effective for a department given the increase in time required for planning and quality assurance is yet to be seen.

3.

4.

5.

6.

7.

8.

9. 10.

Conclusion SBRT is a new radiation technology that offers the delivery of conformal radiation treatments in fraction sizes comparable with HDR brachytherapy. Patients who can be treated with this technology beneﬁt from a short outpatient treatment technique that in the majority of cases is noninvasive. The preceding chapter has highlighted where this technique has been studied in both the curative and palliative settings and has also highlighted the potential role for this technique to treat larger tumor volumes than can currently be targeted with brachytherapy techniques. Both brachytherapy and SBRT endeavor to provide the same beneﬁt to the patient in terms of local control and potential cure of their cancers with minimal toxicity and have the added beneﬁt of being able to deliver high doses per fraction and hence shorten the overall treatment time.

References 1. Kim TH, Cho KH, Pyo HR, et al. Dose-volumetric parameters for predicting severe radiation pneumonitis after three-dimensional conformal radiation therapy for lung cancer. Radiology 2005; 235:208–215. 2. Yorke ED, Jackson A, Rosenzweig KE, et al. Correlation of dosimetric factors and radiation pneumonitis for non-small-cell lung

11.

12.

13.

14.

15.

16.

17.

cancer patients in a recently completed dose escalation study. Int J Radiat Oncol Biol Phys 2005; 63:672–682. Voynov G, Heron DE, Lin CJ, et al. Intraoperative 125I Vicryl mesh brachytherapy after sublobar resection for high-risk stage I nonsmall cell lung cancer. Brachytherapy 2005; 4:278. Mutyala S, Stewart AJ, Cormack R, et al. Toxicity of permanent I-125 interstitial planar seed brachytherapy for close or positive margins for thoracic malignancies. Int J Radiat Oncol Biol Phys 2005; 63:S392–S393. Chen A, Galloway M, Landreneau R, et al. Intraoperative 125I brachytherapy for high-risk stage I non-small cell lung carcinoma. [see comment]. Int J Radiat Oncol Biol Phys 1999; 44:1057– 1063. Raben A, Mychalczak B. Brachytherapy for non-small cell lung cancer and selected neoplasms of the chest. Chest 1997; 112 (4 Suppl):2765–2865. Stewart AJ, O’Farrell DA, Mutyala S, et al. Severe toxicity after permanent radioactive seed implantation for mediastinal carcinoid tumor. Brachytherapy 2007; 6(1):58–61. Kelly JF, Delclos ME, Morice RC, et al. High-dose-rate endobronchial brachytherapy effectively palliates symptoms due to airway tumors: the 10-year M. D. Anderson cancer center experience. Int J Radiat Oncol Biol Phys 2000; 48:697–702. Nag S, Kelly JF, Horton JL, et al. Brachytherapy for carcinoma of the lung. Oncology 2001; 15:371–381. Deodato F, Clemente G, Mattiucci GC, et al. Chemoradiation and brachytherapy in biliary tract carcinoma: Long-term results. Int J Radiat Oncol Biol Phys 2006; 64:483–488. Kennedy AS, Nutting C, Coldwell D, et al. Pathologic response and microdosimetry of (90)Y microspheres in man: review of four explanted whole livers. Int J Radiat Oncol Biol Phys 2004; 60: 1552–1563. Stubbs RS, Cannan RJ, Mitchell AW. Selective internal radiation therapy with 90yttrium microspheres for extensive colorectal liver metastases. J Gastrointest Surg 2001; 5:294–302. Van Hazel G, Blackwell A, Anderson J, et al. Randomised phase 2 trial of SIR-Spheres plus ﬂuorouracil/leucovorin chemotherapy versus ﬂuorouracil/leucovorin chemotherapy alone in advanced colorectal cancer. J Surg Oncol 2004; 88:78–85. Lambert B, Van de Wiele C. Treatment of hepatocellular carcinoma by means of radiopharmaceuticals. Eur J Nucl Med Mol Imaging 2005; 32:980–989. Schuricht AL, Spitz F, Barbot D, et al. Intraoperative radiotherapy in the combined-modality management of pancreatic cancer. Am Surg 1998; 64:1043–1049. Pfreundner L, Baier K, Schwab F, et al. [3D-Ct-planned interstitial HDR brachytherapy + percutaneous irradiation and chemotherapy in inoperable pancreatic carcinoma. Methods and clinical outcome]. Strahlenther Onkol 1998; 174:133–141. Zhang K, Loong SL, Connor S, et al. Complete tumor response following intratumoral 32P BioSilicon on human hepatocellular and pancreatic carcinoma xenografts in nude mice. Clin Cancer Res 2005; 11:7532–7537.

PA R T V

Patient Care and Socioeconomic Issues

647

6 5

Complications and Management in Radiosurgery Isaac Yang, Penny K. Sneed, David A. Larson, and Michael W. McDermott

Introduction Radiosurgery is the precise application of focused radiation to a targeted volume within the brain or spine, in one to ﬁve sessions, identiﬁed on magnetic resonance (MR) or computed tomography (CT) imaging using multiple beams to deposit a large dose to a deﬁned region while limiting the exposure of normal tissue [1–6]. This new deﬁnition of radiosurgery, involving one to ﬁve treatment sessions for the treatment of brain or spinal lesions, was recently approved by the American Society of Therapeutic Radiation Oncology and the American Association of Neurological Surgeons and Congress of Neurologic Surgeons. Although initially conceptualized by Leksell for use in functional neurosurgery, radiosurgery has progressively widened its scope and is now also an option for numerous neoplastic and vascular indications [7, 8]. Differing from standard fractionated radiotherapy (6 to 30 sessions), radiosurgery requires the precise delivery of radiation to smaller volumes without affecting large portions of normal parenchyma, thus allowing for a powerful radiobiologic effect on the targeted volume [2, 9–11]. Radiosurgery is an option when open surgical procedures are prohibited by excessively high risks or when surgical resection is incomplete. In addition, radiosurgery may be an attractive alternative for elderly patients or patients with other severe comorbidities such as respiratory or cardiac dysfunction, which may increase risk for an open procedure [3, 12, 13]. Even without comorbidities, open surgical procedures have inherent risks including damage to the brain (either directly or indirectly by injury to important blood vessels), bleeding (which can require reoperation), blood loss (which can require transfusion), infection, and general anesthetic risks. Furthermore, open procedures require several days of care in the hospital including at least one night in an intensive care unit, which contribute to the economic costs [7, 14]. That having been said, there are certain situations when radiosurgery may not be the best treatment, such as when a patient is symptomatic from a large tumor or when there is abundant surrounding edema. In addition, there may be times when the target deﬁnition may be difﬁcult based solely on imaging and a broader area of coverage

may be required to avoid missing residual tumor and so a standard fraction scheme using three-dimensional conformal radiotherapy techniques may be more appropriate. Radiosurgery can be delivered using several techniques [3]. Gamma Knife and linear accelerator (linac)-based radiosurgery are the most commonly used methods. Gamma Knife utilizes 201 cobalt-60 sources arranged geometrically to deliver a precise dose, whereas linac-based systems use multiple beam angles, arcs, and apertures to achieve the precise application of radiation [3]. Side effects from radiosurgery are dependent on location of treatment, dose of radiation delivered, volume of treatment, and proximity to critical neurologic structures [3, 15]. Larger target volumes have also been associated with higher rates of radiation toxicity [16]. Comorbidities such as diabetes mellitus, hypertension, or immunosuppression may increase the risk of complications from radiation therapy; whether they increase the risks of radiosurgery is not yet well understood.

Deﬁnitions of Radiation Toxicity Acute Complications of Radiosurgery Acute complications of radiosurgery can be deﬁned as those sequelae that occur within hours to days of the radiosurgery treatment. Because radiosurgery is noninvasive, it is not typically accompanied by traditional immediate complications associated with open surgical procedures for intracranial lesions [14]. However, those systems that require stereotactic frame application may be associated with pin site pain, bleeding, swelling, headaches, and, rarely, infection. Chin et al. report that immediate complications of radiosurgery are quite uncommon [14]. The most common immediate sequelae due to radiosurgery are the onset of seizures often due to subtherapeutic levels of blood anticonvulsant [14]. Tago et al. recently reported a case of acute facial neuropathy 2 days after radiosurgery [17]. Acute cranial neuropathy in the ﬁrst several days is uncommon but has been reported [12, 14].

649

650

i. yang et al.

Early-Delayed Effects of Radiation Toxicity Early-delayed radiation toxicities can be deﬁned as sequelae and complications due to radiosurgery within weeks to months after treatment [3, 14, 18]. These acute (early-delayed) complications include headache, local erythema of the skin, alopecia of in-ﬁeld treatment areas, and fatigue [12, 19]. Skin and hair changes are common with larger ﬁeld fractionated treatment, and they also may occur for superﬁcial lesions treated with radiosurgery although the area affected is smaller. Most earlydelayed complications are mild, and severe acute neurologic events are uncommon [3, 14, 18, 20, 21]. Occasionally, cerebral edema can be seen, with the potential to increase intracranial pressure, resulting in headaches, nausea and vomiting, and possible neurologic deﬁcits [3, 14, 15]. If the tumor volume treated is near the middle or outer ear, acute internal or external otitis media may result from treatment [15]. Seizures and vertigo are occasional mild side effects from radiosurgery [3, 20]. In a series of 47 patients undergoing linac radiosurgery, Gelblum et al. reported 15% of patients experiencing acute seizure within the ﬁrst few months after radiosurgery [22]. Unlike the immediate complications, most of these seizures could not be attributed to subtherapeutic anticonvulsant levels but may be sequelae of early-delayed radiation edema and/or necrosis. Hong et al. reported a series of 279 consecutive radiosurgery cases treated with a linac radiosurgery at the University of Wisconsin. In this large study, complications occurred in 34% of patients [3]. The most common side effects were mild, including headache, seizures, and nausea. Severe complications such as cerebrospinal ﬂuid (CSF) leak, aphasia, or hemiparesis only occurred in 2% of the cases [3]. Recently, Majhail et al. prospectively evaluated the early complications after Gamma Knife radiosurgery [18]. In this series of 79 patients, approximately 25% experienced acute complications, which were mild and self-limiting. No acute severe complications were reported in this series [18]. Werner-Wasik et al. in series of 78 patients treated with linac radiosurgery also report a similar rate (35%) of cases with immediate acute side effects with the vast majority of these being mild [23]. In evaluating severe acute complications within 1 week of radiosurgery, Chin et al. report a series of 835 consecutive radiosurgery cases with a severe acute complication rate of 2.2% using Gamma Knife radiosurgery [14]. The most common complication within 1 week of treatment in this series was seizures (1.4%). Five (0.6%) patients experienced a new focal deﬁcit, and three (0.4%) cases were deaths. Most of the complications were mild, and the focal deﬁcits improved in all affected patients in this series [14]. The Radiation Therapy Oncology Group (RTOG) in a series of 102 patients treated with Gamma Knife or linac radiosurgery also report a similar rate of severe acute complications of 2% within the ﬁrst 2 months after treatment [16]. Both of these reported cases presented as unacceptable edema, roughly 2 months after radiosurgery with one case resulting in acute (early-delayed) mortality. This group also reports that acute toxicity due to radiosurgery is rare and usually mild, with most complications occurring after 2 months [16]. Acute complications can occur in roughly 25% to 34% of patients treated with radiosurgery. Most of the sequelae are

mild and self-limited. Severe acute complications are uncommon [3, 14–16, 18, 23–26]. Early risk of mild sequelae should be communicated to patients undergoing radiosurgery procedures.

Delayed (Late-Delayed) Effects of Radiation Toxicity The majority of serious complications due to radiosurgery occur as late-delayed effects of radiation after several months to years after radiosurgery [14, 16]. Delayed complications of radiosurgery include headache, nausea, hydrocephalus, radiation necrosis, seizures, facial spasms, hemiparesis, cyst formation, and cranial nerve deﬁcits [16]. Most delayed complications can be attributed to radiation necrosis within the target volume or late cranial nerve damage [20]. If edema secondary to late-delayed radiation effects cannot be ameliorated with corticosteroids, craniotomy for decompression, tumor resection, or shunting may be required [27]. Most cranial neuropathies after radiosurgery are mild and transient and commonly present as late-delayed complications. The mean latency of trigeminal and facial neuropathies after radiosurgery for acoustic schwannoma has been noted to be roughly 7 months [28]. Ito et al. report a similar latency of 4 to 5 months for the onset of cranial neuropathies [29]. Lee et al. report that most complications after radiosurgery for meningiomas occur in a late-delayed fashion of 5 months to years after treatment [30]. In a series of 183 consecutive patients treated with Gamma Knife radiosurgery, Park et al. reported an 18% rate of complications due to radiosurgery [31]. These complications include radiation necrosis, hydrocephalus, and cranial nerve deﬁcits. Most of these deﬁcits occurred 2 or more months after treatment, but the speciﬁc latency of these complications is not speciﬁed [14, 31]. Radiation necrosis may become symptomatic requiring open surgical decompression. The incidence of symptomatic radiation necrosis is low, usually less than 6% [16, 32, 33]. Although the long-term complications of radiosurgery are not yet fully characterized, it appears that these risks are minimal [34, 35]. There are reported cases of radiosurgery associated “radiation-induced” malignancies, but these reported cases are extremely rare [11, 34–40]. Much longer periods of follow-up must be investigated to fully appreciate the possible long-term complications, as the development of new radiationinduced neoplasms may require decades to develop. A conservative estimate suggests this rate may be 1% to 2% at 20 years after treatment. In most cases, radiosurgery has proved to be safe and has displayed a low rate of delayed, acute, and even lower rate of immediate complications, and the vast majority of complications are mild and self-limited.

Complications for Common Radiosurgery Indications Brain Metastases Metastases to the brain represent the most common indication for radiosurgery [3, 6]. Brain metastases are one of the most

65.

complications and management in radiosurgery

common sites for metastatic spread of solid tumors, most commonly in lung, breast, and melanoma skin cancer [16, 32]. Radiosurgery has shown an actuarial 2-year local control rate of approximately 67% [33]. In a series evaluating radiosurgery for brain metastasis, Shaw et al. report that the most common complication was the symptomatic cerebral edema associated with radiosurgery [16]. Four patients in this series experienced severe cerebral edema, and half of these cases responded to corticosteroids [16]. Six percent of patients in this series required operation for symptomatic radiation necrosis within 1 year after therapy. These investigators also note that larger target volume >8200 mm3 was signiﬁcantly associated with increased acute and delayed complications due to radiosurgery [16]. In a different series of 40 patients with intracranial metastasis, Mehta et al. report a 10% radiation toxicity rate associated with radiosurgery [27]. Symptomatic cerebral edema is the most commonly noted complication of radiosurgery for intracranial metastasis with a range of 3% to 18% [16, 27, 32, 33, 41]. Typically, these complications occur over a broad time period in a late-delayed fashion over several months to years. Whenever patients become symptomatic with headache, seizures, or progressive neurologic deﬁcit, the physician should ﬁrst consider the time interval since radiosurgery to predict the most likely cause. Apart from radiation side effects, tumor recurrence and growth are also to be considered. Acute side effects in the ﬁrst few hours to days after treatment can be treated symptomatically for complaints of pain, nausea, and fatigue. The occurrence of seizures is rare, but anticonvulsant levels should be checked and then an additional half loading dose should be given, orally or intravenously, and the maintenance dose increased pending the results. Follow-up drug levels should be repeated in 7 to 10 days time. If the time interval from radiosurgery is weeks to months, then edema, necrosis, tumor recurrence, and/or hemorrhage are causes to be considered. For acute deterioration, a hospital-based evaluation ﬁrst with CT scanning where appropriate (e.g., melanoma, renal cell, uterine carcinoma metastases) followed by triple dose MR may be necessary. If the symptom complex is mild, then outpatient treatment usually begins with restarting steroids or increasing the dose. If there is no improvement within several days of increasing the dose, then repeat imaging is necessary. Of course MR alone will not be able to fully differentiate radiation necrosis from recurrent tumor, although qualitative changes such as reduced volume of enhancement, a shrunken and irregular margin of the metastasis, and a central ground-glass appearance to the tumor may provide a clue. Currently, we use MR perfusion studies to help make this distinction although MR spectroscopy, positron emission tomography, and thallium single photon emission computed tomography (SPECT) can also be used. For those patients in whom the imaging diagnosis is radiation necrosis and who remain symptomatic, then continued steroids or open operation for removal of the necrotic mass are options (Fig. 65-1). In those patients who recover neurologically, but in whom imaging changes persist, we have used a combination of pentoxifylline 400 mg by mouth three times daily (t.i.d.), vitamin C 500 mg t.i.d., and vitamin E 400 units t.i.d. (“PCE”) for several months as a trial to ward off the need to return to steroid medication.

651

FIGURE 65-1. (A) Patient with prior resection of right parietal occipital renal cell metastasis treated with radiosurgery and operated for necrosis. Now with similar problem on the left, 12 months after radiosurgery. (B) Fluid attenuated inversion recovery (FLAIR) sequence showing perilesional edema unresponsive to long-term steroid use. (C) Postoperative T1-weighted MR image showing resection of necrotic tissue. No viable tumor on pathology. (D) Postoperative FLAIR sequence showing early reduction of edema after surgical removal. Patient remains off steroids with intact visual ﬁelds 18 months later.

Meningiomas Meningiomas are tumors originating from meningothelial cells accounting for approximately 20% of all primary brain tumors [1]. Radiosurgery is indicated as an alternative to surgery for skull base meningiomas or for meningiomas located with a difﬁcult surgical challenge for complete resection with unacceptable risks of neurologic deﬁcits [30, 42–47]. These are most commonly in the petroclival, parasellar, and cavernous sinus regions. It is also an option for elderly patients whose other comorbid conditions may preclude surgical resection [15, 42, 44]. In some cases, radiosurgery may be pursued as an option without a tissue diagnosis [15]. Meningiomas less than 3.5 cm in size, farther than 2 to 5 mm from the optic apparatus, and with well-deﬁned borders may be appropriate for radiosurgical intervention [45]. Radiosurgery complications after treatment for meningiomas include peritumoral edema (which can become symptomatic), seizures, and cranial and optic neuropathies [46, 48, 49]. Late-delayed cranial neuropathies including optic nerve injury were the highest risk complications of skull base meningioma radiosurgery [30, 50]. Singh et al. report a series of 77 patients with meningiomas treated with Gamma Knife radiosurgery [51]. These investigators report complications such as seizure and headaches in 10% of patients. A signiﬁcant number (22%) of parasagittal meningiomas had edema with most of them becoming symptomatic. Caution should be used when radiosurgery is utilized for a parasagittal meningioma, avoiding

652

i. yang et al.

treatment of larger tumors or those associated with vasogenic edema [51, 52]. The optic apparatus is reported to be quite sensitive to radiation [53–55]. Tishler et al. in a series of 42 meningiomas treated with radiosurgery suggest maximum dose to the optic chiasm be less than 8 Gy [55]. In a different series of 50 patients treated with radiosurgery with greater than 3 years of follow-up, Leber et al. report that no patients treated with less than 10 Gy to the optic chiasm experienced optic neuropathy [54]. With doses between 10 and 15 Gy, 27% of patients had optic neuropathy complications, and when doses were greater than 15 Gy, the risk of optic neuropathy was 78% [15, 54]. Duma et al. also report optic nerve injuries in patients receiving more than 10.5 Gy to the optic apparatus [56]. Other cranial nerves appear to be more resistant to radiation than the optic apparatus, as doses up to 20 Gy may be tolerated [15]. It is suggested that the unique cerebrum embryologic etiology and hence myelinization of the second cranial nerve predisposes it to fragility [15]. Kondziolka et al. report a series of 50 patients treated for meningiomas with a mean marginal dose of 17 Gy [49]. In this series from the University of Pittsburgh, the actuarial 2-year tumor control rate was 96%. Three patients exhibited delayed radiosurgery complications of trigeminal hypesthesia, hemiparesis, and oculomotor neuropathy between 3 and 12 months after treatment. These complications gradually improved with time [57]. In a recent series of 200 skull base meningiomas with a median marginal tumor dose of 12 Gy, Kriel et al. report an actuarial progression-free rate of 98% at 5 years and a complication rate of 2.5% [58]. Most of these complications were transient symptomatic peritumoral edema, which was ameliorated with corticosteroids. The only permanent neurologic deﬁcit was optic neuropathy in one patient despite a maximum dose of less than 10 Gy to the optic apparatus [58]. Tishler et al. report a series of 42 cavernous sinus meningiomas treated with radiosurgery [55]. Most of the complications involved delayed radiosurgery associated transient neuropathies of cranial nerves III through VI, which were generally mild and improved with time. Conversely, four patients who received a maximum dose of greater than 8 Gy to the optic chiasm had visual complications. No patient receiving less than 8 Gy to the optic chiasm had optic neuropathies, and these investigators recommend a maximum dose of less than 8 Gy to the optic chiasm [55]. Morita et al. report a series of 88 skull base meningiomas treated with radiosurgery [59]. Progression-free survival at 5 years was 95% in this series. Despite maximum doses up to 16 Gy to the optic chiasm, no visual complications were reported. Trigeminal and other cranial neuropathies were reported in 14% of patients undergoing radiosurgery [59]. This series suggests that short segments of the optic apparatus (<12 mm) may tolerate up to 16 Gy maximum dose without optic neuropathy, but this remains to be conﬁrmed by others. Leber et al. report their series of 50 patients treated with radiosurgery with a mean follow-up of 40 months [54]. Approximately half of the patients in this series had meningiomas and were treated with an average marginal tumor dose of 14 Gy. Using neuroophthalmologic testing, optic neuropathy was detected in 23% of the cases, but no patients with

less than 10 Gy to the optic apparatus developed neuropathy [54]. Stafford et al. evaluated 218 patients with meningiomas treated with Gamma Knife radiosurgery. In this series, although a majority of patients received greater than 8 Gy to the optic apparatus, only 2% of patients had optic neuropathy [60]. The maximum dose tolerated by the optic apparatus is still unclear. Skull base meningiomas can be effectively treated with radiosurgery with a low complication risk [43, 52, 61–63]. Most series report high tumor control rates greater than 95% with a low risk of complications, typically less than 10% [43, 48, 54, 55, 57–59, 61, 62]. These complications typically involve cranial neuropathies, which are mild, except for injury to the optic apparatus [54, 55, 57–59]. Optic neuropathy may be a permanent late complication from radiosurgery for skull base meningiomas, but limiting the maximum radiation dose to the optic apparatus decreases the risk for optic neuropathy. The maximum tolerated dose of the optic apparatus is poorly deﬁned, but 8 Gy is the maximum limit recommended by several authors [15, 43, 53, 55, 60]. Other complications can be seen with radiosurgery for meningiomas. Most cranial neuropathies are mild and improve with time, but uncommon severe complications such as hemiparesis can develop after treatment [50, 63, 64]. Chang et al. report a series of 179 patients treated with Gamma Knife radiosurgery for meningiomas. In this series, 25% of patients had complications, mostly asymptomatic edema found on imaging, and 2% had cranial neuropathies and radiation necrosis [64]. Recently, Hakim et al. report their results with 127 patients treated with linac radiosurgery for meningiomas. These investigators report two deaths (mortality 1.6%) due to posttreatment edema and subsequent herniation. With a morbidity of 3%, they report blindness and hemiparesis as severe delayed complications after radiosurgery treatment [65]. In anther recent series, Shafron et al. report their experience with 70 patients with meningiomas treated with linac radiosurgery [66]. In this series, they report a 3% complication rate of edema and radiation necrosis. Half of these cases resolved with corticosteroids [66]. Subach et al. report their experience with 62 petroclival meningiomas [44]. In this series, 8% of patients had persistent cranial neuropathies, and there were no severe complications from radiosurgery. In a large series of 210 patients, Friedman et al. report a transient and permanent radiation complication rate of only 8% [62]. The complications are not speciﬁed in this study. Lee et al. report a series of 176 cavernous sinus meningiomas treated with radiosurgery [30]. In this series, 6% of patients experienced complications such as optic neuropathy, trigeminal neuropathy, or seizures. Iwai et al. recently described their 43 patients with cavernous sinus meningiomas treated with radiosurgery [67]. Only 5% had cranial nerve deterioration after treatment. DiBiase et al. recently report that in their series of 162 patients with meningiomas treated with radiosurgery, only 8% had radiation toxicities [68]. The most common complication was headache due to edema followed by seizures. Table 65-1 summarizes complications from some recent series [69–71]. Other late-delayed complications from radiosurgery for meningiomas include peritumoral edema and associated symp-

65.

653

complications and management in radiosurgery

TABLE 65-1. Summary of recent radiosurgery for intracranial meningiomas. First author

Year

No. of patients

Chang [69]

1997

55

Subach [44] Hakim [65] Shafron [66] Iwai [43] Kondziolka [70] Singh [51] Roche [46] Kobayashi [50] Shin [71] Stafford [61] Lee [30] Chang [64] Iwai [67] Pollack [49] DiBiase [68] Friedman [62]

1998 1998 1999 1999 1999 2000 2000 2001 2001 2001 2002 2003 2003 2003 2004 2005

62 127 70 24 99 77 92 99 40 190 176 179 43 330 162 210

Complication

Symptomatic radiation necrosis, cranial neuropathies, and hemiparesis Cranial neuropathies Death, blindness, hemiparesis Radiation necrosis and edema Cranial neuropathies Cranial neuropathy and hydrocephalus Symptomatic edema, seizures Cranial neuropathy, seizures Radiation necrosis, edema, and cranial neuropathy Cranial neuropathy Cranial neuropathy, radiation necrosis, and edema Cranial neuropathies and seizures Cranial neuropathies Cranial neuropathies Cranial neuropathy and symptomatic edema Edema, radiation necrosis, and hydrocephalus Transient and permanent radiation complications

tomatic hydrocephalus [45, 48–50, 52, 61, 63, 64, 72–74]. Kim et al. published a case report of late-delayed bleeding after radiosurgery for meningioma occurring 3 years after treatment [75]. Recently, Kwon et al. report four of 173 cases with delayed meningioma bleeding after radiosurgery occurring 1 to 8 years after treatment [76]. Although radiosurgery for meningiomas has high local control rates and has been shown to have low complications rates, parasagittal/falx or cavernous sinus locations appear to contribute most to reports of increased toxicity and complications. We have seen several older patients treated for large meningiomas with radiosurgery in whom edema and features of radiation necrosis have eventually required surgery for tumor removal (Fig. 65-2). The treating physicians must remember that it takes 18 to 24 months before these benign tumors start to shrink after radiosurgery, so a tumor that is symptomatic from mass effect is not an ideal case for radiosurgery. As well, those tumors associated with surrounding edema prior to treatment may be more likely to exhibit wors-

FIGURE 65-2. (A) T1-weighted MR image of right frontal parietal convexity meningioma 12 months after radiosurgery in a patient presenting from another institution with increasing left weakness and focal seizures despite weeks of steroid treatment. (B, C) FLAIR sequence

Incidence (%)

7 8 3 3 5 5 13 5 14 23 13 6 2 5 8 8 8

ening of the edema after radiosurgery than those tumors without preexisting edema. For those who develop seizures then, investigations and management are as those described above for metastases. For those patients who become symptomatic from edema, we tend to use short-term steroids for 2 to 3 weeks and observe the response. If imaging evidence of the T2 signal changes on MR persist along with lingering clinical symptoms, we try the antioxidant/perfusion regimen of pentoxifylline and vitamins C and E. If the edema remains symptomatic and there is contrast enhancement in the brain adjacent to tumor, then open operation for removal of both tumor and necrotic tumor may be necessary. Hyperbaric oxygen may also be of beneﬁt but has not been proved in a controlled study [77].

Acoustic Schwannoma Acoustic schwannoma are tumors derived from the Schwann cells of the vestibular branch of the eighth cranial nerve [78].

images showing extent of vasogenic edema as a complication of treatment. Patient underwent resection, and dissection of interface between tumor and brain was indistinct over a small area at site consistent with area of contrast enhancement in T1 images shown.

654

i. yang et al.

Microsurgery and stereotactic radiosurgery are the main current treatment options for this lesion [4, 6, 12, 13, 29, 79, 80–91]. After radiosurgery or radiation therapy for acoustic schwannomas, adjacent proximal neurologic structures are at risk for radiation toxicity. These structures include the brain stem as well as the trigeminal, facial, and vestibulocochlear nerves [4, 5, 13, 15, 29, 81, 82, 88, 92–96]. It has been suggested that the oligodendrocytes of the brain stem are more sensitive to radiation than the peripheral Schwann cells that myelinate the cranial nerves in the transitional zone and further distally [5, 15, 92]. It has been noted that acoustic schwannomas are at increased risk for mild acute complications versus other common indications for radiosurgery [23]. Miller et al. report a series of 82 patients with acoustic schwannomas treated with high-dose and low-dose radiosurgery [98]. This report suggests that a marginal tumor dose greater than 16 Gy is signiﬁcantly associated with permanent facial neuropathy. Median time for neuropathy onset was approximately 6 months. Although increased marginal tumor dose is an important risk factor for trigeminal and facial neuropathy, half of the patients in this series with neuropathies had symptomatic improvement over time [15, 97]. Other important factors for trigeminal and facial neuropathy after radiosurgery are the length of cranial nerve irradiated and maximum radiation dose to the brain stem. In a series of 149 patients with acoustic schwannomas treated with radiosurgery, Foote et al. report that increased length of trigeminal and facial nerve irradiated correlated with neuropathies [95]. Maximum brain stem dose greater than 17.5 Gy was found to be even more closely associated with trigeminal and facial neuropathies in this series. These investigators recommend 12.5 Gy to the acoustic schwannoma tumor margin as the optimal dose to balance efﬁcacy and risks of complications and that doses greater than 15 Gy at the tumor margin were signiﬁcantly associated with increased risk for cranial neuropathies [95]. Finally, in this series, Foote et al. report that prior surgical resection prior to radiation therapy was also an important risk factor for potential cranial neuropathies [95]. Current radiosurgery practice limits brain-stem maximum dose to 12 to 14 Gy for a small volume, generally right at the edge of the prescription isodose line [53, 81, 92, 93, 97–99]. This practice would also limit trigeminal and facial neuropathies as indicated by Foote et al. [95]. Furthermore, cranial nerve length irradiated may be in important predictor for cranial neuropathy risk [4, 29, 91, 92, 95, 100–103]. Andrews et al. report 69 patients treated with radiosurgery for acoustic schwannomas with 12 Gy at the tumor margins [104]. Facial and trigeminal neuropathies only occurred in 2% to 5% of treated patients with median follow-up of greater than 2 years. However, using audiometric data to assess hearing, 67% of the patients treated with radiosurgery lost functional hearing after treatment [104]. Rowe et al. report a series of 96 consecutive patients with acoustic neuromas treated with Gamma Knife radiosurgery [105]. In this series using audiograms to assess hearing function, 40% of treated cases experienced decreased hearing after treatment, whereas other complications such as persistent trigeminal and facial neuropathies were low, 2% and 5%, respectively [105]. Kobayashi et al. report a series of 44 cases of acoustic neuromas treated with Gamma Knife radiosurgery [106]. In this series,

16% and 7% had facial and trigeminal neuropathies, respectively. The vast majority of these cranial neuropathies resolved, and hearing was preserved in 48% of the patients [106]. Niranjan et al. also reported 29 patients treated with Gamma Knife radiosurgery [107] and suggested that decreasing the marginal dose to less than 14 Gy may improve hearing preservation. In this series, the actuarial hearing preservation rate was 65% [107]. No other cranial neuropathies were reported in this study [105]. Meijer et al. also report 37 patients who received linac radiosurgery for acoustic schwannomas [108]. With mean follow-up of greater than 2 years, trigeminal neuropathy was only seen in 3% of treated patients. Furthermore, 66% of the patients had functional hearing preservation [108]. Hearing was assessed by asking patients whether or not they could use the phone on the affected side and not with audiometrics [108]. The different methods of evaluating hearing preservation, using audiometrics or functional activities, may contribute to the discrepancies in reported hearing loss in these series [15]. Other less common late-delayed complications reported after radiosurgery for acoustic schwannomas include facial spasm and hydrocephalus [15, 88, 104, 108–112]. Shunting may be required to address the late hydrocephalus complication [24, 88, 93, 111, 112]. Prasad et al. reported a 5% complication rate of mild cranial neuropathies, most of which gradually improved. Among 153 patients treated with Gamma Knife radiosurgery for acoustic schwannomas, hearing deterioration was detected in 40% of cases, although audiometrics were not used in all patients in this retrospective study [113]. In another recent series, Flickinger et al. report 190 patients with acoustic schwannomas treated with Gamma Knife radiosurgery [102]. They report trigeminal and facial neuropathy in only 1% to 3% of cases and 71% hearing preservation assessed with audiometrics. These results may be due to improved procedures and technology, as this Pittsburgh group reduced marginal tumor dose to 13 Gy with effective actuarial tumor control rate still high at 97% [102]. Paek et al. report a series of 25 patients with acoustic schwannomas treated with Gamma Knife radiosurgery [114]. They report a trigeminal neuropathy risk of 5% and hearing preservation of 55% [114]. This is similar to the most recent series reported by Combs et al. indicating that trigeminal and facial neuropathies were a risk in only 5% to 8% of cases [115]. Hearing preservation was reported at 55%. The complication risk from radiosurgery using doses of 12 to 13 Gy for acoustic schwannomas is generally quite low [5, 15, 81, 85, 86, 90, 93, 96, 104, 108, 110, 112, 113, 116–119]. Most modern series report trigeminal and facial neuropathy risks at less than 8%, and in most cases these neuropathies improve with time [15, 92, 94, 95, 105, 113, 115, 117–119]. Because the number of studies that rigorously assess hearing preservation after radiosurgery is quite small with limited sample sizes and the various methods of assessing hearing loss have not been standardized, the hearing preservation rate after radiosurgery is not well deﬁned [15, 103, 120]. The rates of hearing preservation vary widely in the literature between 11% and 77%, with most recent reports in the 50% to 70% range [5, 12, 29, 79, 82, 94, 103, 104, 107, 108, 112, 114–118, 120–125]. Table 65-2 summarizes some recent series. Although

65.

655

complications and management in radiosurgery

TABLE 65-2. Summary of recent radiosurgery for acoustic schwannomas. First author

Year

No. of patients

Kondziolka [112] Spiegelmann [12] Prasad [113] Ito [29] Meijer [108] Flickinger [102] Spiegelmann [94] Foote [95] Petit [119] Andrews [104] Ottaviani [79] Rowe [118] Iwai [125] Paek [114] Myrseth [96] Combs [115]

1998 1999 2000 2000 2000 2001 2001 2001 2001 2001 2002 2003 2003 2005 2005 2006

402 44 152 138 37 190 40 149 47 69 30 234 51 25 103 26

Complication

Incidence (%)

Trigeminal and facial neuropathy Trigeminal and facial neuropathy Trigeminal and facial neuropathy Trigeminal and facial neuropathy Trigeminal neuropathy Trigeminal and facial neuropathy Trigeminal and facial neuropathy Trigeminal and facial neuropathy Facial neuropathy Trigeminal and facial neuropathy Trigeminal neuropathy Trigeminal and facial neuropathy Trigeminal neuropathy Trigeminal neuropathy Facial neuropathy Trigeminal and facial neuropathy

unclear, it appears that radiosurgery for acoustic schwannomas is accompanied with some risk of hearing deterioration. For patients who report a sudden decrease in hearing after treatment, we immediately prescribe a 10 day course of oral steroids. As part of our regular routine, we ask all patients over the age of 60 years to take the 81 mg dose of aspirin on Mondays, Wednesdays, and Fridays in the hope that for those with some degree of small-vessel disease already present, an antiplatelet agent may reduce the chances of temporary or permanent cranial neuropathy.

Arteriovenous Malformations Radiosurgery is an effective form of therapy for arteriovenous malformations (AVMs) in any brain location and for both children and adults [126–146]. Obliteration rates depend primarily on AVM size and prescription dose. Complications of treatment also depend on AVM location, size, prescription dose, 12-Gy volume, age, radiographic method for target deﬁnition, perforating artery supply, and era of treatment. Table 65-3 sum-

8 8–24 5 7–25 3 1–3 8–18 10–12 4 2–5 16 1–5 4 5 5 5–8

marizes the temporary and permanent complication rates from selected series in the past 5 years. The treatment of pediatric AVMs poses special considerations of the effect of radiation on the developing brain and the risk of secondary neoplasia. Weighed against these concerns are the mortality and morbidity of hemorrhagic events over the number of years of life left to live assuming a 2% incidence of hemorrhage per year. Kaido [148] reported a 14-year-old boy treated for a periventricular AVM with a marginal dose of 20 Gy who developed a fatal glioblastoma within the irradiated ﬁeld 6.5 years later. Levy et al. [149] reported a series of 53 children treated with single-session radiosurgery followed for a minimum of 36 months. The post-radiosurgery hemorrhage rate was 2.5% per year although there were no deaths. No acute side effects were observed, and there were no cases of cyst formation or encephalomalacia. Complications occurred in only 2 (3.8%) patients: one case each of symptomatic brain-stem edema and one case of pulmonary edema. Smyth et al. [146] in a review of 40 pediatric patients reported a higher postradiosurgery hemorrhage rate (4.3% per year) but again no

TABLE 65-3. Summary of recent complications following radiosurgery for AVMs. First author

Year

No. of patients

Temporary complications (%)

Permanent complications (%)

Lunsford* [136] Schlienger [140] Smyth* [146] Friedman [131] Veznedaroglu [144]

2000 2000 2002 2003 2004

Pollock [138] Bollet [128] Shin [147] Andrade-Souza [126]

2004 2004 2004 2005

53 169 40 269 23 7 56 118 400 136

1.8 3.5 NS 3.7 4.3 28 NS 5 5.5 9.6

1.8 1.8 6 1 8.7 14 12 1.7 1.5 7.4

NS, not stated. *Pediatric series.

Rehemorrhage rate (%)

2.5 2.3 4.3 10 NS NS 12 6 NS Year 1: 4.4 Year 2: 1.5 Year 3: 0

656

i. yang et al.

mortality. There were two new permanent neurologic deﬁcits (6%). The authors believed that the low obliteration rate of 35% was related to a conservative dose prescription practice for children but recommended that with the higher observed rehemorrhage rate, childhood AVMs be treated similar to adults. The report by Flickinger et al. [150] did much to solidify our understanding of the risk of complications after AVM radiosurgery related to the volume of normal brain receiving ≥12 Gy and AVM location. In an analysis of 85 patients with a median follow-up of 45 months, the volume of brain receiving ≥12 Gy allowed for the construction of a post-radiosurgery injury expression score (SPIE). Anatomic location was also associated with complication risk, and the order from low to high risk was frontal, temporal, intraventricular, parietal, cerebellar, corpus callosum, occipital, medulla, thalamus, basal ganglia, and pons/ medulla. In multivariate analysis, the risks of permanent sequelae varied dramatically with location and to a lesser extent with the 12-Gy volume. Larger AVMs are associated with lower obliteration rates, longer time to obliteration after treatment, and an increased incidence of complications. Two approaches have been used to overcome these issues with multisession radiosurgery and volume staged treatment. Lindvall et al. [134] reported on 36 patients treated with multisession treatment and 5.6% developed radiation necrosis. Veznedaroglu et al. [144] reported on two different treatment schemes: 7 patients received 42 Gy using 7-Gy fractions, and 23 patients received 30 Gy using 5Gy fractions. T2 imaging changes were observed in 86% of the 7-Gy group and in 30% of the 5-Gy group. Both transient and permanent side effects were more common in the 7-Gy group at 43% and 14%, respectively, compared with 4.3% and 8.7% in the 5-Gy group. There was one death in the 7-Gy group directly attributable to cerebral edema. Chang et al. [129] using either single-session or multisession radiosurgery depending on AVM size and eloquence of location found that the 3year risk for development of new T2 signal changes was 59% and 64% for single-session and multisession treatment, respectively. Using a volume staged approach, Sirin et al. [142] reported on 37 patients treated with a median follow-up of 50 months. The median AVM volume was 24.9 cm3 treated in two separate stages. During the follow-up period, 14% (4/37) of patients suffered hemorrhagic complications and 1 (2.7%) experienced new neurologic deﬁcit. In 14% of patients, new T2 signal changes appeared on MR. Pollock et al. [138] have calculated that the volume staged approach reduces the 12-Gy volume compared with single-session or multisession treatments covering the entire AVM volume.

Complication Management Most complications from radiosurgery are not immediate but rather early- or late-delayed complications. These toxicities are associated principally with the development of vasogenic edema, demyelination, and/or radiation necrosis. Corticosteroids are the ﬁrst-line treatment in the management of these complications (Fig. 65-3). The initial dose of steroids to be used is not deﬁned and depends on the severity of symptoms. Seizures may respond both to increasing anticonvulsant levels and the addition of steroids.

FIGURE 65-3. (A, B) Two-dimensional time-of-ﬂight MR images of left frontoparietal AVM that presented with headaches and seizures. Dose of 17 Gy prescribed at margin of lesion deﬁned by composite MR and angiographic images. Dose-volume relationship predicting a 3% risk of permanent complications used for dose prescription. (C) CT scan without contrast done for episode of severe headache, nausea and vomiting to rule out hemorrhage. (D) FLAIR sequence showing extent of vasogenic edema around AVM. Patient was treated with short course of steroids with resolution of symptoms.

It has been suggested that corticosteroids can be tapered off in most patients treated with radiosurgery, with only long use in patients with symptomatic edema. Shaw et al. report that only 10% of patients not previously on steroids needed to start corticosteroids within 1 year after radiosurgery treatment [14, 16]. Likewise, Adler et al. found that most patients on steroids were able to discontinue them by 1 month after radiosurgery treatment [14, 41]. If refractory to corticosteroids, open surgical decompression must be considered. In addition, for hydrocephalus exacerbation, shunting may be required to alleviate symptoms. For patients with brain metastases who become symptomatic after radiosurgery, the management of seizures and new or worsening neurologic deﬁcit is described and outlined in Figure 65-4. Many of these patients can survive for a signiﬁcant length of time with proper treatment while maintaining their quality of life. Decisions on retreatment of recurrent symptomatic edema/necrosis needs to be taken in light of the patients’ functional neurologic condition and the status of their systemic disease. Patients who are not independent and those with progressive systemic disease within the prior 3 months can also be considered for supportive care. Meningioma patients can have similar problems with edema but rarely is radiation necrosis a problem as most physicians have been reducing the marginal dose to 15 to 16 Gy (Fig. 65-5). It is not uncommon to see some loss of central enhancement and indistinct margins at 6 to 18 months after radiosurgery. These changes, along with a slight 1- to 2-mm increase in tumor

65.

657

complications and management in radiosurgery Radiosurgery for Brain Metastases

Acute Side Effects

Early Delayed Symptoms

Late Delayed Symptoms

Rx first with steroids

Rx with steroids first

Symptomatic Rx Seizures: -seizure check drug levels -give half loading dose -recheck levels 7–10 days

No Improvement

No Improvement

MR imaging CT scan to R/O bleed -renal cell, melanoma, -choriocarcinoma

MR to R/O recurrence

Tumor: -retreat -operate -WBXRT

edema

?Necrosis Metabolic imaging -PET, MRS, MRP

Continue steroids

Necrosis -steroids -surgery -PCE Rx

Long term -PCE Rx

FIGURE 65-4. Treatment algorithm for patients with brain metastases who become symptomatic after radiosurgery.

dimensions, may be accompanied by surrounding edema. We are careful not to declare these changes as a failure of treatment but recommend repeat 3- to 6-month interval scans. Parasagittal and falx meningiomas seem to be prone to the development of edema, and when symptomatic this is treated ﬁrst with steroids as noted above. For those patients with mild symptoms and who have already had a course of steroid treatment, we use the PCE combination for 3 to 6 months. For patients with unremitting symptoms despite steroids, open operation and tumor removal may be required (Fig. 65-5). For acoustic neuromas, we prescribe 81 mg aspirin for those over the age of 60 years right after radiosurgery. We also remind patients that imaging changes may occur 6 to 12 months after

treatment that might suggest enlargement of the tumor on MR scans but that this is a temporary phenomenon (Fig. 65-6). We advise patients to report immediately any noticeable hearing loss. With respect to hearing loss that may accompany radiosurgery for acoustic schwannomas, Sakamoto et al. reported a small study using corticosteroids to treat hearing loss after radiosurgery [151]. In this series, over 2 years all patients treated with corticosteroids saw an improved hearing recovery compared with a control nonmedicated group [151]. In the long-term for patients with conﬁrmed hearing loss, auditory brain-stem implants may be used to rehabilitate hearing function as a management of this complication [151]. Long-term usefulness of this therapy has yet to be investigated. For those patients who

Meningioma

Late Delayed

Early Delayed

Acute Side Effects -Headache -Nausea -Vomiting -Seizures

MR imaging T2 imaging change No symptoms

Rx Symptomatically For seizures: Observe -check level -give half loading dose -recheck levels 7–10 days

Symptomatic Increased T2

Necrosis

Tumor recurrence

Short course steroids Short course steroids

Persistent symptoms 1. Mild –> PCE Rx 2. Severe –> surgery

Persistent symptoms

PCE Rx

Surgery

FIGURE 65-5. Treatment algorithm for patients with meningiomas who become symptomatic after radiosurgery.

658

i. yang et al. Acoustic Neuroma

Late Delayed

Early Delayed

Acute Side Effects -Headache -Nausea -Vomiting -Vertigo

Imaging change Trigeminal neuralgia No symptoms

Hearing loss Facial weakness

Motor/sensory Sx

Rx Symptomatically Observe

Rx Medically -carbamazapine -gabapentin

Short course steroids

MR imaging

Edema White matter changes

Necrosis

Short course steroids Long term PCE Rx

FIGURE 65-6. Treatment algorithm for patients with acoustic neuromas who become symptomatic after radiosurgery.

develop trigeminal neuropathy, treatment with carbamazepine is indicated when there is associated facial pain. Standard medical therapy starting with carbamazepine or gabapentin is recommended (Fig. 65-6). We have not seen a case of radiosurgery-induced trigeminal neuralgia, although our radiosurgery experience from 1991 to 2005 is approximately 200 cases. The management of radiation complications in AVM patients adds several other factors to the mix (Fig. 65-7). If patients present with a new severe headache, nausea, vomiting, and/or new neurologic deﬁcit, then hemorrhage needs to be ruled out with CT imaging. For those patients with prior hemorrhage who are not surgical candidates and who have an intranidal aneurysm, we recommend endovascular treatment of the aneurysm but not the entire AVM. Patients with known feeding artery aneurysms are recommended to either surgical or endo-

vascular treatment of these aneurysms before radiosurgery. Early-delayed effects of edema are not uncommon with AVMs, although most patients are not symptomatic for the imaged changes. Cyst formation is a rare late complication of radiosurgery for AVMs that can be treated with cyst to peritoneal shunting or open operation for removal of any residual AVM and necrotic tissue.

Conclusion The addition of radiosurgery as a treatment option for several of the common diagnoses discussed above has been a major development in neurosurgery and radiation oncology in the past 20 years. Radiosurgery is now accepted as standard form

AVM Radiosurgery

Acute Side Effects -Headaches -Nausea -Vomiting -Seizures

Early Delayed Effects

Severe headaches New focal deficit

Late Delayed Effects

Mild worsening symptoms Increased seizures

Repeat MRI

Treat Symptomatically Seizures: -drug level CT to R/O hemorrhage -give half loading dose -recheck levels in 7–10 days

Repeat MRI

Edema or Necrosis

Hemorrhage -re-evaluate for surgery No Hemorrhage

Short course of steroids -prolonged symptoms or inaging changes –> PCE Rx -consider surgery for necrosis

Cyst formation

Surgery -cyst shunting –AVM removal –removal necrosis –rule out tumor

FIGURE 65-7. Treatment algorithm for patients with arteriovenous malformations who become symptomatic after radiosurgery.

65.

complications and management in radiosurgery

of therapy and has improved clinical outcomes and quality of life for many conditions. As with any other form of therapy, there remains the risk of treatment-associated complications, although these risks are small and have been reduced as experience with the technique has grown. As software for radiosurgery treatment planning improves allowing more precise applications of radiation while using source-blocking patterns to protect critical structures, the complication and injury to adjacent critical neurologic structures should decrease [153, 154]. As radiosurgery evolves, precise prescription doses that limit toxicities while maintaining efﬁcacy are further being reﬁned with experience and investigations [43, 92, 97, 113, 119]. Furthermore, as radiographic and metabolic imaging improves to further accurately identify critical structures in a more precise manner and the metabolic state of the target tissue before and after treatment, the risk of toxicity of radiosurgery should be lessened [121]. The management of radiosurgery-associated complications can be difﬁcult for both the patient and the treating physician. The diagnosis of imaging-related changes in the target tissue and adjacent brain has been improved by increased access to, and newer, metabolic/physiologic imaging studies. Future investigations of alternatives to steroids in the management of these complications is needed.

References 1. Friedman WA, Foote KD. Linear accelerator radiosurgery in the management of brain tumours. Ann Med 2000; 32:64–80. 2. Nguyen DK, Spencer SS. Recent advances in the treatment of epilepsy. Arch Neurol 2003; 60:929–935. 3. Hong TS, Tome WA, Hayes L, et al. Acute sequelae of stereotactic radiosurgery. In: Kondziolka D, ed. Radiosurgery. Basel: Karger, 2004:38–45. 4. Lunsford LD, Linskey ME. Stereotactic radiosurgery in the treatment of patients with acoustic tumors. Otolaryngol Clin North Am 1992; 25:471–491. 5. Linskey ME, Lunsford LD, Flickinger JC. Radiosurgery for acoustic neurinomas: early experience. Neurosurgery 1990; 26:736–744; discussion 744–735. 6. Shoshan Y, Wygoda M, Umansky F. Stereotactic radiosurgery and fractionated stereotactic radiotherapy: background, deﬁnitions, applications. Isr Med Assoc J 2005; 7:597–599. 7. Kitchen N. Experimental and clinical studies on the putative therapeutic efﬁcacy of cerebral irradiation (radiotherapy) in epilepsy. Epilepsy Res 1995; 20:1–10. 8. Sun B, DeSalles AA, Medin PM, et al. Reduction of hippocampalkindled seizure activity in rats by stereotactic radiosurgery. Exp Neurol 1998; 154:691–695. 9. Kondziolka D, Lunsford LD, Witt TC, Flickinger JC. The future of radiosurgery: radiobiology, technology, and applications. Surg Neurol 2000; 54:406–414. 10. Dillon WP, Barbaro N. Noninvasive surgery for epilepsy: the era of image guidance. AJNR Am J Neuroradiol 1999; 20:185. 11. Regis J, Rey M, Bartolomei F, et al. Gamma knife surgery in mesial temporal lobe epilepsy: a prospective multicenter study. Epilepsia 2004; 45:504–515. 12. Spiegelmann R, Gofman J, Alezra D, Pfeffer R. Radiosurgery for acoustic neurinomas (vestibular schwannomas). Isr Med Assoc J 1999; 1:8–13. 13. Pollock BE, Lunsford LD, Kondziolka D, et al. Outcome analysis of acoustic neuroma management: a comparison of microsurgery and stereotactic radiosurgery. Neurosurgery 1995; 36:215–224; discussion 224–219.

659

14. Chin LS, Lazio BE, Biggins T, Amin P. Acute complications following gamma knife radiosurgery are rare. Surg Neurol 2000; 53:498–502; discussion 502. 15. Smith MC, Ryken TC, Buatti JM. Radiotoxicity after conformal radiation therapy for benign intracranial tumors. Neurosurg Clin N Am 2006; 17:169–180. 16. Shaw E, Scott C, Souhami L, et al. Radiosurgery for the treatment of previously irradiated recurrent primary brain tumors and brain metastases: initial report of radiation therapy oncology group protocol (90-05). Int J Radiat Oncol Biol Phys 1996; 34:647–654. 17. Tago M, Terahara A, Nakagawa K, et al. Immediate neurological deterioration after gamma knife radiosurgery for acoustic neuroma. Case report. J Neurosurg 2000; 93(Suppl 3):78–81. 18. Majhail NS, Chander S, Mehta VS, et al. Factors inﬂuencing early complications following Gamma Knife radiosurgery. A prospective study. Stereotact Funct Neurosurg 2001; 76:36–46. 19. Westbury C, Hines F, Hawkes E, et al. Advice on hair and scalp care during cranial radiotherapy: a prospective randomized trial. Radiother Oncol 2000; 54:109–116. 20. Alexander E 3rd, Loefﬂer JS. Radiosurgery for primary malignant brain tumors. Semin Surg Oncol 1998; 14:43–52. 21. Gannett D, Stea B, Lulu B, et al. 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; 33:461–468. 22. Gelblum DY, Lee H, Bilsky M, et al. Radiographic ﬁndings and morbidity in patients treated with stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1998; 42:391–395. 23. Werner-Wasik M, Rudoler S, Preston PE, et al. Immediate side effects of stereotactic radiotherapy and radiosurgery. Int J Radiat Oncol Biol Phys 1999; 43:299–304. 24. Hudgins WR. Patients’ attitude about outcomes and the role of gamma knife radiosurgery in the treatment of vestibular schwannomas. Neurosurgery 1994; 34:459–463; discussion 463–455. 25. Sakamoto M, Oya N, Mizowaki T, et al. Initial experiences of palliative stereotactic radiosurgery for recurrent brain lymphomas. J Neurooncol 2006; 77:53–58. 26. Cmelak AJ, Abou-Khalil B, Konrad PE, et al. Low-dose stereotactic radiosurgery is inadequate for medically intractable mesial temporal lobe epilepsy: a case report. Seizure 2001; 10:442–446. 27. 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; 24:619–625. 28. Noren G, Greitz D, Hirsch A, Lax I. Gamma knife surgery in acoustic tumours. Acta Neurochir Suppl (Wien) 1993; 58:104– 107. 29. Ito K, Shin M, Matsuzaki M, et al. Risk factors for neurological complications after acoustic neurinoma radiosurgery: reﬁnement from further experiences. Int J Radiat Oncol Biol Phys 2000; 48:75–80. 30. Lee JY, Niranjan A, McInerney J, et al. Stereotactic radiosurgery providing long-term tumor control of cavernous sinus meningiomas. J Neurosurg 2002; 97:65–72. 31. Park YG, Chung SS, Kim DI, et al. Complications following Gamma Knife radiosurgery. Stereotact Funct Neurosurg 1995; 64(Suppl 1):239–248. 32. Engenhart R, Kimmig BN, Hover KH, et al. Long-term follow-up for brain metastases treated by percutaneous stereotactic single high-dose irradiation. Cancer 1993; 71:1353–1361. 33. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multiinstitutional experience with stereotactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol Biol Phys 1994; 28:797–802. 34. Thomsen J, Mirz F, Wetke R, et al. Intracranial sarcoma in a patient with neuroﬁbromatosis type 2 treated with gamma knife radiosurgery for vestibular schwannoma. Am J Otol 2000; 21:364–370.

660

i. yang et al.

35. Shamisa A, Bance M, Nag S, et al. Glioblastoma multiforme occurring in a patient treated with gamma knife surgery. Case report and review of the literature. J Neurosurg 2001; 94:816–821. 36. Yu JS, Yong WH, Wilson D, Black KL. Glioblastoma induction after radiosurgery for meningioma. Lancet 2000; 356:1576–1577. 37. Kaido T, Hoshida T, Uranishi R, et al. Radiosurgery-induced brain tumor. Case report. J Neurosurg 2001; 95:710–713. 38. Ganz JC. Gamma knife radiosurgery and its possible relationship to malignancy: a review. J Neurosurg 2002; 97:644–652. 39. Sanno N, Hayashi S, Shimura T, et al. Intracranial osteosarcoma after radiosurgery—case report. Neurol Med Chir (Tokyo) 2004; 44:29–32. 40. Loefﬂer JS, Niemierko A, Chapman PH. Second tumors after radiosurgery: tip of the iceberg or a bump in the road? Neurosurgery 2003; 52:1436–1440; discussion 1440–1432. 41. Adler JR, Cox RS, Kaplan I, Martin DP. Stereotactic radiosurgical treatment of brain metastases. J Neurosurg 1992; 76:444–449. 42. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radiosurgery for anterior foramen magnum meningiomas. Surg Neurol 1999; 51:268–273. 43. Iwai Y, Yamanaka K, Yasui T, et al. Gamma knife surgery for skull base meningiomas. The effectiveness of low-dose treatment. Surg Neurol 1999; 52:40–44; discussion 44–45. 44. Subach BR, Lunsford LD, Kondziolka D, et al. Management of petroclival meningiomas by stereotactic radiosurgery. Neurosurgery 1998; 42:437–443; discussion 443–435. 45. Chin LS, Szerlip NJ, Regine WF. Stereotactic radiosurgery for meningiomas. Neurosurg Focus 2003; 14:e6. 46. Roche PH, Regis J, Dufour H, et al. Gamma knife radiosurgery in the management of cavernous sinus meningiomas. J Neurosurg 2000; 93 (Suppl 3):68–73. 47. Pamir MN, Kilic T, Bayrakli F, Peker S. Changing treatment strategy of cavernous sinus meningiomas: experience of a single institution. Surg Neurol 2005; 64(Suppl 2):S58–66. 48. Nicolato A, Ferraresi P, Foroni R, et al. Gamma Knife radiosurgery in skull base meningiomas. Preliminary experience with 50 cases. Stereotact Funct Neurosurg 1996; 66(Suppl 1):112–120. 49. Pollock BE. Stereotactic radiosurgery for intracranial meningiomas: indications and results. Neurosurg Focus 2003; 14:e4. 50. Kobayashi T, Kida Y, Mori Y. Long-term results of stereotactic gamma radiosurgery of meningiomas. Surg Neurol 2001; 55:325– 331. 51. Singh VP, Kansai S, Vaishya S, et al. Early complications following gamma knife radiosurgery for intracranial meningiomas. J Neurosurg 2003; 93(Suppl 3):57–61. 52. 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; 43:405–413; discussion 413–404. 53. Kamiryo T, Yamamoto M, Barfod B, Urakawa Y. Dose absorbed by normal brainstem and optic apparatus in gamma knife surgery for ten or more metastasis. In: Kondziolka D, ed. Radiosurgery. Basel: Karger, 2004:77–81. 54. 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. 55. 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; 27:215–221. 56. Duma CM, Lunsford LD, Kondziolka D, et al. Stereotactic radiosurgery of cavernous sinus meningiomas as an addition or alternative to microsurgery. Neurosurgery 1993; 32:699–704; discussion 704–695. 57. Kondziolka D, Lunsford LD, Coffey RJ, Flickinger JC. Stereotactic radiosurgery of meningiomas. J Neurosurg 1991; 74:552– 559.

58. Kreil W, Luggin J, Fuchs I, et al. Long term experience of gamma knife radiosurgery for benign skull base meningiomas. J Neurol Neurosurg Psychiatry 2005; 76:1425–1430. 59. Morita A, Coffey RJ, Foote RL, et al. Risk of injury to cranial nerves after gamma knife radiosurgery for skull base meningiomas: experience in 88 patients. J Neurosurg 1999; 90:42–49. 60. Stafford SL, Pollock BE, Leavitt JA, et al. A study on the radiation tolerance of the optic nerves and chiasm after stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2003; 55:1177–1181. 61. Stafford SL, Pollock BE, Foote RL, et al. Meningioma radiosurgery: tumor control, outcomes, and complications among 190 consecutive patients. Neurosurgery 2001; 49:1029–1037; discussion 1037–1028. 62. Friedman WA, Murad GJ, Bradshaw P, et al. Linear accelerator surgery for meningiomas. J Neurosurg 2005; 103:206–209. 63. Hudgins WR, Barker JL, Schwartz DE, Nichols TD. Gamma Knife treatment of 100 consecutive meningiomas. Stereotact Funct Neurosurg 1996; 66(Suppl 1):121–128. 64. Chang JH, Chang JW, Choi JY, et al. Complications after gamma knife radiosurgery for benign meningiomas. J Neurol Neurosurg Psychiatry 2003; 74:226–230. 65. Hakim R, Alexander E 3rd, Loefﬂer JS, et al. Results of linear accelerator-based radiosurgery for intracranial meningiomas. Neurosurgery 1998; 42:446–453; discussion 453–444. 66. Shafron DH, Friedman WA, Buatti JM, et al. Linac radiosurgery for benign meningiomas. Int J Radiat Oncol Biol Phys 1999; 43:321–327. 67. Iwai Y, Yamanaka K, Ishiguro T. Gamma knife radiosurgery for the treatment of cavernous sinus meningiomas. Neurosurgery 2003; 52:517–524; discussion 523–514. 68. DiBiase SJ, Kwok Y, Yovino S, et al. Factors predicting local tumor control after gamma knife stereotactic radiosurgery for benign intracranial meningiomas. Int J Radiat Oncol Biol Phys 2004; 60:1515–1519. 69. Chang SD, Adler JR Jr. Treatment of cranial base meningiomas with linear accelerator radiosurgery. Neurosurgery 1997; 41:1019– 1025; discussion 1025–1027. 70. Kondziolka D, Levy EI, Niranjan A, et al. Long-term outcomes after meningioma radiosurgery; physician and patient perspectives. J Neurosurg 1999; 91:44–50. 71. Shin M, Kurita H, Sasaki T, et al. Analysis of treatment outcome after stereotactic radiosurgery for cavernous sinus meningiomas. J Neurosurg 2001; 95(3):435–439. 72. Nakamura S, Hiyama H, Arai K, et al. Gamma Knife radiosurgery for meningiomas: four cases of radiation-induced edema. Stereotact Funct Neurosurg 1996; 66(Suppl 1):142–145. 73. Ganz JC, Schrottner O, Pendl G. Radiation-induced edema after Gamma Knife treatment for meningiomas. Stereotact Funct Neurosurg 1996; 66(Suppl 1):129–133. 74. Nakano T, Asano K, Miura H, et al. Meningiomas with brain edema: radiological characteristics on MRI and review of the literature. Clin Imaging 2002; 26:243–249. 75. Kim CH, Kim DG, Paek SH, et al. Delayed bleeding after gamma knife surgery for meningioma. Acta Neurochir (Wien) 2004; 146:741–742. 76. Kwon Y, Ahn JS, Jeon SR, et al. Intratumoral bleeding in meningioma after gamma knife radiosurgery. J Neurosurg 2002; 97: 657–662. 77. Bui QC, Lieber M, Withers HR, et al. The efﬁcacy of hyperbaric oxygen therapy in the treatment of radiation-induced late side effects. Int J Radiat Oncol Biol Phys 2004; 60(3):871–878. 78. Charabi S, Tos M, Thomsen J, et al. Cystic vestibular schwannoma— clinical and experimental studies. Acta Otolaryngol 2000; 543(Suppl):11–13. 79. Ottaviani F, Neglia CB, Ventrella L, et al. Hearing loss and changes in transient evoked otoacoustic emissions after gamma

65.

80.

81. 82.

83.

84.

85.

86. 87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

complications and management in radiosurgery

knife radiosurgery for acoustic neurinomas. Arch Otolaryngol Head Neck Surg 2002; 128:1308–1312. Unger F, Walch C, Haselsberger K, et al. Radiosurgery of vestibular schwannomas: a minimally invasive alternative to microsurgery. Acta Neurochir (Wien) 1999; 141:1281–1285; discussion 1285–1286. Friedman WA, Foote KD. Linear accelerator-based radiosurgery for vestibular schwannoma. Neurosurg Focus 2003; 14:e2. Battista RA, Wiet RJ. Stereotactic radiosurgery for acoustic neuromas: a survey of the American Neurotology Society. Am J Otol 2000; 21:371–381. Sekhar LN, Gormley WB, Wright DC. The best treatment for vestibular schwannoma (acoustic neuroma): microsurgery or radiosurgery? Am J Otol 1996; 17:676–682; discussion 683–679. Kamerer DB, Lunsford LD, Moller M. Gamma knife: an alternative treatment for acoustic neurinomas. Ann Otol Rhinol Laryngol 1988; 97:631–635. Lunsford LD, Kamerer DB, Flickinger JC. Stereotactic radiosurgery for acoustic neuromas. Arch Otolaryngol Head Neck Surg 1990; 116:907–909. Wiet RJ, Micco AG, Bauer GP. Complications of the gamma knife. Arch Otolaryngol Head Neck Surg 1996; 122:414–416. Ramsay HA, Luxford WM. Treatment of acoustic tumours in elderly patients: is surgery warranted? J Laryngol Otol 1993; 107:295–297. Mendenhall WM, Friedman WA, Buatti JM, Bova FJ. Preliminary results of linear accelerator radiosurgery for acoustic schwannomas. J Neurosurg 1996; 85:1013–1019. 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):90–95. Yamamoto M, Hagiwara S, Ide M, et al. Radiosurgery for acoustic neurinoma with rapid growth and relatively high staining indexes for proliferating cell nuclear antigen and MIB-1. Neurol Med Chir (Tokyo) 1996; 36:241–245. Flickinger JC, Lunsford LD, Linskey ME, et al. Gamma knife radiosurgery for acoustic tumors: multivariate analysis of four year results. Radiother Oncol 1993; 27:91–98. 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. J Neurosurg 2003; 98:800–806. Mendenhall WM, Friedman WA, Bova FJ. Linear acceleratorbased stereotactic radiosurgery for acoustic schwannomas. Int J Radiat Oncol Biol Phys 1994; 28:803–810. Spiegelmann R, Lidar Z, Gofman J, et al. Linear accelerator radiosurgery for vestibular schwannoma. J Neurosurg 2001; 94:7–13. Foote KD, Friedman WA, Buatti JM, et al. Analysis of risk factors associated with radiosurgery for vestibular schwannoma. J Neurosurg 2001; 95:440–449. Myrseth E, Moller P, Pedersen PH, et al. Vestibular schwannomas: clinical results and quality of life after microsurgery or gamma knife radiosurgery. Neurosurgery 2005; 56:927–935; discussion 927–935. Miller RC, Foote RL, Coffey RJ, et al. Decrease in cranial nerve complications after radiosurgery for acoustic neuromas: a prospective study of dose and volume. Int J Radiat Oncol Biol Phys 1999; 43:305–311. Yang CC, Ting J, Wu X, Markoe A. Dose volume histogram analysis of the gamma knife radiosurgery treating twenty-ﬁve metastatic intracranial tumors. Stereotact Funct Neurosurg 1998; 70(Suppl 1):41–49. Muacevic A, Jess-Hempen A, Tonn JC, Wowra B. Results of outpatient gamma knife radiosurgery for primary therapy

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113. 114.

115.

116.

117.

118.

661

of acoustic neuromas. Acta Neurochir Suppl 2004; 91:75– 78. Flickinger JC, Kondziolka D, Lunsford LD. Dose and diameter relationships for facial, trigeminal, and acoustic neuropathies following acoustic neuroma radiosurgery. Radiother Oncol 1996; 41:215–219. Linskey ME, Flickinger JC, Lunsford LD. Cranial nerve length predicts the risk of delayed facial and trigeminal neuropathies after acoustic tumor stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1993; 25:227–233. Flickinger JC, Kondziolka D, Niranjan A, Lunsford LD. Results of acoustic neuroma radiosurgery: an analysis of 5 years’ experience using current methods. J Neurosurg 2001; 94:1–6. Ito K, Kurita H, Sugasawa K, et al. Analyses of neuro-otological complications after radiosurgery for acoustic neurinomas. Int J Radiat Oncol Biol Phys 1997; 39:983–988. 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 Radiat Oncol Biol Phys 2001; 50:1265–1278. Rowe J, Radatz M, Walton L, Kemeny A. Gamma knife stereotactic radiosurgery for type 2 neuroﬁbromatosis acoustic neuromas. In: Kondziolka D, ed. Radiosurgery. Basel: Karger, 2004: 100–106. Kobayashi T, Tanaka T, Kida Y. The early effects of gamma knife on 40 cases of acoustic neurinoma. Acta Neurochir Suppl 1994; 62:93–97. Niranjan A, Lunsford LD, Flickinger JC, et al. Dose reduction improves hearing preservation rates after intracanalicular acoustic tumor radiosurgery. Neurosurgery 1999; 45:753–762; discussion 762–755. Meijer OW, Wolbers JG, Baayen JC, Slotman BJ. Fractionated stereotactic radiation therapy and single high-dose radiosurgery for acoustic neuroma: early results of a prospective clinical study. Int J Radiat Oncol Biol Phys 2000; 46:45–49. Shin YJ, Lapeyre-Mestre M, Gafsi I, et al. Neurotological complications after radiosurgery versus conservative management in acoustic neuromas: a systematic review-based study. Acta Otolaryngol 2003; 123:59–64. Chung WY, Liu KD, Shiau CY, et al. Gamma knife surgery for vestibular schwannoma: 10-year experience of 195 cases. J Neurosurg 2005; 102(Suppl):87–96. Thomsen J, Tos M, Borgesen SE. Gamma knife: hydrocephalus as a complication of stereotactic radiosurgical treatment of an acoustic neuroma. Am J Otol 1990; 11:330–333. Kondziolka D, Lunsford LD, McLaughlin MR, Flickinger JC. Long-term outcomes after radiosurgery for acoustic neuromas. N Engl J Med 1998; 339:1426–1433. Prasad D, Steiner M, Steiner L. Gamma surgery for vestibular schwannoma. J Neurosurg 2000; 92:745–759. Paek SH, Chung HT, Jeong SS, et al. Hearing preservation after gamma knife stereotactic radiosurgery of vestibular schwannoma. Cancer 2005; 104:580–590. Combs SE, Thilmann C, Debus J, Schulz-Ertner D. Long-term outcome of stereotactic radiosurgery (SRS) in patients with acoustic neuromas. Int J Radiat Oncol Biol Phys 2006; 64:1341– 1347. Unger F, Walch C, Papaefthymiou G, et al. Radiosurgery of residual and recurrent vestibular schwannomas. Acta Neurochir (Wien) 2002; 144:671–676; discussion 676–677. Rowe JG, Radatz MW, Walton L, et al. Gamma knife stereotactic radiosurgery for unilateral acoustic neuromas. J Neurol Neurosurg Psychiatry 2003; 74:1536–1542. Rowe JG, Radatz MW, Walton L, et al. Clinical experience with gamma knife stereotactic radiosurgery in the management of

662

119.

120.

121.

122.

123. 124.

125.

126.

127.

128.

129.

130.

131.

132.

133.

134.

135.

136.

i. yang et al. vestibular schwannomas secondary to type 2 neuroﬁbromatosis. J Neurol Neurosurg Psychiatry 2003; 74:1288–1293. Petit JH, Hudes RS, Chen TT, et al. Reduced-dose radiosurgery for vestibular schwannomas. Neurosurgery 2001; 49:1299–1306; discussion 1306–1297. Ito K, Kurita H, Sugasawa K, et al. Neuro-otological ﬁndings after radiosurgery for acoustic neurinomas. Arch Otolaryngol Head Neck Surg 1996; 122:1229–1233. Suh JH, Barnett GH, et al. Results of linear accelerator-based stereotactic radiosurgery for recurrent and newly diagnosed acoustic neuromas. Int J Cancer 2000; 90:145–151. Linskey ME, Lunsford LD, Flickinger JC. Tumor control after stereotactic radiosurgery in neuroﬁbromatosis patients with bilateral acoustic tumors. Neurosurgery 1992; 31:829–838; discussion 838–829. Kondziolka D, Lunsford LD. Preservation of hearing in acoustic neurinoma surgery. J Neurosurg 1993; 78:154–156. Hirato M, Inoue H, Nakamura M, et al. Gamma knife radiosurgery for acoustic schwannoma: early effects and preservation of hearing. Neurol Med Chir (Tokyo) 1995; 35:737–741. Iwai Y, Yamanaka K, Shiotani M, Uyama T. Radiosurgery for acoustic neuromas: results of low-dose treatment. Neurosurgery 2003; 53:282–287; discussion 287–288. Andrade-Souze YM, Zadeh G, Ramani M, et al. Testing the radiosurgery-based arteriovenous malformation score and the modiﬁed Spetzler-Martin grading system to predict radiosurgical outcome. J Neurosurg 2005; 103:642–648. Barker FG 2nd, Butler WE, Lyons S, et al. Dose-volume prediction of radiation-related complications after proton beam radiosurgery for cerebral arteriovenous malformations. J Neurosurg 2003; 99:254–263. Bollet MA, Anxionnat R, Buchheit I, et al. Efﬁcacy and morbidity of arc-therapy radiosurgery for cerebral arteriovenous malformations: a comparison with the natural history. Int J Radiat Oncol Biol Phys 2004; 58:1353–1363. Chang TC, Shirato H, Aoyama H, et al. Stereotactic irradiation for intracranial arteriovenous malformation using stereotactic radiosurgery or hypofractionated stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 2004; 60:861–870. Flickinger JC, Kondziolka D, Lunsford LD, et al. 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; 46:1143–1148. Friedman WA, Bova FJ, Bollampally S, Bradshaw P. Analysis of factors predictive of success or complications in arteriovenous malformation radiosurgery. Neurosurgery 2003; 52:296–307. Izawa M, Hayashi M, Chernov M, et al. Long-term complications after gamma knife surgery for arteriovenous malformations. J Neurosurg 2005; 102(Suppl):34–37. 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; 49:1045–1051. Lindvall P, Bergstrom P, Lofroth PO, et al. Hypofractionated conformal stereotactic radiotherapy for arteriovenous malformations. Neurosurgery 2003; 53:1036–1042; discussion 1042–1043. Maruyama K, Kondziolka D, Niranjan A, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for brainstem arteriovenous malformations: factors affecting outcome. J Neurosurg 2004; 100(3): 407–413. Levy EI, Niranjan A, Thompson TP, et al. Radiosurgery for childhood intracranial arteriovenous malformations. Neurosurgery 2000; 47(4):834–841; discussion 841–842.

137. Maesawa S, Flickinger JC, Kondziolka D, Lunsford LD. Repeated radiosurgery for incompletely obliterated arteriovenous malformations. J Neurosurg 2000; 92:961–970. 138. Pollock BE, Gorman DA, Brown PD. Radiosurgery for arteriovenous malformations of the basal ganglia, thalamus, and brainstem. J Neurosurg 2004; 100:210–214. 139. 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– 1296; discussion 1296–1297. 140. Schlienger M, Atlan D, Lefkopoulos D, et al. Linac radiosurgery for cerebral arteriovenous malformations: results in 169 patients. Int J Radiat Oncol Biol Phys 2000; 46:1135–1142. 141. Shin M, Maruyama K, Kurita H, et al. Analysis of nidus obliteration rates after gamma knife surgery for arteriovenous malformations based on long-term follow-up data: the University of Tokyo experience. J Neurosurg 2004; 101(1):18–24. 142. Sirin S, Kondziolka D, Niranjan A, et al. Prospective staged volume radiosurgery for large arteriovenous malformations: indications and outcomes in otherwise untreatable patients. Neurosurgery 2006; 58:17–27; discussion 17–27. 143. Smyth MD, Sneed PK, Ciricillo SF, et al. Stereotactic radiosurgery for pediatric intracranial arteriovenous malformations: the University of California at San Francisco experience. J Neurosurg 2002; 97(1):48–55. 144. Veznedaroglu E, Andrews DW, Benitez RP, et al. Fractionated stereotactic radiotherapy for the treatment of large arteriovenous malformations with or without previous partial embolization. Neurosurgery 2004; 55(3):519–530; discussion 530–531. 145. Zabel-du Bois A, Milker-Zabel S, Huber P, Schlegel W, Debus J. Stereotactic linac-based radiosurgery in the treatment of cerebral arteriovenous malformations located deep, involving corpus callosum, motor cortex, or brainstem. Int J Radiat Oncol Biol Phys 2006 15; 64(4):1044–1048. Epub 2005 Dec 20. 146. Smyth MD, Sneed PK, Ciricillo SF, et al. Stereotactic radiosurgery for pediatric intracranial arteriovenous malformations: the University of California at San Francisco Experience. J Neurosurg 2002; 97:48–55. 147. Shin M, Maruyama K, Kurita H, et al. Analysis of nidus obliteration rates after gamma knife surgery for arteriovenous malformations based on long-term follow-up data: the University of Tokyo experience. J Neurosurg 2004; 101(1):18–24. 148. Kaido T, Hoshida T, Uranishi R, et al. Radiosurgery-induced brain tumor. Case report. J Neurosurg 2001; 95:710–713. 149. Levy EI, Niranjan A, Thompson TP, et al. Radiosurgery for childhood intracranial arteriovenous malformations. Neurosurgery 2000; 47:834–841; discussion 841–842. 150. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multiinstitutional analysis of complication outcomes after arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 1999; 44:67–74. 151. Sakamoto T, Shirato H, Takeichi N, et al. Medication for hearing loss after fractionated stereotactic radiotherapy (SRT) for vestibular schwannoma. Int J Radiat Oncol Biol Phys 2001; 50: 1295–1298. 152. Kalamarides M, Grayeli AB, Bouccara D, et al. Hearing restoration with auditory brainstem implants after radiosurgery for neuroﬁbromatosis type 2. J Neurosurg 2001; 95:1028–1033. 153. Meeks SL, Buatti JM, Bova FJ, et al. Treatment planning optimization for linear accelerator radiosurgery. Int J Radiat Oncol Biol Phys 1998; 41:183–197. 154. Meeks SL, Buatti JM, Foote KD, et al. Calculation of cranial nerve complication probability for acoustic neuroma radiosurgery. Int J Radiat Oncol Biol Phys 2000; 47:597–602.

6 6

Cost-Effectiveness and Quality of Life Minesh Mehta and May N. Tsao

Principles of Economic Appraisal With increased availability of radiosurgical facilities and an increased need for accountability in an era of limited resources and managed care, there is growing interest in understanding the economic impact of such technologies and hence a need for a thorough economic appraisal. The key features pertaining to economic evaluations of medical interventions involve “inputs” that are compared with “outputs.” The inputs include direct costs of providing health care and indirect costs (e.g., production losses when patients are withdrawn from their work as a result of medical interventions). There are also other related and poorly measured intangible costs (although not strictly considered as “inputs”) such as pain and suffering, which may be associated with therapy [1]. Outputs include direct beneﬁt in terms of savings in other direct medical care costs with intervention, the production gains from an earlier return to work, and intangible beneﬁts such as the value of patients feeling better with therapy. Outputs have also been described in utility units (such as quality adjusted life years). The components of economic evaluation have been well summarized in Figure 66-1 adapted from Rutten and Drummond [1]. Explicit deﬁnitions of cost have been described by many authors [2–5] and are listed in Table 66-1. In a cost-analysis or cost-minimization analysis, a simple study is undertaken that only considers costs. This type of analysis is valid only when alternative strategies have been shown to produce equivalent medical outcomes (beneﬁts and risks of injury). Inputs may be conﬁned only to direct costs, indirect costs, or both. An example includes a German study that reported on the direct costs of microsurgical management of radiosurgically amenable intracranial abnormalities such as meningiomas, acoustic neuromas, metastases, and arteriovenous malformations less than 3 cm in diameter [6]. This analysis assumed that the intervention of radiosurgery or microsurgery were equivalent with respect to outcomes and risk of complication. Only direct costs (such as costs of the surgical procedure, ICU care, medical and nursing care on the ward, interclinic bills for services, basic hotel services, and Gamma Knife costs) were considered. Cost of illness analysis examines direct and indirect costs of a particular health condition. These types of analyses allow for economic evaluations on the impact of disease and also provide a baseline estimate of the costs of the disease itself against

which medical interventions used in the treatment of such conditions can be compared. The report by Taylor et al. [7], which deals with the lifetime costs of stroke in the United States, represents an example of a cost of illness study pertinent for arteriovenous malformation patients. In terms of outputs, beneﬁts of medical interventions have been measured by expressing the cost savings of direct medical care costs and the productions gains from earlier return to work. Intangible beneﬁts include quality of life outcomes that are more difﬁcult to quantify in monetary terms. Cost-beneﬁt analyses typically focus on direct and indirect costs and beneﬁt comparisons, excluding intangible costs and beneﬁts that are more difﬁcult to quantify. An example includes the report by van Roijen et al. [8]. In this study, the authors reported the direct and indirect costs and beneﬁts of microsurgery versus radiosurgery in the management of acoustic neuroma. A cost-effectiveness analysis uses cost and beneﬁt comparisons but expresses the analysis as health effects (e.g., cost per life year gained) rather than strictly monetary units. An example of a cost-effectiveness analysis has been reported by Mehta et al. [9]. In this study, the cost-effectiveness of radiosurgery versus resection for single-brain metastases was reported. It was found that the average cost per week of survival was $US310 for whole-brain radiotherapy, $US524 for resection plus wholebrain radiotherapy, and $US270 for radiosurgery plus wholebrain radiotherapy. In a cost-utility analysis, the concept of trade-offs in quality of life and quantity of life is considered. In this analysis, the life-years gained in treatment are adjusted by utility weights that reﬂect the relative values individuals place on different states of health. The measurement used in cost-utility analysis is quality adjusted life year (QALY). A cost utility analysis was performed in the study of radiosurgery versus resection among patients with single brain metastases by Mehta et al. [9]. In this report, the median values for QALYs were 12 weeks for wholebrain radiotherapy, 35 weeks for resection and whole-brain radiotherapy, and 51 weeks for radiosurgery plus whole-brain radiotherapy. Discounting involves a correction for inﬂation over time and also for the value of beneﬁts accrued today being higher than beneﬁts that accrue in the future. With discounting, costs and beneﬁts that accrue in different time periods are adjusted back to their present value [3].

663

664

m. mehta and m.n. tsao TABLE 66-2. Drummond checklist for assessing health-related economic evaluations. 1. 2. 3.

4. 5.

Figure 66-1. Components of economic evaluations. (From Rutten F, Drummond M. Making decisions about health technologies: a costeffectiveness perspective. York: The University of York Centre for Health Economics, Institute for Medical Technology Assessment, 1994:6–36. Used with permission.)

A sensitivity analysis uses a computational model to study how the variation in output can be apportioned (qualitatively or quantitatively) to sources of uncertainty in input variables or model parameters. With an increasing body of medical literature pertaining to economic evaluations regarding the use of radiosurgery, it becomes necessary to judge the methodological quality and relevance of such publications. Drummond et al. [10] have devised a 10-point checklist useful in the evaluation of economic analyses (Table 66-2). It is impossible to produce a ﬂawless economic evaluation of any medical technology. In situations where the conclusions of an economic evaluation are congruent to an individual’s belief, it is unlikely that such studies would be subject to as harsh and close scrutiny compared with those conclusions that are contrary to an individual’s biases. Conﬂicts of interest may also exist with authors and the ﬁndings/interpretation of a particular technology assessment. TABLE 66-1. Deﬁnitions of cost/savings. 1.

2. 3.

4.

Direct health care costs or savings: (a) Costs directly related to the technology (e.g., radiosurgery unit, other equipment, personnel costs, physician fees, imaging, etc.). (b) Test or treatments induced or avoided (e.g., follow-up MRI and angiography for radiosurgery-treated arteriovenous malformations). (c) Treatment side effects or complications (e.g., hospitalization, imaging investigations for side effects and steroid treatment for radionecrosis/edema). (d) Savings due to avoidance of subsequent morbidity (e.g., avoidance of hospitalization and morbidity with AVM hemorrhages). Direct personal costs or savings: (a) Transportation or lodging for treatment. (b) Home care, long-term care, rehabilitation, child care. Indirect costs: (a) Productivity gains or losses (loss of life or livelihood, disability). (b) Opportunity costs (time spent by patients for health care such as travel, appointments for follow-up or treatment, hospitalization, recovery). Intangible costs: (a) Pain and suffering.

6. 7. 8.

9. 10.

Was a well-deﬁned question posed in answerable form? Was a comprehensive description of the competing alternatives given (i.e., can you tell who did what to whom, where, and how often)? Was there evidence that the program’s effectiveness had been established? Was this done through a randomized controlled clinical trial? If not, how strong was the evidence of effectiveness? Were all important and relevant costs and consequences for each alternative identiﬁed? Were costs and consequences measured accurately in appropriate physical units (e.g., hours of nursing time, number of physician visits, days lost from work, or years-of-life gained) prior to valuation? Were costs and consequences valued credibly? Were costs and consequences adjusted for differential timing? Was an incremental analysis of the costs and consequences of alternatives performed? Were the additional (incremental) costs generated by the use of one alternative over another compared with the additional effects, beneﬁts, or utilities generated? Was a sensitivity analysis performed? Did the presentation and discussion of the results of the study include all of the issues that are of concern to users?

Source: From Drummond MF, Brandt A, Luce BR, et al. Standardizing economic evaluations in health care: practice, problems and potential. Int J Tech Assess Health Care 1993; 9(1):26–36. Used with permission.

Furthermore, in order for any economic evaluation to have any impact, it must be timely. Often, new technologies cannot be adequately assessed because follow-up of patients has not matured for a complete understanding of outcomes (e.g., benign diseases such as acoustic neuromas treated with radiosurgery). By the time longer follow-up is acquired, technology changes (e.g., MRI-based radiosurgical planning, reduction of radiosurgery tumor dose, and advanced surgical techniques for acoustic neuromas) and strong biases have developed such that the vehicle for assessing the efﬁcacy and effectiveness of different treatment modalities (e.g., surgery or radiosurgery) in a controlled clinical trial becomes problematic. Buxton [11] argued that because the pace of technological change in medicine is so rapid, health technology assessments are “always too early, until suddenly it’s too late.” Other difﬁculties with technology economic evaluations include the inability to extrapolate results in different settings/ regions where costs and/or outcomes may be different. Other main concerns with economic analyses include (arguably) less developed methods and use of assumptions with modeling approaches that make conclusions perhaps easier to manipulate. One of the intents of economic appraisals is to assist policymakers in the rational diffusion and use of health technologies. This remains a difﬁcult area with respect to the use of radiosurgery as there exists strong conﬂicts of interests among stakeholders, data are imprecise, and issues regarding timeliness of such evaluations and issues regarding external validity exist.

Radiosurgery and Economic Appraisals A literature search was undertaken using the following databases: Medline (1990–2004), CancerLit (1990–2004), CINAHL (1990–2004), EMBASE (1990–2004), and the Cochrane Library (2004, issue 4).

66.

665

cost-effectiveness and quality of life

TABLE 66-3. Radiosurgery and economic appraisal studies. Reference

Disease

Comparison: Radiosurgery modality Epstein et al. [12] Various Konigsmaier et al. [13]

Not speciﬁed

Comparison: Craniotomy versus radiosurgery Ott [14] Various Wellis et al. [6]

Meningioma, acoustic neuroma, metastases, arteriovenous malformations (less than 3 cm)

Comparison: Single brain metastasis management Mehta et al. [9] Single brain metastases

Rutigliano et al. [2] Sperduto et al. [15]

Single brain metastases Single brain metastasis

Interventions

Type of economic appraisal

Gamma Knife versus linear accelerator versus proton radiosurgery Gamma Knife versus linear accelerator radiosurgery

Direct costs

Craniotomy versus Gamma Knife radiosurgery Microsurgery versus radiosurgery

Direct costs

Whole-brain radiotherapy versus resection and whole-brain radiotherapy versus radiosurgery and whole-brain radiotherapy Surgery versus radiosurgery No therapy, whole-brain radiotherapy alone, whole-brain radiotherapy and radiosurgery, whole-brain radiotherapy and surgical resection

Cost-effectiveness Cost-utility

Comparison: Surgery versus radiosurgery for arteriovenous malformations Porter et al. [16] Arteriovenous malformations Surgery versus radiosurgery Comparison: Surgery versus radiosurgery for acoustic neuroma Van Roijen et al. [17] Acoustic neuroma

There were 11 relevant studies that reported on economic appraisals and radiosurgery (summarized in Table 66-3).

Gamma Knife Versus Linear Accelerator (Versus Proton Radiosurgery) Epstein et al. [12] examined the direct costs (cost of equipment and labor costs) for the three dominant radiosurgery technologies (Gamma Knife, linear accelerator, and proton radiosurgery). In this analysis, it was concluded that once more than 100 patients per year are treated with radiosurgery at an institution, the Gamma Knife was found to be the most cost-effective technology by a factor of 100%. The cost of Gamma Knife radiosurgery versus linear accelerator radiosurgery was evaluated in Konigsmaier’s report [13]. Direct costs were considered, namely investment costs, operating costs, and stafﬁng costs. In this analysis, the adapted linac (a linear accelerator used for both fractionated radiotherapy and radiosurgery) was the most favorable in terms of direct costs for small annual numbers of radiosurgery patients. With larger numbers of patients, the Gamma Knife represented the most favorable, from a direct cost point of view.

Surgery Versus Radiosurgery for Various Conditions Ott [14] reported on a comparison of craniotomy versus Gamma Knife direct costs for a variety of health conditions believed to be amenable to either craniotomy or Gamma Knife radiosurgery. The average hospital charge for intracranial procedures

Direct costs

Microsurgery versus radiosurgery

Direct costs

Cost-effectiveness Cost-effectiveness

Cost-effectiveness Direct and indirect costs Cost-effectiveness (quality of life outcomes)

was found to be 14% higher than what the nominal Gamma Knife radiosurgery would have been. Actual hospital net receipts were 55% of charges. When hospital receipts were compared with the estimated cost per procedure of Gamma Knife radiosurgery, Gamma Knife radiosurgery had a 30% cost advantage over surgical resection. Wellis et al. [6] examined the direct costs of microsurgical management of radiosurgically amenable intracranial pathologies such as meningiomas, acoustic neuromas, metastases, and arteriovenous malformations in Germany. Costs of the surgical procedure, ICU care, medical and nursing care on the ward, interclinic bills, and overhead for basic hotel service were compared with Gamma Knife radiosurgery direct costs (global operating cost of the Gamma Knife center divided by the number of patients treated in 1999). Average overall costs per surgical patient including ancillary therapy and unplanned readmissions amounted to *15,242 compared with *7920 for Gamma Knife treatment costs per patient.

Surgery Versus Radiosurgery for Single Brain Metastasis Mehta et al. [9] reported on the cost-effectiveness and costutility of radiosurgery versus resection for single brain metastases. Survival and quality of life outcome for radiation alone or with surgery were obtained from two randomized trials [18, 19]. Radiosurgical results were obtained from a multiinstitutional analysis [20] that included patients meeting surgical criteria for resection of a single brain metastasis. When the

666

m. mehta and m.n. tsao

Mehta et al. [9] economic evaluation was published, the results of RTOG 95–08 [21], a randomized trial of whole-brain radiotherapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases, were not known. The randomized trials reported to date have not directly compared surgery and whole-brain radiotherapy versus radiosurgery and whole-brain radiotherapy. In the Mehta [9] economic analysis, only linear accelerator radiosurgery data were considered. The computerized billing records (direct costs) for all patients undergoing resection or radiosurgery for single brain metastases from January 1989 to July 1994 were assessed. During this time frame, there were 46 resections, 135 radiosurgery procedures, and 454 patients treated with whole-brain radiotherapy. There was a 1.8-fold increase in cost associated with surgical resection compared with radiosurgery. Radiosurgery consistently produced superior cost outcomes even when a sensitivity analysis of up to 50% was performed. Average cost per week of survival was US$310 for whole-brain radiotherapy, US$524 for resection plus whole-brain radiotherapy, and US$270 for radiosurgery plus whole-brain radiotherapy. In this report, the median values for QALYs were 12 weeks for whole-brain radiotherapy, 35 weeks for resection and whole-brain radiotherapy, and 51 weeks for radiosurgery plus whole-brain radiotherapy. Rutigliano et al. [2] reported on the cost-effectiveness of stereotactic radiosurgery versus surgical resection in the treatment of single metastatic brain tumors. Three surgical resection studies [22–24] and one radiosurgery study [25] satisﬁed inclusion criteria. Two of the surgical resection series [22, 24] were phase III randomized trials where the control arms of wholebrain radiotherapy alone were used as baseline for the costeffectiveness analysis. A model was developed for typical resource usage for uncomplicated procedures, reported complications, and subsequent craniotomies (for recurrent tumor or radiation necrosis) for both stereotactic radiosurgery and surgical resection. Costs were estimated using the 1992 Medicare Provider Analysis and Review database (US$) with average cost:charge ratios for surgery and whole-brain radiotherapy. A survey of capital and operating costs from ﬁve radiosurgery centers were used for radiosurgery cost estimates. The authors found that radiosurgery had lower uncomplicated procedure costs ($20,209 vs. $27,587), a lower average complication cost per case ($2534 vs. $2874) and a lower total cost per procedure ($22,743 vs. $30,461). Overall, radiosurgery was more costeffective ($24,811 vs. $32,149 per life year) and had better incremental cost-effectiveness ($40,648 vs. $52,384 per life year) compared with surgery. The analysis was robust to large changes in key assumptions. Some of the assumptions in this report included the assumption of one treatment per patient, although some received multiple surgical resections or multiple radiosurgery procedures. The analysis also assumed that posttreatment courses were equal with regard to quality of survival and subsequent neurologic and nonneurologic morbidity. In addition, the majority of costs were assumed to occur immediately and therefore did not require discounting. The costs incurred during the posttreatment interval were assumed to be equivalent both in terms of amount and timing between the two treatments. Sperduto and Hall [15] examined a cost-effectiveness analysis on no treatment, whole-brain radiotherapy alone, wholebrain radiotherapy and radiosurgery, or surgery plus whole-brain radiotherapy in the management of patients with single brain metastasis. It was concluded that whole-brain radiotherapy and

radiosurgery were more cost-effective than whole-brain radiotherapy and surgery. Local brain control was used as an end point, and Medicare reimbursement was used in the cost analysis. These cost analyses [2, 9, 15] on the use of surgery and whole-brain radiotherapy versus radiosurgery and whole-brain radiotherapy for the management of single brain metastasis could not be based on randomized trials that directly compare these two management strategies, as no such trials exist. Prospective clinical trials that examine the clinical efﬁcacy and costeffectiveness of surgical resection and whole-brain radiotherapy versus radiosurgery and whole-brain radiotherapy are obviously needed. However, attempts at comparing resection and radiosurgery in a randomized setting have been fraught with poor accrual due to strong patient and physician bias and choice.

Surgery Versus Radiosurgery for Arteriovenous Malformations Porter et al. [16] used a decision analysis model to evaluate surgery versus stereotactic radiosurgery for small, operable cerebral arteriovenous malformation from a clinical and cost perspective. Probability estimates for cure and complications for both competing therapies were obtained from the literature. Utility values for minor and major stroke were derived from patients with arteriovenous malformations using a standard gamble method. Sources of cost were derived from several hospitals in Ontario, Canada. Using this type of analysis, surgery was associated with an 0.98 QALY advantage over radiosurgery, at an additional cost of Can$6937 per patient or an incremental cost-effectiveness ratio of Can$7100 per QALY for a patient treated surgically. The beneﬁt of surgery over radiosurgery was related to the immediate protection against hemorrhage associated with surgery. Immediate successful surgery spares follow-up costs (repeat imaging and clinical follow-up associated with radiosurgery) and avoids the risk of stroke during the latency period. Sensitivity analysis revealed that the results were sensitive to estimates of surgical morbidity and surgical mortality. The preferred treatment changed to favor radiosurgery at the extreme high end of possible ranges for these two variables (surgical morbidity exceeding 12% or surgical mortality exceeding 4%). Thus, it was concluded that surgery was optimal for small, operable brain arteriovenous malformations. In this scenario (where surgical morbidity and mortality are likely low), surgical resection conferred a clinical beneﬁt of 0.98 QALYs, which represents a gain of almost a full year of perfect health. The incremental cost associated with surgery (Can$6937 per patient) translates to a cost-effectiveness ratio of Can$7100/QALY, which was deemed to be highly economically attractive. Comparisons were made to other accepted therapies such as three-vessel coronary artery bypass graft surgery $12,000/ life-year, pharmacological cholesterol lowering primary prevention $34,000/life-year, and dialysis for end-stage renal disease $51,000/ life-year [26]. The authors recognized limitations of their model. The probability estimates were imprecise due to the lack of high-quality data in the literature. In addition, the model assumed a one-time attempt at treatment and ignored multiple sequential treatments some arteriovenous malformation (AVM) patients experience as a result of unsuccessful initial management. Furthermore, all costs were not considered. The cost of radiosurgery was limited to capital and maintenance

66.

cost-effectiveness and quality of life

costs of the equipment. The cost did not include hospital costs such as angiography, CT/MRI imaging, nursing, physician, physics, electronics, and radiation therapy personnel costs. Fisher et al. [27] in their decision analysis on the management of AVMs, concluded that surgery was preferable to radiosurgery for all AVMs except those that are large or in patients with very high anesthetic risk. In contrast, Hudgins et al. [28] found that radiosurgery was preferable. The ﬂaws of this study include older, less favorable surgical morbidity and mortality rates. Although mortality with hemorrhage during the latency period was considered, morbidity was not.

Surgery Versus Radiosurgery for Acoustic Neuroma Van Roijen et al. [17] reported on the costs and effects of microsurgery versus radiosurgery in treating acoustic neuroma. Cost and effects estimates of surgery were based on a retrospective study of 53 patients treated in The Netherlands. Data from a similar set of 92 patients treated with radiosurgery in Sweden were also obtained from the period 1990–1995. Production losses and quality of life outcomes were obtained by a mailed questionnaire. Direct costs for microsurgery were Dﬂ 20,072 and for radiosurgery Dﬂ 14,272. Indirect costs were respectively Dﬂ 16,400 and Dﬂ 1020. General health rating was also better for radiosurgery compared with microsurgery. Differences in clinical outcomes were said to be small between the two patient groups. It was concluded that radiosurgery was more costeffective than microsurgery.

Conclusion Although it is clear that radiosurgery is a useful technology in the treatment of many conditions, for many disease states where radiosurgery is offered (such as arteriovenous malformation, acoustic neuroma, meningioma), issues regarding effectiveness and safety compared with competing treatment alternatives are yet to be conclusively determined. As such, it may not be possible to conduct completely unbiased and valid costeffectiveness studies. One of the major limitations of economic analyses is the inability to apply results to different health care regions because of major differences in health resource utilization, health-related outcomes, and costs. Furthermore, many economic evaluations indicate that the number of radiosurgery treatments per year is an important variable in assessing the relative costs of treatment alternatives, and this factor clearly differs among radiosurgery centers. In summary, economic analyses have become inﬂuential for health care policymakers at various levels of decision making TABLE 66-4. Quality of life uses. Economic analyses (e.g., cost-utility analysis) Assessing patient preferences (utilities) Expressing outcomes by combining quality and quantity (e.g., qualityadjusted life years, or QALYs, and quality of time without symptoms and treatment, or Q-TwiST) Choosing among treatments with equal beneﬁts and risks Source: Osoba D. Measuring the effect of cancer on quality of life. In: Osoba D, ed. Effect of Cancer on Quality of Life. Boca Raton: CRC Press, 1991:29. Used with permission.

667

TABLE 66-5. Examples of quality of life instruments. 1.

2.

3.

Global scales Karnofsky performance status ECOG performance status Spitzer quality of life index Psychosocial scales Psychosocial adjustment to illness scale Sickness impact proﬁle Brain tumor multidimensional quality of life instruments FACT-BR EORTC quality of life questionnaire (brain module)

FACT-BR. Functional Assessment of Cancer Treatment-Brain; EORTC, European Organisation for Research and Treatment of Cancer.

(e.g., politicians, health care insurers, health care managers, health care professionals, and medical equipment manufacturers). Such analyses, however, do not form the only basis for decisions. Where comparisons of effectiveness and safety are not well established among different management approaches, patient/physician preferences and accessibility play a major role in decision making at an individual level.

Principles of Quality of Life One type of economic analysis uses quality of life outcomes and expresses results as units of time adjusted for value or quality (e.g., quality adjusted life year gained) rather than dollar amounts. Apart from an economic point of view, the aims of medical management include outcomes that are quantitative (e.g., duration of tumor control) or qualitative (e.g., preservation of function and well-being or quality of life). The World Health Organization [29] deﬁned health as “a state of complete physical, mental and social well-being and not merely the absence of disease or inﬁrmity.” Multidimensional quality of life surveys measure domains such as physical functioning, social and role functioning, mental/emotional wellbeing, and general health perceptions. Examples on the use of quality of life data are presented in Table 66-4. In order to develop multidimensional quality of life instruments, basic steps are undertaken. These are item selection, item reduction, establishing reproducibility, demonstrating validity, and conﬁrming responsiveness. Items that might be important to include in the ﬁnal instrument may be collected using focus groups of patients and/or health care providers. The number of items are then reduced in number to remove duplication of similar items and to limit the number of items with discriminative ability. Questions in which most or all of the respondents give similar or identical answers are not useful. Reproducibility refers to the reliability or precision of the instrument. Serial administration of the test to a group of subjects is undertaken to assess whether the instrument is stable with respect to inter- and intrasubject variation. Statistical tools such as Pearson’s correlation coefﬁcient or intraclass correlation coefﬁcients can be used to measure reproducibility. Validity may be assessed using criterion validity or construct validity. With criterion validity, the developed instrument is tested against a gold standard. Construct validity uses theoretical hypotheses to test whether the developed instrument is consistent with concepts (or constructs) being measured. Responsiveness refers to the ability of the instrument to detect differences when a difference exists. Table 66-5 lists commonly used quality of life instruments.

668

m. mehta and m.n. tsao

Quality of Life and Radiosurgery A literature search was undertaken using the following databases: Medline (1990–2004), CancerLit (1990–2004), CINAHL (1990–2004), EMBASE (1990–2004), and the Cochrane Library (2004, issue 4). There were 14 relevant studies that reported on quality of life and radiosurgery (summarized in Table 66-6).

Radiosurgery for Brain Metastases and Quality of Life In the randomized controlled trial published by Andrews et al. [21] or RTOG 95-08, 164 patients with 1 to 3 newly diagnosed brain metastases were randomized to whole-brain radiotherapy

versus whole-brain radiotherapy and radiosurgery boost. Overall median survival time did not vary between the two groups (6.5 months vs. 5.7 months, respectively, p = 0.1356). However, patients with single brain metastasis had better median survival with radiosurgery boost, 6.5 months versus 4.9 months, than patients treated with whole-brain radiotherapy alone, p = 0.0393. One-year local brain control was also better in the radiosurgery boost arm, 82% versus the whole-brain radiotherapy alone arm, 71%, p = 0.01. Furthermore, patients in the radiosurgery boost group were more likely to have a stable or improved Karnofsky performance source (KPS) at 6 months follow-up than patients allocated to the whole-brain radiotherapy alone arm (43% vs. 27% respectively, p = 0.03). Aoyama et al. [31] reported on a randomized trial examining the use of radiosurgery alone versus whole-brain radio-

TABLE 66-6. Radiosurgery and quality of life studies. Reference

Patient characteristics

Study design

Intervention

Quality of life instrument

Brain metastases Andrews et al. [21]

Brain metastases

Randomized controlled trial

KPS

Aoyama et al. [31]

Brain metastases

Randomized controlled trial

DiBiase et al. [32]

Brain metastases

Prospective

Whole-brain radiotherapy and radiosurgery versus whole-brain radiotherapy Radiosurgery versus radiosurgery and whole-brain radiotherapy Radiosurgery

Auchter et al. [20]

Brain metastases

Retrospective

Whole-brain radiotherapy and radiosurgery

High-grade glioma Souhami et al. [33]

Glioblastoma multiforme

Randomized controlled trial

Spitzer quality of life survey

Mehta et al. [34]

Glioblastoma multiforme

Prospective

Shenouda et al. [35]

Glioblastoma multiforme

Prospective

Shrieve et al. [36]

Glioblastoma multiforme

Retrospective

Radiotherapy and BCNU versus radiosurgery and radiotherapy and BCNU Radiotherapy and radiosurgery boost Accelerated radiotherapy and radiosurgery boost Radiotherapy and radiosurgery boost

Retrospective

Radiosurgery

Modiﬁed Spitzer quality of life scale Pellet [39] functional quality of life questionnaire Glasgow Beneﬁt Inventory

Brain metastases and high-grade glioma Jagannathan et al. [37] Brain metastases Primary malignant glioma

KPS

Spitzer quality of life Survey KPS

KPS KPS KPS

Acoustic neuroma Regis et al. [38]

Acoustic neuroma

Prospective

Microsurgery versus radiosurgery

Sandooram et al. [40]

Acoustic neuroma

Retrospective

Microsurgery versus radiosurgery versus conservative management

Retrospective

Radiosurgery

Duke–University of North Carolina Health Proﬁle

Retrospective

Radiosurgery

Global quality of life scale

Retrospective

Radiosurgery

Global quality of life scale

Arteriovenous malformation Lai et al. [41] Arteriovenous malformation Trigeminal neuralgia Petit et al. [42]

Herman et al. [43]

Trigeminal neuralgia refractory to medical or surgery Repeat radiosurgery for trigeminal neuralgia

66.

cost-effectiveness and quality of life

therapy and radiosurgery as initial therapy in patients with four or fewer brain metastases. An interim analysis was reported to show no difference in survival between the two arms. However, 6-month freedom from new brain metastases was signiﬁcantly favored in the whole-brain and radiosurgery arm, 82% versus 48%, than in the radiosurgery-alone arm, p = 0.019. Actuarial 1-year KPS preservation rate (KPS at least 70) was 25% in the radiosurgery-alone arm and 37% in the whole-brain and radiosurgery arm (p = 0.54). DiBiase et al. [32] reported a prospective study of 20 patients undergoing radiosurgery for brain metastases. All patients responded to the Spitzer quality of life survey before radiosurgery and at each follow-up visit. The median followup time was 7 months. In patients with extracranial tumor progression, Spitzer scores tended to decrease. In patients with no evidence of intracranial or extracranial tumor progression, Spitzer quality of life scores remained unchanged or improved. Not surprisingly, tumor progression (either intracranially or extracranially) adversely affected quality of life scores. Auchter et al. [20] reported on the results from radiosurgery databases from four institutions pertaining to patients with single brain metastases treated with radiosurgery. Whole-brain radiotherapy was used in all but ﬁve patients. The median duration of functional independence (KPS at least 70) was 44 weeks. The authors concluded that radiosurgery with whole-brain radiotherapy can produce substantial functional survival especially among patients with good baseline performance status without active extracranial metastases.

Radiosurgery for High-Grade Glioma and Quality of Life Souhami et al. [33] reported on the randomized trial (RTOG 93–05) that examined patients with glioblastoma multiforme treated with radiosurgery boost followed by radiotherapy and BCNU versus radiotherapy and BCNU. There was no difference in overall survival between the two arms. No improvement in quality of life as measured by the Spitzer index at baseline compared with the end of therapy in patients treated with radiosurgery boost was seen. Furthermore, there was no difference in quality-adjusted survival between the two arms. Mehta et al. [34] reported on 31 patients treated with radiotherapy and radiosurgery boost for glioblastoma multiforme. Median actuarial survivals with the radiosurgery boost cohort were found to be better than historical controls stratiﬁed by recursive partitioning analysis (RPA) class. Median KPS was found to have increased right after radiosurgery, peaking again at 15 to 18 months. Shenouda [35] reported on a prospective cohort of 14 glioblastoma multiforme patients treated with accelerated radiotherapy followed by radiosurgery boost. Unlike Mehta’s [34] report, median survivals were not superior to historical controls. Median KPS at diagnosis was 80. KPS scores were maintained until 48 weeks when the median dropped to 60. Shrieve et al. [36] reported on 78 newly diagnosed glioblastoma multiforme patients treated with radiotherapy and radiosurgery boost. This retrospective analysis found survival to be better than historic controls. Mean KPS at diagnosis was 91.5, which dropped to a mean of 87 at last follow-up.

669

Radiosurgery for Brain Metastases/High-Grade Glioma and Quality of Life Jagannathan et al. [37] examined 22 patients (16 patients with brain metastases and 6 patients with primary malignant glioma) treated with Gamma Knife radiosurgery. Quality of life was assessed in 20 or the 22 patients using a modiﬁed Spitzer scale. These selected patients were found to have high quality of life scores (mean score 8.65 of 10), and only one patient had a KPS score of less than 70.

Radiosurgery for Acoustic Neuroma and Quality of Life Regis et al. [38] used a patient questionnaire [39] to assess functional outcome and quality of life in 97 patients who were more than 3 years post–Gamma Knife radiosurgery compared with 110 patients treated with microsurgery. All patients who underwent radiosurgery had no new facial motor symptoms compared with 63% who underwent microsurgery. Among patients without baseline facial sensory deﬁcit, 29% treated with microsurgery and 4% treated with radiosurgery developed new facial sensory disturbance, p = 0.0009. Among patients with baseline hearing (Gardner Robertson class I), hearing preservation after treatment (Gardner Robertson scale class I or II) was also better in the Gamma Knife group compared with the microsurgery-treated group, 70% versus 37.5%, respectively. Ninety-one percent of patients in the Gamma Knife group reported no change in daily life compared with 61% of microsurgery-treated patients, p = 0.00017. Psychobehavioral problems were reported in 69% of patients after microsurgery compared with 24% after Gamma Knife radiosurgery. Time lost from work was also shorter after Gamma Knife treatment with a mean of 7 days versus 130 days with microsurgery. It was concluded that Gamma Knife radiosurgery was associated with better functional and quality of life outcomes compared with microsurgery. Sandooram et al. [40] reported on quality of life in 165 patients with acoustic neuroma managed with microsurgery, radiosurgery, or conservative management. This was a retrospective study where baseline quality of life assessments before intervention were not obtained. Patients were mailed the Glasgow Beneﬁt Inventory, a validated questionnaire designed to assess quality of life after otorhinolaryngological interventions. The instrument consists of a general subscale (12 items regarding a change in optimism), a physical health subscale (3 items regarding a change in medications), and a social support subscale (3 items regarding a change in family support). The responses are based on a 5-point Likert scale ranging from large deterioration to large improvement in health status. The study reported that quality of life deteriorated after microsurgery, conservative management did not lead to a change, and radiosurgery resulted in a trend toward poorer quality of life.

Radiosurgery for Arteriovenous Malformation and Quality of Life Lai et al. [41] reported on the quality of life of patients with arteriovenous malformations during the latency period between radiosurgery and lesion obliteration. Thirty-nine patients were

670

m. mehta and m.n. tsao

administered the Duke–University of North Carolina Health Proﬁle. The authors concluded that the quality of life in patients with cerebral arteriovenous malformations during the latency period after radiosurgery deteriorated as a result of irreversible physical disabilities rather than the knowledge of persistent hemorrhage risk and bleeding experience.

Radiosurgery for Trigeminal Neuralgia and Quality of Life One-hundred twelve patients with trigeminal neuralgia treated with Gamma Knife radiosurgery were reviewed by Petit et al. [42]. Eighty-six percent of patients completed questionnaires at a median follow-up of 30 months (range, 8 to 66 months). Seventy-seven percent of patients reported pain relief occurring after a median of 3 weeks (range, 0 to 24 weeks) after Gamma Knife radiosurgery. Sixty-percent of patients reported a decrease in medication use. A three-question global quality of life questionnaire was used. (“In general, would you consider the procedure to be a successful treatment in your case? Yes, no, not sure. Has your quality of life been improved since the procedure? Yes, no, not sure. If yes, please choose the amount you feel this procedure improved your quality of life as a percent from 0–100%?”). Patients reported a median 80% improvement in their quality of life associated with pain relief after Gamma Knife radiosurgery. The patients experiencing pain relief at the time of answering the questionnaire reported a median of 100% improvement in their quality of life. Those patients who experienced temporary pain relief (pain recurrence after a median of 8.5 months) reported a median of 80% improvement in quality of life. Eighteen patients [43] with trigeminal neuralgia who underwent repeat Gamma Knife radiosurgery for unsustained pain response were administered a global quality of life instrument where 0 represented no improvement in quality of life and 100% represented 100% improvement in quality of life. After repeat radiosurgery, patients who experienced pain relief reported varying degrees of beneﬁt in quality of life (average 85% range 60% to 90%). Some patients with no pain response still reported some improvement in quality of life (range, 0 to 30%).

2.

3. 4. 5.

6.

7. 8.

9.

10.

11.

12. 13.

14.

15.

16.

Conclusion The beneﬁts of quality of life assessments include an examination of patients’ perspectives relating to all aspects of health. Scientiﬁc methodology exists in the development of quality of life tools, its implementation, and its interpretation. Challenges such as low levels of respondents, attrition rates, missing data, and cultural and educational selection of respondents exist. As the ﬁeld of radiosurgery matures and expands, more studies will be performed examining quality of life end points, which clearly will contribute to our understanding of the multidimensional health outcomes associated with radiosurgery.

17.

18.

19.

20.

References 1. Rutten F, Drummond M. Making decisions about health technologies: a cost-effectiveness perspective. York: The University of

21.

York Centre for Health Economics, Institute for Medical Technology Assessment, 1994:6–36. Rutigliano MJ, Lunsford LD, Kondziolka D, et al. The cost effectiveness of stereotactic radiosurgery versus surgical resection in the treatment of solitary metastatic brain tumors. Neurosurg 1995; 37(3):445–53; discussion 453–455. Penar PL. Cost and outcome analysis. Neurosurg Clin N Am 1996; 7(3):547–558. Eisenberg JM. Clinical economics. A guide to the economic analysis of clinical practices. JAMA 1989; 262:2879–2886. Weinstein MC. Principles of cost-effectiveness resource allocation in health care organizations. Int J Technol Assess Health Care 1990; 6:93–103. Wellis G, Nagel R, Vollmar C, et al. Direct costs of microsurgical management of radiosurgically amenable intracranial pathology in Germany: an analysis of meningiomas, acoustic neurones, metastases and arteriovenous malformations of less than 3 cm in diameter. Acta Neurochir (Wien) 2003; 145(4):249–255. Taylor TN, Davis PH, Torner JC, et al. Lifetime costs of stroke in the United States. Stroke 1996; 27:1459–1466. Van Roijen L, Nijs HG, Avezaat CJ, et al. Costs and effects of microsurgery versus radiosurgery in treating acoustic neuroma. Acta Neurochir (Wien) 1997; 139(10):942–948. Mehta M, Noyes W, Craig B, et al. A cost-effectiveness and costutility anlaysis of radiosurgery vs. resection for single-brain metastases. Int J Radiat Oncol Biol Phys 1997; 39(2):445–454. Drummond MF, Brandt A, Luce BR, et al. Standardizing economic evaluations in health care: practice, problems and potential. Int J Tech Asssess Health Care 1993; 9(1):26–36. Buxton MJ. Problems in economic appraisal of new health technology: the evolution of heart transplants in the UK. In: MF Drummond, ed. Economic Appraisal of Health Technology in the European Community. Oxford: Oxford University Press, 1987: 1–272. Epstein ME, Lindquist C. Cost accounting the gamma knife. Stereotact Funct Neurosurg 1993; 61(Suppl 1):6–10. Konigsmaier H, De Pauli-Ferch B, Hackl A, et al. The costs of radiosurgical treatment: comparison between gamma knife and linear accelerator. Acta Neurochir (Wien) 1998; 140(11):1101– 1110, discussion 1110–1111. Ott K. A comparison of craniotomy and Gamma Knife charges in a community-based Gamma Knife Center. Stereotact Funct Neurosurg 1996; 66(Suppl 1):357–364. Sperduto PW, Hall WA. The cost-effectiveness of alternative treatment for single brain metastasis. In: Kondziolka D, ed. Radiosurgery. Basel: Karger, 1996:180–187. Porter PJ, Shin AY, Detsky AS, et al. Surgery versus stereotactic radiosurgery for small, operable cerebral arteriovenous malformations: a clinical and cost comparison. Neurosurgery 1997; 41(4):757–764; discussion 764–766. Van Roijen L, Nijs HG, Avezaat CJ, et al. Cost and effects of microsurgery versus radiosurgery in treating acoustic neuroma. Acta Neurochir (Wien) 1997; 139(10):942–948. Noordijk EM, Vecht CJ, Haaxma-Reiche J, et al. The choice of treatment of single brain metastasis should be based on extracranial tumor activity and age. Int J Radiat Oncol Biol Phys 1994; 29:711–717. 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. Auchter RM, Lamond JL, Alexander EI, et al. A multiinstitutional outcome and prognostic factor analysis of radiosurgery (RS) for resectable single brain metastases. Int J Radiat Oncol Biol Phys 1996; 35:27–35. Andrews DW, Scott CB, Sperduto PW, et al. Whole brain radiation therapy with or without stereotactic radiosurgery boost for

66.

22.

23.

24.

25.

26.

27. 28. 29. 30.

31.

32.

33.

cost-effectiveness and quality of life

patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004; 363:1665–1672. Patchell RA, Tibbs PA, Walsh JW, et al. A randomized trial of surgery in the treatment of single metastasis to the brain. N Engl J Med 1990; 322:494–500. Sundaresan N, Galicich JH. Surgical treatment of brain metastasis: clinical and computerized tomography evaluation of the results of treatment. Cancer 1985; 55:1382–1388. Vecht CJ, Haaxma-Reiche H, Noordijk EM, et al. Treatment of single brain metastasis: Radiotherapy alone or combined with neurosurgery? Ann Neurol 1993; 33:583–590. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multiinstitutional experience with stereotactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol Biol Phys 1994; 28:797–802. Tengs T, Adams M, Pliskin J, et al. Five-hundred life-saving interventions and their cost-effectiveness. Risk Anal 1995; 15:369– 390. Fisher WS III. Therapy of AVMs: A decision analysis. Clin Neurosurg 1995; 42:295–311. Hudgins WR. Decision analysis of the treatment of AVMs with radiosurgery. Stereotact Funct Neurosurg 1993; 61:11–19. World Health Organization. Constitution of World Health Organization. In: Basic Documents. Geneva: WHO, 1948. Osoba D. Measuring the effect of cancer on quality of life. In: Osoba D, ed. Effect of cancer on quality of life. Boca Raton: CRC Press, 1991:29. Aoyama H, Shirato H, Nakagawa K, et al. Interim report of the JROSG99–1 multi-institutional randomized trial comparing radiosurgery alone vs. radiosurgery plus whole brain irradiation for 1–4 brain metastases. Proceedings from the 40th Annual Meeting of ASCO 2004; 23 (abstract 1506). DiBiase SJ, Chin LS, Ma L, et al. Inﬂuence of gamma knife radiosurgery on the quality of life in patients with brain metastases. Am J Clin Oncol 2002; 25(2):131–134. Souhami L, Seiferheld W, Brachman D, et al. Randomized comparisons of stereotactic radiosurgery followed by conventional radiation therapy with carmustine to conventional radiation therapy with carmustine for patients with glioblastoma multi-

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

671

forme. Report of the Radiation Therapy Oncology Group 93–05 protocol. Int J Radiat Oncol Biol Phys 2004; 60:853–860. Mehta MP, Masciopinto J, Rozental J, et al. Stereotactic radiosurgery for glioblastoma multiforme: report of a prospective study evaluating prognostic factors and analyzing long-term survival advantage. Int J Radiat Oncol Biol Phys 1994; 30:541– 549. Shenouda G, Souhami L, Podgorsak EB, et al. Radiosurgery and accelerated radiotherapy for patients with glioblastoma. Can J Neurol Sci 1997; 24:110–115. Shrieve DC, Alexander E, Black PM, et al. Treatment of patients with primary glioblastoma multiforme with standard postoperative radiotherapy and radiosurgical boost: prognostic factors and long-term outcome. J Neurosurg 1999; 90:72–77. Jagannathan J, Petit JH, Balsara K, et al. Long-term survival after gamma knife radiosurgery for primary and metastatic brain tumors. Am J Clin Oncol 2004; 27:441–444. Regis J, Pellet W, Delsanti C, et al. Functional outcome after gamma knife surgery or microsurgery for vestibular schwannomas. J Neurosurg 2002; 97:1091–1100. Pellet W, Emram B, Cannoni M, et al. Les resultats fonctionnels de la chirurgie des neurinomes de l’acoustique unilateraux. Neurochirurgie 1993; 29:24–41. Sandooram D, Grunfeld EA, Mckinney C, et al. Quality of life following microsurgery, radiosurgery and conservative management for unilateral vestibular schwannoma. Clin Otolaryngol 2004; 29:621–627. Lai EHP, Lun SLC. Impact of quality of life of patients with arteriovenous malformations during the latent interval between gamma knife radiosurgery and lesion obliteration. J Neurosurg 2002; 97(Suppl 5):471–473. Petit JH, Herman JM, Nagda S, et al. Radiosurgical treatment of trigeminal neuralgia: evaluating quality of life and treatment outcomes. Int J Radiat Oncol Biol Phys 2003; 56:1147–1153. Herman JM, Petit JH, Amin P, et al. Repeat gamma knife radiosurgery for refractory or recurrent trigeminal neuralgia: treatment outcomes and quality of life assessment. Int J Radiat Oncol Biol Phys 2004; 59:112–116.

6 7

Regulatory and Reimbursement Aspects of Radiosurgery Rebecca Emerick

Historical Perspective

510K Premarket Approval

When neurosurgeon Lars Leksell and radiobiologist Börje Larsson ﬁrst collaborated to use the proton beam at the University of Uppsala in Sweden, one would doubt that they were concerned whether the procedure would be reimbursed by the health care system. The development of radiosurgery throughout the 1950s to 1970s was largely funded by private and public research monies. According to Professor Erik-Olof Backlund, MD, PhD, during the early era of radiosurgery, procedures by Lawrence’s group (using heavy particles at Berkeley), Kjellberg (using the cyclotron at Harvard University), and Leksell (who traveled from Stockholm to Uppsala to use the proton beam to treat patients) were either free to patients or the physician accepted what the patient was able to pay (E.O. Backlund, personal conversations, 1996–1999). With the installation of the Gamma Knife unit in late 1967 at the private Sophiahemmet Hospital in Stockholm, it is assumed that some payment may have been made by patients toward the costs of their procedures. Further development in Europe and Asia would see most procedures paid for by socialized medicine–a situation that was exactly the opposite in the United States.

The ﬁrst step in the approval process was to obtain the U.S. Food and Drug Administration (FDA) 510K Premarket Approval before the Gamma Knife device could be used on humans. Insurers and Medicare require that a device has 510K approval as a ﬁrst step toward gaining approval for the patient procedure and receiving reimbursement [3]. A 510K Premarket Approval is essentially a private license granted to the applicant for marketing the medical device [4]. Premarket applications to receive 510K approval require the applicant to demonstrate the new device to be marketed is as safe and effective as one or more similar devices currently on the U.S. market that is a legally marketed device not subject to premarket approval. This can be accomplished by providing supporting documentation that the new device is substantially equivalent (SE) to an already legally marketed predicate device. As a neurosurgical device that delivers radiation, the Gamma Knife was found by the FDA to be substantially equivalent to existing approved radiation devices. The medical device 510K process establishes safety and allows the manufacturer to begin marketing the device. The process does not establish that use of the device will result in a positive clinical beneﬁt for those treated. This is unlike the process of receiving FDA marketing approval for a new drug, in which positive clinical outcomes are required to be shown from previously conducted clinical trials. In 1982, the FDA approved marketing of the Gamma Knife unit in the United States [2]. A full decade later, the linear accelerator radiosurgery targeting software for Radionics’ XKnife received its 510K approval in 1993. The Phillips SRS200 soon followed, as did a dedicated X-Knife unit, and radiosurgery was ﬁrmly entrenched in the United States. The 510K Premarket Approval for medical devices has ﬂaws, and manufacturers routinely take advantage of the process. For example, in 1999 a 510K approval was obtained for a robotic radiosurgery device by using an existing gantry radiosurgery device as the compared substantial equivalent [5].

Government Oversight For 4½ years in the 1980s, neurosurgeon L. Dade Lunsford and Swedish manufacturer Elekta Instruments, AB, began the daunting task of obtaining government approval to perform Gamma Knife procedures within the United States [1]. Dr. Lunsford had completed a fellowship in stereotactic surgery at the Karolinska Institute in Stockholm in 1981 [2]. At the time, there were three other Gamma Knife units worldwide in Sweden, Argentina, and the United Kingdom. The approval process was extremely challenging at the time but is largely the same today.

673

674

r. emerick

Although the 510K process was established to ensure safety for patients, it has become stymied as the U.S. Occupational Safety and Health Administration (OSHA) regulates worker safety and the FDA oversees patient safety, and neither agency has addressed the safety of moving radiosurgery units in the medical arena.

U.S. Nuclear Regulatory Commission With the Gamma Knife, the safety and approval of the radioactive cobalt sources fell under the auspices of the U.S. Nuclear Regulatory Commission (NRC), which required regulations to be drafted to cover the training, safety, and operation of the units within hospitals. The NRC deﬁnes the role and training of the Authorized Medical Physicist and the Authorized User during “medical use of by-product materials or radioisotopes” [6]. The NRC’s authority to regulate the medical use of byproduct materials is based on the Atomic Energy Act of 1954 (as amended). Whereas it can be expected that the authority of the FDA and NRC might overlap, the respective agencies have been able to avoid conﬂict by interagency agreement. The NRC relinquishes its individual oversight of most byproduct medical devices to “Agreement States.” Agreement States are those U.S. states that have entered into an agreement with the NRC to regulate this area through their existing government structures or Tribal Programs. Currently, approximately 17 states are not agreement states, preferring to let the NRC itself handle this area. Agreement states have up to 3 years to implement most changes in regulations made by the NRC. The NRC can require immediate adherence to a new regulation. In January 1986, the NRC allowed device registration for the ﬁrst Gamma Knife unit at the Presbyterian University Hospital in Pittsburgh after a multistage review lasting almost 3½ years [2]. The NRC operates under the policy statement published on February 9, 1979 (44 CFR 8242), which essentially is still in effect today. The three-part 1979 policy stated: 1. The NRC will continue to regulate the medical uses of radioisotopes as necessary to provide for the radiation safety of workers and the general public. 2. The NRC will regulate the radiation safety of patients where justiﬁed by the risk to patients and where voluntary standards, or compliance with these standards, are inadequate. 3. The NRC will minimize intrusion into medical judgments affecting patients and into other areas traditionally considered to be a part of the practice of medicine. (A change to this wording was added in the late 1990s: The NRC should not interfere in the practice of medicine.) These simple guidelines were well thought out for the year 1979, when even the idea of radiosurgery in the United States or an understanding of brain surgery with radiation was simply science ﬁction. However, today the guidelines are conﬂicting within the NRC itself. Although the NRC has stated it will minimize intrusion into medical judgments and the practice of medicine, through its Federal Register releases it has regulated which physician specialty will be present during the procedure,

which physician specialty will conduct the radiosurgery procedure, and which physician specialty will sign the medical directive. Perhaps better understanding the safety issues involved, some of the NRC’s own regions that govern Non-Agreement States have required the presence of additional physician specialties as a condition to licensing operations. The NRC is intruding into the practice of medicine, possibly creating an unsafe environment in some instances. The NRC appears to lack an appropriate understanding of the procedure and has recently allowed neurosurgeons to be exempt from radiosurgery procedures and has hampered an already short supply of radiation oncologists by requiring they remain within 20 feet of the treatment console during every procedure [7]. Considering that a physicist and a minimum of a registered nurse are also present at the console where a procedure can be stopped with the simple push of one button, the requirement is burdensome and overly restrictive. The requirement that a radiation oncologist to remain at the console (or within 20 feet) is leading a growing number of radiosurgery providers to comment on the adverse effect of this policy and the impending curtailment of other clinical services during the same period in their radiation oncology departments. In radiation oncology, it is generally accepted that remaining present during a procedure is considered physically being in the department or able to respond within a minute or so by pager if needed. Typically, a technician or therapist might direct the treatment after the physician (the Authorized User) has approved the treatment plan. The NRC is clearly not aware of current operational practices for both neurosurgery and radiation oncology. Perhaps most disturbing in this situation has been a growing “ownership” issue by the specialties involved. Gamma radiosurgery providers are an extremely small group of providers composed of approximately 110 hospitals at this time; the addition of linac radiosurgery providers brings the number to around 300 to 350 total. Considering that there are more than 5000 hospitals, this small subspecialty procedure lacks understanding and garners little regard by its own professional associations. Stereotactic radiosurgery is a small subspecialty that is routinely ignored by the greater needs of both neurosurgery and radiation oncology when routine policy decisions are made without consulting the actual provider members of stereotactic radiosurgery, or by catering to the unprofessional insistence of a few vocal members who see radiosurgery as “crushable” competition within their own ranks for scant reimbursement dollars. In late 2002, the NRC approved a gamma unit in an environment not covered by hospital oversight and credentialing. The unit was licensed in a small plaza to an ownership group composed of ﬁnancial investors without clinical background. As this type of nonhospital environment grows for by-product medical devices, it will further complicate the matter of patient safety. Today, the NRC appears to bow to politics while trying to stay neutral over issues that it has brought upon itself by violating its own tenet concerning intruding into the medical judgment area and by licensing units in unregulated environments. The solution may require more regulatory intrusion in the future to undo issues that the NRC’s own regulations have caused with patient safety. Government regulators will need to decide whether or not they are in charge of just safety and access to by-product materials or if they must intrude into the

67.

regulatory and reimbursement aspects of radiosurgery

area of medical decision making by regulating most aspects of the medical procedure. In direct contrast with the NRC’s regulatory strictness toward gamma-type units that house radioactive cobalt sources, regulation of the radiosurgery linear accelerator industry falls under the department of health in the state in which the medical device resides and receives very little regulatory oversight. The NRC should aspire to be part of the solution by understanding the general practice of the radiation oncology professional and gamma unit operations with its multidisciplinary nature and should work with the industry to provide appropriate regulations for different environments.

Certiﬁcate of Need In May 1985, the ﬁrst Certiﬁcate of Need (CON) approval from Southwestern Pennsylvania Regional Council of the Health Systems Agency was granted to Presbyterian University Hospital for the ﬁrst installation of a Gamma Knife in the United States. Later in September 1985, the Pennsylvania Board of Health approved the CON [2]. Certiﬁcate of Need procedures still exist today, whereby the state evaluates expenditures for medical devices over a certain dollar amount (usually $1 million) to determine whether the state has a need for the device, to assess the need for additional devices or determine whether the device is an unnecessary use of health care dollars. Although a few states have dropped their CON programs, there are still approximately 38 states with a formal CON process. The primary purpose of the CON process has been to limit expenditures and avoid duplication. Certiﬁcate of Need processes are lengthy and involved. They require a formal application from an existing health care provider. The provider must give an extensive and costly analysis of how the new device will be paid for, what services it will replace and/or enhance, estimated usage of the device, and an estimate of how many of the state’s indigent will be able to use the device. All evidence must be provided with some type of documentation. Certiﬁcate of Need processes have the unintended effect of impeding new organizations from entering the market. This protects the existing providers and most likely slows down progress toward a more efﬁcient and cost-effective health care environment. The CON process is by nature adversarial and political [8]. It is generally accepted that academic, politically linked, and well-endowed hospitals are usually granted their application after some scrutiny and a public hearing. Local health care entities are free to oppose a CON application and mount a case for its refusal. This usually serves to deeply divide the local health care community. In pursuit of maintaining a market position, a CON application may be opposed by other providers who hold the position that there is no further need for an additional device. Most states have a dedicated and highly educated staff within their departments of health that diligently analyze and seek appropriate information in order to evaluate an application. Public testimony is normally allowed for the applicant, the opposing providers, manufacturers, and relevant parties including associations.

675

In recent CON applications for stereotactic radiosurgery services, it has been discouraging to see unscrupulous assertions about professionals and unsubstantiated testimony without clinical basis presented by medical device manufacturers in order to gain revenue through sales. Because the presentation exists within a public forum, there is little legal action that can be relied upon to refute suspect testimony or defend a professional. The staff for a deciding department of health weighs much evidence and decides what is best for their state and the patients’ needs within the state. The staff must weigh the total cost of new devices including stafﬁng, installation, maintenance agreements, and costs of their decision to the general public. The staff for a highly contentious CON process can only be held in high esteem. It is typical for the CON staff on the state level to consider what other states are doing and to call on organizations, manufacturers, or associations with interest in the speciﬁc device for their knowledge and experience. The CON process must also consider the excess capacity of existing medical technology and where competing providers can share such technology. The CON process is concerned with positive evidencebased clinical outcomes for a device, the need to serve the population, the copayments and out-of-pocket costs to each patient treated, and the loss of economic work hours to the patient. Stereotactic radiosurgery speaks positively to each of these concerns.

Coding When the Pittsburgh Gamma Knife began treating patients in August 1987, reimbursement issues were paramount. At the time there was no speciﬁc coding for radiosurgery, but more than 10 insurance companies (including Blue Cross of New York, California, and Pennsylvania) had previously reimbursed patients treated at the Gamma Knife units in Sweden and Argentina [2]. After fulﬁlling Blue Cross’ requirements for reimbursement of new technology with appropriate data and research literature, and appealing former denials, reimbursement was approved in November 1987. In December 1987, the American Medical Association established the Current Procedural Terminology (CPT) code of 61793 for all stereotactic radiosurgery with both a professional and technical part to the code. In the 1980s and early 1990s, most stereotactic radiosurgery procedures were conducted as an inpatient stay. At the time, the average length of stay was reported as 2.24 days for radiosurgery and 11.44 days for open skull surgery [2]. Costs were reported to be 30% to 70% less compared with surgical resection. As radiosurgery replaced a craniotomy, Diagnostic Related Grouping (DRG) 1, which was designated for surgical craniotomy, was used to obtain reimbursement from payors and Medicare. To receive reimbursement for an outpatient procedure when linac technology was introduced in 1993, the primary CPT code 61793 was used to represent stereotactic radiosurgery for hospital providers. In early 1995, Elekta Instruments, Inc., and consultants met with the Health Care Financing Administration (HCFA), which is now the Centers for Medicare and Medicaid (CMS), to discuss

676

r. emerick

a code for stereotactic radiosurgery. HCFA proposed and ﬁnalized the procedure code 92.3 for stereotactic radiosurgery to begin use in October 1995. The procedure code would pay at the surgical DRG 1 rate of payment. DRG 1 was used exclusively for inpatient reimbursement of craniotomies. It was the establishment of this new procedure code that allowed HCFA to isolate and track stereotactic radiosurgery’s cost and clinical indications treated, which would result in lower reimbursement in the future. In 1997, HCFA analyzed 18 months of its data on radiosurgery and concluded that it was overpaying the procedure in DRG 1 and moved the procedure to new DRGs (DRGs 7, 8, 292, 293, 401, 402, and 408) to begin use in October 1997. The primary DRGs were now DRG 7 and 8. This meant a substantial reduction in reimbursement for treating Medicare patients. In 2008, inpatient stereotactic radiosurgery procedures continued to be reimbursed from the same DRGs for Medicare patients. HCFA had compared the charges from all radiosurgery procedures to the charges for all other craniotomy procedures and found they were substantially less. Although it resulted in a loss of reimbursement, it also clearly showed that stereotactic radiosurgery procedures were less costly than open skull craniotomies on a national level.

Cost of Stereotactic Radiosurgery The assessment and management of the costs of stereotactic radiosurgery (SRS) technology has always been accepted as vital to health care providers. It became apparent as early as 1995 that regardless of the costs and installation of the type of technology (cobalt-60 multisource, linac dedicated, or linac modiﬁed), at some reasonable patient volume (reported as 150 to 170 annually), the cost per procedure was comparable [9]. Since that time, a number of articles have been published attesting to the health care dollar savings of 30% when SRS can be utilized over open skull surgery [10]. For 2008, the cost of onesession SRS was documented to cost less than hypofractionation with image-guided robotic SRS systems. Robotic SRS systems currently have reported annual maintenance costs that are 400% higher than one-session equipment and require daily quality and assurance testing and reveriﬁcation of the target

planning with each treatment, thus raising the costs per procedure. Although robotic systems have the ability to treat extracranially, the required quality assurance and daily time to treat the patient may limit the number of patients treated each week compared with the capacity of one-session SRS technology. Each potential program must decide for itself the type of radiosurgery program (dedicated intracranial or whole body), the disease indications, and the costs it is willing to assume in making the decision to purchase among the equipment vendors that currently exist. In the past, one was dependent upon various radiosurgery centers to publish information on cost savings of SRS over surgical interventions. Today, we can look to large publicly available data sets maintained by government agencies. The Healthcare Cost and Utilization Project (HCUP) [11], maintained by the Agency for Healthcare Research and Quality in the U.S. Department of Health and Human Services, is an excellent resource for health care data. Included in HCUP data is the Nationwide Inpatient Sample, which is the largest allpayor (insurer) inpatient care database in the United States, containing data from approximately 7 million hospital stays. The sample includes the data of 39 states that report annually for each hospital admission and includes information on discharges, length of stay (LOS), and charges, among other data, by diagnosis procedure codes (ICD-9 codes) and DRGs for all hospital admissions. The Nationwide Inpatient Sample provides highly reliable information for comparison of SRS with other procedures. Currently, about 80% of SRS patients have their procedures in an outpatient setting; however, there will always be the need for some patients to be admitted for the procedure when there are existing comorbidities or the physician believes the patient may have problems with the procedure and must be watched carefully for a period of time. Simple querying of the HCUP ﬁles in Table 67-1 using recent data provides compelling information on the comparison of charges, length of stay and discharges of stereotactic radiosurgery, craniotomies, and selected diagnoses. The HCUP data does not provide cost information. However, mean charges can and are used as a measure of costs in the health care environment. Hospital costs are increased to reﬂect appropriate charges to insurers of the procedures provided. It is shown in Table 67-1 that the mean charges of $37,743 for SRS are substantially lower than for craniotomies ($41,683 and $75,398), and the

TABLE 67-1. HCUP 2005 data comparison of SRS and craniotomy. Mean charge ($)

Average length of stay (days)

Number of discharges

SRS Inpatient Procedure Code 92.30–92.39 with and without comorbidities

37,743

1.9

6,475

0.0

—

DRG 1: Craniotomy >17 years with comorbidities

75,398

8.0

86,476

12.0

99.7

DRG 2: Craniotomy >17 years without comorbidities

41,683

4.0

48,585

53,460

5.6

2,201

ICD-9 or DRG

Example by diagnosis ICD-9-CM Diagnosis Code 225.1—Benign Neoplasm Cranial Nerve (Schwannoma) [Procedure Code 04.01]

Deaths (%)

2.19

Not reported

Percentage greater than SRS mean charge (%)

10.4

42

67.

regulatory and reimbursement aspects of radiosurgery

length of stay of 1.9 days for SRS within the hospital is signiﬁcantly less than for craniotomies (4.0 and 8.0 days). Interestingly, when charges and length of hospital stays by a selected diagnosis are compared with SRS, the results are more revealing. Comparing SRS to ICD-9-CM 225.1 for Benign Neoplasm Cranial Nerve (Schwannomas) diagnosis shows that having SRS is 42% less costly than having an open skull procedure for schwannomas, which results in an average length of stay reported as 5.6 days. Schwannoma patients who do not need immediate relief of symptoms, or can be managed medically, could receive SRS in place of surgery. Although it is recognized that surgery in experienced hands for schwannoma patients may be without long-term sequelae, there is still a long-term recovery period and potential rehabilitation period that results in substantial sacriﬁce to the patient, the family, and the economic work dollars that are lost to this long recuperation period. All too often, surgical vestibular schwannoma patients have extremely long recovery periods, with rehabilitative surgeries and permanent loss to the quality of life. When surgery is not absolutely necessary, each patient should be provided with appropriate, current, and truthful information on SRS and any other appropriate procedure, and the patient and his family should make the choice without duress from the medical community. Organized neurootologists have been slow to embrace technology solutions for vestibular schwannomas. As of mid2007, there were approximately 50 neurootologists or otolaryngologists who have trained in the past 3 years to operate Gamma Knife technology, some of whom are from the prestigious House Ear Institute in Los Angeles, California. Recently, insurers are looking to their own collected data and are coming to the same conclusions that are revealed with the HCUP data that charges/costs are lower for SRS procedures. With this information, some insurers are beginning to require patients to be considered for SRS. Insurers are also aware that all too often, craniotomies have associated sequelae that continue to claim their health care dollars in the months and years after surgery.

Payor Mix Payor mix is the reimbursement composition of the patients who receive a stereotactic radiosurgery procedure over a given period, which is usually 1 year. For example, the following 2007 payor mix of a survey of 25 hospital providers with a radiosurgery unit was found to be Medicare, 22%; Medicaid, 2%; Blue Cross/Blue Shield, 25%; HMOs, 20%; Local Insurer, 29%; and others, 2%. A favorable payor mix includes a high percentage of payors with attractive reimbursement rates. In stereotactic radiosurgery units, commercial payors like Blue Cross and Blue Shield and local insurers are often the most desirable, as the provider can negotiate their technical/facility reimbursement rate. On the other hand, Medicaid is generally the least desirable, with Medicare second least desirable as neither reimburses fully the cost of a radiosurgery procedure. The payor mix for stereotactic radiosurgery varies by the type of institutional afﬁliation, the specialty of the primary physician conducting the procedure, and the type of instrument itself (gamma, linac, or proton unit). As would be expected with

677

academic institutions, the Medicare and Medicaid payor mix is higher (30% to 40%) than in a small community hospital environment (20% to 25%). Where the unit is primarily a part of radiation oncology with little involvement of neurosurgery, there is also a higher Medicare and Medicaid payor mix, which may run as high as 35% to 40% of all procedures, reﬂecting the payor mix of radiation oncology as a whole. Where the instruments are located within surgical suites with neurosurgeons as gatekeepers, the Medicare and Medicaid mix is usually lower (20% to 25%). This occurs because neurosurgeons utilize the machines to treat non-lesion indications such as pain, movement and psycho-neuro disorders. Therefore, one will typically see a payor mix that closely mirrors DRG 1 in these situations. Higher Medicare and Medicaid utilization of stereotactic radiosurgery procedures may affect the ﬁnancial viability of the unit’s operations as neither Medicare nor Medicaid was intended to pay the cost of a procedure. Understanding this type of information may be vital in the future to structuring new stereotactic radiosurgery facilities that are composed of cooperating specialty physicians from both radiation oncology and neurosurgery.

Medicare Reimbursement Medicare ﬁrst became a viable program in 1966, signed into law on July 31, 1965, with President Lyndon B. Johnson’s push for reform of the Social Security Act [12]. Medicare was created to provide health insurance to most American citizens age 65 years and over, certain disabled citizens of any age, and those with end-stage renal disease. Medicare was and is a single federal government program, applied equally to all who qualify. Medicare has two main components. Part A, the hospital insurance program, covers inpatient hospital and surgery services and some posthospital care if needed. Part B, the supplemental insurance program, covers physician services and outpatient medical and surgical services [13].

The Medicare Patient Perspective Part A is ﬁnanced through payroll deductions of active workers. At this time, no premium is collected during retirement for most beneﬁciaries, as they are considered to have previously paid their premiums during their working years for the services provided in Part A. However, there are some Medicare beneﬁciaries who must pay a premium for Part A. In 2008, this premium was set at $423 per month for those who had no work history or did not earn enough during their working years to receive Part A free of premium. Those with limited income may qualify for Medicaid, which will be discussed further in this chapter. Part B is ﬁnanced through patient premiums (which are deducted directly from Social Security checks) and general federal tax revenues. The 2008 Part B monthly premium was $96.40, which is a $2.90 increase from 2007 and totals $1157 in annual payments by the Medicare beneﬁciary. If the beneﬁciary did not sign up for Part B concurrently with signing up for Social Security, he or she is assessed higher premiums at a rate

678

r. emerick

of 10% higher for each 12-month period the beneﬁciary did not apply for Part B. This can be prohibitive for the retiree who missed the appropriate sign-up period and thus this type of coverage may be met with other private insurance called MediGap insurance. Medicare requires beneﬁciaries to pay a deductible and copayments to the provider of service for the care provided. These deductibles and copayments are now a substantial part of the reimbursement for health care services for the Medicare patient. The Medicare Part A hospital deductible is meant to represent the cost of a ﬁrst-day stay in a hospital and is determined by CMS to be $1024 for each admission to a hospital stay. The 2008 copayment for the Part B deductible is $135 plus a minimum of 20% of the reimbursement approved by Medicare for the outpatient physician and the technical procedure [14]. What does all this mean for the radiosurgery Medicare patient? Should the patient have a one-session inpatient stereotactic radiosurgery procedure because of comorbidities or expected reaction to the procedure, the patient will be responsible for the ﬁrst $1024 of the hospital costs, along with 20% of the approved reimbursement for all professionals involved (neurosurgeon, radiation oncologist, and possible anesthesiologist). The total ﬁnancial responsibility of the Medicare recipient in this case will range from $1750 to $2500 out of pocket. If the Medicare beneﬁciary has an outpatient radiosurgery procedure, the copayment will vary by the type of stereotactic radiosurgery. For gamma, linac, and proton beam one-session procedures in 2007, the facility copayment was $1340 plus physician copayments of $1750 to $2000 for a total of $3000 to $3500 out of pocket for the beneﬁciary. For image-guided robotic radiosurgery with ﬁve treatment sessions, the beneﬁciary’s out of pocket copayment for the facility totals $5500 for the ﬁve sessions plus the physicians’ copayments for the procedure. The total outlay for the Medicare patient will range from a low of $6000 to as high as $8000, dependent upon physician charges per day for image-guided robotic radiosurgery. Although the Medicare patient is not the only patient to pay considerable deductibles and copayments for a stereotactic radiosurgery procedure, he may have the least ﬁnancial resources. This type of understanding of the cost of a procedure for a patient should be paramount to physicians when there are multiple types of radiosurgery instruments available. Additional costs of travel, lodging, and economic work loss can also be considerable for the patient. Physicians should remain mindful of their ﬁduciary responsibility to the Medicare patient and all patients with limited ﬁnancial resources, high copayments, or lack of insurance. If the patient’s condition can be appropriately treated in one session, the patient should be offered one-session stereotactic radiosurgery over multisession stereotactic radiosurgery to alleviate ﬁnancial consequences beyond his control. Ultimately, it is the responsibility of the physician to offer the best clinical outcome procedure at the lowest out of pocket cost to the patient. Should a patient need open skull surgery, one-session radiosurgery and radiation treatments, which is a common scenario with a large malignant brain tumor, the resulting copayment and deductibles could amount to more than $12,000 in less than a year for the patient. Should the patient have ﬁve-session

image-guided radiosurgery instead of one-session radiosurgery, the out of pocket fees can rise to more than $15,000.

The Medicare Facility Perspective Hospital facility Medicare reimbursement is currently adequate for all types of stereotactic radiosurgery including gamma, linac, image-guided, and proton beam. Medicare establishes provider facility reimbursement rates by collecting data from Medicare patient billings. A simpliﬁed explanation follows. Using an algorithm, CMS establishes an average charge for a procedure and converts it to cost of the procedure by applying cost to charge ratios for hospital departments taken from the Hospital Medicare Cost Report that is ﬁled yearly by every hospital. After obtaining the data for a period of 2 or more years, CMS may place a procedure in a new DRG or outpatient Ambulatory Payment classiﬁcation (APC). In 2004 and 2005, gamma, linac, and proton beam onesession radiosurgery were reimbursed approximately $6700 for each outpatient Medicare patient. Image-guided radiosurgery was reimbursed $6700 for the ﬁrst treatment and $3750 for each additional treatment up to ﬁve treatments. For the 2006 year, CMS concluded from its collected data that gamma radiosurgery was underreimbursed. Therefore, gamma radiosurgery now receives around $9000 to $10,000 for each outpatient Medicare procedure since January 1, 2006. CMS will assess the rates for linac and image-guided radiosurgery in 2008. CMS did provide preliminary collection data for linac and the ﬁrst session of image-guided radiosurgery, which showed linac radiosurgery to be overreimbursed and image-guided radiosurgery to be adequately reimbursed. It is interesting to note that all equipment providers of radiosurgery instruments (whether gamma or linac based) sell their technology for around the same price of $3 million to $4 million. In addition, installation is similar, except for the CyberKnife, which requires a taller vault in comparison with other linacs and more shielding because the robotic arm can target above and behind, and currently requires a maintenance fee of three times the average at close to $500,000 per year. Therefore, this author believes that the collected data on linac radiosurgery that shows an overreimbursement of linac radiosurgery is caused by the linac radiosurgery provider’s failure to use standard cost accounting and thus charge appropriately. Charging appropriately requires an understanding and application of quality cost accounting and application of costcenter cost to charge ratios. More emphasis and assistance by professional associations should be directed to educating members on how to appropriately charge for a procedure, which includes depreciation of equipment, building and resources, as well as stafﬁng costs. In 2005, CMS relented to the push of organized radiation oncology to eliminate the bundled temporary codes for the hospital facility planning portion of gamma and linac units and to use all available radiation coding in their place [15]. The available radiation coding (CPT 77295, 77370, 77334, and 77470) results in an inadequate professional payment for those conducting stereotactic radiosurgery, which requires exceptional skills and an enormous amount of time compared with other radiation procedures.

67.

regulatory and reimbursement aspects of radiosurgery

In 2007, permanent coding was obtained for the hospital delivery portion of gamma radiosurgery procedures with a radiation designated code (CPT 77371). Permanent radiation coding was also granted for free-standing radiosurgery procedures both gamma and linac (CPT 77371, 77372, and 77373). In 2008, linac and image guided radiosurgery still have temporary coding for the hospital delivery portion of their radiosurgery procedures. Temporary coding is not usually reimbursed by most commercial payors. It is expected that once commercial insurers become aware that the billing codes for stereotactic radiosurgery no longer represent a surgical procedure but a radiation procedure, and that the procedure is being provided outside of a hospital provider, the current premium reimbursements will begin to erode and affect the adequate ﬁnancial situation that now exists. Radiation oncology Medicare technical reimbursement is and has been inadequate and does not serve the radiation oncology ﬁeld well. The addition of stereotactic radiosurgery reimbursement competing for limited funds within radiation oncology will also affect the future viability of the ﬁeld amid depressing reimbursement. As previously stated, the different professionals involved in stereotactic radiosurgery are a small group within large specialty professional groups. Therefore, they are not well represented and their issues are not well understood. The current misguided push by organized professionals to capture, own, and recharacterize stereotactic radiosurgery from a surgical procedure to a radiation procedure will eventually affect the technical reimbursements and thus the viability of providing this valued alternative procedure to open skull surgery. As reimbursements erode, it will in turn affect whether free-standing professionally owned facilities will want to provide the procedure because of both inadequate professional and technical facility reimbursement.

Medicaid Medicaid was created with the same Social Security Act as Medicare in 1965. It was created as a state-operated program to provide publicly ﬁnanced health care coverage for the poor [12]. This means there are really 50 distinctly run state programs. Unfortunately, Medicaid patients are not commonly referred for radiosurgery procedures as they are seldom diagnosed during the window of opportunity that affords them this quality procedure. The state of overall care for the Medicaid patient is lacking, and many do not go to the clinics until they are in need of surgical or radiation therapy intervention. Medicaid was designed to provide the most basic health care, primarily inpatient emergencies and surgeries, pregnancy care and delivery, and clinic and preventive immunizations. There was no foresight that medical technology would make it possible to perform highly surgical procedures on an outpatient basis. Therefore, most Medicaid programs can only reimburse for radiosurgery as an inpatient and might allow as little as $600 total if the procedure is conducted in an outpatient setting. Both facilities and professionals must be aware of how Medicaid operates within their state and be willing to accommodate an inpatient procedure if it is necessary to provide the procedure to the patient.

679

Insurance and Approvals Stereotactic radiosurgery is approved by the insurance industry in lieu of open skull surgery or as an adjunct to craniotomy. This is clear to the insurance industry by the use of CPT code 61793, which is the primary billing code for stereotactic radiosurgery. Traditionally, radiosurgery is considered a surgical strike and is not in competition with radiation procedures. With the increase of stereotactic radiosurgery procedures to treat psycho-neuro, movement, pain, and ocular disorders as well as brain lesions, it is clearly seen as a threat to some in organized neurosurgery. On the other side, organized radiation oncology also perceives a threat that may only exist in the metastatic brain lesion area, as radiation oncology developed this area. In most cases, if a patient needs radiation therapy or other specialized radiation, he will more than likely still need it after stereotactic radiosurgery. The perceived threat may be more of an “ownership issue” than an actual threat. Overall, reimbursement for stereotactic radiosurgery enjoys a premium in extra beneﬁts as the patient will not suffer from infection, surgical complications, or anesthesia side effects. Also, the patient will not need rehabilitation, an extended hospital stay, or require home health services after radiosurgery. Thus, insurers routinely negotiate or “carve out” a reimbursement for the procedure of stereotactic radiosurgery when it is considered in lieu of open skull surgery. The technical fees received in 2007 by hospitals that carve out a rate with an insurer normally ranged from $20,000 to $35,000 (IRSA 2007 Member Survey, unpublished). There are negotiated rates reported in the $46,000 range for Gamma Knife surgery centers. In 2007, approvals and reimbursement for most stereotactic radiosurgery procedures were not an issue and denials were rare. The enjoyment of good reimbursement and approvals may change in the future if the nature of radiosurgery becomes perceived as a high-priced replacement for less costly radiation treatments and not as a replacement for a more expensive surgical option.

Conclusion In many aspects, radiosurgery is still in the infancy of its development. We can expect to see the proliferation of more centers with the addition of free-standing coding and the ability to “rent” equipment cheaply per procedure or joint venture with manufacturers to own the equipment. We must question whether this will enhance quality assurance or leave us with a decline in measurable outcomes. Will both neurosurgeons and radiation oncologists participate in the procedures in these new centers? Will the patient receive the best procedure with the least amount of side effects (vision loss, hearing loss, speech problems, etc.) from an experienced team? It can be reasonably presumed that the current trend for investor groups and some hospitals to advertise and market claims that cannot be substantiated about radiosurgery and individual professionals will eventually bring scrutiny from regulatory bodies with resulting ﬁnes or limitations, as has been seen in the pharmaceutical industry. Recent court cases are examining how informed consent is administered. Is the patient

680

r. emerick

given relevant current information? Is the patient afforded the opportunity to seek other opinions? Or is the patient guided only one way? Is the professional honest about his experience? Stereotactic radiosurgery has brought much to the clinical world. It is the forefather of stereotactic body radiotherapy and stereotactic radiotherapy, where typical hypofractionation regimens of three to ﬁve fractions prevail. Radiosurgery, with the help of diligent professionals, has seen a new treatment area open for patients by providing quality economical treatments within the body. We should be careful to differentiate the descriptions of these procedures for research as well as coding. Although some would have all procedures that involve radiation be called radiation therapy, they do an injustice to themselves and to the time and training they have expended to provide radiosurgery. It also leads insurers and CMS to conclude that everything with radiation is just therapy that results in the eventual erosion of reimbursement through confusion. It can easily be surmised that radiosurgery is the latest subspecialty for professionals. As always, the constantly changing nature of reimbursement and coding is taxing for facility administration and professionals. Understanding the required different billing scenarios for commercial insurers, government payors, and private-pay individuals can at times overwhelm the manager of a center. It is vital when seeking adequate reimbursement from regulatory bodies that we be compelled to work together regardless of our physician specialty or association. There is more power in combining to work for the whole than in ﬁghting individually for small pieces. All involved should strive to see that all facilities and professionals are appropriately reimbursed so that we may continue to advance research and positive clinical outcomes into new areas and diagnoses. Teamwork and joining together will keep the radiosurgery industry strong.

References 1. Backlung EO. The history and development of radiosurgery. In: Lunsford LD, ed. Stereotactic Radiosurgery Update. New York: Elsevier, 1992:3–9.

2. Lunsford LD, Flickinger J, Lindner G, et al. The stereotactic radiosurgery of the brain using the ﬁrst United States 201 cobalt60 source Gamma Knife. Neurosurgery 1989; 24(2):151–159. 3. Code of Federal Regulations, Title 21—Food and Drugs. Subchapter H—Medical Devices. Volume 8. 21 CFR 8. U.S. Government Printing Ofﬁce, 2005. Available at http://www.gpoaccess. gov/cfr/index.html [all dates]. 4. The Food and Drug Administration. Information on Premarket Approval Applications. FDA, 2007. Available at http://www.fda. gov/cdrh/pmapage.html. 5. The Food and Drug Administration. 510K Premarket Notiﬁcation Searchable Database. (510K numbers K984563, K923522, and K913174). FDA, 2007. Available at http://www.accessdata.fda. gov/scripts/cdrh/cfdocs/cfPMN/pmn.cfm. 6. Code of Federal Regulations, Title 10—Nuclear Regulatory Commission. Part 35—Medical Use of Byproduct Materials. 10 CFR 35. U.S. Government Printing Ofﬁce, 2007. Available at http:// www.gpoaccess.gov/cfr/index.html [all dates]. 7. United States Nuclear Regulatory Commission. Ofﬁce of Nuclear Material Safety and Safeguards Communications. RIS 2005–23, October 7, 2005. 8. Getzen TE. Management and regulations of hospital costs. In: Health Economics, Fundamentals and Flow of Funds, 2nd ed. New York: John Wiley & Sons, 2004:175–196. 9. University Hospital Consortium Services Corporation, Clinical Practice Advancement Center. Stereotactic Radiosurgery. Oak Brook, IL: UHCSC, 1995. 10. Ott K. A comparison of craniotomy and Gamma Knife charges in a community-based Gamma Knife center. Stereotact Funct Neurosurg 1996; 66(Suppl 1):357–364. 11. Healthcare Cost and Utilization Project. 2007. Available at http:// hcup.ahrq.gov/. 12. Pearman WA, Starr P. Medicare: A Handbook on the History and Issues of Health Care Services for the Elderly. New York: Garland, 1988. 13. Centers for Medicare and Medicaid Services. Medicare & You Handbook. Washington, DC: U.S. Government Printing Ofﬁce, 2007. 14. Berger SH. Fundamentals of Health Care Financial Management, 2nd ed. San Francisco: Jossey-Bass, 2002. 15. Code of Federal Regulations, Parts 419 and 485. Nov. 10, 2005: Vol. 70. No. 217. 42 68515–69040. Washington, DC: U.S. Government Printing Ofﬁce.

6 8

Medicolegal Issues in Stereotactic Radiosurgery April Strang-Kutay

Introduction Every medical practitioner involved in the administration of radiosurgical treatment beneﬁts from, at minimum, a rudimentary understanding of the concept of medical negligence and how such litigation ﬂavors medical practice. Before thoroughly grasping the nuances of the anatomy of a medical negligence lawsuit, it is important to understand the general concepts that surround the topic. In a civil action, a judgment in favor of the plaintiff is conceptualized as an attempt to make the victim “whole” by awarding monetary damages. In such a civil action, where a plaintiff is contending injury, there are usually allegations of conscious pain and suffering, potential loss of wages, loss of consortium (where the injured party is married at the time of the occurrence of the alleged negligence), payment of medical fees and medical expenses, and diminishment of earning capacity. All of these described usual damages and attendant suffering are then converted to monetary damages. Civil wrongs, a category under which medical negligence falls, are commonly referred to as torts. Fundamentally, a patient must prove that a health care professional’s conduct comprised a tort and must show all of the elements of that particular tort; the patient will then have a causative action for a lawsuit, which may succeed in proving the civil wrong.

Informed Consent An area of medical negligence that has received national attention and increased scrutiny is the surgeon’s legal and moral obligation to gain his patient’s informed consent prior to treatment. In the realm of stereotactic radiosurgery, there is often a menu of treatment choices from which a patient may select to address his skull-based disease. Often, there is no consensus in the health care ﬁeld about which modality of treatment is best or “preferred” for a particular malady. When more than one feasible option of treatment exists to treat an individual patient’s disease process, the onus is on the surgeon to educate and inform with reliable, understandable, and current information. When a patient is considering submitting to radiosurgical management of his cranial disease, his decision is personal, cannot be generalized, and can only be made intelligently after

a thorough discussion of all available options. When appropriate information is conveyed by the physician, the patient is the individual within the partnership who must ﬁrst understand the information offered so that a realistic weighing of the advantages and disadvantages of each option takes place. A paternal or dogmatic approach to patient counseling is not only antiquated, it fails the modern legal objective to inform and educate. If the surgeon does not have direct experience in all categories of treatment relevant to the patient who is making an informed choice, and the surgeon offers inaccurate and/or inadequate information about a particular treatment option, he may ﬁnd himself liable for failure to gain informed consent before having treated the patient. In some states, such a failure of informed consent operates to impose liability on the physician under theories of negligence. In other states, such as Pennsylvania, lack of informed consent imposes liability under a theory of battery, meaning that the surgeon becomes liable and legally accountable for all complications a patient experiences subject to the treatment he chooses, even if that treatment (radiosurgical or otherwise) is carried out competently. Under a scenario where battery is the governing legal concept, a radiosurgeon who treats his patient with stereotactic radiosurgery without the equivalent of having ﬁrst gained his patient’s legal informed consent may be liable for all cranial neuropathies the patient develops, even though the radiosurgical treatment was indicated and performed appropriately, and even though the complications sustained were known risks inherent in the procedure. Thus, the duty to educate and inform the patient is paramount in any treatment paradigm and should never be minimized, condensed, or hurried. Because radiosurgery has become a vital tool in the management of skull-based tumors, it cannot be overstated that optimal education is forthcoming from experienced, multidisciplinary teams. In pertinent part, the Hippocratic oath advises: I will respect the hard-won scientiﬁc gains of those physicians in whose steps I walk. . . . I will apply, for the beneﬁt of the sick, all measures which are required. . . . I will not be ashamed to say “I know not,” nor will I fail to call in my colleagues when the skills of another are needed for a patient’s recovery. . . .

Those physicians involved with the administration of radiosurgery have long understood that all too frequently, surgeons are biased toward conventional surgical modes of treatment, and, purposefully or sometimes unwittingly, allow these biases

681

682

a. strang-kutay

to creep into their presentation of options to patients. In these circumstances, patients hear a one-sided, dangerously tainted analysis that undoubtedly heavily inﬂuences treatment choice. Radiosurgeons, too, must guard against a tendency to permit their own preferences for treatment to dictate the way the options are presented to patients. If a radiosurgeon is not sufﬁciently possessed of an up-to-date understanding of competing surgical alternatives to address a patient’s particular problem, then the patient should be encouraged to visit with a surgeon who can offer appropriate counseling prior to making a treatment decision. A policy of liberal patient referral and sharing of patients among physicians who make up the multidisciplinary team ensures the most thorough and unbiased education of patients. Medical practitioners should breathe a sigh of relief and take solace in the certainty that they need not have encyclopedic knowledge about all modes of treatment available to address a particular disorder. Rather, the law imposes only the burden that the practitioner know of the existence of competing therapies, reveal these alternative methods of treatment to the patient, and point the patient in a direction where he can explore this alternative. One of the worst mistakes a physician can make is to represent himself as an authority on a treatment method about which he does not have competent, up-to-date knowledge. In this situation, by virtue of the doctor’s stature as a physician, the average patient will accord the physician’s explanations great weight and might ﬁnd himself seriously misled, choosing a treatment based on incorrect information. The law is clear that a physician has a duty to do more than merely mention competing modes of treatment; any information he chooses to convey about alternative methodologies must be reasonably accurate and veriﬁable. For example, presenting radiosurgery as an option for treatment, but not its speciﬁc applicability, or what beneﬁts or advantages radiosurgery might have is inadvisable. A patient’s consent to a procedure is valid only if the patient has been informed of all of the risks that a reasonable person in that situation would consider important to his or her decision whether to undergo the procedure. Thus, a physician is required to inform a patient of the alternatives to the proposed treatment and of the risks and chances of success of those alternatives. In other words, if the patient consents to the procedure proposed by the physician without this information, the consent is not informed and not legally valid. When counseling a patient prior to administering treatment, the physician must devote sufﬁcient time for this encounter to encourage meaningful discussion and exchange of information. Different patients have different priorities, and it is important that the physician, during the critical counseling session, make an effort to learn about and understand a given patient’s personal agenda. It is recommended that the physician keep in mind—at all times—that it is the patient and the patient’s family who will have to live with the consequences of the treatment decision: at times, a life-altering event in the patient’s life. Such a period of meaningful communication between physician and patient cannot be accomplished in a harried or hectic atmosphere. To attempt to discuss the goals of radiosurgery for a particular patient, the advantages and disadvantages of radiosurgery in the individual circumstances at hand, as well as to

educate regarding feasible options of treatment in less than 30 minutes is not advisable. The pretreatment time spent with the patient is critical, not only to lay the foundation for a solid education that allows the patient to make an intelligent, informed choice, but also to begin to build a strong interpersonal rapport that will be a positive and pervasive inﬂuence throughout the patient’s treatment. In documenting this encounter with the patient, it is helpful to discuss in the medical record the length of time devoted to counseling. The radiosurgeon who is counseling a prospective patient on treatment choices may supplement his discussion by giving the patient literature to review. The author strongly encourages physicians to share literature with patients, as the content of such materials summarizes and enlarges upon the subjects discussed and acts as a reinforcement of the educational process initiated by the physician. One caution should be observed if literature is disseminated: it is imperative that any literature shared with patients be accurate and current. In the rapidly evolving radiosurgical ﬁeld, pamphlets and brochures not quickly become outdated. There is little evidence more devastating for the physician in the courtroom than when the plaintiff’s attorney can present revised editions of brochures given to the patient (which had been available at the time of counseling) that provide new and different information that the patient should have been able to consider before deciding on treatment. Once again, it is prudent for the medical practitioner to reference in the patient’s chart that literature was given, as well as a description of the particular materials provided. When a radiosurgeon is discussing risks of a given procedure, it is necessary to be reasonably accurate, although there is no legal requirement that risks must be quantiﬁed in numerical form. If, however, the radiosurgeon is inclined to quote precise ﬁgures (for example, likelihood of tumor control), he should be sure to specify if he is describing the statistics gleaned from peer-reviewed literature or the individual statistics of his institution. If individualized, speciﬁc percentages are quoted, the surgeon should point out any existing discrepancies with the national statistics. Of particular importance is the radiosurgeon’s understanding that he may, in fact, make a recommendation for treatment. The law does not forbid a physician from expressing his sincere preference for treatment or from discouraging a patient from pursuing a treatment alternative that the physician does not believe serves the patient’s best interest. The key here is that the physician may not lawfully discourage a patient based on inaccurate or incomplete information. As long as the competing treatment modality has been described fairly and the patient knows where he can turn to further explore this treatment alternative, the physician may make any recommendation he feels appropriate for the presenting circumstances. When discussing risks and complications, particularly with respect to the proposed treatment, physicians are encouraged to portray complications realistically and not to “downplay” or minimize the potential consequences of a given complication, should it occur. For instance, for the radiosurgeon who is educating a patient on the potential for treating his acoustic neuroma via primary radiosurgery, despite the excellent tumor control rates published in recent years, patients must under-

68.

medicolegal issues in stereotactic radiosurgery

stand that the treatment is not foolproof. A small minority of patients will fail this therapy. The prospective patient must understand before submitting to treatment what consequences may exist when radiosurgery fails its objective. Can radiosurgery be attempted on a second occasion? If surgical excision is elected, will surgery prove more difﬁcult as a result of the previously irradiated tumor? In other words, it is not enough for the physician to merely describe the chances of a complication occurring; the nature of the complication and the threat it may pose must be thoroughly explored. Finally, as radiosurgical techniques continue to evolve in this ﬂuid and advancing ﬁeld of medicine, radiosurgeons will need to be ﬂuent in discussing the risks and beneﬁts not only of competing modes of treatment (such as microsurgical excision), but also in educating patients on the pros and cons of the various forms of radiosurgical intervention. For instance, what is the difference between treating a patient’s acoustic neuroma with a one-session system such as Gamma Knife or linac, versus treatment via fractionation with systems like the Peacock, SmartBeam IMRT, or CyberKnife? Similarly, as the popularity of radiosurgery grows, such treatments will gain momentum for treatment of cranial diseases beyond the now “routine” or “ordinary.” Consider that Gamma Knife technology has been found to be very effective in treating patients suffering from refractory trigeminal neuralgia. But, what are the indications for offering radiosurgical treatment to a patient experiencing what is categorized as atypical trigeminal neuralgia? Should patient expectations be different, through informed counseling, when there is no large patient base on which treatment results have been measured? In the ﬁnal analysis, the process of gaining informed consent that will rise to the level of legal sufﬁciency can be challenging. Tips to best ensure that a physician has met his responsibility in this realm can be summarized as follows: • Abandon a dogmatic, paternalistic approach to counseling. • Involve multidisciplinary team members to educate about competing modes of treatment. • Allot sufﬁcient time to the patient encounter in which options for treatment are discussed. • Consider providing patients with up-to-date, accurate literature to supplement the discussion. • Carefully document in the medical record the patient counseling session, including reference to the amount of time spent with the patient, as well as reference to the pertinent literature disseminated. • Do not quote precise statistics unless there is a certainty of accuracy. • If statistics are given, explain whether these statistics reﬂect the experience of the institution in question or national published data. • Do not feel constrained from offering honest, heart-felt recommendations, as long as competing alternatives are described fairly and accurately. • Stay abreast of current developments in the ﬁeld and be certain to refer patients for further counseling if thorough and complete information is lacking. • Give complete information about the effects of a complication, if it should develop.

683

Medical Negligence Whereas the process of informing patients adequately is a subject of scrutiny under the law, negligence is the most common complaint in a medical negligence action and requires that the patient prove the existence of ﬁve basic elements: 1. An act or omission; 2. A duty to act or not to act; 3. A breach of a duty to act or not to act; 4. That the act or omission was the direct, proximate cause of the patient’s injury; and 5. That the act or omission resulted in damage or injuries that can be afﬁrmatively linked to the conduct in question. Negligence is a term of art that may result from an act of commission or omission. A health care professional can be held liable for being negligent by doing something he or she was under an obligation not to do or by not doing something he or she was under an obligation to do competently. The distinction between commission and omission is one between action and inaction. The case for inaction is limited to those cases in which the health care provider fails to competently follow through with a duty imposed upon him by the prevailing standard of care. For example, when a neurosurgeon was not available during the performance of radiosurgical treatment and complicating issues developed, he may be held liable for his failure to respond appropriately to a crisis. A health care professional can also be liable in negligence by doing something he or she was obligated to do and by doing it below the recognized standard of care. If the neurosurgeon responsible for targeting, who has speciﬁc training in the neuroanatomy of the brain, creates a plan that disposes the patient to develop preventable cranial neuropathies, he may be held liable by a theory of negligence.

Standard of Care In order for the health care professional to understand his or her duty to the patient, there must be a basic understanding of the standard of care with respect to the treatment and/or advice being administered. How, then, is the standard of care measured for the radiosurgeon who is intent on conforming to the national standards of practice? Unfortunately, although respected groups in various ﬁelds of medicine publish guidelines that do, in fact, serve as governing concepts for the practitioner, the actual standard of care is not a magic formula embodied in any writing. Common law is applied by the court system in many states to determine the liability of a health care professional for malpractice. The common law is law that is derived from judicial decisions. When a court and/or jury decides the outcome of a case, it sets a precedent for other judges who are hearing similar cases. Often, common law deﬁnes medical negligence as the failure of a health care professional to possess or exercise that degree of learning, skill, care or diligence in the diagnosis and treatment of a patient that would be expected of a health care professional. Health care professionals who are specialists are held to a higher degree of skill that is comparable with other

684

a. strang-kutay

specialists in the same ﬁeld. A neurosurgeon, for example, would be considered a specialist within his ﬁeld of expertise. Similarly, a radiation oncologist would be considered a specialist as well, and be judged by a court of law with respect to his speciﬁc education and ﬁeld of experience. The court arrives at an appreciation of the standard of care at issue in any given case through the beneﬁt of expert witness testimony, which informs it of the practices in the specialty ﬁeld. Both plaintiff and defendant may introduce evidence through an expert witness that the court then uses to decide the applicable standard of care. Frequently, experts from opposing sides of the case argue on many fundamental components comprising the standard of care that applies to the issue or issues in the case at bar. Inevitably, there will be serious and irreconcilable differences of opinions between the experts who testify for plaintiff and defendant, and the jury and/or judge is then charged with the task of determining which expert or experts are most persuasive and convincing. The art of persuasion covers a broad spectrum of facets of expert testimony including the witness’s demeanor on the witness stand, his rapport with the jury/judge, his apparent mastery and/or familiarity with the subject matter at issue, his professional background and experience in the subject discussed, and the bolstering weight of medical literature underscoring the expert’s opinion. The court and/or jury then compares the acts of the health care professional with the applicable standard of care and evaluates whether the third element, a breach of duty, has been proved effectively by plaintiff. The fourth element of the tort action is that the act caused actionable and compensable harm. This is known as causation. A plaintiff must prove that the negligent act or omission complained of was the direct or proximate cause of the patient’s injury. In an ordinary negligence action, the causal connection is relatively easy to prove. If the driver of an automobile is inattentive and collides with a vehicle in front of him, causing impact injuries to the driver of the vehicle struck, the physical injury is often immediately obvious and the causal connection clearly established. Contrast the simple automobile case with one of radiosurgical medical negligence. For example, take the case of a patient who presents for radiosurgical treatment of his pituitary tumor. After radiosurgery, the patient suffers serious complications related to a cranial deﬁcit of the optic nerve. Was the injury sustained because of a risk inherent in the procedure, given the proximity of the pituitary tumor to the optic chiasm? Or was the targeting by the neurosurgeon in error? At times, it could be very difﬁcult to determine whether complications experienced by the patient were unavoidable and inextricably intertwined with the disease process itself, or whether they developed as a consequence of medical negligence. Persuasively linking the patient’s injury to a breach in the standard of care can be an extremely challenging aspect of medical negligence litigation. Establishing the proximate cause of a patient’s injuries may require the use of an expert witness, especially if the medical procedure and/or injury is complex, as in the above-described example. Courts also require the testimony of an expert witness in order to establish the standard of care for a particular injury or illness.

When an expert is called to testify in a case involving medical negligence, his background must demonstrate that training and experience have provided the foundation necessary to enable him to testify as to what a hypothetical health care professional would usually and customarily do in like circumstances. Usually, the expert witness is examined on his or her credentials before giving substantive testimony relating to the case. This examination is designed to determine if the expert can competently testify as a credible expert on the particular subject that forms the basis of a lawsuit. The attorney not offering the witness as an expert may object to the witness’s qualiﬁcations to testify as an expert, which requires a speciﬁc ruling by the court on whether the individual may testify as to his opinions. The judge will then rule on whether or not the witness is an expert for the purpose of offering an opinion at trial. Sometimes, a health care professional does not have to be in the same ﬁeld in order to testify on a particular subject. For example, a radiation oncologist may be qualiﬁed to testify about neurosurgical treatment with respect to the radiosurgical targeting of a cranial mass. The ﬁfth and ﬁnal element a plaintiff must prove in a malpractice action is damages. Damages are measured in a malpractice action by comparing the patient’s condition after treatment with the patient’s hypothetical condition if he or she were not injured as a result of the health care professional’s negligence. The ﬁnal award of money damages to a malpractice victim is conceptualized to make the plaintiff whole by reimbursing him or her for lost wages, medical expenses, conscious pain and suffering, diminishment of earning capacity, possible claim for loss of consortium, and, if the court and/or jury ﬁnds that the health care professional has engaged in reckless or grossly negligent conduct, a potential award of punitive damages as punishment of the health care provider.

The Anatomy of a Lawsuit The Complaint A lawsuit is usually formally initiated by the ﬁling of a complaint with the applicable court. In accordance with local rules, the complaint will be served on the health care professional and all others named as defendants, along with a summons. The summons will inform the health care provider that he or she has a designated period of time to formally answer the complaint by the ﬁling of an answer. If the complaint is not answered in a timely manner, the plaintiff’s attorney may ﬁle a default. Lawsuits must be brought within a certain time after the cause of action has arisen or they are barred by law. The time period for commencing a lawsuit against a health care professional varies from state to state but usually falls within 1 to 4 years since the date of actionable conduct by the health care provider. Some states have adopted a “discovery” rule, which, under particular circumstances, relaxes the time frame under which the suit must be brought and extends the period of time for ﬁling. The complaint document itself contains the general allegations upon which the patient bases his or her claim to damages.

68.

medicolegal issues in stereotactic radiosurgery

The plaintiff may allege more than one legal theory in pursuit of the action, and the theories may also be inconsistent at this juncture of the case. The plaintiff cannot prevail in any of the theories unless the elements of each and every theory are proved to a preponderance of the evidence in court.

The Answer The answer is the health care professional’s response to the plaintiff’s complaint. It is designed to be a document where the health care professional admits or denies each of the allegations made by the plaintiff. Some allegations may be denied because the health care professional simply and legitimately does not have enough knowledge to form a belief as to whether the allegations are true or false. Any allegation that is admitted as true by the health care professional relieves the burden from the plaintiff in proving such allegation at trial. In addition to admitting or denying the allegations made by the patient, the health care professional may assert certain defenses. For example, the health care professional may allege that the action is barred by the statute of limitations (that the ﬁling of the complaint was late) or that the injury, if any, was caused or contributed to by the patient or another entity. Furthermore, the health care professional’s attorney may ﬁle counterclaims or cross-claims. The health care professional may deny any negligence in a case and counterclaim for a sum of money that represents the professional fees incurred or owed by a patient.

Pleadings and Discovery The pleadings may raise factual questions along with questions of law. In a jury trial, the judge will ordinarily settle questions of law, and the jury will decide questions of fact. After initial pleadings are ﬁled, the case meanders into an area known as pretrial discovery. Pretrial discovery is a mechanism through which investigations of the allegations of the patient are made. It is a period of time that permits each party to discover information about the particular positions asserted. Pretrial discovery is available to the patient in order to develop the allegations in the complaint and is similarly available to the defendant health care professional to develop information for support of the denials and afﬁrmative defenses in the health care professional’s answer to the complaint. Pretrial discovery includes the taking of depositions, ﬁling of requests for answers to Interrogatories, demands for admissions of fact, and request for production of documents. In most jurisdictions, discovery is required to have been completed prior to the commencement of trial.

Arbitration: An Alternative to Trial When certain circumstances exist, some states have legislation that requires a medical negligence case to be referred to arbitration rather than to trial by jury. Occasionally, a case may be heard and settled rapidly by arbitration during the litigation process. The decision to arbitrate depends on the particular legislation governing malpractice cases in the state in which the claim has arisen, the unique and peculiar factual circumstances involved in the case, and the opinions of the attorneys and parties involved about how the case may be resolved.

685

The Trial In the climax to litigation, the jury trial is begun by the process of juror selection. Predominately, malpractice trials are decided by panels of jurors rather than heard by a single judge. The number of jurors who sit on a particular panel differs from state to state, with most states requiring between 6 and 12 jurors to hear the action. The voir dire process begins with questions posed by the opposing attorneys designed to select the best potential jurors to hear the case. Speciﬁcally, the questions are designed to determine if the juror has any physical incapacity that would inhibit him or her from serving as a member of the jury and whether the potential juror can be objective in weighing the evidence and in following the judge’s instructions. Jurors may be removed “for cause,” which simply means that an attorney can usually garner the court’s approval to remove a juror who is likely to be biased in the evaluation process. In addition, each party is allowed a number of preemptory challenges. Preemptory challenges allow each attorney to remove potential jurors without stating any reason for their removal. Such preemptory challenges are ordinarily limited in number by the rules of the particular local court. The trial formally starts in most jurisdictions by permitting the attorneys to give opening statements. The opening statement is a vehicle through which each attorney has an opportunity to explain to the jury what the case is about and what each side intends to prove through the presentation of evidence. The jury is thus given an overview of the evidence to be presented: a snapshot of the case. Opening statements are not considered evidence; rather, they are recitations of the expectations of what each attorney plans to develop in presenting evidence to support his or her case. After opening statements by all attorneys, the plaintiff begins his case by presenting evidence. Witnesses are called ﬁrst by the plaintiff and are then subject to cross-examination by defense counsel. At the conclusion of all testimony presented by the plaintiff, the health care professional’s attorney may move for a directed verdict in favor of the health care professional if the defense attorney believes that the plaintiff has failed to prove the necessary elements of the case. If the judge grants a motion for a directed verdict, the trial has concluded, and the health care professional has prevailed. If the motion is overruled, the health care professional’s attorney must then proceed with the defense of the case and present his witnesses in an orderly fashion. After the presentation of all evidence, both attorneys are again granted an opportunity to address the jury, this time with a closing argument. This is the ﬁnal chance for each side to persuade the jury to ﬁnd in its favor. After closing arguments, the judge ordinarily instructs the jury on the law that they may apply to the factual determinations they make during deliberations. The instructions are read to the jury by the judge, and in many jurisdictions, jurors cannot take notes, nor are they provided with written instructions to ponder. After instructions to the jury, the jury then has an opportunity to deliberate and decide the case. When a verdict is reached, the jury returns to the open courtroom and the foreperson of the jury announces the verdict.

686

a. strang-kutay

Appealing the Verdict

Conclusion

In most instances, the losing party may request a new trial by ﬁling a motion for appeal. If the motion is denied by the local court, the decision may be appealed to a higher court. If the motion is granted, the case will be retried in the same court. When a case is appealed to an appellate court, the level above the commonwealth court, the case is then reviewed by an appellate panel requiring written briefs to be prepared and submitted by each side. Frequently, the attorneys are also permitted an oral argument to discuss the merits of their respective positions. The appellate court enters its decision by publishing a written opinion that either afﬁrms or reverses the trial court’s decision. The party who loses the appeal at the appellate level may appeal to the state supreme court; however, the supreme court may determine whether or not the appeal has merit to be heard. If the case is accepted by the supreme court, the same procedure as described in the appellate division is followed. The supreme court then publishes its ﬁnal opinion, which may form precedent by which other cases will be inﬂuenced in the future. Because the ﬁeld of radiosurgery is relatively new, there is little, if any, precedent in the law that helps to deﬁne a health care professional’s liability within this specialty ﬁeld.

While the fear of legal ramiﬁcations issuing from alleged inadequate informed consent or allegations of medical negligence is paramount for many physicians, it remains the author’s steadfast belief that the best weapon against the institution of litigation is healthy physician-patient rapport. The welleducated patient is in the best position to intelligently chart the course of his own treatment, and is best poised to accept complications, when they develop, assuming adequate and complete explanations have been given prior to the treatment elected. Except for the minority of circumstances where untoward complications develop, allegations of malpractice can largely be avoided by conscientious attention to detail, and by embracing a commonsense approach to fostering the physician-patient relationship. Regrettably, it has been said that the path from the treatment room to the courtroom is well trod—representing a direct line from hope to disappointment, from belief to dismay, and from May to October. On the contrary, the author counsels that such pessimism is not an inevitable consequence of practicing medicine, as the majority of allegations of a breach in the standard of care are preventable.

6 9

The Semantics of Stereotactic Radiation Therapy Louis Potters

Introduction Radiation therapy, or the therapeutic use of radiation for the treatment of malignant and benign disease, continues to evolve from the days when radium was ﬁrst discovered by the Curies through to modern linear accelerators with multileaf collimation and conformal treatment planning. The basic tenets of radiation therapy have remained identical over the past 100plus years, namely to improve the therapeutic index by enhancing tumor cell death while preserving normal tissue. To achieve improvements in the therapeutic index, several generations of radiation delivery approaches have been used to various degrees of success and failure. There has been a balance between the competing effects of biological potency against tumor tissues versus that against normal tissue. Imprecise radiation delivery technology as well as greater uncertainties in the existence and location of microscopic disease spread required large treatment volumes with substantial margins to account for previous inaccuracies. Therefore, both the tissue volumes receiving radiation dose and the resultant negative effects in normal tissues were large. By exploiting the differential degree by which sublethal damage is repaired from radiation exposure to normal cells compared with cancer cells, the concept of fractionation was developed [1]. With the advance of diagnostic radiologic technology, tumor location can be better deﬁned. Use of magnetic resonance imaging (MRI) or positron emission tomography (PET) has greatly enhanced our ability to better target tumor locations and, importantly, avoid unnecessary radiation coverage [2]. Linear accelerators with multileaf collimation capable of modulating beam intensity and tight penumbras were coupled with inverse treatment planning software to offer intensitymodulated radiation therapy (IMRT). IMRT has been shown to improve the therapeutic ratio by allowing higher radiation doses (with standard fractionation) without added toxicity. An alternative way in which the therapeutic ratio can be enhanced is for precise, focused radiation delivery, based on image guidance and immobilization that allows higher than usual fractionation doses of radiation [3]. In order for this approach to be successful, the collateral effects of such high

radiation doses need to be greatly reduced, by further minimizing the volume of normal tissue that receives signiﬁcant radiation dose. Technology capable of this has been available for tumors or benign lesions of the brain and skull for years using rigid head frames with ﬁducial markings. Though generically referred to as stereotactic radiation therapy, the use of a single large-dose fraction to the brain is called radiosurgery [4]. With imaging enhancements that focus on the clinical target while also accounting for motion, such as respiration or cardiac, the concept of fractionation can be considered for stereotactic delivery of radiation therapy that extends beyond the conﬁnes of the cranium [5].

Historical Aspects Stereotactic radiation to brain targets is facilitated by the rigidly or semirigidly secure immobilization devices such as halos and masks to the skull. In contrast with targets within the body, intracranial targets do not inherently move and can be feasibly secured in reference to the isocenter of the radiation delivery machine. That way, image assessment can be done in a leisurely fashion with multiple modalities, such as computed tomography (CT) and magnetic resonance (MR) based images. There is assurance that the target does not move relative to the frame system between the planning and delivery phase of therapy. Beneﬁts of stereotactic targeting in the brain, then, are inherently tied to the use of appropriate immobilization including the avoidance of inherent organ motion [6]. With extracranial treatment, organs are not rigidly related to the skeleton. For example, the liver moves several centimeters with each respiration relative to the rib cage due to the quite necessary movement of the diaphragm. As such, without special accommodations, stereotactic treatments in the body will be less precise. Initially, systems employing rigid ﬁxation to the spinal skeleton were utilized to immobilize the patient and guide X-rays from a linear accelerator akin to brain radiosurgery. Eventually, noninvasive techniques have been reported for both head and neck region and spine treatments employing either implanted ﬁducials for tracking or noninvasive masks and body casts [7].

687

688

l. potters

In the early 1990s, use of the Stereotactic Body Frame in Stockholm, Sweden, allowed for stereotactic treatments in solid organs within the body without rigid ﬁxation [8]. Others employed alternate immobilizing ﬁducial frames (e.g., Leibenger, Z-med, and Medical Intelligence) or designed their own systems to conduct extracranial stereotactic treatments. More recently, technology originally developed for performing stereotactic radiation therapy in the brain has either been modiﬁed or recommissioned to carry out stereotactic treatments within the body (e.g., Novalis and CyberKnife). During this period, some clinical experience has been reported based mostly on technique and quality control, and not on outcomes.

Radiobiological Considerations Stereotactic radiation delivery shares many of the principles of targeting and treatment delivery pioneered for similar treatments within in the brain. Logically, it would seem to follow that knowledge of stereotactic treatment from the brain combined with clinical experience in the body would constitute a straightforward basis for conducting stereotactic treatments in the body. Perhaps this logic is appropriate when employing conventionally fractionated dose schedules. Undoubtedly, technology developments have improved our ability to deﬁne targets and spare normal tissue; however, the decision to deliver unconventionally large doses of radiation must be undertaken with a keen appreciation for the fact that biologic effects may be drastically different than those seen with standard fractionation. In particular, late effects (like ﬁbrosis and vascular injury) and overwhelming acute effects (like hepatitis and mucosal sloughing) can lead to disastrous consequences for patients after less than prudent, unconventional large-dose stereotactic delivery.

Semantics One way to conceptualize radiation treatment is from the perspective of the tumor or target lesion. Assuming normal tissue exposure is limited, all the target sees are photons, or packets of energy. Alternatively, distinct technology such as particles using protons or neutrons can be used for some treatment. Manipulation of photons within a ﬁeld (or beam) that covers the target also represents distinct technology referred to as intensity-modulated radiation therapy (IMRT). What then makes stereotactic treatment distinct from the target’s perspective? Actually, not that much. One can use conventional photons (cobalt or linac based), IMRT, or particles to treat benign and malignant conditions. Therefore, it is not “stereotactic” radiation therapy that is distinct from the target’s perspective, but rather the normal tissue’s perspective, where stereotactic radiation can be delivered so precisely that normal tissue is spared. Therefore, SRT is deﬁned as high-dose radiation therapy with a very low tolerance of beam deviation coupled with image guidance technology to assure target coverage. The problem with SRT terminology is that, for better or for worse, new descriptive terms are constantly being deﬁned. Descriptive nouns and adjectives are not infrequently introduced, for any of several reasons: to describe a particular

apparatus or methodology, to differentiate from competitive forms of treatment, or perhaps to carve out niches for reasons as varied as personal, institution, or vendor promotion. As a result, the Centers of Medicare and Medicaid Services (CMS), which manages the Medicare program in the United States, receives regular requests to (re-)deﬁne stereotactic radiation in ever more complex, narrow, and limited terms. These requests are initiated by individuals, company representatives, or representatives of physicians. Of course, each request is meant to inﬂuence coding and reimbursement, to the beneﬁt of some subgroup, but often absent long-term clinical outcomes data. Any single dose of radiation delivered with a single beam or with multiple beams is called a fraction (sometimes called a shot and occasionally called a stage; proponents of this latter term liken “staged” radiation procedures to “staged” surgery, in which different surgical procedures are performed on different days, though when it is applied to teletherapy procedures it inappropriately refers to repeated identical radiation fractions). Teletherapy refers to radiation entering the body from an external source or from multiple external sources. The complex dose distribution in a patient of radiation delivered from either a single beam or from multiple beams—a distribution dependent on beam characteristics and differential tissue absorption of radiation—is called the isodose distribution. When multiple beams are used for any single fraction, the single fraction radiobiological effect of a given isodose distribution does not depend on whether the beams contributing to the isodose plan enter the body sequentially or simultaneously. Given that various teletherapy machines and methods exist, it is not surprising that teletherapy users and vendors have developed a variety of names applied to brand, model, technique, or purported effect. Some terms distinguish between well-established teletherapy methodologies, such as IMRT, proton beam radiation therapy, and stereotactic radiation therapy, and have separate and appropriate codes. The adjective stereotactic is properly applied to a subset of teletherapy treatment techniques called stereotactic radiation therapy, deﬁned by the U.S. Nuclear Regulatory Commission (NRC) as “The use of external radiation in conjunction with a stereotactic guidance device to very precisely deliver a therapeutic dose to a tissue volume.” This term correctly applies no matter the number of fractions delivered and no matter the body site, provided stereotactic radiation techniques are used. In spring 2006, members of the American Society of Therapeutic Radiology and Oncology (ASTRO), the Congress of Neurosurgeons (CNS), and the American Association of Neurosurgery (AANS) met to deﬁne the term stereotactic radiosurgery (SRS). The consensus deﬁnition approved by the Boards of each of these organizations is Stereotactic Radiosurgery is a distinct discipline that utilizes externally generated ionizing radiation in certain cases to inactivate or eradicate (a) deﬁned target(s) in the head and spine without the need to make an incision. The target is deﬁned by high-resolution stereotactic imaging. To assure quality of patient care the procedure involves a multidisciplinary team consisting of a neurosurgeon, radiation oncologist, and medical physicist. Stereotactic Radiosurgery (SRS) typically is performed in a single session, using a rigidly attached stereotactic guiding device, other immobilization technology and/or a stereotactic image-guidance system,

69.

the semantics of stereotactic radiation therapy

but can be performed in a limited number of sessions, up to a maximum of ﬁve. Technologies that are used to perform SRS include linear accelerators, particle beam accelerators, and multisource Cobalt 60 units. In order to enhance precision, various devices may incorporate robotics and real time imaging.

All stereotactic radiotherapy procedures, and thus all radiosurgery procedures, are teletherapy procedures. For SRS procedures (as properly deﬁned above), both a radiation oncologist and a neurosurgeon must jointly participate [9-13], as this reﬂects past, current, and future standard of care. Occasional claims that radiosurgery is “surgery” rather than teletherapy strain credulity [14]. Such claims are divisive and certainly unnecessary given that the participating neurosurgeon is considered a necessary participant. Such claims potentially confuse the greater claim, which radiation oncologists and neurosurgeons agree on, that radiosurgery is an effective and safe treatment modality for a wide variety of brain indications. In addition, the CMS has recognized that both physician specialists are necessary, by stating that “for the safe and effective delivery of Cobalt 60-based multisource photon SRS to typical patients with brain lesions, the contributions of hospital physician and nonphysician staff with expertise in neurosurgery and radiation therapy are essential for both the planning of the treatment and its delivery” [15]. Stereotactic body radiation therapy (SBRT) describes a stereotactic radiation therapy treatment that is not SRS. Any treatment delivered to the body excluding the head and spine delivered in one to ﬁve fractions with image guidance is thus deﬁned as SBRT. SBRT coding was established ﬁrst by the introduction of robotic G codes to reimburse Part A providers initially for CyberKnife machines, as a result of the request by the vendor. However, almost all modern linear accelerators use “robotic” means of some sorts. Computerization of the medical record, the treatment planning equipment, and the function of the linear accelerators with multileaf collimation and intensitymodulated delivery cannot be done without some degree of automation. The small linear accelerator coupled to the movable arm of the CyberKnife system has low output, which protracts treatment delivery. As such, image guidance is needed during the treatment delivery to conﬁrm that stereotaxis is maintained during the treatment delivery. This combination of image guidance coupled with the lack of a gantry-mounted system was presented as a new “robotic” delivery machine for the creation of G-codes. Other vendors offer treatment machines with higher outputs, and therefore intrafraction image guidance may not be needed. As of 2007, the nomenclature and coding deﬁnitions will stabilize. SRS is now recognized as a distinct service from SBRT. Although the CMS will continue to collect data using various G-codes for SBRT delivery, it is expected that the Gcodes will ultimately merge into the CPT code deﬁnitions. Changes in the process of care may require new codes. Nonetheless, the confusion with the nomenclature of SRS and SBRT should go away. Although included to represent the most up-to-date billing CPT deﬁnitions, these examples are not meant as a coding guide. Other codes are necessary to account for the entire process of care but are not included in this example. The fol-

689

lowing codes are only for single-fraction cranial SRS complete course of treatment in one session. 77371: Radiation treatment delivery, stereotactic radiosurgery (SRS), complete course of treatment of cerebral lesion(s) consisting of 1 session; multi-source cobalt-60 based. 77372: Radiation treatment delivery, stereotactic radiosurgery (SRS), complete course of treatment of cerebral lesion(s) consisting of 1 session; linear accelerator based. For 2007, hospital facilities (Part A billing) will be required to report the following G-codes instead of 77372. G0173: Stereotactic radiosurgery, complete course of therapy in one session. G0339: Image-guided robotic linear accelerator-based stereotactic radiosurgery, complete course of therapy in one session, or ﬁrst session of fractionated treatment. For SBRT or fractionated SRS, CMS approved a new CPT code 77373 (“Stereotactic body radiation therapy, treatment delivery, per fraction to 1 or more lesions, including image guidance, entire course not to exceed 5 fractions”) to describe linear accelerator–based SRS treatment delivery. This code can be billed per fraction up to a maximum of ﬁve fractions and cannot be associated as a “boost” with a conventional course of radiation therapy. CMS will continue to request (at least for 2007) that hospital facilities (Part A billing) will use the following G-codes: G0251: Linear accelerator-based stereotactic radiosurgery, delivery including collimator changes and custom plugging, fractionated treatment, all lesions, per session, maximum 5 sessions per course of treatment. G0339: Image-guided robotic linear accelerator-based stereotactic radiosurgery, complete course of therapy in one session, or ﬁrst session of fractionated treatment. G0340: Image-guided robotic linear accelerator-based stereotactic radiosurgery, delivery including collimator changes and custom plugging, fractionated treatment, all lesions, per session, second through ﬁfth sessions, maximum ﬁve sessions per course of treatment. Based on the process of care presented in this chapter, it is clear that the G-codes do not best deﬁne the delivery of SRS or SBRT. With time, the CPT codes will better reﬂect the current technology.

Conclusion The term stereotactic radiation therapy represents a movement within radiation oncology to deliver safe, high doses of radiation to a target with little to no normal tissue exposure. Understanding that the application of this new technology to treat human diseases may include patients that currently may not be considered for radiation therapy, the application of SRS and SBRT may expand the ﬁeld of radiation oncology. With that, there remains a fundamental need that all physicians work cooperatively to achieve the potential good for all patients. There remains a distinct need to understand the principles of radiation biology and physics, as well as the pathophysiology of the disease process. If cooperative trials expand and conﬁrm

690

l. potters

that stereotactic radiation treatment offers better and safer outcomes than conventional surgery, or conventional fractionation schemes, as it has for some cerebral conditions, then that achievement will beneﬁt society as a whole. I believe we getting beyond the past war of terminology and technology ownership with the new CPT codes that better lay a foundation for this technology.

References 1. Douglas RM, Beatty J, Gall K, et al. Dosimetric results from a feasibility study of a novel radiosurgical source for irradiation of intracranial metastases. Int J Radiat Oncol Biol Phys 1996; 36: 443–450. 2. Adams EJ, Cosgrove VP, Shepherd SF, et al. Comparison of a multi-leaf collimator with conformal blocks for the delivery of stereotactically guided conformal radiotherapy. Radiother Oncol 1999; 51:205–209. 3. Rosenthal DI, Glatstein E. We’ve got a treatment, but what’s the disease? Or a brief history of hypofractionation and its relationship to stereotactic radiosurgery. Oncologist 1996; 1:1–7. 4. Shepherd SF, Childs PJ, Graham JD, et al. Whole body doses from linear accelerator-based stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 1997; 38:657–665. 5. Timmerman R, Papiez L, McGarry R, et al. Extracranial stereotactic radioablation: results of a phase I study in medically inoperable stage I non-small cell lung cancer. Chest 2003; 124: 1946–1955. 6. Kim KH, Cho MJ, Kim JS, et al. Isocenter accuracy in frameless stereotactic radiotherapy using implanted ﬁducials. Int J Radiat Oncol Biol Phys 2003; 56:266–273.

7. Sperduto PW. A review of stereotactic radiosurgery in the management of brain metastases. Technol Cancer Res Treat 2003; 2: 105–110. 8. Phillips MH, Stelzer KJ, Grifﬁn TW, et al. Stereotactic radiosurgery: a review and comparison of methods. J Clin Oncol 1994; 12:1085–1099. 9. Brown PD, Wald JT, McDermott MW, et al. Oncodiagnosis panel: 2002. Metastatic NSCLC. Radiographics 2003; 23:1591– 1611. 10. Wulf J, Haedinger U, Oppitz U, et al. Stereotactic radiotherapy for primary lung cancer and pulmonary metastases: a noninvasive treatment approach in medically inoperable patients. Int J Radiat Oncol Biol Phys 2004; 60:186–196. 11. Fuss M, Thomas CR Jr. Stereotactic body radiation therapy: an ablative treatment option for primary and secondary liver tumors. Ann Surg Oncol 2004; 11:130–138. 12. Fuss M, Salter BJ, Cavanaugh SX, et al. Daily ultrasound-based image-guided targeting for radiotherapy of upper abdominal malignancies. Int J Radiat Oncol Biol Phys 2004; 59:1245– 1256. 13. Larson DA, Bova F, Eisert D, et al. Current radiosurgery practice: results of an ASTRO survey. Task Force on Stereotactic Radiosurgery, American Society for Therapeutic Radiology and Oncology. Int J Radiat Oncol Biol Phys 1994; 28:523–526. 14. Kondziolka D, Lunsford LD, Loefﬂer JS, et al. Radiosurgery and radiotherapy: observations and clariﬁcations. J Neurosurg 2004; 101:585–589. 15. Department of Health and Human Services, Centers for Medicare & Medicaid Services, 42 CFR Part 419 [CMS-1427-FC]. Available at http://www.cms.hhs.gov/providers/hopps/2005fc/cms1427fc.pdf (p. 103).

7 0

Building a Radiosurgery Program N. Scott Litofsky and Andrea D’Agostino-Demers

One of the major trends in medicine over the past decade is the development of minimally invasive techniques for performance of surgical procedures. Patients desire these types of treatments because they are usually associated with less pain and quicker recovery. Insurance companies like these procedures because they are usually associated with short lengths of hospital stay and therefore less cost. Key issues that have had to be established include the development of the needed technologies, identiﬁcation of appropriate patient proﬁles and disease processes that lend themselves to these treatments, and education of physicians—both those that treat the disease processes and those who refer patients for such treatments—about the procedures. In neurosurgery, one of the most prevalently used minimally invasive techniques is stereotactic radiosurgery. Originally described by Lars Leksell in 1951, stereotactic radiosurgery is a technique in which a single, high dose of ionizing radiation is focused on a three-dimensional target volume using stereotactic guidance. A variety of radiation sources, imaging technologies, and stereotactic localizing devices have been developed to make this procedure readily available to patients throughout the world. As hospitals, clinics, and physicians develop radiosurgery programs, they must determine which of these available technologies will be most appropriate for their practice setting. They must determine how to coordinate the various disciplines involved into a cohesive working group. They must let physicians and patients inside and outside the community know about the available treatment options. This chapter discusses many of the issues that should be considered when building a radiosurgery program from the ground up.

ﬁrst patient in 1987; the Pittsburgh unit was the ﬁfth in the world. By 1999, 123 Gamma Knife units were in use worldwide [1], and by 2004, 170 Gamma Knife units had been set up [2]. Linear accelerator (linac) radiosurgery programs developed later, with the ﬁrst unit established in Barcia-Solario, Spain, in 1982. Lutz and Winston started the ﬁrst United States linac radiosurgery program in Boston in 1986, and subsequent growth has exceeded that seen with the Gamma Knife; by 1999, more than 200 linac programs were present in the United States alone [1]. A third form of delivery of ionized radiation for radiosurgery—heavy charged particles—has also been available since the 1950s (initially at Lawrence Livermore in Berkeley). A handful of other programs are available in the United States and throughout the world (14 in 1997), but this technology has been limited by the cost of building a cyclotron to produce the protons [1], currently of the order $40 million to $50 million. Clinical indications for treatment with stereotactic radiosurgery have expanded as availability has increased. The largest group of patients is those with metastatic tumors to the brain, where radiosurgery has been shown to be more cost-effective than open surgery in several studies [3, 4]. Patients with gliomas who have been treated with radiosurgery have been shown to have improved survival compared with others [1]. “Benign” tumors, such as meningiomas, particular those involving the skull base or dural sinuses, are frequent targets, as are schwannomas and pituitary adenomas. Arteriovenous malformations can be effectively treated with radiosurgery. Although Leksell initially developed radiosurgery to create lesions for functional neurosurgery, it has only been in the past several years that such use has become more routine. Epilepsy, trigeminal neuralgia, pain, and movement disorders are all treatable with radiosurgery [1, 5].

Current Utilization

Future Growth of Radiosurgery

Lars Leksell, of the Karolinska Institute, introduced the term stereotactic radiosurgery in 1951; he treated his ﬁrst patient with the Gamma Knife, a cobalt-60 source of ionizing radiation, in Sweden in 1967. Dade Lunsford, at the University of Pittsburgh, introduced the Gamma Knife to the United States, treating his

Current clinical indications create a theoretical patient base of about 180,000 patients a year in the United States who could be candidates for stereotactic radiosurgery. The number actually treated is not nearly this high because other forms of therapy are often utilized because of lesion location, size,

Introduction

691

692

n.s. litofsky and a. d’agostino-demers

or numbers. In actuality, extrapolating data from the Innovations Center, about 72,000 patients could be treated with radiosurgery each year [6]. Because of increased cost-effectiveness, less procedural morbidity, and patient choice, radiosurgery is used increasingly as the ﬁrst-line treatment for its indications. As such, craniotomy as a treatment is expected to slightly decline over the next decade, with radiosurgery volumes increasing by almost double. For instance, radiosurgery cases for neurooncologic indications are expected to increase from about 18,000 in 2004 to 32,000 in 2014 [6]. Furthermore, many patients previously thought to have inoperable lesions by location or patient health are candidates for stereotactic radiosurgery. Population aging and population growth will also increase potential patient volumes. Lastly, as patients become more informed about health care options, they are more and more frequently requesting treatment with minimally invasive techniques, such as radiosurgery [7]. Clearly, the demand for stereotactic radiosurgery treatment will require the development of new programs where availability is low to improve patient access. Radiosurgery programs, as part of a national trend to development centers for neuroscience care, are and will be a “hot” topic for the next several years. Radiosurgery is also an important component of neurologic surgery residency training. The Residency Review Committee for neurologic surgery is proposing expanding the specialty deﬁnition of neurologic surgery to include treatment of diseases of the brain and spinal cord and their coverings and blood supply with stereotactic radiosurgery. This expanded deﬁnition will more formally address the need for neurosurgeons to have clinical and didactic training in radiosurgery during residency training (ACGME—Program Requirements for Residency Education in Neurological Surgery, personal communication). Concerns have been raised about potential market saturation as new programs are built, which can limit individual program patient volumes [6]. Despite these concerns, the role of stereotactic radiosurgery as a tool to treat diseases of the nervous system is quite high and should remain so for the foreseeable future.

Program Development Program development should proceed in a fairly stepwise fashion. It is wise to keep a broad overview in mind throughout the process. Some issues may need to be addressed simultaneously; others depend on decisions made in the preceding steps. Table 70-1 lists the recommended steps, which are discussed in greater detail below. TABLE 70-1. Radiosurgery program development steps. 1. 2. 3. 4. 5. 6. 7. 8. 9.

Decision to proceed Identiﬁcation of key participants Identiﬁcation of space Identiﬁcation of ﬁnancial resources Selection of hardware and software Integration of vendors into the program Credentialing physicians Marketing to the community Treating patients

Decision to Proceed The ﬁrst step in developing a radiosurgery program is making the decision to proceed with such a program. It is important to remember that because a radiosurgery program is a multidisciplinary program, input will be required from a variety of viewpoints, including representation from neurologic surgery, radiation oncology, medical oncology, hospital or clinic administration, and nursing. In some settings, such input may be required from competing institutions, in order to maximize efﬁciencies of limited ﬁnances and population base. Although one group could make a unilateral decision for proceeding, and decide how they wish to proceed, most of these disciplines do not function in a vacuum; decisions of one group frequently affect the other parties, sometimes signiﬁcantly. For instance, in a medical community with a 500,000 catchment area, it probably does not make much sense for two institutions to develop independent programs; each institution would probably be more successful if they collaborate. Similarly, if neurologic surgery decides that they would prefer to purchase a Gamma Knife for their program, but the institution’s administration does not have space or ﬁnancial resources to build a new structure to house the device, then a different direction will be required. Early dialogue can steer the development in the most appropriate direction. A team approach, in which issues are discussed before expenditure of signiﬁcant time and money, leads to more expedient program development. Sharing of information in a close working group helps ensure that all parties can develop a shared philosophy and work toward a common goal. Whereas individuals or small groups can beneﬁt from looking out for their own self-interests, the radiosurgery program will do best if the greatest-good concept is adopted; that is, how can the group develop the program so that the program, rather than one speciﬁc discipline, will be most successful. In some communities, two or more institutions or groups may feel that it is in their best interests to collaborate to develop the radiosurgery program. These groups may be multiple academic health centers, physician groups, or community hospitals. Working together, these groups may be able to achieve economies of scale that a single institution cannot realize. Overall patient population base should be larger in a combined effort. Areas of strength, which often differ between institutions, may be combined to create a sum greater than each of the individual parts. Financial resources, when combined, may allow purchase of equipment and other items that may not have been affordable by a single entity. This type of joint venture is outlined by Hession and Brown of the Advisory Board Company [8]. To best integrate multiple institutions into a functioning working group, the following recommendations, devised for enhancing collaboration for radiation oncology services, are helpful in this setting, as well [8]: 1. Form a joint executive task force to coordinate strategic planning. 2. Incorporate ideas of all parties to ensure universal support. 3. Ensure all parties are comfortable with terms of the contract. 4. Create a successful marketing strategy. 5. After establishing the arrangement, continue to service the relationship.

70.

building a radiosurgery program

Identiﬁcation of Key Participants Once the institution decides to proceed with building a radiosurgery program, key individuals from each discipline should be identiﬁed to participate in planning. Obvious involvement is required by neurologic surgery. Neurosurgeons play a key role in patient selection, head immobilization by halo ring if required by the type of hardware that is chosen, and anatomic deﬁnition of treatment targets. Because many patients may also be candidates for open surgery as an alternative treatment, neurosurgery needs to be sold on the concept that radiosurgery can play a signiﬁcant role in the management of their patients. Otherwise, conﬂicts such as that illustrated by Brada and Cruickshank [9] will prevent development of a successful program. Similarly, radiation oncology should be represented during program planning. Radiation oncology participates with neurosurgery for patient treatment in most programs. The radiation oncologist is usually responsible for the radiation dose, as well as supervising the physics personnel involved with developing the treatment plan. Radiation oncology is often, but not always, the responsible party for upkeep and quality assurance of the devices used to generate and deliver the ionizing radiation. Furthermore, radiation oncology participates in patient referral for radiosurgery treatment. Although medical oncology does not directly participate in stereotactic radiosurgery treatment, their participation in program development is probably wise. The largest group of patients who are candidates for stereotactic radiosurgery are those with metastatic disease to the brain. Medical oncology, therefore, serves as the greatest referral source of patients. Better patient referral can be expected when the medical oncologists understand the issues related to stereotactic radiosurgery treatment. Hospital or clinic administration should also be involved during the planning process. The chief executive ofﬁcer may chose to represent hospital administration himself or herself, or rather designate another administrator to assist in program development. Administration usually controls available space and ﬁnancial resources for whatever new construction or purchases will be required. Resources for developing a business plan are also available through hospital or clinic administration. Questions that these leadership individuals typically will ask include the following: 1. What is the most appropriate radiosurgery system for [their] institution? What are its advantages and disadvantages? 2. What are present clinical volumes and future projected volumes? 3. Who will be responsible for operating the radiosurgery system and training physician users and technical support staff? 4. What is local and regional competition for such a program? 5. Who should be involved on the steering committee for planning the program? Administration usually has access to some of this information from market analyses. They may require some assistance understanding current practice patterns, as well as medical issues, but administration usually understands the long-term ﬁnancial picture for the program.

693

Although other individuals may also be asked to participate in planning, a representative from nursing administration is also helpful. Patients need nursing care during pretreatment evaluation and during treatment. Nursing personnel may participate during patient education, coordination of pretreatment tests, administration of conscious sedation during ring placement, and patient care throughout the treatment day. Commitment by nursing administration to provide for necessary manpower makes the overall program run more ﬂuidly and efﬁciently. Although anesthesiology has little to do with facility planning, their involvement during patient treatment may be necessary when using conscious sedation for head ring placement. Most neurosurgeons can easily learn conscious sedation techniques for adult patients; however, sedation of pediatric patients is more complicated and usually involves sophisticated anesthesia techniques requiring either an anesthesiologist/nurse anesthetist or a pediatric critical care physician. Because pediatric patients can account for 10% to 15% of patients in some centers [10], participation in the planning process by anesthesiology is warranted.

Identiﬁcation of Space As with any program, space will be required. New space may be necessary to house the radiosurgery device. Such space may involve building a shielded bunker in the hospital or building a new, free-standing facility [10, 11]. Different devices have different space requirements. Alternatively, conversion of previously existing space or replacement of old equipment with new can be used. These choices will depend on the collaborators involved in the project and the spaces available for such an endeavor. Other items that will require space will include computer planning stations, equipment storage facilities, and radiation delivery computer systems. Space to apply and remove the immobilization head ring is an issue, too. Advantages of an in-hospital site include ease of access to neurosurgeons who use that hospital as their primary site of activity. Many academic neurosurgeons have an ofﬁce in or connected to the hospital, so an in-hospital site can assist with their continued academic productivity in terms of research, writing, and teaching. They can also be otherwise clinically active, seeing patients in clinic, supervising concurrent activity in the operating room (as long as the key portions of the operation do not overlap with the key portion of the radiosurgery), and providing emergency consultative coverage as well. Conversely, an off-site facility can increase logistical difﬁculties to perform these activities concurrently. However, if the facility is equipped with exam rooms, concurrent patient visits can be scheduled. Also, the electronic era permits access to e-mail and academic ofﬁce computer programs, so the neurosurgeon can do other work when not directly involved during the radiosurgery procedure. A free-standing facility may be desirable when the radiosurgery program is a collaborative effort between two otherwise competing institutions. Pride and possession can make collaboration difﬁcult if the radiosurgery facility is in one institution versus another, unless one institution clearly has more space available. Occasionally, patient access can be better with a freestanding facility. Alternatively, patients may be unfamiliar with the location of an off-site center.

694

n.s. litofsky and a. d’agostino-demers

Identiﬁcation of Financial Resources Part of decision to proceed with a radiosurgery program will be based on the ﬁnancial picture of the institution. Although radiosurgery programs can generate signiﬁcant revenues for the institution, as well as the physicians, an initial ﬁnancial outlay is necessary. Purchase of hardware and software can run of the order several million dollars, depending on the vendor and device. Other anticipated costs include preparation of the site for the radiation delivery device, maintenance contracts, computer upgrades, and additional staff requirements. An appropriate business plan and examination of available capital are necessary at this point. Many vendors sponsor programs in which they “partner” with institutions, placing hardware and software in a “lease” type setting, with a portion of generated revenues returning to the vendor. In some arrangements, the institution pays a portion of the generated revenues to the vendor per treatment. The institution must still prepare the site for the equipment, pay staff, and pay for maintenance but does not need to actually buy the radiation delivery device [6]. Some institutions ﬁnd these types of arrangements to be unacceptable because longterm proﬁts for the institution can be signiﬁcantly limited. For others, such a “lease” program may be the only way that radiosurgery equipment can be made available for patient care.

Selection of Hardware/Software A number of options for radiosurgical devices are available on the market. Each option has its own pros and cons. Factors such as patient population base, institutional ﬁnancial resources, and physician preference all play a role in determining which device is most appropriate. Generally speaking, though each device has its advantages and disadvantages, the treatment efﬁcacy is of each is relatively equivalent [12]. Reimbursements to the facility may differ between devices and between third-party payors. These differences can complicate the development of a business plan. Proton beam systems are by far more expensive, by an order of magnitude. Costs may run $40 million to 50 million [1, 12, 13]. Unless a synchrocyclotron is nearby, such a delivery system is really not an option for the vast majority of radiosurgery programs to consider. Therefore, proton beam issues will not be discussed further here. Gamma Knife, which has the longest track record for radiosurgical treatment, is probably the best technology from a marketing perspective. As a radiosurgical tool, Gamma Knife has the best and most widespread name recognition. When patients enquire about radiosurgery, they usually ask, “Is that the Gamma Knife?” Many physicians also generically refer to radiosurgery as “Gamma Knife.” Gamma Knife planning times are fairly short, so more than one patient can be treated in 1 day. It has less moving parts than other systems, so maintenance is less and down-time is less; more patients per year can theoretically be treated [14, 15]. It has several limitations, however. Initial costs for Gamma Knife are higher than for linac systems [14]. The Gamma Knife unit is approximately $3.2 million. Because of a radiation decay half-life of 5.25 years, treatment time doubles by 5 years; therefore, the cobalt sources must be replaced every 5 years, at considerable expense, estimated at

approximately $750,000. An upgraded radiation vault is also required, running about $700,000 new. Maintenance contracts (which all systems require) are about $100,000 [6]. Another drawback of the Gamma Knife is that it can only be used for intracranial applications. Because break-even volume is estimated at 86 patients per year [6], small-volume centers may have difﬁculty recouping their investment. Gamma Knife also requires rigid head immobilization; some new technologies can be used without application of a halo ring, which improves patient comfort. Purchase of a Gamma Knife has allowed some programs to grow greatly, exceeding expectations of program directors [16]. For greater than 200 radiosurgery patients per year, Gamma Knife has lower overall costs than a linac from an accounting perspective [15]. In a setting in which a large treatment population is available, the Gamma Knife is a frequent choice because of its ease of operation, ability to provide conformal dose plans for a wide variety of target shapes and sizes, and general familiarity of the device [17]. The requirement for rigid head immobilization, however, may be a signiﬁcant long-term limitation with this technology. CyberKnife is a robotic linac equivalent used for stereotactic radiosurgery. It is a relatively new delivery system, developed in 1997 at Stanford University, and approved by the U.S. FDA in 2001 [18, 19]. It does not require rigid head immobilization, and it can be used for intracranial and extracranial indications, but only for stereotactic or intensity-modulated radiation therapy (IMRT) indications; it is not able to be used for standard linac applications. Cost is about $3.5 million. An upgraded vault is also required. Annual hardware and software upgrades may run $225,000 to $450,000. Most facilities will break even ﬁnancially treating 109 patients per year [6]. Because CyberKnife is a small linear accelerator, treatment times are somewhat longer than those systems using a standard linac. CyberKnife also is developing name recognition, which helps its marketing to physicians and patients. Furthermore, because a head frame is not required for treatment, some patients speciﬁcally request treatment with CyberKnife, as opposed to Gamma Knife or other frame-based linac treatments. A number of linac-based systems are available. Vendors include Nomos, Varian, Radionics, and BrainLAB. Each system has somewhat different features. New system costs run about $2.5 million to $3.2 million. If used solely for radiosurgery, break-even patient volume for linac systems is 122 patients per year [6]. But because the linac can be used for other radiation indications, radiosurgery volume does not need to be nearly that high. Retroﬁtting an existing linac to perform radiosurgery with the usual cone collimators is probably the most inexpensive means of beginning a radiosurgery program. Software planning systems are available for several hundred thousand dollars. One feature that can be found in a linac system is a minimultileaf collimator to enhance treatment options. Mini-multileaf collimators allow shaping of X-ray beams to conform more precisely to the shape of the lesion in the beam’s-eye view. They permit more homogenous radiation dosing to the lesion than can be obtained with either the Gamma Knife or circular collimators using non-coplanar arcs [20]. Using a mini-multileaf collimator, each lesion can be treated using a single isocenter, or target, which permits faster treatment time. Frequently, a mini-multileaf collimator can be added to an existing linac.

70.

building a radiosurgery program

Mini-multileaf collimators tend to be a bit bulkier than circular cone collimators, so physics staff and radiation therapists must be aware of the possibility of collision between the linac and the patient/couch [21]. A mini-multileaf collimator and its accompanying software can be added to an existing linac for about $700,000. Novalis, made by BrainLAB, is an example of a standard linac-based device. With its software and hardware additions, it can be used for intracranial and extracranial radiosurgery. An additional feature is that it can also be used for standard radiation therapy and IMRT. Therefore, it can function as a backup linear accelerator. Novalis, like CyberKnife, does not require rigid head immobilization for most of its radiosurgery indications; such immobilization is available for functional indications such as trigeminal neuralgia to improve accuracy. Novalis also includes a mini-multileaf collimator within its platform. Beams can be shaped in either static or arcing ﬁelds to increase ﬂexibility. Treatment times with Novalis (7 Gy/min) are faster than for CyberKnife (4 Gy/min) because it is similar to a standard linac. For programs with relatively small anticipated radiosurgery volumes, a linac-based program may make the most sense [17].

Integration of Vendors into the Program Once a decision is made for which hardware will be best for the individual program (one system will not be best for all programs), purchase should proceed. It is important to be sure that the planned system will integrate with existing hardware. Some institutions may chose to purchase some hardware and software to add to an existing linac as a relatively inexpensive way to begin a radiosurgery program. But not all computer programs for the various devices speak the same language (are compatible), and signiﬁcant linkage issues can slow utilization of the purchases. Furthermore, once the purchase has occurred, many vendors do not respond as quickly as programs may desire to requests for assistance in solving these problems. Sometimes, proprietary issues must be resolved before a solution to linkage issues will be forthcoming. These issues can delay by many months the planned start date for programs. This lack of computer communication tends to occur most often when incorporating a new mini-multileaf collimator into a linac manufactured by a different vendor. Anticipation of these problems in advance can lead to a more seamless transition to utilizing new hardware and software. Vendors can frequently play a signiﬁcant role in the early stages of program function. Vendors frequently offer courses to physicians and other staff so that they can gain familiarity with new equipment. As previously discussed, they can work with institutions to develop the most appropriate purchase plan. Vendors can review the available space on-site in which the equipment will be placed to be sure that appropriate speciﬁcations are met. Vendors can also provide technical assistance during the installations, quality assurance, and operation of the equipment. Computers at the site of operation can be linked to the vendor’s technical support ofﬁces to assist with troubleshooting. Evaluation of these issues prior to purchase is essential. It is recommended that program representatives speak to other users of the equipment, as well, to be sure that the information presented by a vendor’s salespeople is accurate in practice.

695

Credentialing Physicians Institutions and patients both want to be certain that physicians who care for patients are qualiﬁed to do so. Radiosurgery is no different than any other treatment in that regard. For neurosurgeons, radiosurgery is not as technically demanding as other procedures (microvascular surgery, for instance), but it does require a signiﬁcant cognitive skill set. The neurosurgeon needs to know which patients and lesions are appropriate for treatment, what structures are potentially in harm’s way during treatment of a particular patient, what is an appropriate target volume, and so forth. The radiation oncologist must frequently incorporate past radiation treatments into a radiosurgery prescription to achieve an effective dose without undue long-term sequela. Therefore, it is important that physicians participating in the radiosurgery program are able to be credentialed. Such credentialing is particularly important if the radiosurgery program is open to multiple physicians from outside the hospital/clinic in order for the program to maintain quality control. A few credentialing protocols have been described for neurosurgeons [11, 16]. A protocol similar to one of these is reasonably effective without being burdensome and will give managers and patients conﬁdence that the physicians are competent. A credentialing protocol should probably include the following features: 1. Neurosurgeons should complete a radiosurgery course, ideally sponsored by the vendor of the software/hardware being used in the program. The course should include simulated cases. Most vendors include such a course for a number of physicians as part of the purchase of their devices. 2. Neurosurgeons should have experience in stereotactic surgery. Such experience demonstrates competence and understanding of volumetric localization of targets in threedimensional space and ensures that halo ring mobilization, which many systems require, will be well accomplished. 3. The neurosurgeon should perform a number of radiosurgery cases under the direct supervision of an experienced neuro/radiosurgeon. This supervised experience may occur during residency training or as an attending neurosurgeon. Documentation of the experience is important, as well.

Marketing to the Community Much of the success of a radiosurgery program is related to marketing to the medical community as well as to patients. Neurosurgeons, neurologists, radiation oncologists, endocrinologists, otolaryngologists, and, perhaps most importantly, medical oncologists need to know that radiosurgery capabilities are available. Because treatment of functional disorders, such as trigeminal neuralgia, is an expanding indication, dentists and ophthalmologists should also be included [6]. Helpful information includes a list of the clinical indications for radiosurgery, inclusion and exclusion criteria, the sequence of events involved in treatment of a patient, and how to refer patients for treatment. A variety of means of communicating this information are available. Presentations at medical staff meetings, mailed brochures or letters, grand rounds presentations, and conferences are all useful for marketing purposes, particularly as they also provide appropriate education about radiosurgery. Web

696

n.s. litofsky and a. d’agostino-demers

sites and the Internet also increase physician as well as patient access to information about radiosurgery. Marketing to patients directly can also be important. Because patients are using the Internet more often as their source of medical information, programs that do not utilize Web sites are at a distinct disadvantage as patients look for options for treatment of their ailments. Therefore, information on a Web site should be written at a level that patients can comprehend. Reduction or simpliﬁcation of medical vocabulary is helpful to make the information more understandable. Television and radio spots can also be used. Special events, such as groundbreaking ceremonies and an opening day, are often covered by the local media, especially if the media is notiﬁed in advance. Two different strategies for marking are available [6]. In the ﬁrst, marketing revolves around the radiation delivery device as the key point of focus. The institution’s advertisements indicate that it has the capability of treating patients with a Gamma Knife, a CyberKnife, or whatever delivery system has been purchased for the program. This type of marketing appeals to the name-recognition of the particular device. An alternative strategy focuses on the particular disease process—like trigeminal neuralgia, or arteriovenous malformations, or brain tumors—and notiﬁes the public that a noninvasive treatment modality is available for that disease. Both strategies, especially if used together, can be very effective. Even though marketing is important to increase physician and patient awareness about radiosurgery as a tool to treat diseases, it is important to avoid exaggeration and misrepresentation in information provided, particularly on the Internet, where “there are no truth- in-advertising regulations, and where ‘data’ can be promulgated directly to patients without scientiﬁc peer review” [22]. A number of years ago, such concerns were raised by Brada and Cruickshank [9] in an editorial in the British Medical Journal. Unfortunately, these concerns were lost in the authors’ own misrepresentations [23–26]. As centers and programs market their expertise and capabilities, they must provide accurate information to maintain their credibility.

Issues Involved in Treating Patients Stafﬁng Even before the ﬁrst patient is scheduled for treatment, it is important to identify what departments will be involved in the actual care of the patient. If a frame-based system will be used, more staff will likely be required, as will more space. Staff may be from the operating room, neurological surgery, anesthesiology, radiation oncology, or other departments, depending on the venue. Table 70-2 depicts roles typically assigned in a radiosurgery program. Multidisciplinary ancillary staff involvement is critical in supporting the ﬂow and care of the stereotactic radiosurgery patient. Each staff member plays a vital role in the care of the patient as the patient travels through his or her treatment day. Because of the specialized patient care requirements, some departments need to identify speciﬁc members to be involved in the care of these patients. At the very least, these staff members should have some training. This training could include

TABLE 70-2. Roles in the radiosurgery program. Radiation oncologist

Neurosurgeon

Coordinator (NP)

Radiation nurses Radiation physicists Anesthesiologists/ anesthetists

Evaluation and selection of SRS patients Identiﬁcation of target volume and adjacent anatomies Radiation treatment and planning Calculation of radiation dose to target volume Evaluation of radiation dose to adjacent normal anatomies Evaluation and selection of SRS patients Frame application Identiﬁcation of target volume and adjacent Anatomies Image fusion of CT and MR Radiation treatment planning Administration of conscious sedation Assist in frame application and removal Coordination of patient through-fare Preprocedure education and discharge instruction Fielding patient and family questions Care of patient while patient in department Quality assurance of equipment Radiation treatment planning Administration of conscious sedation

preparation of the patient in the holding unit, assistance in frame application, medication administration and sedation, computed tomography (CT) scanning protocols and requirements, and special oncologic nursing care issues. It would also be prudent for participating staff to have a general idea of the potential beneﬁts of radiosurgery to patients. Individuals who have a better understanding of these treatment issues will feel more like a part of the team. They will be better able to provide emotional and knowledgeable support to minimize patient fear and anxiety [2]; they will also be less likely to make uninformed off-handed remarks that send the wrong message to patients. Nursing staff is an integral part for the success of the program [27]. It is best to designate a program coordinator who not only guides the patients’ throughfares but also can assist in procedures, prescribe medications, discharge patients from services, and provide follow-up care. This individual allows the neurosurgeon and radiation oncologist to focus on planning and treatment, ultimately allowing for more efﬁcient use of their time. An advanced practice nurse, or nurse practitioner (NP), is probably the best option to serve as the program coordinator, as an NP has the appropriate credentials and training to be able to function in this role. The coordinator should keep a checklist on each patient to be sure that all pretreating planning and education has been completed and that the patient is ready on the day of treatment. Nursing assistants, if available, can help with patient transportation; radiation oncology nursing staff are usually involved with patient care when the patient is being treated [27], particularly if the patient needs medications or assistance with using toilet facilities while wearing the head ring.

Halo Ring Application Halo ring application generally requires a room with monitoring capabilities. If conscious sedation will be used for such ring

70.

building a radiosurgery program

application, the neurosurgeon will need to decide if he or she is comfortable supervising its administration or if anesthesiology will be needed. For out-of-hospital centers, anesthesiology is usually not available, so the neurosurgeon and available nursing personal will usually need to work together to safely administer agents and monitor the patient. Patients are generally much more comfortable when conscious sedation is used. While doses may vary from patient to patient, Versed (adazolam) 1 to 2 mg followed by fentanyl 50 to 100 μg is usually very effective and well tolerated. ECG rhythm, blood pressure, and oxygen saturation are appropriate parameters to monitor. The ring can be applied with the patient sitting; a stroke chair allows rapid placement of the patient into a supine position in the rare event of a hypotensive episode. A resuscitation cart with appropriate medications and an ambu bag (artiﬁcial respirator) should also be readily available. A designated room in radiation oncology would be ideal, but frequently monitoring is not routinely available. Portable monitoring devices can rectify this situation. Other sites that can be considered are the operating room holding area or recovery room, an ambulatory chemotherapy infusion room, or a procedure room in or near the neurosurgery intensive care unit. Staff may need to be borrowed from the site chosen, so obviously discussion with site managers will be necessary. The same site should be used regularly to avoid confusion and duplication of resources. A mobile cart containing supplies and small equipment is helpful for making ring application more efﬁcient. The cart should contain local anesthetics, needles, syringes, gauze sponges, Betadine, and gloves. The halo ring and localizing attachments can also be stored in the cart. Band-Aids and antibiotic ointment for after the halo ring is removed should be available as well. Data sheets, neurodiagnostic images, and other materials can be carried in the cart as the patient moves from location to location.

Treatment Day Timeline One of the roles of the program coordinator is to ensure smooth ﬂow of the patient through the system, especially during the treatment day. Patients are frequently anxious, and the presence of the coordinator helps reassure the patient that the various components of the system are in appropriate communication. The coordinator therefore helps the day ﬂow efﬁciently for staff and the patient. Generally speaking, from the time of frame application to frame removal and discharge, a typical patient length of stay is approximately 6 to 8 hours, depending on the complexity of the treatment plan. A timeline for planning and treatment, such as seen in Table 70-3, helps keep all on track. If more than one patient will be treated in a single day, overlapping timelines will be required. Coordination becomes more essential so that multiple patients are not in the same place at the same time and so that each patient has all of his or her individual needs met. Timelines will be shorter for systems not requiring frame placement. In these settings, much of the planning can be done before patients even arrive for treatment. Planning and treatment time may take about 2 to 4 hours. It is important to remember that staffs of different departments are frequently on different schedules. A timeline that is

697

TABLE 70-3. Frame-based radiosurgery patient care protocol/timeline. 7:00 AM 7:10 AM

7:15 AM

7:45 AM 8:00 AM

10:00 AM

11:30 PM 12:00 PM 2:30 PM 2:45 PM

Patient arrival at ring placement site Intravenous (IV) access started Maintain IV ﬂuid, O2 by nasal canula 2 L/min Monitor pulse oximetry, telemetry Administer Versed and fentanyl IV Stereotactic frame applied with assistance of NP and nursing assistant Patient transported to CT scan Maintain IV ﬂuid Patient transported to holding area Patient monitoring and observation for comfort & safety; continue every 1 hour until treatment time Patient may have breakfast Treatment planning begins Review treatment plan with patient and family members using 3D planfrom computer; answer questions. Quality assurance on equipment Lunch for patient Decadron IV Treatment starts Frame removal Discharge instructions given

good for one department may be at odds with stafﬁng schedules of a collaborating department. Complicated treatment plans may require additional time for both planning and treatment. Examples of these types of plans include irregular lesions involving the cavernous sinus with multiple isocenters and multiple brain lesions being treated at the same sitting. These logistic constraints can create a situation where overtime may be required—a situation that most centers would like to avoid as it increases costs. When multiple patients are being treated on the same day, it might be wise to avoid two complicated plans on the same day, if at all possible.

Special Considerations Early identiﬁcation of patient comfort issues for those requiring stereotactic frame placements considerably reduces patient anxiety. Monitoring comfort is the most important factor in oncologic nursing management of the patient [28, 29, 30]. Often, the frame causes some discomfort throughout the day primarily due to the length of time the patient is required to “wear” the frame. The patients frequently complain of “peak” discomfort approximately 4 hours after application. Although oral narcotics are mildly effective, immediate relief is obtained with local anesthetics to the pin site(s) [27]. Other special considerations below are noted based on patients’ reports: 1. Eyeglasses are difﬁcult to maintain with the frame in place. Without them, patients often complain of a headache. A tool kit for glasses can be helpful so that one or both hinges may be removed, depending on the frame ﬁt, if a patient wishes. Glasses should be placed in a container for safekeeping. Glasses can be stabilized on the patient’s head by taping them to the frame. 2. A neck roll provides comfort for the patient while waiting in a recliner chair with frame attached. The roll helps

698

n.s. litofsky and a. d’agostino-demers

support the neck during the wait for treatment. Patients, however, still remain “ﬁdgety.” 3. Although most patients prefer to wait for treatment in a recliner chair, a stretcher nearby is helpful for those who prefer full recumbency. A private waiting area/room is also helpful so family may be present and participate in the patient’s care (i.e., staying with patient during ambulation). 4. The patient should have a call light or bell for safety and nursing needs if the patient is waiting in a private waiting area/room. 5. A TV and video center available to patients helps them pass the time. A patient education video, previously offered at time of neurosurgical consultation, can be viewed again, as can entertainment options.

Conclusion Each radiosurgery program should be constructed to attend to the speciﬁc needs of the institution and community that the program will serve. Not all needs will be the same. What works well in one place may not be the best solution of another setting. Consideration of this issue before proceeding with the plan to build the radiosurgery program should lead to the best longterm success.

References 1. Niranjan A, Lunsford LD. Radiosurgery: Where we were, are, and may be in the third millennium. Neurosurgery 2000; 46:531–543. 2. Nesbitt J. Gamma knife radiosurgery: a patient-friendly procedure. Axone 2004; 25:23–27. 3. Ott K. A comparison of craniotomy and gamma knife charges in a community-based gamma knife center. Stereotact Funct Neurosurg 1996; 66(Suppl 1):357–364. 4. Rutigliano MJ, Lunsford LD, Kondziolka D, et al. The cost effectiveness of stereotactic radiosurgery versus surgical resection in the treatment of solitary metastatic brain tumors. Neurosurgery 1995; 37:445–455. 5. Niranjan A, Jawahar A, Lunsford LD, et al. Radiosurgery: future directions and new frontiers. In: Germano IM , ed. LINAC and Gamma Knife Radiosurgery. Park Ridge: IL, AANS, 2000:3–10. 6. Innovations Center, Health Care Advisory Board. 2004–2005 National Member meeting: Future of Neurosciences. Strategic Forecast and Investment Blueprint. Washington, DC: The Advisory Board Company, 2004. 7. Ward-Smith P. Stereotactic radiosurgery for malignant brain tumors: the patients perspective. J Neurosci Nurs 1997; 29:117–122. 8. Hession M, Brown K. Providing radiation oncology services at health system community hospitals. Original Inquiry Brief. Washington, DC: The Advisory Board Company, 2004. 9. Brada M, Cruickshank G. Radiosurgery for brain tumors. Triumph of marketing over evidence based medicine [editorial]. Br Med J 1999; 318:411–412. 10. Barnett GH. Evolution and organization of a regional Gamma Knife center. Stereotact Funct Neurosurg 1996; 66(Suppl 1): 365–369.

11. Groetsch SJ, Hardy T, Hodgens D, et al. The open Gamma Knife Center concept. Stereotac Funct Neurosurg 1996; 66(Suppl 1):296– 301. 12. Luxton G, Petrovich Z, Joszef G, et al. Stereotactic radiosurgery: principles and comparison of treatment methods. Neurosurgery 1993; 32:241–259. 13. Adler JR. Stereotaxic radiosurgery. Surgical Rounds 1989; 12:42– 46. 14. Stieber VW, Bourland JD, Tome WL, et al. Gentleman (and ladies), choose your weapons: Gamma Knife vs. linear accelerator radiosurgery. Technol Cancer Res Treat 2003; 2:79–86. 15. Konigsmaier H, de Pauli-Ferch B, Hackl A, et al. The costs of radiosurgical treatment: comparison between gamma knife and linear accelerator. Acta Neurochir (Wein) 1998; 140:1101– 1111. 16. Suh JH, Barnett GH, Miller DW, et al. Successful conversion from a linear accelerator-based program of a Gamma Knife radiosurgery program: the Cleveland Clinic experience. Stereotac Funct Neurosurg 1999; 72(Suppl 1):159–167. 17. Pollack BE, [Comment] to Yu C, Jozsef G, Apuzzo MLJ, et al. Dosimetric comparison of CyberKnife with other radiosurgical modalities for an elipsodal target. Neurosurgery 2003; 53:1163. 18. Adler JR, Chang SD, Murphy MJ, et al. The CyberKnife: a frameless robotic system for radiosurgery. Stereotact Funct Neurosurg 1997; 69:124–128. 19. Kuo JS, Yu C, Petrovich Z, et al. The CyberKnife stereotactic radiosurgery system: description, installation, and an initial evaluation of use and functionality. Neurosurgery 2003; 53:1235– 1239. 20. Shiu AS, Kooy HM, Ewton JR, et al. Comparison of miniature multileaf collimation (MMLC) with circular collimation for stereotactic treatment. Int J Radiat Oncol Biol Phys 1997; 37:679– 688. 21. Urie MM, Lo YC, Litofsky S, et al. Miniature multileaf collimator as an alternative to traditional circular collimators for stereotactic radiosurgery and stereotactic radiotherapy. Stereotact Funct Neurosurg 2001; 76:47–62. 22. 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):90–95. 23. Adams CBT. Editorial was wrong to denigrate radiosurgery so strongly [letter]. Br Med J 1999; 318:1489. 24. Ganz JC. Radiosurgery for brain tumours. Not all practitioners of this technique can have succumbed to marketing [letter]. Br Med J 1999; 318:1490. 25. Ganz JC. Presentation based on extremely selective use of references [letter]. Available at http://bmj.bmjjournals.com/cgi/ eletters/318/7181/411. 26. Loefﬂer J, Lindquist C. Radiosurgery for brain tumors [letter]. Br Med J 1999; 318:7181. 27. Browner, CM, Hendrickson, K. A nursing perspective of gamma knife treatment. Barrow Neurological Institute Quarterly 1997; 13(1):1–6. 28. Gnanadurai A. Nursing care of patients undergoing radiosurgery. Nurs J India 2001; 92:129–131. 29. Gnanadurai, A, Purushothamam, L, Rajshekhar, et al. Stereotactic radiosurgery for brain lesions: an observation and follow-up. J Neurosci Nurs 2004; 36:225–227. 30. Wheatley R. Nursing management of the patient undergoing stereotactic radiosurgery. Br J Theatre Nurs 1995; 5:5–6, 8–9.

7 1

Patient Care in Stereotactic Radiosurgery Terri F. Biggins

Introduction The intention of this chapter is to provide basic guidelines surrounding the patient process in relation to stereotactic radiosurgery, with the goal of providing a seamless and pleasant experience for the patient and staff. Although the role of the registered nurse (RN) is highlighted, additional disciplines may gain insight into the patient perspective. Expectantly, valuable information will be gained not only for those opening a new center but for the established site as well. Various methods of stereotactic radiosurgery have been developed over the past several years, but essentially the mechanics are similar. This author’s experience is with the Leksell Gamma Knife, and therefore this will be the equipment referenced in this chapter.

Historical Perspective More than 50 years ago, a Swedish neurosurgeon pioneered a method of treating brain disorders without opening the skull. Approximately 17 years later in 1968, this same neurosurgeon, Dr. Lars Leksell, installed the ﬁrst Gamma Knife using cobalt60 at the Karolinska Institute in Stockholm. This invention was considered the ﬁrst stereotactic device. Stereotactic radiosurgery delivers a single high dose of ionizing radiation to a radiographically well-deﬁned, small intracranial target without delivering a signiﬁcant portion of the prescribed dose to the surrounding brain tissue [1]. Some of the ﬁrst patients treated were afﬂicted with pain, movement, or behavioral disorders [1]. The ﬁrst Gamma Knife center in the United States opened in 1987 [2]. It is difﬁcult to imagine what patient care must have entailed in 1968. Nurses were limited in what role they played and the duties they were allowed to perform. Short-acting intravenous sedation and portable monitoring for patient travel were not readily available. Although patients today generally live longer and require much care, patients in the 1960s presented their own challenges. In 1969, Dr. Leksell delivered stereotactic radiosurgery to many people suffering with severe pain. Parkinson disease was another disorder treated in those early years [3]. Maintaining an acceptable comfort level while undergoing a lengthy procedure must have been trying. Unlike the frames

in use today, the ﬁrst patients treated by Dr. Leksell and his staff needed to endure the placement of a plaster helmet [3]. The plaster had to dry before it could be used. All of the dose planning needed to be completed by drawing on the pneumoencephalogram. A trip across town in an ambulance was required in some cases in order to deliver the radiation [3]. One patient treatment could typically take many hours to complete. Stereotactic radiosurgery recipients beneﬁt greatly from today’s advancements in technology. Radiographic imaging in association with the use of computed tomography (CT) and magnetic resonance imaging (MRI), as well as updated computerized treatment planning has greatly reduced patient waiting time. Advancements in patient monitoring makes the delivery of sedation drugs safer. Access to education through home computers enhances a patient’s knowledge base. Today’s focus on pain control allows for easier monitoring and control of discomfort. “Frameless” systems are beginning to emerge, promising even greater tolerance of treatment in the near future. Clearly, the life expectancy in the United States and many regions throughout the world has increased dramatically. As the world’s population lives longer and technology advances, one looks for less invasive and debilitating health care treatments. Dr. Lars Leksell’s vision some 50 years go has helped achieve this very goal [4].

Training and Stafﬁng The training required to perform radiosurgery can be very specialized. In the case of Gamma Knife therapy, there are less than 100 machines in the United States. Therefore, if an institution is planning to open a new radiosurgery practice, training will likely occur at one of the existing sites in the United States. It is highly recommended that new staff attend a course at an accredited center. Ultimately, each independent radiosurgery center with input from the state radiation safety department will set rules for professional staff training. Certiﬁcates of education completion will be presented to professional staff at the conclusion of training. Physician staff must be credentialed before independently using the radiosurgery tool. The nurse should have access to this list and ensure its strict adherence.

699

700

t.f. biggins

The registered nurse is a valuable member of the radiosurgery team and must not be forgotten during the training phase. A background in various nursing specialties prepares one to adequately care for the individual undergoing stereotactic radiosurgery. A background in neurosurgery, neurology, radiation-oncology, medical oncology, and pain management can be great assets. Certainly, nursing certiﬁcation in any one or more of the above specialties would be advantageous. The volume of patients treated at each center will help determine the number of nurses and supporting staff

needed to staff the center at any one time. If one nurse typically staffs the area, it is strongly recommended that at least two other nurses are trained and ready to step in secondary to planned or unplanned absences or for an increase in patient acuity. These “back-up” nurses should complete an orientation program with the “permanent” nurse and reorient themselves by assisting during a patient treatment day every 6 to 8 weeks (Table 71-1). Supporting staff such as nurse extenders should also receive a thorough orientation period.

TABLE 71-1. Sample orientation checklist for nursing staff. 1.

Introduction to facility Date: Initial: a. Physical layout b. Supply areas c. Location of ofﬁces d. Location of hospital and unit manuals e. Medication stock and narcotics f. ECG, crash cart, respiratory supplies 2. Flow of patients: introduction to areas a. Admitting area b. Minor surgery suite: frame placement c. Radiology: MRI, CT, angiography suite d. Gamma Knife: waiting area, treatment room, family waiting e. Transfer to recovery suite f. Discharge 3. Introduction to interdisciplinary staff a. Physicians: neurosurgeons, radiation oncologist, physicist, medical director b. Gamma Knife secretary c. Radiology technicians and scheduling assistant d. Neurosurgery anesthesiologist for pediatric cases 4. Equipment a. G-frame, tray, and contents b. Emergency equipment c. Gamma Knife couch, video and audio monitoring d. Physiologic monitoring e. Fiducial indicators: MRI, CT, angiography, instructions for use f. Filling of indicators g. Assisting with “skull measurement” 5. Charting a. Admission note b. Special Procedure Nurses Note c. Time-out signatures d. Physician orders e. Patient treatment log f. Discharge documentation g. Patient teaching: pre-and postprocedure 6. Policies, protocols, and procedures a. Unit safety manual b. Credentialing manual c. Nursing procedure/policy manual (i.e., conscious sedation policy) d. Radiation dosimetry badges and monthly reports e. Routine patient care: adult f. Patient care: pediatric g. Radiation emergency procedures 7. Manage patient ﬂow-through procedure a. Readies necessary equipment in treatment and prep room b. Greets patient, preps for procedure c. Assist with frame placement d. Assist with radiology procedures e. Assist with skull measurements f. Reunite family with patient g. Provide ongoing education as well as postprocedure instructions h. Administer medications (preprocedure such as steroid, anticonvulsant, pain) i. Assist with positioning during treatment j. Monitor vital signs, chart treatment records in appropriate book k. Assist with frame removal, apply dressings l. Ready chart and transport patient to recovery m. Clean, reassemble, and store G-frame, ready ﬁxations pins for sterilization n. Complete discharge instructions Date and initial each item after completion during orientation

71.

patient care in stereotactic radiosurgery

Quality Assurance Every radiosurgery center needs a quality assurance program in place. All patients should be interviewed after treatment, possibly during a follow-up phone call from the nurse. Questions regarding patient satisfaction and outcome expectations are essential. Quarterly meetings involving the entire professional staff should be held to discuss the results of the patient interviews. Patient satisfaction goals can be set and constant reevaluation conducted to ensure quality improvement. Meetings can also be utilized to discuss any additions or changes to current protocol. Radiation safety classes and review of emergency patient procedures for all existing and new staff can be conducted in this medium. Documentation of minutes and attendance will ensure all staff have been properly included in the quality assurance process.

The Patient Flow Process Intake Patients will enter the radiosurgery program either by physician or self-referral. In the age of the Internet, more individuals are educating themselves about health care options in the comfort of their own homes in front of computers. Eventually, a phone call will be made to set up consultation appointments. All the usual information regarding the patient’s demographic data is collected at this time. Referring and primary care physician information is gathered as well. Past hospitalizations and locations of treatment records, particularly pertaining to the diagnosis that is relevant to the radiosurgery treatment, should be gathered. Radiographic imaging such as MRI, CT, and angiography prints as well as reports are an absolute must for the consultation process. A preprinted comprehensive intake form will ensure consistent data collection. A follow-up letter should be sent to each patient after this phone call. The letter should contain directions, appointment dates and times, and concise instructions on what materials the individual is responsible for bringing to the appointment. Many patients are eager to ask questions about radiosurgery and what it entails during this initial intake call. It is never too early to begin the education process. Having a clinical person available to begin dialogue at the time of intake is invaluable. Inappropriate referrals can be redirected at this time as well. The RN is an excellent resource for this role. The earlier the involvement of the nursing staff, particularly the staff that will be present at the time of treatment, the better the patient satisfaction and outcome with the entire experience.

701

a deﬁnite treatment plan, the patient’s case should be presented at a weekly neurosurgery-oncology team conference for ﬁnal approval. It is imperative that the patient records are gathered prior to the appointment so as to expedite the scheduling of the appropriate treatments. After consulting with the appropriate medical personnel, the patient should meet with a representative of the nursing staff. Viewing of a patient education video, tours of the radiosurgery facility, and the answering of questions can all help to ease anxiety. Appropriate consent forms should be obtained at this time as well. The patient should leave at the end of this consultation period with a ﬁrm treatment plan in place.

Preauthorization and Collection of Preoperative Data The preauthorization process should be started when the patient makes the decision and is scheduled for the radiosurgery treatment. Designated personnel from the treating institution contacts the patient’s insurance company. Information to have handy at the time of contact may include: 1. Diagnosis 2. Procedure code 3. Outpatient versus inpatient visit 4. Date of procedure 5. Referring physician consultation note 6. Number of lesions to be treated 7. List of treatments that have been completed related to the diagnosis Allow several days for preauthorization to be ﬁnalized. Each institution has a policy regarding what preoperative evaluations are required prior to procedures. The following list should be considered as a guide regarding data collection and chart completeness: 1. Preauthorization obtained from insurance company 2. Preoperative check list completed by RN (Table 71-2) 3. History and physical complete 4. Anesthesia consult (select patient’s receiving deep or general anesthesia) 5. ECG 6. Serum results (electrolytes, complete blood count, coagulation ) 7. Pregnancy testing on women of child-bearing age 8. Anticonvulsant serum levels (Dilantin, Tegretol, phenobarbital, etc.) TABLE 71-2. Preoperative nursing check list.

Preprocedure Consultation At the time of initial intake, the patient will be given appointment(s) with all the team members necessary to properly evaluate the appropriateness of radiosurgery. This may include consultations with neurosurgery, radiation-oncology, neuroradiology, medical oncology, and nursing. In that most patients need to travel a distance and also rely on others for transportation, attempt to combine as many appointments as possible into one visit. The neurosurgery-oncology multidisciplinary clinic is ideal for this type of referral. Prior to rendering

• Review patient allergy history • Assess any contraindication to imaging studies (contrast allergies, MRI metal concerns) • NPO status • Review of appropriate medications patient is to take prior to procedure • What to bring to hospital (music, appropriate clothing, snack, medications) • Assess anxiety level (help plan for proper sedation) • Describe ﬂow of the procedure and when and where family members will be allowed to wait and visit • Review postprocedure course

702

t.f. biggins

9. All consents complete (including angiography, MRI, CT, etc.) 10. Acquisition of imaging and reports necessary for localization Completing the above list will minimize the risk of postponement, cancellation, or a negative outcome on the day of the procedure.

Patient Arrival and Preparation The majority of radiosurgery centers require individuals to arrive early on the morning of the procedure. Patient registration should be complete prior to arrival to minimize any added anxiety. Patients should ﬁrst change into comfortable clothing as outlined in preprocedure instructions. Intravenous access is obtained, as well as any last-minute blood work. Diabetic patients should be checked for baseline serum glucose to determine any intervention. Vital signs are obtained, and a baseline neurologic assessment is completed by nursing staff. Medications taken in the past 12 hours are reviewed, and a ﬁnal check of consent forms occurs.

Frame Application After all pertinent paperwork is completed and approximately 60 minutes prior to frame placement, LMX-4 (Ferndale IP, Inc., Ferndale, MI) cream (lidocaine 4%) should be applied to the forehead and to the back of the head in the estimated region of insertion of the local anesthetic. Cover the LMX-4 cream with a clear adhesive covering to keep from dripping, particularly into the patient’s eyes (Fig. 71-1). Early application of LMX-4 cream will yield the best patient satisfaction during frame placement. Any oral sedation medications should be administered 30 minutes prior to frame placement. After directing family members to the appropriate waiting area, the patient is moved to a procedure room for application of the stereotactic frame. The room should contain monitoring capabilities (ECG, pulse oximetry, blood pressure), as well as access to oxygen. The room lighting should be adequate and room temperature comfortable for the patient. Environmental stimuli should be kept to a minimum. When the treating physician is present, a “time out” should occur, verifying correct patient, correct diagnosis, and appropriate site of treatment veriﬁcation.

Figure 71-1. Apply LMX-4 cream at least 1 hour prior to frame placement.

While sitting erect on a gurney the stereotactic frame is suspended on the patient’s head by use of a Velcro strap (Fig. 71-2). One may opt to use the magnetic resonance (MR) indicator box during frame ﬁtting to ensure a comfortable, accurate ﬁt during imaging (Fig. 71-3). If conscious sedation is to be used, it should be administered just prior to frame application with strict adherence to policy surrounding this competency. Anesthesia staff may be utilized to administer deeper sedation for appropriate individuals. Before the ﬁxation screws are advanced, local anesthetic is injected. A small (25-gauge) needle and 10mL syringe is ﬁlled with lidocaine HCL 1% and epinephrine 1 : 1000,000. Two milliliters of 8.4% sodium bicarbonate is combined with 20 mL of the lidocaine solution to help decrease the “burning” sensation commonly associated with lidocaine. As a rule, 2 to 5 mL of anesthetic is injected at each site to achieve satisfactory comfort (Fig. 71-4). The majority of patients tolerate frame application with minor discomfort (Fig. 71-5). Preprocedure education is a key component. Describing what to anticipate in regard to sensations such as pressure during tightening of ﬁxation screws will assist in managing anxiety. Allow individuals to view photos of previous patients wearing the device, or coordinate patientto-patient contact with previous radiosurgery candidates. Time required for frame application rarely requires longer than 15 minutes.

Figure 71-2. (A) Apply Velcro head holder to frame to create a sling. (B) Sit frame with holder easily onto the head.

71.

patient care in stereotactic radiosurgery

703

Figure 71-3. Checking ﬁt of indicator box during frame placement.

Figure 71-5. Patient immediately after completion of frame placement.

Skull measurements will be conducted after the frame is securely in place (Fig. 71-6). The patient should be sitting erect with support behind the back allowing for easy access to the head. There is absolutely no discomfort associated with this process.

is encouraged in the claustrophobic individual. The MR ﬁducial should be ﬁlled with copper sulfate 0.25 L solution (Fig. 71-9). Air bubbles that collect between the plastic panels could make image acquisition impossible and may require a second trip to the imaging center and a delay in the treatment. The imaging ﬁducials are very fragile and costly and should be carefully stored and cared for by a designee of the radiosurgery team. Educate the patient regarding length of imaging and what sensations may be experienced. The patient requiring angiography will need explicit instructions speciﬁcally concerning postprocedure monitoring and care. Radiology staff also need to be educated about quick removal of the stereotactic equipment subsequent to a medical emergency. Before imaging is complete and prior to patient removal from the table, the radiation physicist should ascertain that the images have been completed correctly and that the transfer to the planning computer can be successfully achieved. Neuroradiology should be monitoring and reading all images prior to patient treatment. Carefully remove the patient from the imaging table. Great care is taken when ﬁtting or removing the

Radiographic Imaging After the frame is in place, the patient is moved via gurney to the appropriate imaging machine(s). Individuals will undergo MRI, CT, and/or angiography imaging (Fig. 71-7). Each of these machines requires use of individualized ﬁducial localization equipment. Prior screening regarding allergies, metal risk factors, and claustrophobia will expedite safe, efﬁcient imaging. Care is taken to place the appropriate ﬁducial box securely onto the frame seating all connections and to ensure the side clips are tightly fastened (Fig. 71-8). Providing the patient with comfort measures such as warm blankets and knee support makes this a more pleasant experience. Pressure points, particularly at the shoulders, should be properly padded to reduce risk of impaired skin integrity. Administration of conscious sedation

Figure 71-4. Neurosurgeon injects lidocaine.

Figure 71-6. The taking of skull measurements prior to treatment planning.

704

t.f. biggins

Figure 71-7. Patient within the head coil awaiting start of MRI.

Figure 71-8. Check that MRI ﬁducial is without air bubbles and correctly attached to frame prior to imaging.

ﬁducial in that sound is magniﬁed via the frame attached to the skull. After imaging, the patient is transported to the stereotactic center to await the treatment.

logic assessment, and the reinforcement of education is completed at this time. Assess and treat pain prior to the initiation of treatment.

Treatment Planning

The Stereotactic Treatment

Treatment planning times vary according to the complexity of the case. Typically, the patient will be waiting 1 to 2 hours before treatment begins. Allowing family to visit during this time promotes a decrease in patient anxiety. Providing a light snack is appropriate if conscious sedation is not expected during the treatment phase. Ongoing monitoring of vital signs, neuro-

Comfortable positioning during the treatment phase can be achieved by using a combination of pillows and blankets to provide support. The supine position is typically used for treatment so elevating the knees will support the lower back. Shoulder and head support may be used as long as it does not interfere with machine operation. Meet individual patient needs as nec-

Figure 71-9. (A) A spinal needle, syringe, and copper is gathered. (B) Extracting the sulfate. (C) Spinal needle is narrow and makes for easy ﬁlling of ﬁducial.

71.

patient care in stereotactic radiosurgery

705

TABLE 71-3. Treatment room supplies. • • • • • • • • • •

Oxygen set-up including ambu bag Suction set-up Crash cart Emergency patient removal equipment supplied by manufacturer Suture material/suture tray for uncontrolled bleeding from pin sites Pillows, blankets, etc., for patient support, comfort Gauge, Band-Aids, and all supplies for frame removal IV supplies, gloves Bedpans, urinals, emesis basins, tissues ECG, BP, oxygen monitoring

essary. ECG/BP/oxygen monitoring should be used as appropriate. The availability of suction equipment will help ensure patient safety. Assess all supplies needed the day prior to treatment (Table 71-3). Educate the patient prior to each step. Ensure that close contact is maintained via audio and video monitoring. Describe the number of exposures to anticipate and the amount of time each exposure will last. The use of two-way microphones during treatment will assist in warning when movement will occur. Most patients choose to rest or sleep during treatment, and the playing of music into the room has proved to be very relaxing. Keep the patient and family informed of treatment progression. During helmet or equipment changes, allow for bathroom breaks or change of position. Total treatment time can vary greatly depending on diagnosis.

Figure 71-11. Fixation screw sites after the manipulation of skin.

Immediately upon completion of the procedure, the stereotactic frame should be removed. Light massaging with manipulation of the pin sites will help close the skin and decrease scarring (Figs. 71-10 and 71-11). Antibiotic ointment and some sort of dressing may be applied to the pin sites (Fig. 71-12). Typically, patients experience discomfort at the pin sites within 15 minutes of frame removal. Encourage pain medications either for prophylaxis or at initial onset of discomfort. Have antiemetics on hand for the occasional nausea. The convalescence period varies depending on patient general condition and length of

procedure. The patient requiring cerebral angiography prior to treatment may require a slightly longer recovery time. Characteristically, patients who have undergone stereotactic radiosurgery are prepared to return home within an hour or two of procedure completion. Seizure-prone individuals may require an overnight stay. Remind families that driving is not recommended the day of treatment and prohibited if any conscious sedation has been administered. Encourage the consumption of food and drink prior to discharge. Follow-up instructions should be carefully reviewed with the patient and family. If follow-up appointment times are not assigned on discharge, provide appropriate contact individuals and phone numbers. Ensure prescriptions are made available as needed, although over-thecounter pain relievers almost always are sufﬁcient for any discomfort related to the procedure. Patients taking anticonvulsant therapy should be reminded of the importance of frequent monitoring of serum levels of these drugs. Ensure that patients who are taking steroid therapy understand which physician ofﬁce they are to contact regarding any concerns with these drugs. Check to ensure that a gastrointestinal support drug has been prescribed for anyone on steroid therapy. Stress the

Figure 71-10. Fixation removal.

Figure 71-12. Band-Aids are commonly used to cover pin sites overnight.

Postoperative Care

screw

sites

immediately

after

frame

706

t.f. biggins

importance of providing radiology imaging (MRI, CT, etc.) at the time of follow-up with the radiosurgery team. Assist with referrals so imaging can be completed in a timely manner prior to the visit. Provide the imaging centers with physician phone and fax numbers.

Home Follow-up Phone contact should be made within 24 hours of discharge. The RN is the best individual to complete this call. Reafﬁrm preoperative teaching as necessary. If home care or therapy (nursing, PT, OT, etc.) was established at the time of the stereotactic treatment, assess that these referrals have been completed and the patient has been contacted by the appropriate persons. Assess pain status and make recommendations as appropriate. Assist with scheduling follow-up visits as needed. The follow-up phone call is the ideal time to assess patient satisfaction regarding the entire stereotactic process.

Care of the Pediatric Patient A small percentage of all stereotactic candidates fall under the category of the pediatric population. Typically this group is categorized as the following: Newborn: birth to 3 months Infant: 3 months to 1 year Toddler: 1 year to 4 years Preschool: 4 years to 5 years School age: 6 years to 12 years Adolescent: 13 years to 18 years Brain tumors and arteriovenous malformations are some of the conditions in this population that can be treated via radiosurgery. The central nervous system continues to develop throughout childhood, therefore careful consideration is given to the appropriateness of the delivery of radiation. Consultations are scheduled with the patient and parents involving all the pertinent members of the radiosurgery team. Risks and beneﬁts of the procedure are discussed at length. Once a decision has been made to proceed with radiosurgery, the nursing staff should make a thorough assessment of the patient and family needs. Preoperative blood work is generally not necessary in the young pediatric patient. The application of LMX-4 cream is recommended before any needle sticks, such as an intravenous (IV) start. Whenever possible, have the child in an ageappropriate environment. The use of pediatric specialty pre-and postoperative suites will help make transitions more ﬂuent. Classically, all children below the age of 13 years require general anesthesia. Most adolescent children are able to tolerate radiosurgery with some light to moderate sedation. Nurses can help determine the maturity level of the adolescent patient through patient and parent interview. Pediatric anesthesia physicians can help determine the optimal sedation plan for the individual child. Clearly, the younger child will require parent involvement until anesthesia has occurred. Encourage the young child to take a favorite tool or stuffed animal and keep it with the child until they are anesthetized. Allow family members to be with a child of any age during appropriate intervals. As with

any patient, provide frequent updates to parents throughout the procedure to decrease anxiety. The identical stereotactic frame that is used for an adult can be used for the child. Certainly, the longer ﬁxation screws will need to be used to compensate for a smaller head circumference. The use of lidocaine at the pin sites may or may not be used for the anesthetized patient. Supporting the frame during application of a patient under general anesthesia can be difﬁcult and may require two people, one to support the frame and the second to support the head. In that the child may be immobile for several hours during planning and treatment, monitor pressure points on the skin, and a Foley catheter may be practical. Monitor body temperature and keep warm blankets near. During the delivery of radiation, the anesthesiologist will need to have visualization of key components. Certainly the cardiac, BP, and oxygen monitor should be in plain view, and the ability to change monitoring parameters accessible. If possible, two cameras should be available in the treatment room. One camera should be focused on the patient for visualization of respirations and the other focused on the ventilator, speciﬁcally the CO2 waveform. Prior to the start of treatment, evaluate the length of all tubing and account for movement into the machine. Ensuring that these features are in place will promote a safe atmosphere. Offering the adolescent music of their choosing can decrease anxiety during treatment. Removal of the stereotactic frame prior to reversal of anesthesia is recommended. Moving the patient to a pediatric PACU (Post-anesthesia care unit) setting will allow for safe recovery. Allow family to visit in the PACU as soon as possible. Observe the very young child for signs of discomfort speciﬁcally immediately after frame removal and administer analgesics as needed. Encourage the older child to request pain medications as well. Admit to age-appropriate setting until discharge. Educate parents and the child regarding pain management, pin site care, and follow-up appointments. Provide phone numbers including emergency contact. After discharge, it is recommended that nursing staff make phone contact with families within 24 hours.

Tricks of the Trade (Table 71-4) Recognized colleagues have all learned the radiosurgery process through some trial and error. Beginning with consultation, preoperative analysis, frame ﬁtting, gathering of equipment, proper imaging techniques, much can be learned from the established center.

Conclusion Care of the radiosurgery patient can initially be challenging for the clinician. Proper training, communication with established centers and clinical experts, as well as hands-on experience will support a thriving center. A support group for radiosurgery nurses allows for the constant ﬂow of ideas and solutions to challenges that only this population can appreciate. A strong quality assurance program with ongoing patient satisfaction evaluation is crucial. Establishing concise guidelines beginning with the intake process and ﬂowing through the follow-up patient visit will assist in making the experience gratifying for patients and clinicians.

71.

patient care in stereotactic radiosurgery

707

TABLE 71-4. Tricks of the trade. Preprocedure • Arrange initial consultations with all physicians/nurse for the same day. • When scheduling patients for the procedure, attempt to schedule the simplest treatment for the ﬁrst case of the day progressing to the most complex treatment for the last patient of the day. Frame placement • Use Velcro strap instead of ear bars. The strap can be positioned to ﬁt each patient individually. This increases comfort for patient and individual supporting the frame. • Use LMX-4 cream on forehead prior to lidocaine injection. Strive for application at least 1 hour prior to frame placement. • Add sodium bicarbonate to lidocaine to decrease “sting” at injection site. • If using CT scan, place ﬁxation screws as far from the targeted area as possible to decrease artifact. • Always have patient ﬁlms available for frame placement. • Use “curved’ front posts for extreme lateral or very low (cerebellar) lesions. • Place front posts low on forehead to decrease risk of collision with treatment helmet. • Test all ﬁducials during frame placement. This will avoid the need to adjust the frame later in the process. • Know your patient’s surgical history, speciﬁcally past craniotomies, shunt placements, etc. Fixation screws need to be carefully placed to avoid these areas. • Place ﬁxation screws as “ﬂush” as possible with the post to avoid collision. • While on a gurney, prop a pillow(s) behind patient’s back to make access to the back of the head adequate for frame placement. • Follow correct site protocol by marking side of treatment involving unilateral site, such as trigeminal neuralgia. Ask the patient to verbalize side of pain prior to premedication. Radiographic imaging • Place earplugs in patient prior to MRI. • Elevate knees during scans. • Provide padding anywhere pressure is a concern due to the frame or ﬁducial cradle touching shoulders, etc. (foam pads, 4 × 4s). • Do not dismiss a patient’s complaint of heating at a ﬁxation screw site. Immediately check the screw for any magnetic components and the patient’s skin for any burning. • Attempt to ensure that the imaging is acceptable to all involved (radiation physicist, neurosurgeon, radiologist) before leaving the imaging area. • Check that MRI ﬁducial is ﬁlled with solution and free of air bubbles the day before needed. Treatment • Suggest music be played while undergoing treatment. • Offer support for knees. • Make sure the mattress is at an acceptable height to ensure patient comfort throughout the entire procedure. • If plug patterns will be needed on different helmets, roll the subsequent helmet out of the room and prepare plug pattern while the patient is being treated. Frame removal • Massage or pinch screw sites closed immediately after frame removal to assist in the quick closure of screw openings and to decrease scarring. Patient-speciﬁc issues • CT cisternogram can be successfully used to visualize the trigeminal nerve when MRI is contraindicated. • Anticoagulant therapy will need to be stopped as per physician recommendation when angiogram or cisternogram are to be performed. • Use hardware that will cause the least amount of artifact when using CT imaging (i.e., carbon-ﬁlled posts, nonmetallic ﬁxation screws).

References 1. Mughaw SB. An overview of methods in stereotactic radiosurgery. Am Soc Radiologic Technol 1992; 63:402–405. 2. Browner CM, Hendrickson K. A nursing perspective of Gamma Knife treatment. Barrow Neurological Institute: Quarterly 1997; 13:41–48.

3. Lindquist C, Kihlstrom L, Hellstrand E. Functional neurosurgery-a future for the Gamma Knife? Funct Radiosurg 1991; 57:72–88. 4. A tribute to Dr. Lars Leksell. Another Perspective–The Publication for the International Radiosurgery Support Association. 1998:2–3.

Index A N-Acetylaspartate, as tumor marker, 17 Acquired immunodeﬁciency syndrome patients, cavernous sinus lymphoma in, 242 Acromegaly, 116, 142, 299, 302, 304, 309, 322–323, 329 Adenocarcinoma esophageal, 184 pancreatic, 176–177 skull base, 121 thyroid, 421 Adenoid cystic carcinoma, 418, 421 Adrenocorticotropic hormone deﬁciency, 348 Adrenocorticotropic hormone-secreting adenoma, 116–117, 301, 311–312, 323–324, 328 Akhenaten, Pharoah, 299 Alcohol use, effect on essential tremor, 561 Alopecia, radiosurgery-related, 650 Amantadine, as Parkinson disease treatment, 542 American Association of Neurological Surgeons, 6, 649, 688 American Association of Physicists in Medicine, 46–47 American College of Radiology, 101, 455 American Society for Therapeutic Radiology and Oncology, 455, 611, 649, 688 Amygdalohippocampectomy, 584–585 Anesthesia for halo ring application, 696–697 pediatric, 332 Aneurysms arteriovenous malformations-related, 461, 463, 481 pineal, 377 of vein of Galen, 501 Angioﬁbroma, 383–384 Angiography, 15–17, 703 of arteriovenous malformations, 20, 136 of cavernous malformations, 492–493 computed tomography, 15, 16, 17 magnetic resonance, 15, 16, 17 planar, 71, 72 use with Gamma Knife radiosurgery, 109, 111 Anticholinergic medications, 542 Anticonvulsant/antiepileptic medications, 223, 519, 535–537, 583, 651 Anxiety, 566, 568–569 Apparent diffusion coefﬁciency (ADC), 14, 15 Arachnoid cap cells, 258

Aristotle, 163 Arteriovenous malformations aneurysms associated with, 461, 463, 481 of basal ganglia, 459, 461, 463, 464 brain-stem, 459, 461, 464 of cerebellum, 459 deep-brain location of, 475 deﬁnition of, 15, 459, 473, 479 embolization (neurovascular) therapy for, 459, 464–467, 479–489 aneurysms associated with, 481 complications of, 481, 485, 486, 487 cure rate in, 483–484 nidal volume in, 481 as palliative therapy, 484 partial, 484 preoperative/preradiation, 465, 466–467, 480–481 as targeted therapy, 481 epileptogenic, 461, 585–586 fractionated radiation therapy for, 459, 465, 466 grading/staging systems for, 484–485 Pollock-Flickinger score, 461, 468 radiosurgery-based, 467–468 Spetzler-Martin, 459, 460, 461, 467, 473–474, 479–480 as headache cause, 459 hemorrhage associated with, 15, 16, 112, 152, 459, 460, 473, 474, 476, 479 aneurysm-related, 481 effect of partial resection on, 480 in pediatric patients, 459 post-radiosurgery, 482–483 re-hemorrhage risk, 479 risk factors for, 459 imaging of, 15–17, 20, 72, 473 incidence of, 459 movement disorders associated with, 544–545 observation of, 459, 460 in pediatric patients, 336–338, 655–656 as seizure cause, 459, 461 spinal-cord, 176 staged-volume radiosurgery for, 465, 467 stereotactic radiosurgery for, 15, 459–469, 474–475 comparison with surgical treatment, 468–469, 666–669 complications of, 55, 462, 463–465, 482–483, 655–656 cost-effectiveness of, 666–669 CyberKnife, 81, 174 disadvantages of, 475 dose planning in, 461–462

efﬁcacy of, 20 Gamma Knife, 108, 111, 112, 113, 337–338, 497–498, 544–545 historical perspective on, 459–460 isodose line in, 464 K index in, 462, 467, 468 of large malformations, 465–467 linear accelerated-based, 130, 136–138, 460, 465 linear accelerator-based, 460 multisession, 656 as neurological deﬁcit cause, 17 obliteration prediction index (OPI) in, 462 patient selection for, 460–461 proton beam, 87, 142, 152 quality-of-life outcomes, 669–670 radiobiological considerations in, 54–55 repeat, 464–465 for small-volume lesions, 482, 483 surgical treatment for, 15, 459, 460, 473–478 complications of, 460 cost-effectiveness of, 476 multimodality, 475–477 open surgery, 473–475 thalamic, 459 treatment strategy formulation for, 479–480 Astrocytes, 355 Astrocytoma anaplastic, 207, 225, 226, 362, 381 brain-stem, 367, 377–378 Gamma Knife radiosurgery for, 118 grading of, 378 high-grade, 207. See also Astrocytoma, anaplastic in pediatric patients, 333 in pediatric patients, 333, 334 pilocytic, 346 radiation therapy for, 341, 346, 347 pilocytic, 367 juvenile, 346 pineal, 371, 377–378 pituitary, 300, 301 spinal-cord, 377–378 Atlas d’Anatomie Stereotaxique (Talairach), 26 Automated Robotic System for Optimal Position (AESOP), 164 B Baclofen, as trigeminal neuralgia treatment, 519, 537–538

709

710 Basal ganglia arteriovenous malformations of, 459, 461, 463, 464 cavernous malformations of, 504, 509 Basophilism, pituitary, 299 Bile duct tumors, brachytherapy for, 645 Biopsy of brain tumors, 11 of cavernous sinus tumors, 242, 243 of pineal tumors, 355, 357, 365–366, 368–369, 372, 377 endoscopic, 357, 359, 363 open, 357 spinal, 447 Bleomycin, as pineal tumor treatment, 372–373, 378–379 Blood-brain barrier, as obstacle to drug delivery, 223, 333 Blood volume, cerebral, 13, 14, 18 Brachytherapy comparison with stereotactic body radiation therapy, 643–646 cost-effectiveness of, 646 endoluminal, 644–645 historical perspective on, 643 interstitial, for high-grade gliomas, 209 for liver cancer, 645 for nasopharyngeal carcinoma, 412, 413, 418 patient selection for, 645 in pediatric patients, 335 selective internal radiation therapy (SIRT) technique in, 645 for skull base tumors, 388 Bragg, William Henry, 141 Bragg peak, 28–29, 45, 46, 141, 142, 149 native, 142, 143 “pristine,” 85? spread-out, 142, 143, 144, 145 Brain, radiosensitivity of, 183, 564 Brain metastases incidence of, 181 local control of, 195–196 multiple, 181 in pediatric patients, 332 prognosis for, 181 recurrent, 184–187, 195–196 single, 181 stereotactic radiosurgery for, 181, 183–185, 186–189 comparison with surgical resection, 665–666 complications of, 188, 196–197, 650–651, 656 cost-effectiveness of, 665–666 Gamma Knife, 118, 119 linear accelerated-based, 134–135 quality-of-life outcomes in, 668–669 for recurrent metastases, 184–185, 186 as whole-brain radiation therapy adjunct, 183 surgical treatment for, 181–183, 193–199 comparison with stereotactic radiosurgery, 184, 193–199 comparison with whole-brain radiation therapy, 182–183 complications of, 196

index whole-brain radiation therapy for, 181–183, 187–188, 193, 194–195, 197, 198, 201–206 complications of, 202, 203, 205 disadvantages of, 187 dose and scheduling of, 201–202 efﬁcacy of, 201, 203–204 for recurrent metastases, 184–187, 185–187 Brain stem arteriovenous malformations of, 459, 461, 464 astrocytmas of, 367, 377–378 cavernous malformations of, 504, 509–510 gliomas of, 333, 346 radiosensitivity of, 654 Brain tumors benign, 55, 131–134, 333 fractionated stereotactic radiotherapy for, 54, 55, 215–216, 342, 343 Gamma Knife radiosurgery for, 108, 114–117 malignant, 333 linear accelerator-based radiosurgery for, 134–136 metastatic. See Brain metastases in pediatric patients, 332–336, 341–349, 345–347 chemotherapy for, 351–354 fractionated stereotactic radiotherapy for, 342–343 high-grade, 335–336 incidence of, 332 low-grade, 333–335 radiation therapy for, 341–342 stereotactic radiosurgery for, 342–343 pharmacotherapy for, obstacles to, 223 recurrent, 215 Breast cancer chemotherapy for, 616 meningioma as risk factor for, 257 as meningioma risk factor, 249, 257, 272 metastatic to the brain, 181, 650–651 ocular, 605 to the pituitary gland, 306 to the spine, 438, 439–440, 443, 445 radiation-related, 249 Bremsstrahlung, 35–36 Brown-Roberts-Wells stereotactic device, 26 Bruch membrane, 597 C Cabergoline, 329 Calciﬁcation of cavernous malformations, 491 pineal, 357 Callosotomy, 578, 583, 587, 588 Cancer. See also speciﬁc types of cancer radiation-related, 306, 650 Capillary blush, 493 Capsaicin, as trigeminal neuralgia treatment, 538 Capsulotomy, 564, 565–566

Carbamazepine side effects of, 536 as trigeminal neuralgia treatment, 519, 535–536, 657–658 Carbidopa, as Parkinson disease treatment, 559, 560 Carboplatin as low-grade glioma treatment, 352 as pineal tumor treatment, 378–379, 380 Carmustine, 214, 224 side effects of, 227 Carotid arteries, in pituitary adenoma surgery, 309, 312 Carotid body tumors, 406, 421 Cataracts, ophthalmic radiosurgery-related, 601 Catechol-O-methyltransferase (COMT) inhibitors, 542, 560 Caudatotomy, 123–124, 542–543 Cavernous malformations, intracranial, 491–499 anatomic distribution of, 504 of basal ganglia, 504, 509 brain-stem, 504, 509–510 calciﬁcation of, 491 clinical presentation of, 492, 505 deﬁnition of, 491, 503 developmental venous anomalies associated with, 492, 493 epidemiology of, 491–492, 503–504 epileptogenic, 585–586 estrogen receptors in, 503 familial form of, 491, 492, 503–504, 505 Gamma Knife radiosurgery for, 112–113, 509 heavy-charged particle radiosurgery for, 509 hemorrhage associated with, 112–113, 491, 492, 503, 504, 505, 513–515 clinical presentation of, 505 postsurgical, 495, 496, 497, 498–499, 509 imaging of, 492–493, 505–507 incidence of, 503 infratentorial, 504, 505, 508–509 microsurgical resection of, 507–509 natural history of, 493, 504–505, 513–514 observation of, 493 pathophysiology of, 491 radiologic features of, 492–493 as seizure cause, 503, 504, 505 stereotactic radiosurgery for, 493, 495–499, 509, 514–515 dose planning techniques in, 495, 497 Gamma Knife, 495, 496, 497 linear accelerator-based, 496, 497 proton beam, 495, 496 radiobiological considerations in, 497–499 results of, 495–499 supratentorial, 507–508 surgical resection of, 493, 494 in pediatric patients, 337 Cavernous sinus, dural arteriovenus ﬁstulas of, 499–501 Cavernous sinus tumors chemodectomas, 405

index hemangiopericytomas, 133–134 lymphoma, in AIDS patients, 242 meningiomas, 133, 134, 236, 238, 239, 652 needle biopsy of, 242, 243 in pediatric patients, 335 pituitary tumor-related, 115–116, 311, 312, 313–314, 3152 proton beam radiosurgery for, 87 Central nervous system tumors, pediatric, 332 Cerebellar tumors, imaging of, 10 Cerebrospinal ﬂuid analysis of malignant germinomas, 357 of pineal tumors, 374, 377 Certiﬁcate of Need (CON) procedures, 675 Cetuximab, 426–427 Charged-particle radiation therapy, 28–29, 167, 403. See also Proton beam stereotactic radiosurgery Charged-particle radiation therapy, 28–29, 167. See also Proton beam stereotactic radiosurgery Chemodectoma, 422, 423. See also Glomus tumors general features of, 405 location of, 401 pathology of, 405 treatment for, 405–408 Chemotherapy. See also under speciﬁc types of cancer antitumor activity measurement of, 224 as stereotactic body radiation therapy alternative, 616 Children. See Pediatric patients Chlorambucil, as pineal tumor treatment, 372–373 N-(2-Chloroethyl)-N¢-cyclohexyl-Nnitrosourea (CCNU), 372–373, 380 1,3-bis-(2-Chloroethyl)-1-nitrosourea (BCNU) as anaplastic astrocytoma treatment, 226 as glioma treatment, 214–215, 217, 224 as pineal tumor treatment, 372–373, 381 Cholangiocarcinoma, 615, 627 Cholesteatoma, 386 Choline, as tumor marker, 13, 18 Chondroma, 387–388, 389 radiosurgery for, 403 surgical resection of, 403 Chondrosarcoma, 121, 387–388, 389, 448 fractionated stereotactic radiosurgery for, 297, 403–405 general features of, 402 local control of, 383 postoperative/adjunctive radiotherapy for, 403 proton beam radiosurgery for, 148, 149, 150, 153 skull base, 149, 150, 396, 397 stereotactic radiosurgery for, 403–405 surgical treatment for, 396, 397, 402–403 Chordoma, 121, 387–388, 389, 418, 448 clival, 296–297, 309 general features of, 401–402 of the head and neck, 418 histologic subcategories of, 401

intracranial, 401–402 local control of, 383 pathology of, 401, 402 pituitary, 305 postoperative/adjunctive radiotherapy for, 403 sacral, 401 skull base, 149, 150, 395–396, 401–405 spinal, 401 stereotactic radiosurgery for, 403–405 fractionated, 403–405 proton beam, 148, 153 surgical treatment for, 395–396, 402–403 Choriocarcinoma, 305, 356, 371, 377 Choristoma, 305 Choroid plexus tumors, 341, 347 Circadian rhythms, role of pineal gland in, 355 Cisplatin, as pineal tumor treatment, 372– 373, 380 Clivus, chemodectomas of, 405 Clonozepam, as trigeminal neuralgia treatment, 538 Cobalt-60-based stereotactic radiosurgery, 25, 33, 34, 41–42, 331, 597, 649, 689. See also Gamma Knife stereotactic radiosurgery Cobalt Gray equivalents (CGEs), 83, 85 Cochlear nerve neuroﬁbromatosis of, 115 proximity to acoustic neuromas, 292, 293, 297 Coding, for stereotactic radiosurgery procedures, 675–676, 679, 689 Colon/colorectal cancer, 181, 615, 616 Common law, 683–684 Complex regional pain syndrome, 535 Compton scattering, 36, 37, 38 Computed tomography (CT), 69, 72, 649, 703 of cerebellar masses, 10 in combination with with Gamma Knife radiosurgery, 109, 110, 111, 115 with linear accelerator-based radiosurgery, 131 magnetic resonance imaging, 11, 12, 110 cone beam, 165, 166, 621–622 development of, 9 ﬁducial systems in, 71 isodose lines in, 9 limitations to, 9 Computed tomography (CT) angiography, 15, 16, 17 Computer-assisted surgery (CAS), 164, 169 Cone beam tomography, 165, 166, 621–622 Conformity index, 75 Congress of Neurological Surgeons, 649, 688, 689 Continuous hyperfractionated accelerated radiation therapy (CHART), 636 Contrast media administration of, 11–12 post-treatment uptake of, 18 Corpus callosotomy, 583, 587, 588 Corpus callosum, cavernous malformations of, 504

711 Corticosteroids, 18, 223, 448 effect on magnetic resonance imaging ﬁndings, 18 Cosman-Roberts-Wells frame, 26 Cost, of stereotactic radiosurgery units, 101 Cost-effectiveness, of stereotactic radiosurgery, 663–671 quality-of-life outcome-based, 663, 667–670 CPT (Current Procedural Terminology) coding, 675–676, 679, 689 Cranial fossa tumors, middle, 386–387 Cranial nerves in cavernous malformations, 504 external beam radiation therapy-related deﬁcits in, 265 head and neck cancer, 423 in meningiomas, 249 paralysis of, 239 acoustic neuroma treatment-related, 279 radiosensitivity of, 17, 652 in schwannomas, 120 stereotactic radiosurgery-related deﬁcits in, 262, 652 V, in trigeminal neuralgia, 19–20 VII, in fractionated stereotactic radiosurgery, 293 VIII, in fractionated stereotactic radiosurgery, 292, 293 Craniopharyngioma, 120, 311–312, 334, 386–387 chemotherapy for, 352–353 clinical presentation of, 318 diagnosis of, 327 histopathologic categories of, 318 incidence of, 300, 301 pathophysiology of, 301 in pediatric patients, 318, 341, 342, 346, 347, 352–353 radiotherapy for, 324 sellar, 309 stereotactic radiosurgery for CyberKnife, 175 GammaKnife, 305 surgical treatment for, 295–296, 320 Craniotomy Gamma Knife-based, 665 for meningioma, 254 for pineal region tumors, 366 for pituitary tumors, 20, 300 pterional, 252, 253 retromastoid, 519 transsphenoidal, 299 Creatine, as tumor marker, 17, 18 Credentialing, of neurosurgeons, 695 Current Procedural Terminology (CPT) coding, 675, 679, 689 Cushing, Harvey, 249, 299, 300 Cushing’s disease, 116–117, 323–324 Cushing’s syndrome, 299–300, 328 CyberKnife stereotactic radiosurgery, 80–83, 167, 171–178, 433–434 accuracy of, 171, 173 for arteriovenous malformations, 176 for body cavity targets, 173–174 components and functions of, 44–45, 91, 94–96, 167, 171, 172

712 CyberKnife stereotactic radiosurgery (Cont.) deﬁnition of, 80 ﬁducial-free tracking system of, 173 frameless targeting system of, 171 future developments in, 177 for head and neck cancer, 423 history and development of, 5, 44 imaging in, 81 installation of, 45, 434 for intracranial lesions, 174–175 for intrathoracic and intraabdominal lesions, 176–177 multileaf collimators of, 177 for pancreatic cancer, 627–628 procedure in, 30, 171, 172, 173 radiation delivery in, 434 robotic components of, 167, 689 for spinal tumors and metastases, 175–176, 432, 433–441, 458 procedure, 434–436 results, 436–438 target localization in, 29–30 treatment planning in, 5, 30, 81, 82–83, 174 for trigeminal neuralgia, 530 Cybernetics, 163 Cyclooxygenase-2 inhibitors, 272 Cyclophosphamide, as pineal tumor treatment, 372–373, 379, 380 Cyclotrons, 4, 33, 34, 45–46, 141, 142 Cysts arteriovenous malformation treatmentrelated, 658 hemangioblastoma-related, 121, 516 pineal, 371, 377 Rathe cleft, 305, 309 D Dactinomycin, as pineal tumor treatment, 372–373 Decompression microvascular, for trigeminal neuralgia, 519, 523, 524, 529–533, 567 of the spinal cord, 447, 448–450 Deep-brain stimulation therapy, 549 comparison with radiosurgical ablation, 549–558 for essential tremor, 561 as hemorrhage cause, 550, 553, 554 for movement disorders, 541–548, 559–561 pallidal, 552–553 for Parkinson disease, 541, 543–544, 549–550, 552–553, 554–555, 560 for psychiatric and pain disorders, 566, 567 subthalamic, 554–555 thalamic, 549–550 Deﬁbrillators, cardiac, 521, 522 Dementia, whole-body radiation therapyrelated, 202 Depression, 566, 569 Dermoid tumors, 305, 357 Diabetes insipidus, 306 Diagnostic Related Grouping (DRG), 675, 676 Diﬂuoromethylornithine (DFMO), as anaplastic astrocytoma treatment, 226 Dopamine receptor agonists, 117, 542, 560

index Dose and dosimetry, in stereotactic radiosurgery, 37–41, 46 calculations in, 73–74 measurement-based, 38–39 model-based, 39–40 multifractionated, 56–58 dose-volume histograms (DVHs), 41, 72, 74–75, 77, 146–147 in Gamma Knife radiosurgery, 46 isodose line, 40–41 in linear accelerator-based radiosurgery, 46 in ophthalmic stereotactic radiosurgery, 595–596 output, 39 pencil-beam kernel, 40 percent depth dose, 38–39 point kernel, 40 proﬁle or off-axis ratio, 39 in proton beam radiosurgery, 83, 85 in psychiatric and pain disorders, 567 quality assurance guidelines for, 46–47 radiobiological principles underlying, 54 relationship to lesion size, 17 treatment planning and, 73 Dose-volume histograms (DVHs), 41, 72, 74–75, 77, 146–147 Doxorubicin, 272 Doxtil, 272 Dynamic conformational arcs, 80 Dyskinesia, Parkinson disease-related, 543–544, 559–560 Dystonia, pallidal deep brain stimulation treatment of, 553 E Edema cerebral, 656 in arteriovenous malformation patients, 464, 465 brain-stem, 655 external beam radiation therapy-related, 263, 265 parasellar meningioma-related, 296 radiosurgery-related, 263, 464, 650, 651, 652–653 pulmonary, 655 Elderly patients acoustic neuromas in, 283, 285–286 meningiomas in, 242–243, 651 stereotactic radiosurgery in, 649 telozolomide use in, 227 Electroencephalography, for epilepsy evaluation, 583 Embolization therapy for arteriovenous malformations, 473, 475–477, 479–489 complications of, 481, 485, 486, 487 cure rate, 483–484 nidal volume in, 481 as palliative therapy, 484 partial, 484 preoperative, 480–483 as targeted therapy, 481 for dural arteriovenous ﬁstulas, 499, 500, 501, 515–516

for meningiomas, 271 for vein of Galen malformations, 501 Embryonal cell carcinoma, 344, 345, 356, 371, 377 Enchondroma, 386 Endodermal sinus tumors, 377 Ependymoblastoma, 335 Ependymoma, 332, 333, 335, 336, 341, 342 chemotherapy for, 351–352 pineal, 367 radiation therapy for, 346, 347–348 stereotactic radiosurgery for, 351–352 Epidermal growth factor receptor, 636 Epidermal growth factor receptor inhibitors, 272, 426–427 Epilepsy. See also Seizures arteriovenous malformation-related, 475 deﬁnition of, 583 hypothalamic hamartoma-related, 573, 574–575, 578, 586, 588–589 medication-resistant, 583 mesial temporal lobe, 575–576, 577, 578, 583–585, 588, 589 pharmacotherapy for, 583 stereotactic radiosurgery for, 573–581, 588–589 expermental models of, 64–65 surgical treatment for, 583–591 corpus callosotomy, 583, 587, 588 focal resections, 583–587 lesionectomy, 585–586 multiple subpial transection, 583, 587 vagus nerve stimulation, 583, 587–588 Esophageal cancer, 157, 184 Esthesioneuroblastoma, 384–385, 397 Estrogen receptors, 257, 503 Etoposide, 352, 372–373, 378–379 External beam radiation therapy (EBRT) for angioﬁbromas, 384 for germinomas, 361 for high-grade gliomas, 210, 212–214 for meningiomas, 259, 261, 262 for atypical meningioma, 265–266 comparison with stereotactic radiosurgery, 266–267 complications of, 265, 267 dural tail targeting in, 264–265 postoperative, 264 primary, 263–264 radiation dose in, 264 for recurrent meningioma, 266 target volume in, 264–265 for nasopharyngeal carcinoma, 412, 415–416 for pineal region tumors, 375 as pneumonitis cause, 643 for uveal melanoma, 597 Extratemporal resection, 586 Eye tumors. See Ocular and orbital lesions F Facial nerve function, effect of acoustic neuroma surgery on, 19, 114, 115, 283, 284 Facial pain. See also Trigeminal neuralgia atypical, 524–525

index Fentanyl patch, as trigeminal neuralgia treatment, 538 α-Fetoprotein, as pineal region tumor marker, 365, 371, 377, 379 Fiducial systems, 71, 72, 687, 688, 703–704 Fistulas dural arteriovenous, 499–501, 515–516 esophageal, 644 510K Premarket Approval, 674–674 Fixation frames/systems, stereotactic, 25, 26, 70–71, 687, 688 complications of, 649 discomfort associated with, 697–698 Gill-Thomas-Cosman (CTG) frames, 5, 294 for ophthalmic stereotactic radiosurgery, 593–594 placement of, 702–703 on pediatric patients, 332 removal of, 705 Fixed ﬁelds, 80 Flexner-Wintersteiner rosettes, 356, 371–372, 377 Fluoroscopy, 165 Foramen magnum tumors, 174 Fotemustine, 224 Fractionated stereotactic radiosurgery frameless, 71 for head and neck cancer, 422 for high-grade gliomas, 210–211, 215–217, 218 history of, 5 hypofractionated, 217 for nasopharyngeal carcinoma, 416–417 for recurrent gliomas, 218 for spinal metastases, 455–458 Fractionated stereotactic radiotherapy for acoustic neuromas, 289, 290–294, 291 for arteriovenous malformations, 459, 465, 466 for brain tumors, 54, 55, 215–216, 342, 343 conventional, 295 differentiated from radiosurgery, 649 for head and neck cancer, 427–428 for intracranial meningiomas, 257–270 for lung cancer, 635, 636–637 as memory loss cause, 265 multifactionated dose calculation in, 56–58 for optic nerve sheath meningiomas, 289–290, 294–295 for pineal region tumors, 375–376 for pituitary tumors/pituitary region tumors, 317–326 for craniopharyngiomas, 318 diagnostic workup and staging in, 319 for meningiomas, 294–295, 317–318 for pituitary adenomas, 317, 318 radiobiological principles underlying, 51–60 for skull base tumors, 265 technique, 293–294 Fractionation, 687 Frames, ﬁxation. See Fixation frames/ systems, stereotactic Frontal bone, arteriovenous malformations of, 465, 466

Frontal fossa, meningiomas of, 240 Frontal lobe, cavernous malformations of, 504 Frontal lobe resection, 586 Functional brain disorders, Gamma Knife radiosurgery for, 121–124 G Gabapentin, 536, 561, 657–658 Gadolinium texaphyrin, 218 Gamma-aminobutyric acid (GABA)-related agents, 535, 537–538 Gamma Knife radiosurgery facilities, 83 Gamma Knife stereotactic radiosurgery, 4, 26–27, 107–127, 166–167 for acoustic neuromas, 114–115, 131, 277 comparison with fractionated stereotactic radiotherapy, 290, 291, 292, 293 complications of, 290, 291, 654 control rate in, 291 isodose prescription in, 291, 292 for age-related macular degeneration, 607 for angioﬁbromas, 384 for arteriovenous malformations, 15 complication rate in, 15 histopathologic ﬁndings after, 497–498 in pediatric patients, 337–338 radiation dose/obliteration relationship in, 462 for atypical facial pain, 524–525 automatic patient positioning system of, 107, 109, 111, 112 brain imaging techniques in, 110–111 for brain tumors, 333–336 for cavernous malformations, 495, 496, 497 for chemodectomas, 407, 408 for chondromas, 404 for chondrosarcomas, 404, 405 in combination with pallidotomy, 553, 555 subthalamotomy, 555 thalamotomy, 551–552, 555 comparison with linear accelerator-based stereotactic radiosurgery, 69, 70, 663 complications of, 650 components and functions of, 27, 41, 43, 91–92, 97, 99, 100, 101 conformational dose planning in, 109, 111 cost-effectiveness of, 663 cost of, 692 for craniopharyngiomas, 386–387 development and history of, 3, 25, 26, 41, 51, 75, 107–109, 129, 563, 691, 699 dose planning in, 27 dosimetric characteristics of, 46 for dural arteriovenous ﬁstulas, 499–500 for epilepsy, 573–581 hypothalamic hamartoma-related, 573, 574–575, 578 mesial temporal lobe, 575–576, 577, 578 patient selection for, 577–578 technical questions in, 576–577 for esthesioneuroblastomas, 385 ﬁducial markers in, 71, 72 ﬁxation frames in, 70

713 for functional brain disorders, 121–124 for germinomas, 360, 361–362 government approval for, 673–675 intraocular, 593–595 isodose calculations for, 73, 74 limitations to, 27, 167 long-term outcomes after, 112–121 for meningiomas, 237, 238, 240–241, 250, 251, 652 model A, 107 model B, 107 model C, 107–109, 111 for movement disorders, 541–548 pallidotomy, 546 in pediatric patients, 331–340 for pineal region tumors, 355, 375–376 for post-herpetic neuralgia, 524 preventive maintenance of, 110 procedures in, 27, 75–76, 109–112, 331 application of stereotactic guiding devices, 109, 110 stereotactic brain imaging techniques, 109, 110–111 for prolactinomas, 328 for psychiatric and pain disorders, 563, 565, 566, 567, 568, 569 quality assurance in, 109–111 radiation delivery in, 75, 76, 109, 111–112, 129, 331, 649 radiation leakage in, 42 radiation source for, 33, 41 for recurrent gliomas, 211, 212 for retinoblastoma, 605 robotic, 167 skull-scaling devices of, 72, 73 target localization with, 27, 166–167 target volume determination in, 109, 111 thalamotomy, 545–546 treatment helmets in, 27 treatment planning in, 69–70, 73, 75–79 inverse planning, 77–79 as tremor treatment, 551–552 for trigeminal neuralgia, 121–122, 519–525, 530–531, 670 complications of, 520–521, 523 dose in, 521, 522–523 dose ratio in, 521 targeting in, 520 treatment response in, 522–523 for uveal melanoma, 598–602, 603 for vein of Galen malformations, 501 Gastrointestinal disorders, stereotactic body radiation-related, 628, 629, 630 Germ cell tumors biomarkers for, 365, 377 central nervous system, 377 chemotherapy for, 378, 379–380 in combination with radiation therapy, 358 immunohistochemistry of, 356 incidence of, 356 mixed, 359, 371 nongerminomatous, 368 chemotherapy for, 373, 379 Gamma Knife radiosurgery for, 360, 362 histologic variants of, 371

714 Germ cell tumors (Cont.) radiotherapy for, 373, 374, 375 surgical resection of, 373 in pediatric patients, 342, 356 pineal, 371 radiation therapy for, 342, 378–379 spinal seeding of, 373–374 surgical resection of, 373, 378 systemic, 377 Germinoma, 355, 371 age factors in, 356 chemotherapy for, 358, 373, 379, 380 Gamma Knife radiosurgery for, 360, 361–362 nongerminomatous, 357, 358, 377 pineal, 368 pure, 357, 358 radiation therapy for, 358, 373, 374, 375, 378 radiosensitivity of, 355 recurrent, 380 survival rates in, 374 Giant cell tumors, 153, 405, 448 Gill-Thomas-Cosman (GTC) frame, 5, 294 Glaucoma ophthalmic radiosurgery-related, 601–602, 603 pathophysiology of, 607 treatment for, 607–609 Glial tumors, 117–118, 355 Glioblastoma, 225, 228 Glioblastoma multiforme, 207 chemotherapy for, 381 fractionated stereotactic radiotherapy for, 217 proton beam radiosurgery for, 151–152 stereotactic radiosurgery for, 207–209, 214, 669 Glioma brachytherapy for, 134 brain-stem, 333, 346 classiﬁcation of, 207 high-grade external-beam radiation therapy for, 210, 212–214 fractionated stereotactic radiosurgery for, 210–211, 215–217, 218 linear accelerator-based stereotactic radiosurgery for, 135–136, 211, 212 malignant, 227–228 in pediatric patients, 342 proton beam radiosurgery for, 151–152 radiation cell-sensitizing agent treatment for, 209 radiation therapy for, 342 recurrent, 211, 214–216 single-fraction radiosurgery for, 214–216 stereotactic radiosurgery for, 207–221, 669 stereotactic radiotherapy for, 218 treatment planning for, 210 linear accelerator-based stereotactic radiosurgery for, 135–136 low-grade chemotherapy for, 352 in pediatric patients, 154, 352 proton beam radiosurgery for, 154

index optic in pediatric patients, 153, 154, 334 proton beam radiosurgery for, 153, 154 in pediatric patients, 153, 154, 334, 342 pineal, chemotherapy for, 381–382 proton beam radiosurgery for, 148 radiosurgery for, 18 recurrent, 223 pharmacotherapy for, 223–231 of the tectal plate, 334 Glomus jugulare tumors, 388–390, 395, 406, 422 Glomus tumors, 120, 383, 405, 421 Glomus tympanicum tumors, 406, 422 Glossary, 169 Gold grain implantation therapy, 412, 413 Gold seed implantation, 173 Gonadotroph pituitary tumors, 329–330 Growth hormone deﬁciency, radiation therapy-related, 348 Growth hormone-secreting adenomas, 116, 117, 309, 322–323, 329 H Hadron therapy. See Proton beam stereotactic radiosurgery Hair loss, whole-body radiation therapyrelated, 202 Halo ring, application of, 696–697 Hamartoma diagnosis of, 327 hypothalamic, 334, 573, 574–575, 586, 588–589 Head and neck cancer combined chemotherapy/radiation therapy for, 425–430 conventional treatment for, 421–422 epidermal growth factor receptor overexpression in, 636 recurrent squamous cell, 428–429 stereotactic radiosurgery for, 411–419, 422–423 proton beam, 155–156 surgical resection of, 421–422 Health, deﬁnition of, 667 Hearing loss acoustic neuroma treatment-related, 657 Gamma Knife radiosurgery-related, 114, 115 proton beam radiosurgery-related, 158 Hearing preservation in acoustic neuroma radiosurgery, 284, 286, 654–655 in fractionated stereotactic radiosurgery, 290, 291, 292, 293, 294, 297 Gardner-Robertson scale for, 290 Helium-ion therapy, for uveal melanoma, 598, 603 Hemangioblastoma, 311–312, 335 natural history of, 516 radiosurgery for, 121, 516 spinal, 433 Hemangioma, Gamma Knife radiosurgery for, 119, 120–121 Hemangiopericytoma, of the cavernous sinus, 133–134

Hemiballism, 555 Hemidystoxia, 544 Hemispherectomy, 587 Hemorrhage arteriovenous malformation-related, 15, 16, 112, 152, 473, 474, 476, 479 aneurysm-related, 481 effect of partial resection on, 480 in pediatric patients, 459 post-radiosurgery, 463, 467, 480, 482–483, 485, 656, 658 re-hemorrhage risk, 479 risk factors for, 459 cavernous malformation-related, 112–113, 491, 492, 503, 504, 505, 513–515 post-operative, 495, 496, 497, 498–499, 509 choriocarcinoma-related, 357 deep brain stimulation-related, 550, 553, 554 gastrointestinal, stereotactic body radiation-related, 629 hemangioma-related, 120–121 in nasopharyngeal carcinoma patients, 418 stereotactic radiosurgery-related, 383 Hepatocellular carcinoma hyperfractionated radiation therapy for, 615 stereotactic body radiation therapy for, 613–614, 625–626, 627, 639, 645 transarterial chemoembolization (TACE) of, 616 Heterotopia, periventricular, 579 Hippocampal stereotactic radiosurgery, for epilepsy, 64–65 Hippocratic Oath, 681 Homer-Wright rosettes, 356, 377 Homogeneity index, 75 Hormone deﬁciencies, radiation therapyrelated, 348, 349 Horsley-Clarke device, 25, 26 β-Human chorionic gonadotropin, as pineal tumor marker, 365, 371, 377, 379, 385 Huntington’s chorea, 25 Hydrocephalus pineal region tumors-related, 357, 359, 365, 372 stereotactic radiosurgery-related, 19, 650, 652–653, 654 vein of Galen malformation-related, 501 Hydroxyurea, 272 Hypercortisolism, 328 Hyperfractionated accelerated radiation therapy (HART), 636 Hyperfractionated radiotherapy, for head and neck cancer, 427–428 Hyperprolactinemia, 327 Hyperthyroidism, 328, 329 Hypofractionated stereotactic radiosurgery, 175 Hypofractionation, 5 Hypopharyngeal cancer, 422 Hypophysopexy, 313–314 Hypopituitarism, 117, 306 Hypothalamic tumors, 334 Hypoxia, 51, 52, 54, 637

index I Image-guided radiation therapy (IGRT), 164–165 Imaging. See Neuroimaging Informed consent, 681–683 Infratemporal fossa tumors, 423 Insulin-like growth factor-1, 116, 323, 329 Intensity-modulated radiation therapy (IMRT), 80, 164 for angioﬁbromas, 384 in combination with image-guided radiation therapy, 166 deﬁnition of, 688 for esthesioneuroblastomas, 385 intraocular, 595 with isocentric linear accelerators, 44 for meningiomas, 240 for nasopharyngeal carcinoma, 48, 412, 417–418, 427 for pancreatic adenocarcinoma, 177 for skull base tumors, 239, 240, 296, 388 for spinal tumors, 431 treatment planning in, 80 Interferon-alpha, as meningioma treatment, 272 Interleukin-6, 272 International Stereotactic Radiosurgery Society (IRRS), 6 Intracarotid amobarbital test, 583 Intrathoracic tumors, CyberKnife radiosurgery for, 176, 177 Ionizing radiation, 33 Isodose curve, in trigeminal neuralgia, 76 Isodose distribution, 688 Isodose line, 9, 40–41, 461 Isodose plot, 74 J Jugular foramen, schwannomas of, 120 K Kaplan-Meier product-line method, 275 Keen, William W., 249 Kjellberg, Raymond, 3, 4, 142, 673 Kyphoplasty, 176, 448, 451–452 L Lacrimal gland adenoid cystic cancer, 422 Laminectomy, decompressive posterior, 449 Lamotrigine, as trigeminal neuralgia treatment, 537 Larsson, Borje, 4 Laryngeal cancer, 421 Lawrence, Ernest, 3, 4, 673 Leksell, Lars, 3, 4, 5, 25–26, 41, 46, 75, 91, 107, 129, 131, 519–520, 521, 635, 699. See also Gamma Knife stereotactic radiosurgery Leksell Gamma Knife radiosurgery. See Gamma Knife stereotactic radiosurgery Leksell GammaPlan (LGP) software, 111–112 Lennox-Gastaut syndrome, 587, 588 Lesioning/lesionectomy, 585–586 Leukemia, 332, 333, 606

Levodopa, as Parkinson disease treatment, 541–542, 559, 560 Linear accelerator-based stereotactic radiosurgery, 27–28, 129–140 accuracy of, 69, 130 for acoustic neuromas, 277, 279, 654 advances in, 27–28 advantages of, 4 for arteriovenous malformations, 136–138, 462 for benign tumors, 131–134 for cavernous malformations, 496, 497 with circular cones, 79 comparison with Gamma Knife radiosurgery, 69, 70, 663 complications of, 650 components and functions of, 43–44, 91, 92–96, 98, 99 cost-effectiveness of, 663 deﬁnition of, 129 dose planning in, 28 dosimetric characteristics of, 46 for high-grade gliomas, 209 history of, 3, 129 intraocular, 595 isocentric, 43–44 isodose calculations for, 73–74 for malignant tumors, 134–136 for meningiomas, 237, 238 with micro-multileaf collimators, 28, 79–80, 167, 694–695 procedure and technique in, 28, 131 quality assurance for, 47, 48 radiation delivery technology of, 129, 130, 649 for recurrent gliomas, 211, 212 robot-assisted, 26, 29–30 for spinal tumors, 431–432 treatment applications of, 131–138 treatment planning for, 70, 73 for trigeminal neuralgia, 530–531 volumetric imaging system of, 44 Linear accelerators (linac), 27 X-ray production in, 33–36 Lipoma, 357 Liver cancer ablation treatment for, 616 brachytherapy for, 645 hyperfractionated radiation therapy for, 615 liver transplantation for, 615–616 stereotactic body radiation therapy for, 625–627, 638 complications of, 628–630 for metastatic liver cancer, 619, 625–627, 639, 645 target volume in, 618 surgical treatment of, 616 Liver disease, radiation-induced (RLIND), 628–629, 630 Liver transplantation, 615–616 Local anesthetics, as trigeminal neuralgia treatment, 535, 538 Lomustine, 224–225 Lung cancer brachytherapy for, 644–645 conventional radiation therapy for, 615

715 CyberKnife radiosurgery for, 5, 81–82, 176 fractionated radiation therapy for, 635, 636–637 metastatic, 435 to the brain, 181, 195, 196, 650–651 to the spine, 438, 439–440, 443 stereotactic body radiation therapy for, 623, 624, 625, 638, 639 proton beam radiosurgery for, 156–157 radiofrequency ablation of, 615 radiosurgical ablation of, 176 squamous cell, 623 stereotactic body radiation therapy for, 613, 615, 618, 619, 623–625, 635, 638, 639, 643 complications of, 630, 638, 639 surgical treatment for, 615 Lymphangioleiomyomatosis, pulmonary, 317–318 Lymphoma cavernous sinus, in AIDS patients, 242 orbital and ocular, 422, 606 pediatric, 332, 333 M Macular degeneration, age-related, 606–607 Magnetic resonance angiography, of arteriovenous malformations, 15, 16, 17 Magnetic resonance imaging (MRI), 10–11, 69, 649 accuracy of, 11 of cerebellar masses, 10 in combination with computed tomography, 11, 12, 71, 110 with Gamma Knife radiosurgery, 109, 110–111, 115 with linear accelerator-based radiosurgery, 131, 132 deﬁnition of, 169 development of, 9 diffusing weighted imaging techniques in, 13–14, 15 ﬁducial systems in, 71, 72, 703, 704 limitations to, 10–11, 71–72 procedure of, 703, 704 Magnetic resonance perfusing imaging, 13, 14, 651 Magnetic resonance spectroscopy, 10, 13, 14 Malcaverin, 492 Malpractice lawsuits, 681–686 components of, 684–686 Manhattan Project, 141 MAPS system, for spinal surgery decision making, 445, 447 Mastoid, meningioma of, 254 Maxillary sinus adenoid cystic cancer, 422 Medicare, 688 Medicolegal issues, in stereotactic radiosurgery, 681–686 Medulloblastoma, 153, 335 chemotherapy for, 352, 380 incidence of, 341, 342 in pediatric patients, 332, 333, 352 proton beam radiosurgery for, 154 radiation therapy for, 342–345

716 Melanoma metastatic to the brain, 181, 650–651 ocular, 148–149 uveal brachytherapy for, 597, 603 choroidal, 597 of the ciliary body, 597 comparison of treatment methods for, 603 external beam radiation therapy for, 597 helium-ions therapy for, 598, 603 imaging of, 597 of the iris, 597 location of, 596 pathophysiology of, 596 proton beam therapy for, 597, 603 radiotherapy for, 597 staging of, 596 stereotactic radiosurgery for, 598–603 stereotactic radiotherapy for, 598, 603 surgical treatment for, 597 treatment planning for, 595 Melatonin, 371 Melphalan, as pineal tumor treatment, 372–373 Memory loss, external beam radiation therapy-related, 265 Meningioma, 233–248 aggressiveness of, 318 anaplastic/malignant, 258 atypical, 258 differentiated from typical meningioma, 245 external beam radiation therapy for, 265–266 with mastoid invasion, 254 as percentage of all meningioma, 258, 265 recurrence of, 265 stereotactic radiosurgery for, 240–241, 266 surgical resection of, 266 benign, 258, 271 biopsy of, 243–244 breast cancer-associated, 249, 257, 272 cavernous sinus, in pediatric patients, 335 chemotherapy for, 272 classiﬁcation of, 233, 245 clinical presentation of, 318 conventional radiation therapy for, 234–235 convexity, 250, 251 cranial radiation-related, 317–318 deﬁnition of, 651 diagnosis of, 327 dural tail of, 249–250 en plaque, 249 epidemiology of, 257–258 estrogen receptors in, 257 etiology of, 257 external beam radiation therapy for, 259, 261, 262 for atypical meningioma, 265–266 comparison with stereotactic radiosurgery, 266–267

index complications of, 265, 267 dural tail targeting in, 264–265 postoperative, 264 primary, 263–264 radiation dose in, 264 for recurrent meningioma, 266 target volume in, 264–265 falx, 335 fractionated radiotherapy for, 324 of the frontal fossa, 240 grading of, 249, 258 as headache cause, 249 histologic dedifferentiation of, 233 histopathology of, 258 imaging of, 249–250, 258 incidence of, 257–258, 317 location of, 317 neuroﬁbromatosis type 2 associated with, 249, 257, 317–318 observation of, 263, 320 of the olfactory groove, 240 of the optic nerve, 240, 271, 272, 294–295 parasagittal, 250, 251 preoperative embolization of, 252 radiosurgery for, 250, 651–652 surgical resection of, 250, 251 parasellar, 290, 296, 305 pathophysiology of, 249 in pediatric patients, 257, 335 petroclival, 238, 253–254, 652 pineal, 367, 371 pituitary, 305 progesterone receptors in, 257 pulmonary lymphangioleiomyomatosisrelated, 317–318 radiation-related, 249, 257 radiation therapy for, 261 radiosensitivity of, 263 radiotherapy for, 319, 324 recurrent, 245, 249–250, 251, 265, 266 as seizure cause, 249 sellar, 305 sex factors in, 257, 272, 317 signs and symptoms of, 249 sites of, 260 skull base, 651 in pediatric patients, 335 radiosurgery-related complications of, 652 of the sphenoid wing, 250, 251 spinal, CyberKnife radiosurgery for, 433, 436–437 staging system for, 319 stereotactic radiosurgery for, 235–238, 261–263, 651–653 complications of, 262–263, 656–657 CyberKnife, 175, 433, 436–437 Gamma Knife, 115, 116 linear accelerated-based, 133 local control rates in, 261, 262 optimal radiation dose in, 261–262 proton beam, 87, 150–151 recommendations regarding, 242–245 single-fraction, 261 for small lesions, 261 toxicity of, 271

stereotactic radiotherapy for, 235–245 avoidance of complications in, 244–245 surgical resection of, 233–234, 243–244, 249–255, 259, 271 grade zero resection, 259 gross total, 259–260 recurrence after, 258, 259 Simpson resection, 250, 259, 260 subtotal, 258–259, 260 transspheroidal approach in, 309 systemic therapy for, 271–273 Metastases. See also under speciﬁc types of cancer oligo, stereotactic body radiation therapy for, 611, 614 orbital and uveal, 605, 606, 607 palliative therapy for, 615 pituitary, 305–306 surgical treatment for, 309 Methylguanine methyltransferase promoter, 381–382 Middle cranial fossa tumors, 386–387 Mifepristone, 272 Misonidazole, 201, 637 Modeling, radiobiologic, 612 MOMOS Radiation Oncology Batcam, 166 Monoamine oxidase inhibitors, 542, 560 Movement disorders, 541–548, 559–561 deep-brain stimulation treatment for, 549 comparison with radiosurgical ablation, 549–558 pallidal, 552–553 subthalamic, 554–555 thalamic, 549–550 hyperkinetic, 559 hypokinetic, 559 stereotactic radiosurgery for complications of, 546 expermental models of, 65 Gamma Knife radiosurgery, 122–124 history of, 25 pallidotomy, 549, 553 subthalamotomy, 555 technical considerations in, 545–546 thalamotomy, 549, 550–552 Moya-moya phenomenon, 335 Multiple endocrine neoplasia type 1, 317 Multiple sclerosis, 524 N Narcotic agents, as trigeminal neuralgia treatment, 538 Nasal cancer, 422 Nasal cavity, esthesioneuroblastoma of, 384–384 Nasopharyngeal carcinoma, 411–419, 422 angioﬁbromas, 383–384 metastatic, 306 treatment for brachytherapy, 412, 413, 418 combined chemotherapy/radiation treatment, 426–430 conventional salvage treatment, 411–412 CyberKnife radiosurgery, 174 intensity-modulated radiation treatment, 48, 427

index proton beam radiosurgery, 156 stereotactic radiosurgery salvage treatment, 411, 412–418 surgical resection, 412 Necrosis, radiation, 118, 383, 650 in arteriovenous malformation patients, 464 in brain metastases patients, 188 deﬁnition of, 17, 18 differentiated from tumor recurrence, 17, 18 effect on neuroimaging ﬁndings, 17–18 in glioblastoma muliforme patients, 217–218 proton beam radiosurgery-related, 152 treatment for, 651 Negligence, medical, 681–686 Nephromas, pediatric, 333 Neuralgia post-herpetic, 524, 535 trigeminal. See Trigeminal neuralgia Neuroanatomic structures, radiosensitivity of, 17, 654 Neuroblastoma, 154, 422 Neuroendocrine tumors of the head and neck, 421 metastatic, 121 Neuroﬁbroma, spinal, 433, 436, 437 Neuroﬁbromatosis linear accelerator-based radiosurgery for, 132 type 1, 335 type 2, 115, 275, 276, 279 Neuroimaging, in radiosurgery, 9–23, 71–72, 687, 703–704 combined modalities in, 15 contrast media administration in, 11–12 evaluation of treatment efﬁcacy in, 17–21 image-guided radiation therapy (IGRT), 164–165 imaging modalities, 9–17 radiologic considerations in, 17 Neurologic deﬁcits, radiosurgery-related, 17, 650, 651 Neuroma, acoustic, 5, 115 conservative management of, 283 cost-effectiveness of treatment for, 667 deﬁnition of, 19, 275 fractionated stereotactic radiosurgery for, 297 imaging of, 19 incidence of, 19 local control of, 275, 276–277, 279 management of, 275–281 neuroﬁbromatosis type 1-associated, 335 neuroﬁbromatosis type 2-associated, 275, 276, 279 observation of, 275, 276, 283 in pediatric patients, 335 radiotherapy for, 277–278 recurrent, 286 stereotactic radiosurgery for, 19, 275, 276, 278–280, 654, 677 complications of, 19, 657–658 CyberKnife, 175 facial nerve preservation in, 19

Gamma Knife, 114–115, 131, 290, 291, 292, 293 hearing preservation in, 19 linear accelerator-based, 131–133 proton beam, 87, 152–153 quality-of-life outcomes in, 669 surgical treatment for, 19, 275, 276–277, 283–284, 283–287, 654, 667 comparison with radiosurgical management, 283–287 comparison with single-fraction radiosurgery, 297 disadvantages of, 677 hearing preservation in, 284, 286 surgical approaches in, 284–285, 286 treatment advances for, 279–280 treatment options for, 275–278 treatment planning for, 278, 279–280 Neuropathic pain, 535 Neuropathic pain agents, 535–538 Neuropathy cranial, radiosurgery-related, 262, 290, 291, 649 diabetic, 535 facial, 650, 654 optic, 238, 240, 265, 289, 290, 652 glaucoma-related, 607–608 ophthalmic radiosurgery-related, 601–602 radiosurgery-related, 651 stereotactic radiation therapy-related, 262 trigeminal, 238, 265, 650, 654, 657–658. See also Trigeminal neuralgia, 265, 650, 654, 657–658 Nimustine, 224, 225 Nitrosoureas, 224 Non-coplanar arcs, 595, 618 Non-coplanar conformal static ﬁelds, 595 Novalis, 695 Nuclear Regulatory Commission, 109–111, 674–675, 688–689 Nurses, 696 O Obesity, hypothalamic radiosurgery for, 66 Obsesssive-compulsive disorder, 565–566, 568–569 Occipital lobe resection, 586 Ocular and orbital lesions, stereotactic radiosurgery and stereotactic radiotherapy for, 593–610 CyberKnife radiosurgery for, 174–175 proton beam radiosurgery for, 148–149 technical aspects of, 593–596 Olfactory groove, meningiomas of, 240 Olfactory neurogenic tumors, 385 Oligoastrocytoma, anaplastic, 226–227 Oligodendroglioma, anaplastic, 226–227 Oligometastatic disease, 448, 451 Optic apparatus distance from tumors, 17 radiosensitivity of, 652 Optic chiasm, meningiomas of, 115 Optic nerve gliomas of, 334

717 meningiomas of, 240, 271, 272, 294–295 radiosensitivity of, 289–290, 320, 602 Optic pathway tumors, in pediatric patients, 334 Oral cavity cancer, 421, 422 Oral glucose tolerance test, 116 Orbit lymphoma and leukemia of, 606 metastases to, 605 Osteoblastomas, 448 Oxcarbazepine, as trigeminal neuralgia treatment, 536 P Pacemakers, 521, 522 Pain management in metastatic bone disease, 448, 449 with radiation therapy, 448 with stereotactic radiosurgery, 541, 563–572 Palamotomy, as essential tremor treatment, 561 Pallidotomy, 541 radiosurgical, 123, 124, 543, 544, 546, 549, 553 Pancoast tumors, 448 Pancreatic cancer brachytherapy for, 645 CyberKnife radiosurgery for, 81–82, 627–628 stereotactic body radiation therapy for, 614, 627–628, 630, 645 Paraganglioma. See Chemodectoma Parasellar tumors, 305 diagnostic evaluation of, 319 fractionated stereotactic radiosurgery for, 295–296 meningiomas, 252–253 radiation therapy for, 320 Paraspinal tumors, stereotactic body radiation therapy for, 614, 627 Parietal bone, arteriovenous malformations of, 465, 466 Parkinson disease deep-brain stimulation therapy for, 541, 543–544, 549–550, 552–553, 554–555, 560 diagnosis of, 559 differentiated from parkinsonism, 541, 542 Gamma Knife radiosurgery for, 123–124 meningiomas associated with, 243 pallidotomy for, 541 radiosurgical, 543, 544, 553 pharmacologic therapy for, 541–542, 559, 560–561 radiosurgical caudatomy for, 542–543 radiosurgical subthalamotomy for, 543–544 signs and symptoms of, 559–560 surgical treatment for, 560 thalamotomy for, 541 radiosurgical, 542–543, 555 Parkinsonism, 542, 559 Particle accelerators, 33, 34 Particle beam accelerators, 3, 4

718 Patient care, in stereotactic radiosurgery, 699–707 for pediatric patients, 706 Patient positioning, 164, 168 Pecchiolo, Zanobi, 249 Pediatric patients angioﬁbromas in, 383–384 arteriovenous malformations in, 112 follow-up angiography in, 463 as hemorrhage cause, 459 radiosurgery for, 112 radiosurgery-related complications in, 655–656 brain metastases in, 332 brain tumors in, 341–349 chemotherapy for for, 351–354 fractionated radiation therapy for, 342–343 radiation therapy for, 341–342 stereotactic radiosurgery for, 342–343 craniopharyngiomas in, 318 Gamma Knife radiosurgery in, 331–340 germ cell tumors in, 356, 378 gliomas in, 352, 381 hippocampal sclerosis in, 583–584 medulloblastomas in, 342–345 meningiomas in, 257 patient care for, 706 pineal region tumors in, 317, 355 proton beam radiosurgery in, 148, 153–154, 158 radiation therapy-related side effects in, 153, 154–155 cognitive effects, 342–344, 348 endocrine effects, 348–349 radiosensitivity in, 331–332 retinoblastomas in, 605 vagus nerve stimulation in, 588 Pegvisemant, 329 Pelvic cancer proton beam radiosurgery for, 142, 153, 154, 155 stereotactic body radiation therapy for, 638, 640 Pencil-beam dose calculations, 74 Perfusing imaging, magnetic resonance, 13, 14, 651 Pharyngeal cancer, 421 Phenytoin, as trigeminal neuralgia treatment, 537 Photocoagulation, as macular degeneration treatment, 606 Photoelectric absorption, 36–37 Photon beam delivery systems, 41–45 Photon/matter interactions, 36–37 Photons, 33 Physics, of stereotactic radiosurgery, 33–50 dose and dosimetry, 37–41, 46 photon beam delivery systems, 41–45 photon/matter interactions, 36–37 proton/matter interactions, 45–46 quality assurance, 46–47, 48 radiation sources, 33–36 Pineal apoplexy, 357 Pineal gland anatomy of, 355–356, 371

index calciﬁcation of, 357 location of, 355, 356 Pinealocytes, 355 Pinealocytoma. See Pineocytoma Pineal parenchymal tumors, 355, 367–368, 371, 377. See also Pineoblastoma; Pineocytoma chemotherapy for, 373, 380–381 classiﬁcation of, 377 Gamma Knife radiosurgery for, 375–376 histologic diagnosis of, 357 immunocytochemistry of, 356–357 incidence of, 356 of indeterminate differentiation, 357, 377, 380–381 pathology of, 356–357 radiotherapy for, 373, 374, 375 spinal seeding of, 373–374 whole-brain radiation therapy for, 376 Pineal region, anatomy and function of, 355–356 Pineal region tumors, 355–364. See also Pineal parenchymal tumors; Pineoblastoma; Pineocytoma benign, 367 chemotherapy for, 355, 372–373, 374, 375, 377–382 in pediatric patients, 355 classiﬁcation of, 366 clinical presentation of, 356 debulking of, 358 germ cell, 367, 368 glial cell, 366, 367 histologic diagnosis of, 374 hydrocephalus associated with, 365 imaging of, 357, 372 incidence of, 355–356, 371, 377 metastatic, 374 natural history of, 356 parenchymal, 367–368 pathophysiology of, 356–357 radiotherapy for, 372–376, 373, 374–375 “second-look” surgery for, 358 spinal seeding of, 372, 373–374, 375, 376 stereotactic radiosurgery for, 358–362, 365 fractionated, 371–376, 375–376 Gamma Knife, 118–119, 355, 375–376 surgical treatment for, 355, 357–358, 365–370, 372 symptoms of, 372 treatment of, 357–363 types of, 355 Pineoblastoma, 367–368, 371–372, 377 chemotherapy for, 380–381 Gamma Knife radiosurgery for, 360, 362, 375 in pediatric patients, 335, 341, 342 Pineocytes, 371 Pineocytoma, 367, 371, 377 chemotherapy for, 373 Gamma Knife radiosurgery for, 359, 360, 361, 375–376 pathology of, 356–357 radiotherapy for, 373 survival rates in, 374 Pituicytoma, 305

Pituitary adenoma, 317 adrenocorticotropic hormone-secreting, 116–117, 301, 311–312, 323–324, 328 clinical presentation of, 318 contrast-related volume increase of, 19 corticotroph, 328 functioning, 299, 317 growth hormone-secreting, 116, 117, 301, 309, 322–323, 329 macroadenomas, radiotherapy for, 322 nonfunctioning, 295, 296, 299, 300, 301, 302, 317, 322, 329–330 radiotherapy for, 320 of the sella turcica, 299 sex factors in, 300–301 somatotroph, 329 stereotactic radiosurgery for CyberKnife, 175 Gamma Knife, 115–117, 300, 302–305 proton beam, 300 surgical treatment for, 309–316 with adjuvant radiotherapy, 313–314 comparison with radiosurgery, 314–315 historical perspective on, 299–300 surgical approaches in, 309–313 with transsphenoidal craniotomy, 299 thyrotroph, 328–329 Pituitary hormone deﬁciency, 117 Pituitary/pituitary region tumors, 299–308. See also Pituitary adenoma classiﬁcation of, 318, 327 diagnosis of, 327 diagnostic workup of, 398 fractionated radiotherapy for, 317–326 pathophysiology of, 300–301 in pediatric patients, 334, 341, 342 pharmacotherapy for, 327–330 proton beam radiosurgery for, 86, 87 radiation therapy for, 320–324 staging of, 398 stereotactic radiosurgery for, 299–308 proton beam, 86, 87 surgical treatment for, 301–302, 320 endoscopic, 300 Pituitary transposition, 313–314 Pneumonitis, 629, 630, 639, 640, 643 Pons, cavernous malformations of, 504, 507 Positron emission computed tomography (PET), 12–13, 26, 165 deﬁnition of, 169 of tumor recurrence, 18 Posterior fossa, arteriovenous malformations of, 461 Posterior fossa tumors, 387–390, 527. See also Chondroma; Chordoma in pediatric patients, 154 radiation therapy for, 344–345, 346–347 proton beam radiosurgery for, 154 Postoperative care, 705–706 Prednisone, as pineal tumor treatment, 372–373 Pregabalin, as trigeminal neuralgia treatment, 537 Pregnancy, meningioma during, 272 Primidone, as essential tremor treatment, 561

719

index Primitive neuroectodermal tumors (PNETs), 335, 356, 367–368, 371, 380 chemotherapy for, 352 Gamma Knife radiosurgery for, 362, 363 incidence of, 342 supratentorial, 344 Procarbazine, 224–225, 372–373 Progesterone receptors, in meningioma, 257 Progesterone receptors antagonists, 272 Prolactinoma, 117, 300–301, 323, 327–328 Propranolol, as essential tremor treatment, 561 Prostate cancer CyberKnife radiosurgery for, 81–82 metastatic to the spine, 443, 445 proton beam radiosurgery for, 155 Proton beam accelerators, 3, 4 Proton beam radiosurgery facilities, 142–143 Proton beams, 33, 34 production of, 44–45 Proton beam stereotactic radiosurgery, 83–88, 141–161 for age-related macular degeneration, 606 for arteriovenous malformations, 87, 142, 152 for bronchial and esophageal cancers, 156–157 for cavernous malformations, 495, 496 for central nervous system cancer, 149–153 for chondromas, 403 for chondrosarcomas, 403 cost of, 694 early history of, 141–142 for eye tumors, 148–149 for head and neck tumors, 155–156 patient setup for, 145–146 patient treatment planning and simulation for, 146–147 for pediatric cancers, 153–154 for pelvic cancers, 155 physical properties of, 143–144 procedure in, 29 proton production and delivery systems in, 144–145 radiobiological considerations in, 145 for skull base tumors, 388 socioeconomic aspects of, 158 target localization in, 29 toxicity of, 157–158 treatment planning in, 29 for uveal melanoma, 597, 603 Proton/matter interactions, 45–46 Protons deﬁnition of, 141 physical properties of, 143–144 Pseudotumor, orbital, 418–419, 422 Psychiatric disorders, stereotactic radiosurgery for, 563–572 Q Quality-adjusted life years (QALYs), 663, 666 Quality assurance, 46–47, 48, 101, 109–110, 701 Quality evaluation, of treatment plans, 74–75

Quality of life in brain metastases patients, 197–198 following proton beam radiosurgery, 157–158 following stereotactic body radiation therapy, 640, 646 R Radiation cell-sensitizing agents, 201–202 gadolinium texaphyrin, 218 for high-grade glioma treatment, 209 Radiation exposure ALARA principle of, 98 monitoring of, 110 Radiation safety programs, 98 Radiation shielding, 97–98 Radiation therapy. See also Intensitymodulated radiation therapy (IMRT); Stereotactic body radiation therapy (SBRT) as meningioma cause, 257 for meningiomas, 234–235 for pain management, 448 in pediatric patients cognitive effects of, 342–344, 348 endocrine effects of, 348–349 for spinal metastases, 448–449 Radiation Therapy Oncology Group (RTOG), 201, 202 brain metastases prognostic scale of, 181 brain metastases studies of, 181, 182, 188 central nervous system radiotoxicity criteria of, 188 homogeneity index deﬁnition of, 75 radiosurgery quality assurance guidelines of, 46 Radioactive decay, 33 Radiobiological principles, underlying stereotactic radiosurgery, 51–60 clonogen repopulation, 53, 54 reoxygenation, 51–52, 54, 637 tissue repair, 52–53, 54 RadioCameras Treatment Guidance System, 71, 72 Radioisotopes, 33, 34 Radioprotective agents, 63 Radiosurgery. See Stereotactic radiosurgery Rathe cleft, cysts of, 305, 309 Rectal cancer, proton beam radiosurgery for, 155 Reimbursement, for stereotactic radiosurgery, 677–679 Renal cell carcinoma metastatic, 181, 183, 438, 439–440, 628, 651 stereotactic body radiation therapy for, 614, 628 Reoxygenation, 51–52, 637 Retinoblastoma, 154, 605 trilateral, 355, 372 Retinopathy diabetic, 142 ophthalmic radiosurgery-related, 601–602, 603, 604 Retroperitoneal tumors, 638 Rhabdomyosarcoma, 422

Robotics, 163–170, 689 Roentgen, Wilhelm Konrad, 3 S Sagittal sinus, meningiomas of, 250, 251 Salivary gland tumors, 421, 422 Sarcoma chondroid, 387 skull base, 158 spinal, 448 Schiller-Duvall bodies, 356, 371 Schwannoma facial, 119–120 of the jugular foramen, 120 non-acoustic, 119–120 spinal, 433, 435, 436–437 traumatic, 435 trigeminal, 120 vestibular. See Neuroma, acoustic Sclerosis biliary, 629 hippocampal, 583–585 reactive, 250 tuberous, 335 Seizures arteriovenous malformation-related, 475 cavernous malformation-related, 503, 504, 505 cavernous malformations-related, 491, 492, 493 deﬁnition of, 583 radiosurgery-related, 649, 650, 651 Sellar tumors, 252–253, 299, 300, 305, 319, 327 Semantics, of stereotactic radiosurgery, 687–690 Seminomas, 378 Shunting cerebrospinal ﬂuid, 372 ventriculo-peritoneal, 365, 374 Single-fraction radiosurgery for acoustic neuromas, 286 for high-grade glioma, 214–216 Single-positron emission computed tomography (SPECT), 165, 651 Skull base tumors, 401–409 benign, 383 complex, 296–297 head and neck cancer-related, 418, 422, 423 invasive, 121 malignant, 383 meningiomas, 238, 239, 245 of anterior cranial ﬂoor, 252–253 in pediatric patients, 335 surgical resection of, 241–242, 252–253, 271 pituitary tumor-related, 310–311, 312 radiotherapy for, 320 staging system for, 319 stereotactic radiosurgery for, 383–392 CyberKnife, 174 fractionated, 296–297 Gamma Knife, 118–121 intensity-modulated, 240 proton beam, 149–150

720 Skull base tumors (Cont.) surgical treatment for, 393–399 types of, 401 Skull-scaling devices, 72, 73 Sphenoid wing, meningiomas of, 250, 251 Sphere packing, in treatment planning, 69–70 Spinal cord arteriovenous malformations of, 176 cavernous malformations of, 504 radiosensitivity of, 630 Spinal cord tumors, in pediatric patients, 341, 342 Spinal fusion, in cancer patients, 448 Spinal stabilization, in cancer patients, 447–448 Spinal tumors benign, 436–437 conventional radiation treatment for, 431 intensity-modulated radiation therapy for, 431 location of, 443 metastatic CyberKnife radiosurgery for, 437–440 fractionated stereotactic radiosurgery for, 455–458 palliative treatment of, 447 stereotactic body radiation therapy for, 627 surgical management of, 443–454 pediatric, 332 staging system for, 445 stereotactic body radiation therapy for, 614 stereotactic radiosurgery for, 431–442, 444 CyberKnife, 81, 171, 173, 432, 433–441 frameless, 168 surgical treatment for, 443–454 Spondylectomy, en bloc, 448, 451 Spongioblastoma multiforme, 207 Squamous cell carcinoma of the head and neck, 418, 421, 422–423, 425–430 recurrent, 428–429 of the skull base, 121 SRS. See Radiosurgery Standard of care, 683–684 Stereoelectroencephalography (SEEG), 577–578, 579 Stereoencephalotomy, 25 Stereotactic, deﬁnition of, 91, 611 Stereotactic Body Frame, 688 Stereotactic body radiation therapy (SBRT), 611–633, 635–642 advantages of, 611 alternative local options to, 611, 615–616 avoidance/partial volume effects in, 629–630 clinical experience in, 637–640 clinical outcomes in, 622–628 comparison with brachytherapy, 643–646 complications of, 614, 617, 628–629 correction for breathing motion during, 616–617, 618 cost-effectiveness of, 640, 646 deﬁnition of, 455, 611, 689 effect on quality of life, 640, 646 historical perspective on, 612

index hypofractionated, 455, 456, 457, 639 image guidance in, 619–622, 635 three-dimensional, 621–622 two-dimensional, 620–621 multifraction, 641 patient immobilization during, 616–617 patient selection for, 645 prognostic factors in, 628 radiobiology of, 612–613 rationale for, 613–615 single-dose, 637 single-fraction, 641 for spinal metastases, 455, 456, 457 comparison with fractionated stereotactic radiosurgery, 455–458 treatment planning in, 617–619 Stereotactic radiation therapy, deﬁnition of, 688 Stereotactic radiosurgery programs, development of, 691–698 Stereotactic radiosurgery (SRS). See also CyberKnife stereotactic radiosurgery; Gamma Knife stereotactic radiosurgery; Linear acelerator-based stereotactic radiosurgery; Proton beamstereotactic radiosurgery antiepileptic effects of. See Epilepsy, stereotactic radiosurgery for biologic effects of, 236 comparison with open surgery, 649 complications of, 649–662 acoustic neuroma-related, 653–655, 657–658 acute, 649 arteriovenous malformation-related, 655–656, 658 brain metastases-related, 650–651, 656 early-delayed, 650 late-delayed, 650 management of, 656–658, 659 meningioma-related, 651–653, 656–657 components and functions of, 91–97 cost of, 101, 676–677, 694 deﬁnition of, 9, 169, 209, 275, 649, 688–689 development of, 25 dose inhomogeneity-related complications of, 237 dose proﬁle measurements in, 98, 99 errors in, 97 experimental models of, 61–67 focal radiation dose distribution in, 9 frameless, 6, 71, 72, 167–168. See also CyberKnife Stereotactic Radiosurgery System future of, 691–692 history of, 3–7, 5–6, 25–26, 91, 687–688 indications for, 71 installation and acceptance of, 3–4, 98–100 integrated units, 96–97 operating personnel training and qualiﬁcations for, 101 photon-based, 91–103 probable risk analysis of, 100 quality control, quality assurance, and quality management for, 101 radiation shielding in, 97–98

radiobiological considerations in, 688 experimental models of, 61–67 regulatory aspects of, 673–675, 679–680 reimbursement for, 673, 675–679, 680 as secondary tumor cause, 20–21 semantics of, 688–689 service and maintenance of, 100–101 single-fraction, misadministration of, 101 technological advances in, 69 training in, 101, 692, 695, 699 treatment accuracy of, 97 tumoricidal enhancement of, 63–64 Stereotactic radiosurgery systems, vendors of, 6, 694, 695 Stroke embolic, 476 hemorrhagic, 473 Subpial transection, multiple, 583, 587 Subthalamic nucleus, as Parkinson disease therapeutic target, 543–544 Subthalamotomy, radiosurgical, 55, 123, 543–544 Sudden death radiosurgery-related, 650, 652 stereotactic body radiation-related, 629 Sudden unexplained death in epileptic patients (SUDEP), 578 Synchrocyclotrons, 144 Synchrotrons, 33, 34, 44–45 T Talairach, Jean, 26, 573? Tectal plate, gliomas of, 334 Teletherapy, 688, 689 Temozolomide, 209 as anaplastic astrocytoma treatment, 226 as glioblastoma treatment, 225 as glioma treatment, 225, 226, 227–228, 352 as pineal glioma treatment, 381 as recurrent high-grade glioma treatment, 214–216, 218, 224, 225 toxicity of, 227–228 Temporal bone, arteriovenous malformations of, 465, 466 Temporal bone tumors, 253, 386, 405, 406 Temporal fossa, meningiomas of, 237 Temporal lobe resection, 583–585 Teratocarcinoma, 305, 371 Teratoma, 377 immature, 357, 371, 373 with malignant transformation, 357 mature, 357, 371, 373, 374 pathology of, 357 pineal, 367 Tetratoid rhabdoid tumors, 344 Thalamotomy, radiosurgical, 542–543, 549, 550–552 Gamma Knife, 123, 124, 568 magnetic resonance imaging in, 545–546 target planning in, 545–546 ventralis intermedialis, 545, 552 Thalamus arteriovenous malformations of, 461, 464 cavernous malformations of, 504, 509 Therapeutic index, 687 Thiotepa, as pineal tumor treatment, 372–373

721

index Thyroid carcinoma, 421 Thyrotropin-secreting tumors, 301, 328–329 Todd-Wells stereotactic frame, 26 TomoTherapy, 640 MV, 622 Tomotherapy units, 96, 97 Topiramate as essential tremor treatment, 561 as trigeminal neuralgia treatment, 537 Torticollis, 544–545 Torts, 681 Transarterial chemoembolization (TACE), 616 Transsphenoidal resection, endoscopic, 121 Trauma, as meningioma cause, 249, 257 Treatment planning, for stereotactic radiosurgery, 69–90, 704. See also under speciﬁc types of stereotatic radiosurgery background on, 69–70 basic steps in, 70–75 contouring of structures, 72, 73 deﬁnition of prescription, 73 design of treatment plans, 73 dose calculations, 73–74 evaluation of plan quality, 74–75 forward versus inverse planning, 74, 77–79, 83 patient imaging, 71–72 registration of images, 71, 72 stereotactic localization, 70–71 neuroimaging in, 9–23 in proton beam radiosurgery, 85–87, 146–147 sphere packing in, 69–70 Tremor arteriovenous malformation-related, 544–545 deep-brain stimulation treatment for, 549– 550, 552, 561 essential deep-brain stimulation treatment for, 561 deﬁnition of, 544 pharmacotherapy for, 561 stereotactic radiosurgery for, 123, 544, 561 Parkinson disease-related, 542, 543, 559 radiosurgical thalamotomy for, 544, 550–552 Trigeminal nerve function, 115 Trigeminal neuralgia, 46, 519–526 ablative percutaneous treatment for, 519 atypical, 683 clinical features of, 535 CyberKnife radiosurgery for, 81, 174, 175, 530 deﬁnition of, 19

Gamma Knife radiosurgery for, 121–122, 519–525, 530–531, 670 complications of, 520–521, 530–531 dose in, 521, 522–523 dose ratio in, 521 targeting in, 520 treatment response in, 522–523 imaging of, 19–20 imaging studies of, 19–20 isodose curves for, 76 linear accelerator-based stereotactic radiosurgery treatment of, 79–80, 530–531 pathophysiology of, 519, 535 pharmacotherapy for, 19, 519, 535–539 recurrent, 523–524 secondary, 524 stereotactic radiosurgery for, 563, 564–565, 566, 567–568 expermental models of, 65 surgical treatment for, 527–533 balloon compression, 528–529, 532 glycerol rhizotomy, 527–528, 532 microvascular decompression, 519, 523, 524, 529–531, 531, 567 radiofrequency rhizotomy, 528, 532 Trigeminal neuropathy, 238, 265, 650, 654, 657–658 Tuberous sclerosis, 335 Tumor control probability (TCP), 637 Tumor necrosis factor-α, 63–64 Tumor recurrence cerebral blood volume decrease in, 18 differentiated from radiation necrosis, 18 Tumor size, implication for radiation dose, 17 U Ultrasound, 165, 166, 169 Uvea melanoma of. See Melanoma, uveal metastases to, 605 V Vagus nerve stimulation, 566, 567, 583, 587–588 Varian 600SR radiosurgery device, 289, 295 Varian Trilogy Unit, 96, 97 Vascular malformations epileptogenic, 588 Gamma Knife radiosurgery for, 081, 107, 108, 112–113 pineal, 377 Vein of Galen malformations, 501 Ventralis intermedius in radiosurgery thalamotomy, 545, 552 therapeutic stimulation of, 549–550, 552, 561

Ventricles, cavernous malformations of, 504, 506 Ventriculostomy, third, 359, 363, 365, 366 Vertebrectomy, 451 Vertebreplasty, percutaneous, 448, 451–452 Vinblastine, 372–373 Vincristine, 224–225, 372–373, 380 Vision loss pituitary adenoma-related, 306–307, 309 prolactinoma-related, 301 proton beam radiosurgery-related, 157–158 radiotherapy-related, 320 Visual preservation, in optic nerve sheath meningioma patients, 295 von Hippel-Lindau disease, 121, 335, 516 W Wada test, 583 Whole-brain radiation therapy (WBRT), 201–206 as adjuvant therapy, 202, 203, 204 adverse effects in children, 332 for brain metastases, 134, 181–183, 187– 188, 193, 194–195, 197, 198, 201–206 complications of, 202, 203, 205 disadvantages of, 187 dose and scheduling in, 201–202 efﬁcacy of, 201, 203–204 for recurrent metastases, 184–187, 185–187 complications of, 202, 203, 205 cost-effectiveness of, 666 dose and scheduling of, 201–202 efﬁcacy of, 201, 203–204 for germinomas, 375 for glioblastoma multiforme, 209 for gliomas, 271 for metastatic brain tumors, 193, 194–195, 197, 198 for multiple brain metastases, 202 for nongerminomatous germ cell tumors, 375 for pineal region tumors, 376 for recurrent brain metastases, 204–205 for single brain metastases, 203–204 tumor recurrence after, 203, 204 Winston-Lutz units, 96–97 World Health Organization, meningioma grading system of, 249, 258 X X-rays, 129, 130, 165, 166 discovery of, 3, 9 production in linear accelerators, 33–36 therapeutic, 33 Y Yolk sac tumors, 305, 357, 371