OCNA Sept 2009 Radiosurgery and Radiotherapy for Benign Skull Base Tumors

Contents

Preface xiii

Robert A. Battista

Glossary e1

The History of Stereotactic Radiosurgery and Radiotherapy 593

John M. Lasak and John P. Gorecki

Stereotacticneurosurgery originated from the pioneeringworkofHorsley andClarke, who developed a stereotactic apparatus to study the monkey brain in1908. Spiegel and Wycis applied this technology to the human brain in 1947,which ultimately lead to the development of multiple stereotactic neurosurgi-cal devices during the 1950s. It was Lars Leksell of Sweden, however, whoenvisioned stereotactic radiosurgery. Leksell developed the gamma knifeto treat intracranial lesions in a noninvasive fashion. His work stimulatedworldwide interest and created the field of stereotactic radiosurgery.

Basic Principles of RadiobiologyApplied to Radiosurgery and Radiotherapyof Benign Skull Base Tumors 601

Christopher J. Anker and Dennis C. Shrieve

Various types of ionizing radiation may be used therapeutically for benignskull base tumors. Treatment may involve single-dose radiosurgery, ormay be fractionated into multiple doses. Designing and implementinga radiotherapy plan that maximizes the therapeutic ratio requires knowl-edge of the biophysical and radiobiological principles involved in thesetreatments. These basic radiobiological tenets are discussed in this chap-ter, with the focus on radiotherapy of benign skull base tumors. Animal andclinical data, however, acquired from the radiation of malignant tumors arenecessarily included, as they comprise much of our knowledge of fraction-ation schedules, central nervous system (CNS) toxicity, and CNS volumeeffects.

Radiation Effects on theAuditory andVestibular Systems 623

Niranjan Bhandare, William M. Mendenhall, and Patrick J. Antonelli

Definitive or postoperative radiation therapy (RT) is commonly used for themanagement of intracranial and extracranial head and neck tumors. Be-cause of the variability of tumor location and dimensions, sparing of non-target normal tissue and organs may not be possible. Treatment modalitiesthat deliver the highest doses of radiation to the auditory system includestereotactic radiosurgery (SRS) and fractionated stereotactic radiotherapy(FSRT) for the treatment of vestibular schwannomas (VS), and fractionatedradiotherapy (FRT) or intensity-modulated radiation therapy (IMRT) for thetreatment of head and neck malignancies. Radiation therapy for VS is

Radiosurgery and Radiotherapy for Benign Skull Base Tumors

unique because of its involvement of the inner ear and preexisting auditoryand vestibular dysfunction. Auditory and vestibular dysfunction followingRT for VS may be limited by limiting the total dose of cranial nerve VIIIirradiation and by fractionation.

Gama Knife Radiosurgery for Vestibular Schwannoma 635

Robert A. Battista

In 1951, Dr. Lars Leksell from Sweden conceived of what is now known asgamma knife radiosurgery (GKRS). Since Leksell first treated a patient whohad a vestibular schwannoma in 1967, there has been a year-to-year in-crease in the number of patients treated with the gamma knife for vestib-ular schwannoma. This article outlines the technique of GKRS anddiscusses the current results of its use to treat vestibular schwannomas.Other topics discussed include tumor control, treatment of recurrent/re-sidual and cystic vestibular schwannomas, and the results of treatmentof neurofibromatosis type 2.

Stereotactic Radiotherapy forVestibular Schwannoma 655

Patrick Sweeney, Santosh Yajnik, William Hartsell, George Bovis,and Jagannath Venkatesan

Vestibular schwannomas are benign tumors of the Schwann cells of theeighth (VIII) cranial nerve. Precision radiotherapy techniques used to man-age these tumors include stereotactic radiotherapy (SRT), which can bedelivered with either a conventional or hypofractionated regimen. The ra-dio–biologic rationale and reported clinical outcomes of patients treatedwith SRT are reviewed.

Cyberknife Radiotherapy forVestibular Schwannoma 665

Gordon T. Sakamoto, Nikolas Blevins, and Iris C. Gibbs

Stereotactic radiosurgery is a well-established treatment modality for ves-tibular schwannoma. Initial reports using single-stage radiosurgery havedemonstrated excellent tumor control rates. Many patients now elect toundergo radiosurgery given the potential for tumor control while avoidingthe morbidity associated with microsurgical resection. In attempt to im-prove hearing preservation rates of single-state radiosurgery, stagedframe-based radiotherapy using a 12-hour interfraction interval wasused at the authors’ institution and has shown a hearing preservationrate of 77% at 2 years of follow-up. With the arrival of the Cyberknife,a frameless, image-guided radiotherapy system, staged stereotactic ra-diotherapy for vestibular schwannoma became more practical. This articleoutlines the rationale and treatment protocols developed at Stanford Uni-versity (California) and reports the authors’ initial experience using the Cy-berknife to treat vestibular schwannoma.

Contentsviii

Stereotactic Radiosurgery and Stereotactic Radiotherapy in the Treatmentof Skull Base Meningiomas 677

John M. McGregor and Atom Sarker

Meningiomas are the most common nonglial brain tumors. They tend to beslow growing and benign and can reach substantial sizes before becomingsymptomatic. Complete surgical resection of intracranial meningiomas re-mains the treatment of choice. Location of a meningioma within the cranialvault may make complete surgical resection unlikely; tumors arising fromthe dura of the skull base are particularly challenging. Advances in radia-tion therapy, including stereotactic techniques, can expand the optionsfor treatment available in these situations. They may be used as adjunctsto surgery or as alternative modalities in the treatment of these complextumors.

Radiosurgery for Glomus Jugulare Tumors 689

Jonathan P. Miller, Maroun T. Semaan, Robert J. Maciunas,Douglas B. Einstein, and Cliff A. Megerian

Glomus jugulare tumors arise from adventitial chemoreceptor tissue in thejugular bulb. Although histologically benign, these tumors can be locallyaggressive because of their proximity to the lower cranial nerves and majorvascular structures. Traditional treatment involves microsurgical removalwith or without endovascular embolization, but morbidity following total re-section can result in injury to the facial and lower cranial nerves. Radiosur-gery has recently emerged as a promising alternative to older therapeuticstrategies for treatment of glomus jugulare tumors. This article reviews thelatest benefits of radiosurgery and demonstrates how this modality repre-sents an effective treatment option for glomus jugulare tumors with excel-lent tumor control and low risk for morbidity. In addition, this article willdetail the role of minimally invasive sub-total resection of glomas jugularetumors as a surgical complement to gamma knife therapy.

MicrosurgeryAfter Radiosurgery or Radiotherapy forVestibular Schwannomas 707

William H. Slattery III

Radiosurgery or radiotherapy for vestibular schwannomas has becomea common practice with a high chance for tumor control. Despite thehigh rate of tumor control, there are some tumors that cannot be con-trolled with radiation therapy. Surgical treatment after radiosurgery or ra-diotherapy may be necessary for tumors that continue to grow, or forpatients who develop brainstem compressive symptoms, disabling hem-ifacial spasm, or hydrocephalus. The House Ear Clinic (Los Angeles,California) experience with microsurgery after irradiation has demon-strated that the facial nerve is different once it has been radiated. An ir-radiated facial nerve’s regeneration potential is diminished, and therecovery from microsurgical trauma is not as robust. It is recommendedthat patients who require microsurgical excision following radiosurgeryor radiotherapy have a more conservative approach compared to non-irradiated cases.

Contents ix

Neoplastic Transformation After Radiosurgery or Radiotherapy: Risk and Realities 717

Ajay Niranjan, Douglas Kondziolka, and L. Dade Lunsford

In recent years, the use of radiosurgery or radiotherapy for benign brain tu-mors has increased significantly. Although long-term follow-up from sev-eral centers suggests that radiosurgery or radiotherapy is effective andsafe, there are particular concerns regarding development of radiation-in-duced tumors. This article reviews the use of radiosurgery and fractionatedradiation therapy with particular regard to new tumor induction and malig-nant transformation. The authors have found that the risk of radiation as-sociated tumors after radiosurgery or radiotherapy for benign braintumors is very low. All patients should be informed about the risks and con-sequences of radiation and microsurgery. The current practice standardsfor radiosurgery should not be modified because of this very low risk.

Index 731

Contentsx

Radiosurgery and Radiotherapy for Benign Skull Base Tumors

Preface

Robert A. Battista, MD

Guest Editor

Benign tumors of the base of the skull, such as vestibular schwannoma, glomus jugu-lare, and meningiomas commonly cause symptoms referable to the realm of an otolar-yngologist. Specifically, the symptoms of these types of tumors may include hearingloss, dizziness, tinnitus, facial numbness, facial paralysis, or difficulty swallowing.For this reason, otolaryngologists are the physicians most frequently involved in thecare of these patients.

Over the last two decades, radiosurgery and radiotherapy have been used withincreasing frequency to treat benign tumors of the skull base. Stereotactic radiationis now used for both primary management and secondary management of recurrentor planned residual disease. While traditional microsurgery remains the main treatmentfor skull-base tumors, The North American Skull Base Society predicts that, in the nearfuture, a greater percentage of vestibular schwannomas will be treated with radiosur-gery than by traditional microsurgery.

In order to prepare for this paradigm shift in treatment, otolaryngologists mustbecome familiar with the different types of radiation treatments, as well as the potentialcomplications associated with radiation treatment for benign skull-base tumors. It isfor this reason that this edition of Otolaryngologic Clinics of North America wascompiled.

The issue begins with the fascinating history of stereotactic radiosurgery and radio-therapy in an article by Drs. Lasak and Gorecki. The next article, by Drs. Anker andShrieve, is a thorough discussion of the biologic effects of radiation. The article byDr. Bhandare and colleagues narrows the discussion of radiation effects by focusingon radiation’s effects on the auditory and vestibular systems. Articles by Drs. Battista,Sweeney and colleagues, Sakamoto and colleagues, McGregor and Sarkar, Miller andcolleagues, are devoted to various radiation techniques to treat vestibularschwannoma, glomus tumors, and meningiomas. Because there are stereotactic radi-ation failures, the article by Dr. Slattery is devoted to the results of microsurgeryfollowing stereotactic radiation. The final article, by Dr. Niranjan, discusses the unfor-tunate—but rare—chance of malignancy following stereotactic radiation. For eachcontributing author, I would like to extend my many thanks for your efforts.

Otolaryngol Clin N Am 42 (2009) xiii–xivdoi:10.1016/j.otc.2009.05.001 oto.theclinics.com0030-6665/09/$ – see front matter ª 2009 Elsevier Inc. All rights reserved.

http://oto.theclinics.com

Prefacexiv

Finally, a note regarding terminology used throughout the book:Radiosurgery is radiation delivered with a single dose of radiation, commonly with

rigid immobilization of the head.Radiotherapy is radiation delivered in more than one fraction.Isocenter is the precise mathematical location where the radiation dose is aimed.Conformal is the close approximation of radiation delivery to tumor.Fractionated therapy is radiation delivered in multiple sessions.Marginal dose is the dose of radiation delivered to the outer edge of tumor.Tumor size, unless noted otherwise, is reported as the maximal tumor diameter of

its cerebellopontine angle component.I hope that you enjoy and learn from this issue of the Otolaryngologic Clinics of

North America.

Robert A. Battista, MDThe Ear Institute of Chicago, LLC

Department of OtolaryngologyNorthwestern University Medical School

303 East Chicago AvenueChicago, IL 60611-3008, USA

E-mail address:[email protected]

mailto:[email protected]

Radiosurgery and Radiotherapy for Benign Skull Base Tumors xi

FORTHCOMING ISSUES

Sialendoscopy and LithrotripsyMichael Fritsch, MD, Guest Editor

Technical Innovations in RhinologyRaj Sindwani, MD, Guest Editor

Cough: An Interdisciplinary ProblemKenneth W. Altman, MD, PhD, andRichard S. Irwin, MD, Guest Editors

Thyroid and Parathyroid SurgeryRalph Tufano, MD, and Sara Pai, MD,Guest Editors

RECENT ISSUES

June 2009

Surgical Management of Nasal Obstruction:Facial Plastic Surgery PerspectiveDaniel G. Becker, MD, Guest Editor

April 2009

Surgical Management of Nasal Obstruction:Rhinologic PerspectiveSamuel S. Becker, MD, Guest Editor

February 2009

PalliativeTherapy in Otolaryngology^Headand Neck SurgeryKenneth M. Grundfast, MD, FACS,and Geoffrey P. Dunn, MD, FASC,Guest Editors

RELATED INTEREST

Neuroimaging Clinics, May 2009Interventional Head and Neck ImagingDheeraj Gandhi, Guest Editor

THE CLINICS ARE NOW AVAILABLE ONLINE!

Access your subscription at:www.theclinics.com

http://www.theclinics.com

Glossary

Radiosurgery Radiation delivered with a single dose ofradiation, commonly with rigidimmobilization of the head

Radiotherapy Radiation delivered in more than onefraction

Isocenter Precise mathematical location whereradiation dose is aimed

Conformal Close approximation of radiation deliveryto tumor

Fractionated therapy Radiation delivered in multiple sessions

Marginal dose Dose of radiation delivered to the outeredge of tumor

Otolaryngol Clin N Am 42 (2009) e1doi:10.1016/j.otc.2009.05.002 oto.theclinics.com0030-6665/09/$ – see front matter ª 2009 Elsevier Inc. All rights reserved.


The Historyof StereotacticRadiosurgeryand Radiotherapy

JohnM. Lasak, MDa,b,c,d,*, John P. Gorecki, MDb,d,e

KEYWORDS

� Stereotactic radiosurgery � Stereotactic radiotherapy� Gamma knife � Linear accelerator (LINAC) � Cyberknife� Historical review

Stereotactic radiosurgery evolved from the pioneering work reported in 1908 byHorsley, a neurophysiologist and neurosurgeon, and his associate Clarke, a mathema-tician. Horsley and Clarke1 developed a tool that could localize an intracranial struc-ture in three dimensions, enabling the insertion of a needle electrode for studyinga desired locale within the monkey brain. These early investigators developed a stereo-tactic atlas of the monkey brain based on a cartesian coordinate system relative tolandmarks (the inferior orbital rim and internal auditory canal) on the monkey’s skull.2,3

Similar to today’s stereotactic head frames, their stereotactic apparatus was fixed tothe monkey’s skull. Using this apparatus, Horsley and Clarke were the first to describethe stereotactic destruction of an intracranial target using electrode electrocoagula-tion.2,3 Despite the fact that Clarke patented the use of this device in humans, itwas never applied outside the animal model.4,5

Spiegel, a neurologist and director of experimental neurology at Temple MedicalSchool in Philadelphia, began developing the first stereotactic device for humanuse. Spiegel and colleagues,6 a neurosurgeon who worked in Spiegel’s laboratoryas a medical student, reported on their human stereotactic apparatus in 1947. Theirdevice was fixed to the patient’s head in much the same way as Horsley and Clarke’ssimian model; however, intraoperative radiographs were used for localizing intracra-nial structures during the procedure. By 1952, Spiegel and Wycis7 had developed

a Department of Pediatrics, Kansas University School of Medicine-Wichita, Wichita, KS 67214, USAb Department of Surgery, Kansas University School of Medicine-Wichita, Wichita, KS 67214,USAc The Wichita Ear Clinic, 9350 East Central Avenue, Wichita, KS 67206, USAd The Jack B. Davis Gamma Knife Center, Wesley Medical Center, Wichita, KS 67214, USAe Department of Surgery, The Wichita Surgical Specialists, PA, 818 North Emporia, S-200,Wichita, KS 67214, USA* Corresponding author. The Wichita Ear Clinic, 9350 East Central Avenue, Wichita, KS 67206.E-mail address: [email protected] (J.M. Lasak).

Otolaryngol Clin N Am 42 (2009) 593–599doi:10.1016/j.otc.2009.04.003 oto.theclinics.com0030-6665/09/$ – see front matter ª 2009 Elsevier Inc. All rights reserved.



Lasak & Gorecki594

a stereotactic atlas of the human brain, and they coined their technique stereoence-phalotomy. Their methods paved the way for functional human stereotactic neurosur-gery. By this time, it was understood that sectioning the extrapyramidal system couldbe used to treat movement disorders such as Parkinson’s disease; however, openneurosurgical operative mortality was quite high at the time. Russell Meyers,8,9

a neurosurgeon at the University of Iowa and a leader in functional neurosurgery formovement disorders, reported that his open neurosurgical techniques carrieda 15.7% mortality rate. By 1958, Spiegel and colleagues10 were reporting the opera-tive mortality for stereotactic surgery for movement disorders to be 2%. Their method-ology and reduced mortality rate were met with enthusiasm by the worldwideneurosurgical community. Multiple stereotactic apparatus designs were developedin the 1950’s, and it was estimated that nearly 25,000 functional stereotactic proce-dures were done worldwide before the introduction of L-dopa therapy for movementdisorders.11,12

Lars Leksell is known as the father of stereotactic radiosurgery for his pioneeringwork applying the stereotaxic technique to radiation delivery. Leksell was born inFassberg, Sweden, in 1907, and attended medical school at the Karolinska Institute.By 1935, he began his neurosurgical training under Herbert Olivecrona, and ultimatelysucceeded him as chairman of neurosurgery in 1961. Leksell understood that much ofthe morbidity and mortality associated with neurosurgery during this time was a conse-quence of invasive procedures, and therefore, he dreamed of developing a minimallyinvasive approach to treat intracranial lesions. Building on the principles produced byHorsley and Clarke and applied by Spiegel and Wycis, Leksell13 developed his arc-centered stereotactic apparatus for intracerebral surgery in 1947. The device enabledthe precise placement of a needle or electrode into a desired location within thehuman brain. Leksell first described the concept of stereotactic radiosurgery in1951.14 The initial device used a collimated x-ray beam (gamma rays) that couldmove along the semicircular arch of his stereotactic apparatus to strike an intracranialtarget.14 Leksell14 discussed the potential of using stereotactic radiosurgery toproduce a lesion within the human brain, and noted the technology might be well-suited for functional neurosurgery. As exciting as the concept was, the effect of radi-ation on the central nervous system was poorly understood. Further research wasneeded, and Leksell14 developed a feline stereotactic radiosurgical device to deter-mine the effects on the cat brain.

During the late 1950s and early 1960s Leksell searched for an ideal form of radiantenergy convenient for clinical stereotactic radiosurgery. Physicists Liden and Larrsonof Uppsala University in Sweden were integral in the development of a system thatused a cyclotron to direct proton beams at a target.15–17 These physicists determinedin the animal model that a sharply delineated stereotactic lesion was produced withhigh-energy protons.15 In 1960, Leksell and colleagues16 performed their first humanstereotactic proton beam operation (a bilateral anterior capsulotomy) at the GustafWerner Institute in Uppsala. About the same time, Woodruff and colleagues18 atUniversity of California at Berkeley introduced a similar cyclotron-based radiosurgerysystem and began irradiating pituitary lesions. The cyclotron ultimately was deter-mined to be too cumbersome and impractical for clinical use. Leksell eventuallysettled on gamma rays as a practical compromise for stereotactic radiosurgery. Thefirst gamma unit was installed in the Sophiahemmet Hospital in Stockholm, Sweden,in 1968. The device used 179 sources of cobalt-60 distributed with collimators tocreate a sharply circumscribed disc-shaped lesion, and the device initially was in-tended for use during functional neurosurgery.19–21 The second gamma unit wasinstalled at the Karolinska Hospital in Stockholm, Sweden, in 1974 (Fig. 1). This

Fig.1. Lars Leksell and the 2nd gamma Knife in the world; 1974. Courtesy of Dan Leksell, sonof Lars Leksell; Stockholm, Sweden.


updated version was designed specifically for stereotactic radiosurgery and produceda more practical spherical radiation field.22 Plain radiographs and air encephalographytypically were used to identify the target during these early procedures.20 In 1972,Steiner and colleagues23 applied angiography to radiosurgery when treating the firstarteriovenous malformation.

Target localization improved dramatically when CT was used in conjunction with thestereotactic apparatus, as reported by Bergstom and Grietz24 in 1976. Leksell25

described the application of CT during radiosurgery and noted the technique enabledrapid and accurate target localization. With the advent of better imaging methods,solid tumors became radiosurgical targets. Leksell and Jernberg’s26 rationale fortreating vestibular schwannomas with radiosurgery is detailed in his A Note on theTreatment of Acoustic Tumours. Leksell26 discussed how Harvey Cushing referredto the cerebellopontine angle as the bloody angle, and that a noninvasive modalityto treat these tumors was needed. Furthermore, he quoted Pool,27 who statedcomplete acoustic tumor removal was ‘‘ not only one of the most exacting and labo-rious, but also one of the most dangerous and unpredictable operations in the entireneurosurgical repertoire.’’ Leksell further discussed the evolution of vestibularschwannoma surgery and the operative morbidity and mortality reported by vestibularschwannoma surgeons of the time (Pool, Olivecrona, House, and Hitselberger).27–30

He discussed that despite improving mortality rates, incomplete tumor removal wascommon and often had unsatisfactory long-term results. Leksell believed stereotacticradiosurgery was a logical approach to decrease the morbidity and mortality associ-ated with vestibular schwannoma surgery.

From 1968 to 1982, Leksell22 treated 762 patients who had stereotactic radiosurgeryusing the Gamma Knife (Elekta, Stockholm, Sweden), and over half of these patientshad arteriovenous malformations or benign tumors, including 94 vestibular schwanno-mas. By 1985, Leksell and colleagues31 reported the potential application of MRI to ra-diosurgery, pointing out the many advantages of this detailed imaging modality. TodayCT, MRI, and angiography are used commonly during radiosurgical planning.

After 4.5 years of intensive regulatory review, the first Gamma Knife unit wasinstalled in the United States at the University of Pittsburgh in 1987 under the

Lasak & Gorecki596

leadership of Lunsford, a neurosurgeon.32 Lunsford’s effort paved the way for theinstallation of Gamma Knife units throughout the United States. The first updatedmodel C Gamma Knife unit was installed in Krefeld, Germany, in 1999.33 This unitincluded sophisticated treatment planning software and an automated positioningsystem (APS) that was less cumbersome and time-consuming than the model Bmanual trunnion system. The APS offered more conformal dose plans and less radia-tion exposure to personnel.34,35 The model C Gamma Knife is the most commonlyused unit throughout world today. The latest version of the Gamma Knife, the LeksellGamma Knife Perfexion was introduced in 2006 and the first unit became operationalat the Timone University Hospital of Marseille, France. The Perfexion is reported tohave a wider treatment range, enabling the collision-free management of hard-to-reach targets in a shorter period of time.36

Although Leksell considered employing linear accelerator (LINAC) technology forstereotactic radiosurgery, technical and accessibility limitations led him to use gammarays. The first report using a LINAC- based radiotherapy system was in 1983 by Bettiand Derechinsky.37 Their early device employed several isocentric fixed radiationfields arrayed on different planes, and the patient’s head then was rotated arounda horizontal lateral axis.37 Shortly thereafter, Colombo and colleagues38 reported theirtechnique based on multiple converging arc irradiations. Hartman and colleagues re-ported a similar technique in 1985.39 It was Lutz and Winston, however, who createdthe technology to quantify and calculate dosimetry and improve the accuracy of theLINAC- based system.40,41 Most commercially available LINAC systems today arean evolution of Lutz and Winston’s original work. Since linear accelerators weremore available and less expensive, the LINAC became a popular radiosurgery system.Sophisticated and easy-to-use dosimetry planning software has been produced, andstarting in the mid-to-late 1980s, the Brigham and Women’s Hospital in Massachu-setts and the University of Florida began treating patients with their LINAC stereotacticradiotherapy systems.42,43 Today, over 150 LINAC radiotherapy systems are in useworldwide, and many different models are emerging to meet the needs of this increas-ingly popular form of radiotherapy.

Perhaps the most innovative LINAC based system developed in recent years wasproduced by Adler and colleagues, a neurosurgeon at Stanford University (Stanford,California) (Fig. 2). In 1997, Adler44 introduced the Cyberknife (Accuray Incorporated,Sunnyvale, California) at the 12th meeting of the World Society of Stereotactic Func-tional Neurosurgery. Cyberknife is a frameless stereotactic radiotherapy system thatmounts a lightweight (130 lb) 6 mV linear accelerator on a highly mobile roboticarm. This revolutionary design incorporates a real-time guidance system, which obvi-ates the need for rigid fixation as used in frame base systems. Cyberknife determinesthe location of the skull or spine in the coordinate frame of the radiation deliverysystem by comparing digitally reconstructed CT phantoms obtained from the patient’streatment planning images with real-time oblique radiographs obtained during theprocedure.45 If needed, fiducials may be implanted within the target and used duringimage registration. The dose placement accuracy has been determined to comparefavorably with frame-based systems.46,47 When the patient (target) moves, the systemdetects the change, and the robot makes appropriate adjustments to maintain accu-rate targeting. It is also possible, through sophisticated infrared tracking, to accom-modate target movement during respiration.48 This innovative system has enabledstereotactic radiotherapy to be expanded to include the head, spine, chest, abdomen,and pelvis.

The future role of stereotactic radiosurgery and radiotherapy is certain to expand.Technological and clinical advancements have occurred in every decade since the

NNMHPlaptop

Line

NNMHPlaptop

Line

NNMHPlaptop

Line

Fig. 2. John Adler and the Cyberknife. Courtesy of Stanford University, Department ofNeurosurgery; Stanford, CA.


1940s. The medical community and patients have a well-founded combination ofanticipation and expectation for noninvasive treatment methods. The potential fordecreased morbidity and improved quality of life with the use of radiosurgery andradiotherapy is being realized. The growing involvement of medical disciplines outsideof neurosurgery and radiation oncology, such as neurotologists, thoracic surgeons,general surgeons, and urologists undoubtedly will lead to fundamental shifts in theuse of radiosurgery and radiotherapy. Stereotactic radiosurgery and radiotherapywill continue to advance treatment options for patients, and the future of this evolvingfield should be interesting.

REFERENCES

1. Horsley V, Clarke RH. The structure and functions of the cerebellum examined bya new method. Brain 1908;31:45–124.

2. Clarke RH, Horsley V. On a method of investigating the deep ganglia and tracts ofthe central nervous system (cerebellum). Br Med J 1906;2:1799–800.

3. Horsley V. The function of the so-called motor area of the brain. The LinacreLecture. Br Med J 1909;2:125–32.

4. Levy R. A short history of stereotactic neurosurgery. Park Ridge (IL): AmericanAssociation of Neurological Surgeons; 1992.

5. Schaltenbrand G. Personal observations on the development of stereotaxy.Confin Neurol 1975;37:410–6.

6. Spiegel EA, Wycis HT, Marks M, et al. Stereotactic apparatus for operations onthe human brain. Science 1947;106:349–50.

Lasak & Gorecki598

7. Spiegel EA, Wycis HT. Stereoencephalotomy, part I. New York: Grune & Stratton;1952.

8. Meyers R. The modification of alternating tremors, rigidity, and festination bysurgery of the basal ganglia. Res Publ Assoc Res Nerv Ment Dis 1942;21:602–65.

9. Meyers R. Historical background and personal experiences in the surgical reliefof hyperkinesias and hypertonus. In: Fields W, editor. Pathogenesis and treatmentof parkinsonism. Springfield (IL): Thomas; 1958. p. 229–70.

10. Spiegel EA, Wycis HT, Baird HW, et al. Long-range effects of electropallido–an-trostomy in extrapyramidal and convulsive disorders. Neurology 1958;8:734–40.

11. Gildenberg PL. Principles of stereotaxis and instruments. In: DeSalles AAF,Goetshe SJ, editors. Stereotactic surgery and radiosurgery. Madison (WI):Medical Physics; 1993. p. 17–28.

12. Spiegel EA. Methodological problems in stereoencephalotomy. Confin Neurol1965;26:125–32.

13. Leksell L. A stereotaxic apparatus for intracerebral surgery. Acta Chir Scand1947;99:229–33.

14. Leksell L. The stereotactic method and radiosurgery of the brain. Acta Chir Scand1951;102:316–9.

15. Larsson B, Leksell L, Rexed B, et al. The high-energy proton beam as a neurosur-gical tool. Nature 1958;182:1222–3.

16. Leksell L, Larsson B, Andersson B, et al. Lesions in the depth of the brainproduced by a beam of high-energy protons. Acta Radiol 1960;54:251–64.

17. Leksell L, Larsson B, Rexed B. The use of high-energy protons for surgery inman. Acta Chir Scand 1963;125:1–7.

18. Woodruff KH, Lyamn JT, Lawrence JH, et al. Delayed sequelae of pituitary irradi-ation. Hum Pathol 1984;15:48–54.

19. Leksell L. Stereotaxis and radiosurgery. An operative system. Springfield (IL):Charles Thomas; 1971.

20. Larsson B, Liden K, Sarby B. Irradiation of small structures through the skull. ActaRadiol Ther Phys Biol 1974;13:512–34.

21. Wennerstrand J, Ungerstedt U. Cerebral radiosurgery II. An anatomical study ofgamma radiolesions. Acta Chir Scand 1970;136:133–7.

22. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatr 1983;46:797–803.

23. Steiner L, Leksell L, Greitz DM, et al. Stereotactic radiosurgery for cerebral arte-riovenous malformations. Acta Chir Scand 1972;138:459–64.

24. Bergstrom M, Greitz T. Stereotaxic computed tomography. Am J Roentgenol1976;127:167–70.

25. Leksell L, Jernberg B. Stereotaxis and tomography. A technical note. Acta Neuro-chir 1980;52(1–2):1–7.

26. Leksell L. A note on the treatment of acoustic tumours. Acta Chir Scand 1971;137:763–5.

27. Pool JL, Pava AA. The early diagnosis and treatment of acoustic tumors. Spring-field (IL): Thomas; 1957.

28. Olivecrona H. The surgical treatment of intracranial tumors. In: Handbuch derNeurochirurgie, vol. 4. Heidelberg (Germany): Springer; 1967.

29. House W. Management of the large acoustic tumor. In: Hamberger CA, Wersaall J,editors. Nobel symposium 10. Disorders of the skull base region. Stockholm(Sweden): Almqvist & Wiksell; 1969.


30. Hitselberger W. Management of the large acoustic tumor. In: Hamberger CA,Wersaall J, editors. Nobel symposium 10. Disorders of the skull base region.Stockholm (Sweden): Almqvist & Wiksell; 1969.

31. Leksell L, Leksell D, Schwebel J. Stereotaxis and nuclear magnetic resonance.J Neurol Neurosurg Psychiatr 1985;48:14–8.

32. Lunsford L, Flickinger J, Lindner G, et al. Stereotactic radiosurgery of the brainusing the first United States 201 cobalt-60 gamma knife. Neurosurgery 1989;24(2):151–9.

33. Horstmann GA, Schopgens H, van Eck AT, et al. First clinical experience with theautomatic positioning system and Leksell gamma knife model C. Technical note.J Neurosurg 2000;93(3):193–7.

34. Tlachacova D, Schmitt M, Novotny J, et al. A comparison of the gamma knifemodel C and the automatic positioning system with Leksell model B. J Neurosurg2005;102:25–8.

35. Horstmann GA, Van Eck AT. Gamma knife model C with the automatic positioningsystem and its impact on treating vestibular schwannomas. J Neurosurg 2002;97(5):450–5.

36. Regis J, Manabu T, Guillot C, et al. Radiosurgery of the head and neck with theworlds first fully robotized 192 Cobalt-60 source Leksell gamma knife perfectionin clinical use. Available at: www.elekta.com. Accessed November 15, 2008.

37. Betti O, Derechinsky V. Irradiation stereotaxique multifasceaux. Neurochirurgie1983;29:295–8.

38. Colombo F, Benedetti A, Pozza F, et al. External stereotactic irradiation by linearaccelerator. Neurosurgery 1985;16:154–60.

39. Hartmann G, Schlegel W, Sturm V, et al. Cerebral radiation surgery using movingfield irradiation at a linear accelerator facility. Int J Radiat Oncol Biol Phys 1985;2:1185–92.

40. Lutz W, Winston K, Maleki N. A system for stereotactic radiosurgery with a linearaccelerator. Int J Radiat Oncol Biol Phys 1988;14:373–81.

41. Winston K, Lutz W. Linear accelerator as a neurosurgical tool for stereotacticradiosurgery. Neurosurgery 1988;22:454–64.

42. Kooy H, Nedzi L, Loeffler J, et al. Treatment planning for stereotactic radiosurgeryof intracranial lesions. Int J Radiat Oncol Biol Phys 1991;21:683–93.

43. Friedman W, Bova F. The University of Florida radiosurgical system. Surg Neurol1989;32:334–42.

44. Adler JR, Chang SD, Murphy MJ, et al. The Cyberknife: a frameless roboticsystem for radiosurgery. Stereotact Funct Neurosurg 1997;69:124–8.

45. Chang SD, Adler JR. Robotics and radiosurgery. Stereotact Funct Neurosurg2001;76:204–8.

46. Murphy MJ, Cox RS. The accuracy of dose localization for an image-guidedframeless radiosurgery system. Med Phys 1996;23:2043–9.

47. Maciunas RJ, Galloway RL, Latimer JW. The application of accuracy of stereo-tactic frames. Neurosurgery 1994;35:682–95.

48. Schweikard A, Glosser G, Bodduluri M, et al. Robotic motion compensation forrespiratory movement during radiosurgery. Comput Aided Surg 2000;5(4):263–7.

http://www.elekta.com

Basic Principles ofRadiobiology Appliedto Radiosurgery andRadiotherapy of BenignSkull BaseTumors

Christopher J. Anker, MD*, Dennis C. Shrieve, MD, PhD

KEYWORDS

� Radiobiology � Radiotherapy � Radiosurgery � Benign tumors� Normal tissue tolerance

Various benign tumors arise in the base of the skull including epithelial, fibro-osseous, mesenchymal, neurogenic, and vascular varieties. Despite the slow growthof benign skull base tumors, significant clinical symptoms, and even death, mayresult from compression and dysfunction of adjacent vital structures. The mainstayof treatment of benign skull base tumors is surgical resection. Radiation may be indi-cated in nonoperable cases or in the postoperative or recurrent setting. The goal ofradiotherapy is to provide local control without a high risk of treatment-relatedtoxicity.

Radiotherapy of tumors of the skull base may involve one of several types ofionizing radiation, given as single fraction (most commonly referred to as stereotacticradiosurgery or SRS) or fractionated radiotherapy. An understanding of the biophys-ical and radiobiological principles involved in these treatments is essential to thedesign and delivery of safe and efficacious treatment. The purpose of this chapteris to discuss the basic radiobiological principles applied to radiotherapy of skullbase tumors.

TYPES OF IONIZING RADIATIONGamma Rays and X-Rays

Gamma rays and X-rays are electromagnetic radiation with energies in the range of100 to 2 billion electron volts (eV). X-rays are produced when electrons transitionfrom a higher to lower energy level, usually in the outer shell of heavy atoms and

Department of Radiation Oncology, Huntsman Cancer Hospital, University of Utah, 1950 Circleof Hope, Room 1570, Salt Lake City, UT 84112-5560, USA* Corresponding author.E-mail address: [email protected] (C.J. Anker).




Anker & Shrieve602

are thus produced outside the nucleus. X-rays may be the products of radioactivity(electron capture) or may be man made from X-ray tubes or linear accelerators, whichaccelerate electrons onto a heavy metal target producing a continuous spectrum ofphoton energies up to that of the accelerated electrons. Gamma rays are photonsemitted by radioactive nuclei (eg, Cobalt-60) and have a much narrower range of ener-gies than X-rays, 10 keV to 10 MeV. Gamma rays and X-rays are otherwise identicaland once produced are indistinguishable.

Protons

The applicability of protons to radiotherapy is based on the physical properties ofthese particles and the related characteristics of dose deposition in irradiated tissues.1

Quantitatively, the entrance dose for particle beams is low compared to photons. Anunaltered beam will deposit more than 50% of its energy over a 2- to 3-cm narrow pathat a depth in water that depends on the beam energy. This peak in energy depositionat depth is referred to as the ‘‘Bragg Peak.’’ The beam may be altered to spread theBragg peak to conform to the thickness and depth of the volume to be treated.However, the entrance dose is significantly increased in this case (Fig. 1).

X-rays, gamma rays, and protons are considered to be low linear energy transfer(LET) radiation with roughly equivalent biological effectiveness. Protons have onlya slightly higher radiobiological effectiveness (RBE) than Cobalt-60. In practice thissmall difference is taken into account when prescribing treatment by calculatingdose for protons in cobalt Gray equivalents (CGE), whether for single or multiplefractions.

BASIC PRINCIPLES OF RADIOBIOLOGYDirect Versus Indirect Effects of Radiation

When tissues are irradiated with low-LET radiation, the vast majority of photonsinteract with intracellular water molecules by stripping an electron from a hydrogenatom. This results in a fast electron and an ionized water molecule through Comptonscattering. The fast electrons interact with surrounding water molecules throughfurther ionizing events. The positively charged water molecules are extremely unstableand dissociate into an H1 ion and an OH$ free hydroxyl radical. The hydroxyl radical ishighly chemically reactive with sufficient energy to break chemical bonds in nearby

Depth Dose Distribution in Water

0

20

40

60

80

100

120

0 4 8 12 16 20

Depth (cm)

Re

la

tiv

e D

os

e (%

)

Unmodulated Proton Beam

10 MVp X rays

Spread-out Bragg Peak

Fig. 1. Depth dose curves for a 160 MeV proton beam. Both unmodulated and spread-outBragg Peak curves are shown. A 10 MVp X-ray curve is shown for comparison.

Basic Principles of Radiobiology 603

(within 2 nm) molecules. This indirect effect of radiation acting through the free radicalintermediary is responsible for 70% of the damage in irradiated tissues. The remainingdamage is caused by the direct effect resulting from fast electrons interacting with thebiologically important molecules.2

Mammalian Cell Survival Curves

Survival of cells following single doses of ionizing radiation is a probability function ofabsorbed dose measured in the unit Gray (1 Gy 5 1 J/kg absorbed dose). Mammaliancell survival curves obtained following single-dose irradiation in culture (Fig. 2) havea characteristic shape including a low-dose shoulder region followed by a steeplysloped, or more continuously bending, high-dose region. The shoulder region isa reflection of the accumulation of sublethal cellular damage at low doses with lethalityresulting from the interaction of two or more such sublethal events.3,4 One modelconsiders that DNA is the target molecule for cell killing by ionizing radiation andthat a double-strand break in the DNA is necessary and sufficient to cause cell death(defined as loss of ability to divide). Double-strand breaks may result from the passageof a single particle track or by two separate particle tracks causing two single-strandbreaks occurring closely in space and time (Fig. 3). Cellular mechanisms of single-strand break repair are efficient, and therefore these represent sublethal damage.By definition, in this model, double-strand breaks are irreparable and lethal. Sucha model is described by the linear-quadratic formula

SF 5 e�ðaD1bD2Þ

where SF is the fraction of cells surviving a dose, D, of radiation expressed in Gy.5 a isthe coefficient related to single-event cell killing, and b the coefficient related to cellkilling through the interaction of sublethal events. a/b is the ratio of the relative contri-butions of these two components to overall cell kill. The b component, related to theamount of reparable damage, is the more variable of the two coefficients among tissuetypes. Cell types and tissues may vary in the a/b, resulting in slightly differently shaped

1

0.1

0.01

0.001

0 5 10 15

Dose (Gy)

D

D2

Su

rvivin

g F

ractio

n

SF = e-(αD+βD2)

Fig. 2. Curve for mammalian cell survival as a function of dose of radiation (solid line) givenas a single fraction. The a/b is 10 Gy, a dose at which the contributions to cell killing by singleevents (aD, dashed line) and the interaction of sublethal events (bD2) are equal.

Fig. 3. Double-strand break production by single events (aD) or interaction of events (bD2).

Anker & Shrieve604

response curves (Fig. 4A). The ratio a/b is the single dose at which overall cell killing isequally attributed to these two components (see Fig. 2).

aD 5 bD2

or

D 5 a=b

Most mammalian cell survival curves are well fit to the linear-quadratic model.6,7 Cellsurvival following a single dose of radiation in vitro reflects the intrinsic radiosensitivityof a particular cell type to a particular type of radiation. There are few in vitro modelsystems for the study of the radiobiology of benign skull base tumors.

Radiobiology of Fractionated Radiotherapy

A spectrum of fractionation schedules are used to treat benign skull base tumors,ranging from single-fraction radiosurgery to fully fractionated courses of radiotherapy.For fractionated radiotherapy, each dose (fraction) produces similar biological effects

Dose

Effect

High

Low

Dose

Effect

High

Low

α/βα/β

α/βα/β

α/βα/β

α/βα/β

A B

Fig. 4. Comparison of single-dose effect curves (A) and fractionated dose-effect curves (B) forlow and high a/b tissues. The small advantage seen in the low dose region sparing low a/btissues (A) is amplified through dose fractionation (B).


(eg, cell survival) given sufficient interfraction interval (Fig. 4B). The linear quadraticformula for fractionated doses thus becomes

SF 5�

e-ad-bd2�n

where d is the dose per fraction and n is the total number of fractions.A basic principle of radiobiology and radiotherapy is that dose fractionation spares

virtually all cell and tissue types. ‘‘Sparing’’ means that for a given total dose, there willbe a lower level of biological effect associated with multiple fractions compared toa single dose. As the number of fractions increases, the total dose (n � d) requiredto achieve a certain level of biological effect also increases (Fig. 5A). The magnitudeof the sparing effect of dose fractionation varies, however, and depends on a/b. Thebiologically effective dose is represented by

BED�

Gya=b

�5 nd

�11

d

a=b

�

where BED is expressed in Gya/b to indicate that it should be used only to compareeffects in tissues with the same a/b, n is the number of fractions of dose d and ndis, therefore, the total dose (D). BED can be expressed as

BED�

Gya=b

�5 D� F

where F is a fractionation factor

F 5

�11

d

a=b

�:

F increases with increasing dose per fraction d, but this effect is dampened bya/b and may be negligible for very high a/b, where F approaches 1.

The linear-quadratic formulation provides a means of estimating the effects of dosefractionation. Other factors, such as a rapid doubling time, may be accounted for byadditional terms.8 In the context of benign skull base tumors, which typically growslowly, a time factor is not likely to be important and has not, therefore, been includedfor the purposes of this discussion.

Total Dose Required to Acheive BED

Equivalent to 20 Gy in a Single Fraction

01020304050607080

Number of Fractions

To

ta

l D

ose (G

y)

Relative BED of Fractionated vs Single-

Dose RT

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20 25 0 5 10 15 20 25

Number of Fractions

Relative B

ED

A B

Fig. 5. The effect of dose fractionation on the biological effectiveness of radiation for lowa/b (solid lines) versus high a/b (dashed lines) tissues. (A) Isoeffect curves show the increasein total dose required to maintain biological effectiveness with increasing number of frac-tions. (B) Decreasing biological effectiveness with increasing fraction number while main-taining total dose.

Anker & Shrieve606

A larger a/b indicates little contribution from the interaction of sublethal events. Alower a/b indicates a greater contribution from this type of damage, the potential formore interfraction repair, and a greater degree of sparing by fractionation than forcell types with a larger a/b (see Figs. 4 and 5). This principle is the basis for fraction-ated radiotherapy.

Malignant tumors and rapidly proliferating normal tissues (eg, skin, mucosa, andbone marrow) demonstrate a high a/b (range: 8–12) and obtain modest sparing throughdose fractionation. Other normal tissues, including those of the central nervous system(CNS) have a lower a/b (range: 2–4) and exhibit marked sparing with dose fraction-ation.4,9 This effect is demonstrated by comparing the total dose required to maintaina certain BED when various numbers of fractions are used (see Fig. 5). The magnitudeof this effect of dose fractionation is quantitatively very different for low compared tohigh a/b tissues. This forms the basis for simultaneously maintaining treatment efficacyfor tissues with high a/b, while decreasing toxicity for tissues with low a/b through dosefractionation.

Application of the Linear-Quadratic Equation to Benign Brain Tumors

Little information is available regarding the a/b of benign brain tumors. Using clinicaldata, it may be possible to estimate a/b based on isoeffective fractionation schedules.If two fractionation schedules result in an equivalent clinical effect, they may beassumed to have the same BED and the linear quadratic model may be used to calcu-late the a/b.10 Since

BED�

Gya=b

�5 D

�11

d

a=b

�

by setting BED1 5 BED2, the unknown a/b is calculated as follows

BED1 5 BED2

or

D1

�11

d1

a=b

�5 D2

�11

d2

a=b

�

or

a=b 5ðD2�d2Þ � ðD1�d1Þ

D1 � D2

This approach has been used to estimate a/b for benign meningioma, assuming,based on clinical data, that a single radiosurgery dose of 15 Gy and fractionated treat-ment to 54 Gy in 30 fractions result in equivalent local control. The resulting a/b is3.3.10

While meningiomas have been almost exclusively treated with radiosurgery or fullyfractionated radiotherapy, vestibular schwannomas have been treated with a numberof fractionation schedules (Table 1). Historically, radiosurgery doses have decreasedfrom approximately 18 Gy to 12–13 Gy without compromise of local tumor control andwith significant decrease in morbidity, including hearing loss and trigeminal and facialneuropathies.11 Equivalent local control and levels of hearing preservation have beenachieved with schedules between 1 and 32 fractions without significant trigeminal orfacial neuropathies.12–17 Using these data, the a/b may be estimated using the recip-rocal plot method of Douglas and Fowler (Fig. 6A).18 This method involves

Table 1Fractionation schedules used to treat vestibular schwannomas

AuthorReference#

TotalDose(Gy) Fractions

Calculateda/ba

BED(Gy3)

LocalControl

HearingPreservation

Flickingeret al

11 13 1 – 69 99% 6yr 70% 6yr

Poenet al

15 21 3 2.75 70 97% 2yr 77% 2yr

Williams 17 25 5 3.67 67 100% 70% 3yr

Williams 17 30 10 4.65 60 100% 100% 3yr

Wallneret al

16 45 25 2.75 72 94% 15yr –

Andrewset al

12 50 25 1.86 83 97 3yr 70 3yr

Chanet al

13 54 30 1.75 86 98% 5yr 88% 3yr

Combset al

14 57.6 32 1.46 92 93% 5yr 94% 5yr

a a/b calculated assuming isoeffectiveness with 13 Gy in one fraction.


rearrangement of the linear-quadratic equation so that inverse total dose (1/D) may beplotted against dose per fraction (d):

�1

D5

a

InðSFÞ1�

b

InðSFÞ

�d

The intercept on the abscissa provides an a/b estimate of 2.3 Gy. What is clear fromthis analysis is that a/b for vestibular schwannomas is not near the 10 Gy value assumedand measured for many malignant tumors (Fig. 6B). None of these fractionation sched-ules appear to have a clear advantage in terms of therapeutic ratio, since each result inexcellent tumor control and equivalent rates of useful hearing preservation. The

Acoustic Neuroma Local Control

00.010.020.030.040.050.060.070.080.09

-4 -2 0 2 4 6 8 10 12 14

Dose per fraction, d (Gy)

1/T

otal D

ose, 1/D

(1/G

y)

0

10

20

30

40

50

60

70

0 5 10 15 20 25 30 35

Number of Fractions

To

tal D

ose (G

y)

10

32

A B

Fig. 6. (A) Reciprocal of the total dose as a function of dose per fraction for various reportedtreatment regimens used to successfully treat vestibular schwannomas (see Table 1). Theslope is proportional to b. The intercept on the ordinate axis is proportional to a. The inter-cept on the abscissa gives a value of d equal to -a/b, which equates to 2.3 Gy. (B) Total doseversus number of fractions for vestibular schwannoma treatment regimens. Symbols repre-sent data points. Curves represent modeling for a/b of 2, 3 and 10 Gy anchored at 13 Gy ina single fraction. These data points to fit an a/b between 2 and 3.

Anker & Shrieve608

decision to fractionate or use single-dose radiosurgery for vestibular schwannomas isbased on tumor size and volume effects on normal tissue risk for brain, brain stem, andnerve.

Time Interval Required for Maximal Repair Between Fractions

Allowing for sufficient time to pass between fractions thus permitting maximal repair ofsublethal damage in normal tissues is crucial to take full advantage of the sparingeffects of dose fractionation. Information on the kinetics of repair of sublethal radiationdamage in the CNS comes from the work of Ang and colleagues.19 They found biex-ponential repair kinetics with half times of 0.7 and 3.8 hours for the fast and slowcomponents, respectively. This work indicates that even a 6- to 8-hour interval willallow accumulation of unrepaired sublethal damage and lower the tolerance dose ofspinal cord.20 This prediction has been supported by increased rates of myelopathyin patients treated for head and neck cancers with three fractions per day.21

Model Predicting Tumor Control Probability Based on Cell Survival

Tumor control probability (TCP) is a function of the likelihood of inactivating all tumorcells in a given tumor following Poisson statistics. In Poisson statistics the probabilityof a specific number of events (eg, no surviving tumor cells) occurring is governed by

PðnÞ5 ðe�xÞðxnÞ

n!

where x is the average number of events and n is the specific number of events forwhich the probability (P) is calculated. Therefore, the probability of a tumor containingno surviving cells following treatment is

Pð0Þ5 ðe�NÞðx0Þ

0!5 e�N

where N is the average number of tumor cells remaining per tumor, determined by thepretreatment number of cells per tumor (N0) and the surviving fraction (SF). Assumingevery tumor cell must be killed to control a tumor, the probability of tumor control isgiven by

TCP 5 e�SF�N0

This model leads to a sigmoid dose response curve for TCP (Fig. 7). Radiosensitivityis not equivalent to radio responsiveness. Some CNS tumors are very radioresponsivebut inevitably recur (eg, CNS lymphoma), while others may show little or no radio-graphic evidence of response but are well controlled by modest radiation doses (eg,meningioma and vestibular schwannoma).

Model Predicting Normal Tissue Complications

The normal tissues of particular interest in the treatment of benign skull base tumorsare the optic apparatus and other cranial nerves, the spinal cord and brainstem, andbrain parenchyma. The effects on the vasculature within normal and tumor tissue arealso of interest.

The probability of normal tissue complication (NTCP) following radiation is, liketumor control probability (TCP), a function of dose and dose per fraction, the tissueat risk (radiosensitivity) and the volume irradiated. NTCP has been shown to be wellrepresented by the model

00.10.20.30.40.50.60.70.80.9

1

0 5 10 15 20 25 30

Dose (Gy)

TC

P/N

TC

P

TCP

NTCP

UncomplicatedCures

00.10.20.30.40.50.60.70.80.9

1

0 20 40 60 80 100 120

Dose (Gy)

TC

P/N

TC

P

TCP

NTCP

UncomplicatedCures

A B

Fig. 7. Curves comparing the probability of tumor control (TCP) with the probability ofa normal tissue complication (NTCP). (A) The curves are positioned close to one another.Normal tissue complications may be avoided only by minimizing the dose to the criticalnormal structure. Such a situation may occur when a normal structure, such as optic nerve,lies adjacent to a benign tumor being treated with single-dose radiosurgery. (B) Dose frac-tionation separates the TCP and NTCP curves allowing for a higher probability of tumorcontrol without significant risk of normal tissue complication. The Uncomplicated Curecurve is TCP�NTCP.


NTCP 5 1� expR

where R is the variable related to dose and volume

R 5 ��d=d0

�k

with d being the administered dose, d0 determining the slope of the NTCP versus dosecurve and k being a constant accounting for volume effects.22,23 This model leads toa sigmoid-shaped curve similar to that obtained for tumor cure (see Fig. 7). Curvesfor a wide variety of normal tissue endpoints have been generated. Although eachhas a similar shape, the relative placement of these curves along the dose axis maybe quite different. In clinical radiotherapy, the relative positions of the curves for tumorcure and normal tissue complication defines what is known as the therapeutic ratio.The therapeutic ratio may be calculated as

Probability of Tumor Cure

Probability of Complication

An ideal therapeutic ratio would be described by curves that allow 100% tumor curewithout appreciable probability of normal tissue complication. The opposite extremewould occur when a tumor requiring high-dose radiation for cure was located withina critical normal structure with a low tolerance to radiation. For the situation wherethe a/b for a tumor is higher than that for critical normal tissue, dose fractionationwill always serve to separate the TCP and NTCP curves and increase the therapeuticratio (see Fig. 7). Although in general the benefit from fractionation for benign skullbase tumors is less than for malignant tumors that possess a higher a/b ratio, multipletreatments may still be essential for their sparing properties in various instances, suchas those involving the optic apparatus.

NORMALTISSUE TOLERANCES

For a particular tissue, the tolerance dose is a function of the chosen toxicityendpoint, volume irradiated, total dose, dose per fraction, and an acceptable risk

Anker & Shrieve610

level.24–27 One method to express tolerance doses is by estimation of the D5/5, or thedose expected to produce complications in 5% of patients within 5 years of treat-ment.26 This concept may be useful for effects such as necrosis or pituitary dysfunc-tion, but it is not useful for effects such as optic neuropathy or paralysis, where 5%could be considered an unacceptably high risk. When planning treatment for benigntumors, a dose regimen thought to be safe would certainly be preferable. Ideally, thechosen dose and fractionation scheme would also be effective in tumor control.10,28

Models have been developed to help determine NTCP, and will be described belowalong with other time-dose-volume data reported in the literature. Along with thesemodels, the formula for BED can be used to compare regimens of varying total dosesand dose per fraction in a particular tissue. The equation may also be manipulated todetermine isoeffective total doses (D) associated with different doses per fraction (d)

D1=D2 5 ða=b1d2Þ=ða=b1d1Þ

From the data which follow, Tables 2 and 3 have been constructed with recom-mended dose limits for use in clinical practice with fractionated and single-doseradiotherapy.

Table 2Recommendednormal structure dose constraints to avoid complicationsusing standard fractionation(1.8 Gy/d)

StructureStandard FractionationDose Constraint Reference#

Brain %60 Gy Marks et al24 Sheline et al29

Brainstem surface %64 Gy Debus et al31

Brainstem center %54 GyBrainstem parenchyma <1.0 cc R60 Gy

Spinal Cord %45–50 Gy Wara et al39 Rades et al40

Marcus and Million41

Schultheiss et al42

Pituitary Dmin %50 Gy & Dmax %70 Gy Pai et al47

Hypothalamus Dmax %50 Gy Pai et al47

Optic Chiasm/Nerves %54–59 Gy Goldsmith et al28 Parsonset al51 Urie et al57

Cranial Nerves III, IV, VI %60 Gy recommended,but MTD unknown

Selch et al58 Uy et al59

Morita et al60

Cranial Nerve V(Trigeminal)

MTD unknown Andrews et al12 Chan et al13

Combs et al14 Williams17

Uy et al59 Morita et al60

Cranial Nerve VII (Facial) MTD unknown Andrews et al12 Chan et al13

Combs et al14 Williams17

Merchant et al73 Millionand Cassisi88

Cochlea 75% Volume <45 Gy Merchant et al73

Cranial Nerves IX-XII MTD unknown Johnston et al83 Zhang et al84

Dose constraint is a maximum unless otherwise specified.Abbreviations: Dmin, minimum dose received by entire structure; Dmax, maximum dose to any

part of a structure.

Table 3Recommended normal structure dose constraints to avoid complications using a single fraction

Structure Single Fraction Dose Constraint Reference#Intraparenchymal brain

lesion %2 cm%24 Gy (MTD unknown) Shaw et al30

Brain lesion >2 cm & %3 cm %18 GyBrain lesion >3 cm & %4 cm %15 Gy

Brainstem %16 Gy if prior WBRT%20 Gy if no prior WBRT

Fuchs et al32 Yen et al33

Hussain et al34 Fuenteset al35 Huang et al36

Spinal Cord %10 Gy to 10% cord volume(defined as 6 mm above &below target)

MTD unknown

Ryu et al45

Pituitary %15 Gy (mean) Jezkova et al48 Vladyka et al49

Optic Chiasm/Nerves %8–10 Gy Tishler et al52 Leber et al53

Stafford et al 54

Cranial Nerves III, IV, VI MTD unknown Leber et al53 Kuo et al61

Cranial Nerve V (Trigeminal) %12.5–13 Gy Flickinger et al11 Flickingeret al65 Flickinger et al66

Beegle et al67 Hasegawaet al 68 Lunsford et al69

Meijer et al70

Cranial Nerve VII (Facial) %12.5–15 Gy Beegle et al67 Hasegawaet al68

Cochlea %3.7 Gy (mean) Rowe et al80

Cranial Nerves IX–XII MTD unknown Martin et al85 Flickinger et al86

Dose constraint is a maximum unless otherwise specified.Abbreviation: WBRT 5 whole brain radiation therapy.


Brain Parenchyma

Sheline and colleagues reviewed the literature and found 80 cases of cerebral necrosisoccurring 4 months to 7.5 years following cranial irradiation. From this, the authorswere able to derive a model for predicting the risk of brain necrosis as a function oftotal dose, number of fractions, and treatment time.29 They defined the neuret, similarto BED, as

Neuret 5 D� N�0:44 � T0:06

where D is the total dose in cGy, N the number of fractions, and T the overall time indays. The N exponent reflects the amount of sublethal damage repair that occurswith fractionation while the T exponent reflects recovery from repopulation of neuralcells. Repair of sublethal damage is the primary recovery mechanism resulting ina strong dependence on N, a surrogate for fraction size. The relatively weak depen-dence on T reflects the near negligible effect of repopulation on neural recovery.Based on an approximate suggested threshold of 1000 neuret, necrosis may beavoided with a limit of 35 Gy in 10 fractions, 60 Gy in 35 fractions, and 76 Gy in 60 frac-tions for treatments given 5 d/wk. These data may also be well fit to the linear-quadratic model using an a/b of 2.0 Gy without a time factor. Marks and colleaguessupported daily dose and total dose as the most significant factors in determining

Anker & Shrieve612

the risk of radionecrosis. With standard fractionated whole-brain radiotherapy, theyfound no incidence of brain necrosis in 51 subjects receiving total doses less than57.6 Gy. The incidence increased to 3.3% and 17.8% for total doses of 57.6 to 64.8Gy and 64.81 to 75.6 Gy, respectively. They recommended a threshold dose biolog-ically equivalent to 54 Gy in 30 fractions, and thus treating benign tumors with thisregimen carries minimal risk of necrosis of brain parenchyma.24

A study undertaken by the Radiation Therapy and Oncology Group (RTOG) exam-ined the maximum tolerated dose (MTD) of single-fraction radiosurgery as a functionof irradiated volume30 (see Table 3). All tumors were recurrent following previousradiotherapy. The single fraction MTD was 15 Gy for lesions with a diameter of 4cm, the maximal size evaluated, but the MTD was not reached for tumors smallerthan 4.2 cc (<2 cm diameter) at 24 Gy. Larger lesions require a dose reduction,because as tumor volume increases, so does the volume of surrounding normal tissuethat is irradiated leading to increased toxicity.

Brainstem

Debus and colleagues31 in their analysis of 367 subjects treated for base of skulltumors with a photon/proton combination using standard fractionation, found toxicity-free survival at 10 years was 96% if less than 1.0 cc of brainstem received greater thanor equal to 60 CGE, but it decreased sharply to 79% above this volume. Additionaldose constraints used by the authors included a maximum dose of 64 CGE to thebrainstem surface and 54 CGE to the center.

In regards to brainstem radiosurgical tolerance, a total of 38 subjects were treated intwo separate reports for brainstem gliomas to an overall mean marginal dose of 12.4Gy (range, 9–20 Gy), with only one transient complication at 13 Gy.32,33 Subjectsreceiving radiosurgery for brainstem metastases may additionally receive a courseof whole-brain radiation confounding the true brainstem single-dose tolerance.Hussain and colleagues reported one case of hemiparesis following a marginaldose of 18 Gy preceded by 30 Gy in ten fractions of whole-brain radiotherapy. Theother 22 subjects were without sequelae after receiving a median tumor margindose of 16 Gy (range, 14–23 Gy).34 In two other series, both of which includedmany subjects who also received whole brain radiation, no permanent sequelaewere seen in 28 subjects treated to a mean dose of 19.6 Gy (range, 11–30 Gy)35 orin 26 subjects receiving a median dose of 16 Gy (range, 12–20 Gy).36 Metastases ineach of these two series had an average diameter of 17 mm (range, 10–36 mm).Despite the lack of volume effect in these studies, a report from the University ofCalifornia, San Francisco (UCSF) found a higher complication rate with lesions greaterthan or equal to 1 cc in size,37 which corresponds to a diameter of 12.5 mm assuminga spherical shape.

Spinal Cord

The neuret model has been applied to the spinal cord in multiple studies,38,39 alsoshowing the importance of fractionation for this structure. Wara and colleagues39

described the time-dose relationship for radiation induced spinal cord injury. Theyused an effect single dose (ED), where

ED 5 DðcGyÞ � N�0:377 � T�0:058

An ED of about 1000 rets was estimated to result in a 1% incidence of myelopathyfor the thoracic cord, although the risk may actually be much lower. Commonly usedregimens including 2000 cGy in five fractions, 3000 cGy in 10 fractions, and 5000 cGy


in 25 fractions were described as safe. Similar regimens retrospectively evaluated byRades and colleagues40 for metastatic spinal cord compression revealed no latetoxicity. Data from Marcus and Million suggest that 5000 cGy in conventional fraction-ation would yield an incidence of myelopathy less than 0.2%, as the only two of 1112total subjects who experienced complications received between 45 and 50 Gy.41

These data led Schultheiss and colleagues42 to suggest that the commonly acceptedradiation limit to the spinal cord of 45 Gy in 25 fractions is conservative and it may berelaxed if treating to a higher dose could improve tumor control. They calculated a 5%NTCP of 57 to 61 Gy in the absence of chemotherapy. In regards to the anatomic levelirradiated, early reports that proposed the thoracic spinal cord is more radiosensitivethan the cervical cord43,44 have not since been substantiated.39,42

In the first report of partial volume spinal cord tolerance to SRS, Ryu andcolleagues45 found that the partial volume tolerance of the human spinal cord is atleast 10 Gy to 10% of the circumferential spinal cord volume defined as 6 mm aboveand below the target. For a total of 230 metastatic lesions treated in 177 subjects, onlyone case of myelopathy was reported. Although rat and primate models show a defi-nite volume effect on spinal cord tolerance to radiation,27,46 the partial volume toler-ance in the human spinal cord is unknown.42

Pituitary

In a retrospective review of 107 patients treated for base of skull tumors, Pai andcolleagues47 found that a minimum pituitary dose (Dmin) greater than or equal to50CGE was associated with a higher rate of hypoprolactinemia (p50.045), hypothy-roidism, and hypogonadism. Significantly increased rates of hypogonadism and hypo-adrenalism were noted for maximum pituitary doses (Dmax) greater than or equal to 70CGE. Dmax greater than or equal to 50 CGE to the hypothalamus was associated withhigher rates of endocrine dysfunction. Deficits could occur after many years, with the10-year rates for hyperprolactinema, hypothryroidism, hypogonadism, and hypoa-drenalism at 84%, 63%, 36%, and 28%.

In two series of subjects treated with single fraction radiotherapy, no T4 or sexhormone deficiency occurred with a median pituitary dose less than or equal to 15Gy and no cortisol deficiency developed with a median dose less than or equal to18 Gy.48,49 In other studies margin doses up to 40 Gy caused no complications.50

Optic Apparatus

Attempts to model isoeffective dose regimens for the risk of optic neuropathyfollowing fractionated radiotherapy have lead to a model published by Goldsmithand colleagues.28 They defined the optic ret as

Optic ret 5 D� N�0:53

but found that the linear-quadratic model did not fit the data well. They identified lessthan or equal to 890 optic ret as safe to avoid optic neuropathy, corresponding tocommonly used fractionation schedules of 5400 cGy in 30 fractions, 3750 cGy in 15fractions and 3000 cGy in 10 fractions. This model emphasizes fraction size, a crucialfactor in determining optic nerve tolerance to radiotherapy.51 Although not based onsingle fraction data, the optic ret model predicts that a single fraction of 8.9 Gy wouldbe safe to the optic apparatus. This coincides very well with single fraction tolerancedoses proposed in the literature, which range from 8 to 10 Gy.52–54 In these reports themajority of optic neuropathies developed in under 2 years, but in some cases they pre-sented over 3 years later.54 It has been noted that small volumes of optic nerve or

Anker & Shrieve614

chiasm may tolerate higher doses of radiation,54–56 but clear dose-volume guidelinesare unavailable.

For fractionated radiation to the optic apparatus, Parsons and colleagues51 foundno injuries following doses less than 59 Gy in 215 optic nerves treated in 131 subjects.Fraction size was more important than total dose among nerves that received greaterthan or equal to 60 Gy, as the 15-year actuarial risk of optic neuropathy was 11% whenadministered in fraction sizes less than 1.9 Gy versus 47% when given at 1.9 Gy orgreater. In a series of 20 subjects treated at the University of Michigan, all moderateor severe optic nerve complications occurred at max doses greater than or equal to64 Gy, and the severe chiasm complication involved a max dose of 59.5 Gy. However,two mild complications occurred at 48 and 57 Gy, underscoring the fact that stayingwithin the perceived dose tolerance does not eliminate the chance of complications.55

Extraocular Nerves (Cranial Nerves III, IV & VI)

Empiric equations such as neuret defining safe fractionation and dose do not seem toextrapolate to other cranial nerves. The tolerance of cranial nerves III, IV and VIappears to be substantially higher than for other cranial nerves. Urie and colleagues57

analyzed dose-response data for cranial neuropathies after mixed photon/protonradiotherapy. They found no neuropathies for doses less than 59.3 CGE among 594cranial nerves and nuclei examined, and estimated a risk of 5% at 70 CGE. Recentreports of cavernous sinus doses up to 60 Gy using stereotactic radiotherapy58 andintensity modulated radiotherapy (IMRT)59 described no cranial neuropathies.

Regarding radiosurgery, data discussing complications involving these cranialnerves are rare. No clear dose tolerance was apparent in 150 patients treated byTishler and colleagues52 and Morita and colleagues60 receiving up to 40 Gy in areassuch as the cavernous sinus. Leber and colleagues53 studied 210 nerves among 50subjects who received single doses of up to 30 Gy to the cavernous sinus with nosubject developing neuropathy. Although multiple series of benign cavernous sinustumors treated with radiosurgery report transient or permanent extraocular nervecomplication rates at approximately 2%,61,62 the majority of series find no harm tothese nerves.63,64

Trigeminal (Cranial Nerve V)

Patients with vestibular schwannoma treated with fractionated radiotherapy have hadtrigeminal preservation rates above 93%,12–17 and there are multiple reports of frac-tionated regimens to cavernous sinus tumors with no discernable trigeminalcomplications.58,59

Using CT treatment planning and a marginal vestibular schannoma dose of 17 Gy,Flickinger and colleagues65 noted a facial numbness complication rate of 33%following radiosurgery. With the use of MRI planning and a marginal dose prescriptionless than or equal to 13 Gy, the trigeminal complication rate has decreased to under5% in the treatment of various skull base lesions.11,64,66–69 Aside from one retrospec-tive series showing better preservation of the trigeminal nerve for those receivingfractionated treatment versus SRS for a vestibular schwannoma (98% vs 92% pres-ervation, P 5 .048),70 there is no clear evidence that fractionation spares the trigeminalnerve. Beegle and colleagues67 reviewed subjects treated for schwannoma andfound, on multivariate analysis, that prior tumor growth and treatment volume weresignificantly associated with facial numbness. Dose was the most important factor,with the risk of cranial neuropathy increasing by a factor of six with each 2.5 Gyincrease over 12.50 Gy.


Facial Nerve (Cranial Nerve VII)

Facial nerve injuries are rarer than trigeminal nerve complications, which is partlyattributed to the fact that the trigeminal is a somatosensory nerve. In standard fraction-ation head and neck cases, the facial nerve routinely receives over 70 Gy withoutcomplication.71 Multiple fraction regimens treating cases of vestibular schwannomahave resulted in facial nerve preservation rates of 94% to 100%.12,14,17,72

With modern stereotactic radiosurgery planning techniques employing MRI anda margin dose less than or equal to 13 Gy, the risk of permanent facial weaknesshas been reduced to less than 1% in most series.11,67–69 This is a clear improvementover older series showing complication rates of approximately 30% for patients whoreceived marginal doses as high as 20 Gy using CT planning.65 As with the trigeminalnerve, Beegle and colleagues found the facial nerve to have an approximate six-foldincrease in complication risk for every 2.5 Gy above 12.5 Gy prescribed to the tumormargin. Only marginal dose correlated with facial weakness in a review of 190 subjectsdescribed by Flickinger and colleagues66 involving modern radiation techniques, withno subject experiencing facial weakness with doses less than 15 Gy.

Cranial Nerve VIII

Hearing loss may follow radiation to benign skull base tumors, and it is of particularconcern in the treatment of vestibular schwannomas. In general, high-frequency rangehearing is the most sensitive to radiation damage.73,74 Despite the effects of radiation,hearing preservation still appears superior to observation75 and microsurgical tech-niques.76 For these tumors fractionated radiotherapy has resulted in preservation ofuseful hearing from 68% to 94%.12–15 Hearing preservation for patients treated withradiosurgery ranges from 40% to 75%.11,66,77,78 With the marginal dose prescriptiondecreases over time from 16–20 Gy79 down to 12 Gy,11,77 hearing loss incidenceshave decreased but may still potentially occur in 40% of patients.11,77 As a group,patients with neurofibromatosis type 2 have less favorable hearing preservation ratesranging from 33% to 67%.14,80 As there has been no randomized trial comparing frac-tionated regimens versus SRS, there is no clear evidence of the superiority of onetreatment over another in terms of tumor control and toxicity.

For subjects treated with radiosurgery, Massager and colleagues81 found that radi-ation dose to the cochlea and hearing outcome were very strongly correlated, with theaverage cochlear dose at 3.7 Gy for those with hearing preservation versus 5.29 Gy forthose with hearing loss. Other volumetric and dosimetric parameters found to portendhearing loss include the intracanicular volume and the integrated dose delivered to theintracanicular tumor volume.79 A recent study of subjects treated with fractionatedradiation for schwannomas identified cochlear radiation dose as the only prognosticvariable regarding hearing deterioration. They found if greater than or equal to73.3% of the cochlea volume received greater than or equal to 45 Gy, there was amedian hearing loss of 25 dB compared to only 10 dB below this volume.72 The hypo-thesis of cochlear sensitivity is supported by other fractionated series, with variousinner ear thresholds for sensorineural hearing loss ranging from 32 to 65 Gy.73,74

Radiation also has an apparent effect on the vestibular portion of cranial nerve VIII,although it is less often reported. In the irradiation of 125 subjects for vestibularschwannoma by Andrews and colleagues,12 the most common post-treatmentsymptom was a gait ataxia that occurred 4 to 6 months after treatment with eithersingle or multiple-fractionated radiation. If found objectively on examination, thepresumed cause was vestibular dysfunction or hydrocephalus and for one SRSsubject these symptoms abated after placement of a shunt. Post-treatment

Anker & Shrieve616

symptoms of vertigo or gait ataxia persisted in less than 5% of subjects regardless offractionation scheme.

Cranial Nerves IX to XII

Damage to the ninth, tenth, eleventh and twelfth cranial nerves have been only rarelyreported, usually in the context of radiation for a head and neck carcinoma.82,83 Ina report by Kong and colleagues82 of long-term survivors of nasopharyngeal carci-noma, they found increasing incidence of posterior cranial nerve palsy over timewith 37.3% subjects exhibiting a deficit at 20-year follow-up. Older treatment tech-niques, soft tissue fibrosis potentially compressing nerves, and radiation dose wereall found on multivariate analysis to be risk factors.

Lower cranial nerves appear to have a high tolerance to radiosurgery. In multipleseries of jugular foramen tumors involving these nerves, no new complications werenoted following SRS.84,85

SUMMARY

Various types of ionizing radiation may be used in the treatment of base of skulltumors, which all cause DNA damage by way of direct and indirect pathways. Tumorcontrol probability and normal tissue complications may be estimated from the linear-quadratic formula, which is dependent on the alpha and beta values of a tissue relatedto single and multiple event killing, respectively. CNS complications may also be pre-dicted with dose-fractionation models of tolerance, which emphasize the importanceof the number of fractions, or fraction size, over the insignificant effect of overall treat-ment time. Dose fractionation has an established history of safe and efficacious treat-ment of benign skull base and brain tumors, helping to maximize the therapeutic ratio.Single-dose radiosurgery may be preferable for smaller tumor volumes when normaltissue tolerance is respected. Fractionated radiotherapy should be considered forlarger volumes and when the use of high doses per fraction carries an unacceptablerisk for normal tissue complication.

A summary of accepted dose tolerances using standard 180 cGy/fraction are listedin Table 2, and single-fraction doses in Table 3. Equivalent biologically effective dosesmay be estimated from BED calculations based on the linear quadratic formula. Indetermining the dose constraints of a particular structure for a specific patient, it isimportant not to take specified dose limits as an absolute guide below which radiationdelivery is completely safe. Other variables deserving consideration include hostfactors such as age, diabetes, or hypertension, and prior or concomitant therapyincluding surgery, chemotherapy, or radiation. Based on experimental and retrospec-tive clinical data, compelling evidence exists that long-term repair of radiation damageoccurs in the CNS, in particular the spinal cord and optic apparatus.19,86–88 However,caution must be exercised when extrapolating from this information.

The best approach to avoiding radiation-induced CNS toxicity is always to minimizethe dose and volume of irradiated tissue. Modern treatment planning techniques,including the use of CT and MRI and methods to attain accurate and precise patientimmobilization, play important roles to minimize the volume treated and to accuratelymeasure doses to normal structures. An essential element in the optimization of treat-ment planning is dose-volume histogram analysis. Conformal treatment delivery tech-niques have reduced the total dose and dose per fraction received by a normalstructure, as well as the volume irradiated. A regimen that maximizes the probabilityof an uncomplicated cure is optimal.


REFERENCES

1. Munzenrider JE, Crowell C. Charged particles. In: Mauch PM, Loeffler JS, editors.Radiation oncology: biology and technology. Philadelphia: W.B. SaundersCompany; 1994. p. 34–55.

2. Hall E, Giaccia AJ. Radiobiology for the radiologist. 6th edition. Philadelphia: Lip-pincott Williams & Wilkins; 2005.

3. Elkind MM, Sutton H. X-ray damage and recovery in mammalian cells in culture.Nature 1959;184:1293–5.

4. Withers H. Biologic basis for altered fractionation schemes. Cancer 1985;55(9):2086–95.

5. McBride WH, Withers HR. Biological basis of radiation therapy. In: Perez CA,Brady LW, Halperin EC, et al, editors. Principles and practice of radiationoncology. 4th edition. Philadelphia: Lippincott, Williams and Wilkins; 2004. p.96–136.

6. Taghian A, Suit H, Pardo F, et al. In vitro intrinsic radiation sensitivity of glioblas-toma multiforme. Int J Radiat Oncol Biol Phys 1992;23:55–62.

7. Fertil B, Malaise EP. Intrinsic radiosensitivity of human cell lines correlated withradioresponsiveness of human tumors. Int J Radiat Oncol Biol Phys 1985;11:1699–707.

8. Travis EL, Tucker SL. Isoeffect models and fractionated radiation therapy. Int JRadiat Oncol Biol Phys 1987;13:283–7.

9. Thames HD, Withers HR, Peters LJ, et al. Changes in early and late radiationresponses with altered dose fractionation: Implications for dose-survival relation-ships. Int J Radiat Oncol Biol Phys 1982;8:219–26.

10. Shrieve DC, Hazard L, Boucher K, et al. Dose fractionation in stereotactic radio-therapy for parasellar meningiomas: radiobiological considerations of efficacyand optic nerve tolerance. J Neurosurg 2004;101:390–5.

11. Flickinger JC, Kondziolka D, Niranjan A, et al. Acoustic neuroma radiosurgerywith marginal doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys 2004;60:225–30.

12. Andrews DW, Suarez O, Goldman HW, et al. Stereotactic radiosurgery and frac-tionated stereotactic radiotherapy for the treatment of acoustic schwannomas:Comparative observations of 125 patients treated at one institution. Int J RadiatOncol Biol Phys 2001;50:1265–78.

13. Chan AW, Black PM, Ojemann RG, et al. Stereotactic radiotherapy for vestibularschwannoma: favorable outcome with minimal toxicity. Neurosurgery 2005;57:60–70.

14. Combs SE, Volk S, Schulz-Ertner D, et al. Management of acoustic neuromas withfractionated stereotactic radiotherapy (FSRT): long-term results in 106 patientstreated with a single institution. Int J Radiat Oncol Biol Phys 2005;63:75–81.

15. Poen JC, Golby AJ, Forster KM, et al. Fractionated stereotactic radiosurgery andpreservation of hearing in patients with vestibular schwannoma: a preliminaryreport. Neurosurgery 1999;45:1299–305.

16. Wallner KE, Sheline GE, Pitts LH, et al. Efficacy of irradiation for incompletelyexcised acoustic neurilemomas. J Neurosurg 1987;67:858–63.

17. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas. IntJ Radiat Oncol Biol Phys 2002;54(2):500–4.

18. Douglas BG, Fowler JF. The effect of multiple small doses of x rays on skin reac-tions in the mouse and a basic interpretation. Radiat Res 1976;66(2):401–26.

19. Ang KK, Price RE, Stephens LC, et al. The tolerance of primate spinal cord to re-irradiation. Int J Radiat Oncol Biol Phys 1993;25:459–64.

Anker & Shrieve618

20. Ang KK, Jiang GL, Guttenberger R, et al. Impact of spinal cord repair kinetics onthe practice of altered fractionation schedules. Radiother Oncol 1992;25:287–94.

21. van den Bogaert W, van der Schueren E, Horiot JC, et al. Early results of theEORTC randomized clinical trial on multiple fractions per day (MFD) and misoni-dazole in advanced head and neck cancer. Int J Radiat Oncol Biol Phys 1986;12(4):587–91.

22. Alber N, Nusslin F. A representation of an NTCP function for local complicationmechanisms. Phys Med Biol 2001;46:439–47.

23. Lyman JT, Wolbarst AB. Optimization of radiation therapy. III. A method of assess-ing complication probabilities from dose-volume histograms. Int J Radiat OncolBiol Phys 1987;13:103–9.

24. Marks JE, Baglan RJ, Prassad SC, et al. Cerebral radionecrosis: Incidence andrisk in relation to dose, time, fractionation and volume. Int J Radiat Oncol BiolPhys 1981;7:243–52.

25. Marks LB, Spencer DP. The influence of volume on the tolerance of the brain toradiosurgery. J Neurosurg 1991;75:177–80.

26. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irra-diation. Int J Radiat Oncol Biol Phys 1991;21(1):109–22.

27. Ang KK. Radiation injury to the central nervous system: Clinical features andprevention. In: Meyer JL, editor. Radiation Injury: Advances in Managementand Prevention. Vol 32. Basil: Karger; 1999. p. 145–54.

28. Goldsmith BJ, Rosenthal SA, Wara WM, et al. Optic neuropathy after irradiation ofmeningioma. Radiology 1992;185:71–6.

29. Sheline GE, Wara WM, Smith V. Therapeutic irradiation and brain injury. Int JRadiat Oncol Biol Phys 1980;6(9):1215–28.

30. Shaw E, Scott C, Souhami L, et al. Single dose radiosurgical treatment of recur-rent previously irradiated primary brain tumors and brain metastases: Final reportof RTOG protocol 90-05. Int J Radiat Oncol Biol Phys 2000;47:291–8.

31. Debus J, Hug EB, Liebsch NJ, et al. Brainstem tolerance to conformal radio-therapy of skull base tumors. Int J Radiat Oncol Biol Phys 1997;39(5):967–75.

32. Fuchs I, Kreil W, Sutter B, et al. Gamma Knife radiosurgery of brainstem gliomas.Acta Neurochir Suppl 2002;84:85–90.

33. Yen CP, Sheehan J, Steiner M, et al. Gamma knife surgery for focal brainstemgliomas. J Neurosurg 2007;106(1):8–17.

34. Hussain A, Brown PD, Stafford SL, et al. Stereotactic radiosurgery for brainstemmetastases: Survival, tumor control, and patient outcomes. Int J Radiat Oncol BiolPhys 2007;67(2):521–4.

35. Fuentes S, Delsanti C, Metellus P, et al. Brainstem metastases: managementusing gamma knife radiosurgery. Neurosurgery 2006;58(1):37–42.

36. Huang CF, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for brain-stem metastases. J Neurosurg 1999;91(4):563–8.

37. Kased N, Huang K, Nakamura JL, et al. Gamma knife radiosurgery for brainstemmetastases: the UCSF experience. J Neurooncol 2008;86(2):195–205.

38. van der Kogel AJ. Central nervous system radiation injury in small animals. In:Gutin PH, Leibel SA, Sheline GE, editors. Radiation injury in the nervous system.New York: Raven Press, Ltd.; 1991. p. 91–111.

39. Wara WM, Phillips TL, Sheline GE, et al. Radiation tolerance of the spinal cord.Cancer 1975;35:1558–62.

40. Rades D, Stalpers LJ, Veninga T, et al. Evaluation of five radiation schedules andprognostic factors for metastatic spinal cord compression. J Clin Oncol 2005;23(15):3366–75.


41. Marcus RB Jr, Million RR. The incidence of myelitis after irradiation of the cervicalspinal cord. Int J Radiat Oncol Biol Phys 1990;19(1):3–8.

42. Schultheiss TE, Kun LE, Ang KK, et al. Radiation response of the central nervoussystem. Int J Radiat Oncol Biol Phys 1995;31(5):1093–112.

43. Glanzmann C, Aberle HG, Horst W. The risk of chronic progressive radiationmyelopathy. Strahlentherapie 1976;152(4):363–72.

44. Kramer S. Radiation effect and tolerance of the central nervous system. FrontRadiat Ther Oncol 1972;6:332–45.

45. Ryu S, Jin JY, Jin R, et al. Partial volume tolerance of the spinal cord and compli-cations of single-dose radiosurgery. Cancer 2007;109(3):628–36.

46. Hopewell JW, Morris JH, Dixon-Brown A. The influence of field size on the latetolerance of the rat spinal cord to single doses of X-rays. Br J Radiol 1987;60:1099–108.

47. Pai HH, Thornton A, Katznelson L, et al. Hypothalamic/pituitary function followinghigh-dose conformal radiotherapy to the base of skull: demonstration of a dose-effect relationship using dose-volume histogram analysis. Int J Radiat Oncol BiolPhys 2001;49(4):1079–92.

48. Jezkova J, Marek J, Hana V, et al. Gamma knife radiosurgery for acromegaly–long-term experience. Clin Endocrinol (Oxf) 2006;64(5):588–95.

49. Vladyka V, Liscak R, Novotny J Jr, et al. Radiation tolerance of functioning pitui-tary tissue in gamma knife surgery for pituitary adenomas. Neurosurgery 2003;52(2):309–16.

50. Choi JY, Chang JH, Chang JW, et al. Radiological and hormonal responses offunctioning pituitary adenomas after gamma knife radiosurgery. Yonsei Med J2003;44(4):602–7.

51. Parsons JT, Bova FJ, Fitzgerald CR, et al. Radiation optic neuropathy after mega-voltage external-beam irradiation: Analysis of time-dose factors. Int J RadiatOncol Biol Phys 1994;30(4):755–63.

52. Tishler RB, Loeffler JS, Lunsford LD, et al. Tolerance of cranial nerves of thecavernous sinus to radiosurgery. Int J Radiat Oncol Biol Phys 1993;27(2):215–21.

53. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the visual pathwaysand cranial nerves of the cavernous sinus to stereotactic radiosurgery. J Neuro-surg 1998;88:43–50.

54. Stafford SL, Pollock BE, Leavitt JA, et al. A study on the radiation tolerance of theoptic nerves and chiasm after stereotactic radiosurgery. Int J Radiat Oncol BiolPhys 2003;55(5):1177–81.

55. Martel MK, Sandler HM, Cornblath WT, et al. Dose-volume complication analysisfor visual pathway structures of patients with advanced paranasal sinus tumors.Int J Radiat Oncol Biol Phys 1997;38:273–84.

56. Ove R, Kelman S, Amin PP, et al. Preservation of visual fields after peri-sellargamma-knife radiosurgery. Int J Cancer 2000;90(6):343–50.

57. Urie MM, Fullerton B, Tatsuzaki H, et al. A dose response analysis of injury tocranial nerves and/or nuclei following proton beam radiation therapy. Int J RadiatOncol Biol Phys 1992;23:27–39.

58. Selch MT, Ahn E, Laskari A, et al. Stereotactic radiotherapy for treatment ofcavernous sinus meningiomas. Int J Radiat Oncol Biol Phys 2004;59(1):101–11.

59. Uy NW, Woo SY, Teh BS, et al. Intensity-modulated radiation therapy (IMRT) formeningioma. Int J Radiat Oncol Biol Phys 2002;53(5):1265–70.

60. Morita A, Coffey RJ, Foote RL, et al. Risk of injury to cranial nerves after gammaknife radiosurgery for skull base meningiomas: experience in 88 patients. J Neu-rosurg 1999;90(1):42–9.

Anker & Shrieve620

61. Kuo JS, Chen JC, Yu C, et al. Gamma knife radiosurgery for benign cavernoussinus tumors: quantitative analysis of treatment outcomes. Neurosurgery 2004;54(6):1385–93.

62. Pollock BE, Stafford SL. Results of stereotactic radiosurgery for patients withimaging defined cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys2005;62(5):1427–31.

63. Nicolato A, Foroni R, Alessandrini F, et al. Radiosurgical treatment of cavernoussinus meningiomas: experience with 122 treated patients. Neurosurgery 2002;51(5):1153–9 [discussion: 1159–61].

64. 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.

65. Flickinger JC, Lunsford LD, Linskey ME, et al. Gamma knife radiosurgery foracoustic tumors: multivariate analysis of four year results. Radiother Oncol1993;27(2):91–8.

66. Flickinger JC, Kondziolka D, Niranjan A, et al. Results of acoustic neuroma radio-surgery: an analysis of 5 years’ experience using current methods. J Neurosurg2001;94(1):1–6.

67. Beegle RD, Friedman WA, Bova FJ. Effect of treatment plan quality on outcomesafter radiosurgery for vestibular schwannoma. J Neurosurg 2007;107(5):913–6.

68. Hasegawa T, Fujitani S, Katsumata S, et al. Stereotactic radiosurgery for vestib-ular schwannomas: analysis of 317 patients followed more than 5 years. Neuro-surgery 2005;57(2):257–65.

69. Lunsford LD, Niranjan A, Flickinger JC, et al. Radiosurgery of vestibular schwan-nomas: summary of experience in 829 cases. J Neurosurg 2005;102(Suppl):195–9.

70. Meijer OW, Vandertop WP, Baayen JC, et al. Single-fraction vs. fractionated linac-based stereotactic radiosurgery for vestibular schwannoma: a single-institutionstudy. Int J Radiat Oncol Biol Phys 2003;56(5):1390–6.

71. Million RR, Cassisi NJ. Management of head and neck cancer: a multidisciplinaryapproach. Philadelphia: JB Lippincott; 1984.

72. Thomas C, Di Maio S, Ma R, et al. Hearing preservation following fractionatedstereotactic radiotherapy for vestibular schwannomas: prognostic implicationsof cochlear dose. J Neurosurg 2007;107(5):917–26.

73. Merchant TE, Gould CJ, Xiong X, et al. Early neuro-otologic effects of three-dimensional irradiation in children with primary brain tumors. Int J Radiat OncolBiol Phys 2004;58(4):1194–207.

74. Oh YT, Kim CH, Choi JH, et al. Sensory neural hearing loss after concurrentcisplatin and radiation therapy for nasopharyngeal carcinoma. Radiother Oncol2004;72(1):79–82.

75. Charabi S, Thomsen J, MantoniM, et al.Acousticneuroma (vestibular schwannoma):growth and surgical and nonsurgical consequences of the wait-and-see policy.Otolaryngol Head Neck Surg 1995;113(1):5–14.

76. Regis J, Pellet W, Delsanti C, et al. Functional outcome after gamma knife surgeryor microsurgery for vestibular schwannomas. J Neurosurg 2002;97(5):1091–100.

77. Iwai Y, Yamanaka K, Shiotani M, et al. Radiosurgery for acoustic neuromas:results of low-dose treatment. Neurosurgery 2003;53(2):282–7 [discussion:287–8].

78. Combs SE, Thilmann C, Debus J, et al. Long-term outcome of stereotactic radio-surgery (SRS) in patients with acoustic neuromas. Int J Radiat Oncol Biol Phys2006;64(5):1341–7.


79. Massager N, Nissim O, Delbrouck C, et al. Role of intracanalicular volumetricand dosimetric parameters on hearing preservation after vestibular schwannomaradiosurgery. Int J Radiat Oncol Biol Phys 2006;64(5):1331–40.

80. Rowe JG, Radatz M, Walton L, et al. Stereotactic radiosurgery for type 2 neuro-fibromatosis acoustic neuromas: patient selection and tumor size. StereotactFunct Neurosurg 2002;79(2):107–16.

81. Massager N, Nissim O, Delbrouck C, et al. Irradiation of cochlear structuresduring vestibular schwannoma radiosurgery and associated hearing outcome.J Neurosurg 2007;107(4):733–9.

82. Kong L, Lu JJ, Hu C, et al. Cranial nerve palsy after definitive radiation therapy fornasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 2006;66(3):S413.

83. Johnston EF, Hammond AJ, Cairncross JG. Bilateral hypoglossal palsies: a latecomplication of curative radiotherapy. Can J Neurol Sci 1989;16(2):198–9.

84. Zhang N, Pan L, Dai JZ, et al. Gamma knife radiosurgery for jugular foramenschwannomas. J Neurosurg 2002;97(5 Suppl):456–8.

85. Martin JJ, Kondziolka D, Flickinger JC, et al. Cranial nerve preservation andoutcomes after stereotactic radiosurgery for jugular foramen schwannomas.Neurosurgery 2007;61(1):76–81.

86. Flickinger JC, Deutsch M, Lunsford LD. Repeat megavoltage irradiation of pitui-tary and suprasellar tumors. Int J Radiat Oncol Biol Phys 1989;17(1):171–5.

87. Nieder C, Grosu AL, Andratschke NH, et al. Update of human spinal cord reirra-diation tolerance based on additional data from 38 patients. Int J Radiat OncolBiol Phys 2006;66(5):1446–9.

88. Million RR, Cassisi NJ. Management of head and neck cancer: a multidisciplinaryapproach. 2nd edition. Philadelphia: Lippincott Williams & Wilkins; 1993.

Radiation Effectson the Auditoryand VestibularSystems

Niranjan Bhandare, MSa,WilliamM.Mendenhall, MDa,PatrickJ. Antonelli, MDb,*

KEYWORDS

� Radiation � Auditory � Vestibular � Toxicity � Complication

Definitive or postoperative radiation therapy (RT) is commonly used for the manage-ment of intracranial and extracranial head and neck tumors. Because of the variabilityof tumor location and dimensions, sparing of nontarget normal tissue and organs maynot be always possible. Parts of the auditory and vestibular systems often receive highdoses of radiation and exhibit radiation-induced morbidity. Treatment modalities thatdeliver the highest doses of radiation to the auditory system include stereotactic radio-surgery (SRS) and fractionated stereotactic radiotherapy (FSRT) for the treatment ofvestibular schwannomas (VS), and fractionated radiotherapy (FRT) or intensity-modulated radiation therapy (IMRT) for the treatment of head and neck malignancies.Radiation therapy for VS is unique because of its involvement of the inner ear and pre-existing auditory and vestibular dysfunction. Direct otologic involvement is rarelyobserved in other neoplasms that are routinely treated with RT. The assessment ofpostirradiation effects thus differs in these two situations and is more complex inthe treatment of VS.

RADIATION BIOLOGY

A comprehensive discussion of the tissue effects of radiation is beyond the scope ofthis article. However, a brief overview of this topic is necessary to fully understandradiation effects on the auditory and vestibular systems.

Tissue effects of radiation are dependent on a number of factors. MegavoltageX rays in the therapeutic range interact with tissue primarily by way of the Compton

a Department of Radiation Oncology, University of Florida College of Medicine, 1600 SWArcher Road, Gainesville, FL 32610, USAb Department of Otolaryngology, University of Florida College of Medicine, 1600 SW ArcherRoad, Gainesville, FL 32610, USA* Corresponding author.E-mail address: [email protected] (P.J. Antonelli).




Bhandare et al624

effect. The Compton mass attenuation coefficient is independent of the atomicnumber and depends only on the number of the electrons per gram of the interactingmaterial. The attenuation of beam is related to density thickness (density of materialmultiplied by the thickness) expressed as g/cm2, thus a relative decrease in attenua-tion leads to increased penetration in air-filled spaces or air cavities, such as the lungsand temporal bones.

Deposition of radiation energy in tissue results in cell injury and death. Most of radi-ation’s tissue effect is thought to be a result of the damage to DNA. This occurs bothdirectly and indirectly. The latter, the predominant mechanism, involves ionization ofsurrounding water molecules to form free radicals, which, in turn, result in double-strand breaks in DNA. This injury may result in cell death during mitosis, inductionof apoptosis (ie, programmed cell death), recovery, and cell cycle arrest or terminaldifferentiation through activation of repair pathways that may also play a role in tumorsuppression (eg, activation of p53).

The radio-response of a tissue depends on the inherent sensitivity of the cells, thekinetics of cell population, total dose, dose per fraction, and time-dose fractionation.Cells with fast turnover rate or higher mitotic activity exhibit more sensitivity to radia-tion, subjecting to cell death in attempting subsequent mitosis. This is the basis fortherapeutic RT (ie, relatively greater damage to highly reproductive tumor cells).Further, RT fractionation offers the potential for greater differential sparing of normaltissues and killing of tumor cells.1,2

These factors also determine the unwanted manifestations of RT. Skin and mucosa,which cycle quickly, manifest more significant early, transient, and inflammatorychanges. Organ dysfunction are often manifest by cell lines with slow turnover (eg,radionecrosis of bone). In some tissues, such as inner-ear hair cells, functionalprogenitor cells may be lacking, resulting in greater organ system dysfunction. It isunclear how RT leads to long-term dysfunction in cells, such as neurons and inner-ear hair cells, which lose mitotic activity after differentiation. Such cells are dependenton supporting cells (eg, glia) and small blood vessels. Though a wide range of morpho-logic changes in neural tissue in response to RT have been observed,3,4 the effect ofRT on neurons cannot be distinguished from the effect on the supporting cells and thevasculature.

Finally, radiation dose to the target and surrounding tissues is controlled by thechoice of modality (eg, conventional external beam, IMRT, or SRS) and treatmentplan. In SRS, doses are prescribed to the tumor margin. The maximal dose withinthe target area may vary tremendously, depending on the specifics of the treatmentplan. For example, single or multiple isocenters may be used.

For further information, please see the Treatment Planning/Radiation DeliverySection in the Gamma Knife Radiosurgery for Vestibular Schwannoma chapter ofthis publication.

RADIATION AND THE EAR

The entire auditory-vestibular system is vulnerable to RT injury. Nearly half of allpatients that have undergone RT for head and neck tumors demonstrate evidenceof auditory or vestibular system pathology.3 RT-induced injury may be manifest aschronic otitis externa, stenosis of the external auditory canal, chronic otitis mediawith effusion, tympanic membrane perforation, osteonecrosis, chronic suppurativeotitis media, middle ear fibrosis, conductive, mixed, or sensorineural hearing loss, lab-yrinthitis, vestibular paresis, and vertigo. The focus of this discussion will be on theinner ear and central pathways rather than the middle and external ear.

Radiation Effects 625

Measurement or scoring of ototoxicity after fractionated RT for head and necktumors has historically been qualitative and descriptive. Comparing results fromsuch assessments has been difficult. Newer scoring systems have been developed,but none have been widely used and validated. The Radiation Therapy OncologyGroup criteria are applicable for retrospective analyses of acute toxicity, but not latecomplications. Detailed prospective assessment of delayed RT-induced ototoxicityhas been addressed by the Late Management of Normal Tissue/Somatic ObjectiveManagement Analytic (LENT/SOMA) scoring system (Table 1). This system doesnot distinguish between external, middle, and inner ear toxicity and has a narrow cate-gorization of hearing loss. The National Cancer Institute Common Toxicity Criteria (NCICTC) includes auditory side effects, but only gross changes in hearing are addressed(Table 2). The NCI CTC system has been mainly applied to chemotherapeutic studies.Note that these classification systems correspond to RT of nonacoustic tumors. Theseclassification systems have not been used for SRS or FSRT of VS. The evaluations ofpost-RT hearing status of patients with VS have consistently been presented accord-ing to the Gardner-Robertson scale.

The unique nature of the inner ear cell lines leads to unique manifestations of RTinjury. Acute dysfunction after RT is a result of transient alterations in the homeostasisof endolymph and perilymph,5 whereas delayed sensorineural hearing loss (SNHL)most commonly exhibits a chronic, progressive, and irreversible evolution.6 Hair cellsand the stria vascularis have been implicated as the two major sites of inner ear radi-ation toxicity.5

Table 1Late radiation ear morbidity according to the LateManagement of NormalTissue/Somatic ObjectiveManagement Analytic Scoring System Scale

Grade1 Grade 2 Grade 3 Grade 4Subjective

1. Pain Occasional andminimal

Intermittent andtolerable

Persistent andintense

Refractory andexcruciating

2. Tinnitus Occasional Intermittent Persistent Refractory

3. Hearing Minor loss, noimpairment indaily activities

Frequentdifficulties withfaint speech

Frequentdifficulties withloud speech

Completedeafness

Objective

4. Skin Drydesquamation

Otitis externa Superficialulceration

Deep ulceration,necrosis

5. Hearing <10 dB loss inone or morefrequencies

10–15 dB loss inone or morefrequencies

>15–20 dB loss inone or morefrequencies

>20 dB loss inone or morefrequencies

Management

6. Pain Occasionalnon-narcotic

Regularnon-narcotic

Regular narcotic Parenteralnarcotics

7. Skin Occasionallubrication/ointments

Regular eardropsor antibiotics

Eardrums Surgicalintervention

8. Hearing Hearing aid

Score the 8 SOM parameters with 0 to 4 (0 5 no toxicity); total the score and divide by 8 5 LENTscore.

Table 2Ear morbidity according to the National Cancer Institute CommonToxicity Criteria

Grade1 Grade 2 Grade 3 Grade 4External

auditorycanal

External otitiswith moist ordrydesquamation

External otitiswith moistdesquamation

External otitiswithdischarge,mastoiditis

Necrosis of canal,soft tissue orbone

Inner ear/hearing(includingconductivehearing loss)

Hearing loss onaudiometryonly

Tinnitus orhearing lossnot requiringhearing aid ortreatment

Tinnitus orhearing loss,correctablewith hearingaid ortreatment

Severe unilateralor bilateralhearingloss(deafness),notcorrectable

Middle ear/hearing

Serous otitiswithoutsubjectivedecrease inhearing

Serous otitis orinfectionrequiringmedicalintervention;subjectivedecreasein hearing;rupture oftympanicmembranewithdischarge

Otitis withdischarge,mastoiditis orconductivehearing loss

Necrosis of thecanal, softtissue or bone

Bhandare et al626

In animal studies, the earliest post-RT changes with lower doses were observed inthe stria vascularis. With increasing dose, shriveling of hair cells and distention ofReissner’s membrane were observed.7 Experimental fractionated radiation of chin-chillas led to loss of myelinated nerve fibers in the osseous spiral lamina, inner andouter hair cells, and supporting cells.8 Hair cell damage and compound action poten-tial changes have been linked to stria vascularis degeneration,9 but direct damage tohair cells is feasible at high- radiation doses. Similar findings have been reported forthe vestibular system.7,10,11

Post-mortem observations of human temporal bones after RT have included loss ofinner and outer hair cells and spiral ganglion cells in the basal turn of the cochlea,atrophy of the stria vascularis, and changes in vessels of the facial nerve.11–13 Furtherhistologic studies have shown that the greatest damage to inner ear is the result ofinjury to the vessels of stria vascularis.14 Vestibular damage has also been observedin cases manifesting with vestibular complaints. An autopsy study after a high dose ofradiation demonstrated absence of the organ of Corti, macula of the utricle and thecristae of the semicircular canals.14 Overall, however, the vestibular apparatus ismore resistant to effects of radiation than the cochlea.7

Sensorineural hearing loss following fractionated RT may begin early or may bedelayed. Transient SNHL begins early.15,16 Recovery usually occurs in 6 to 12 months,but may be delayed.17 This delay has been attributed to an inner ear vasculitis, and hasbeen associated with auditory recruitment.18 Permanent post-RT SNHL may occur inup to 54% of patients receiving high doses to the inner ear.15 A review by Jerekczeck-Fossa and colleagues19 suggested that post-RT SNHL occurs in one third of thepatients treated with definitive RT with the inner ear receiving high radiation doses.Raaijmakers and Engelen’s review20 reported a pure tone average loss of greater


than 10 dB occurred in 18 (� 2%) of patients and a 4 kHz loss occurred in one of threepatients receiving a dose of 70 Gy at 2 Gy per fraction to the inner ear. The pooled inci-dence of post-RT SNHL is 44% and 36% for treatment of the nasopharynx andparotid, respectively.21

Ho and colleagues22 prospectively studied the latency for SNHL after fractionatedRT and found it to be 1.5 to 2.0 years after RT. Kwong and colleagues15 also reportedthe progression of SNHL to plateau within 2 years of treatment. Most other studies arein agreement with these observations.15,16 Onset of hearing deterioration has beenreported to be as early as 3 months after completing RT.23 In one study, the stabilityof postlatency SNHL was followed up for a period of 13 years.24 Though cumulativerisk of persistent SNHL (>15 dB) has been reported to stabilize at 2 years, the cumu-lative risk for severe SNHL (>30 dB) continues to increase through the third and fourthyear after RT.25

Vestibular dysfunction occurs in roughly 25% to 30% of patients treated with RT tothe temporal bone.26 Caloric weakness has been reported in 9% to 36%.27–29 Youngand colleagues30 attributed post-RT vertigo mainly to a peripheral labyrinthinedisorder(69%), followed by central vestibular lesions (31%). The authors proposedtwo mechanisms that contribute to the peripheral labyrinthine dysfunction: (1) directinjury to the inner ear, or (2) RT-induced otitis media with effusion (OME) witha secondary labyrinthitis. In the study by Young and colleagues,30 the mean intervalfrom completion of RT to occurrence of vertigo was 10 years. The relationshipbetween OME and vertigo, however, has not been supported in other studies.16 RT-induced degenerative changes in the vestibular sensory epithelia has been observedin experimental animal studies10 and human temporal bones.14

TREATMENT-RELATED PARAMETERSDose

Honore and colleagues24 used linear and logistic regression methods to evaluate thedose-response relationship of RT and inner ear dysfunction. They observed a generalincrease in hearing loss with increased dose to the inner ear, but this was not a linearrelationship. A statistically significant relationship between dose and the incidence ofSNHL was only observed at 4 KHz. Probability of SNHL increased above 40 Gy to theinner ear when delivered at 2.0 Gy per fraction. Others have reported SNHL at totaldoses above 45 to 50 Gy.31,32 Delayed toxicity increases as a function of dose perfraction. The dose per fraction and the total dose must be considered for the develop-ment of late ear toxicity.

Patient Age

Grau and colleagues31 found that the raw data suggested a correlation betweenpatient age and post-RT SNHL, but this disappeared when RT dose and age weretaken into account. Other reports have indicated an increase in the incidence ofpost-RT SNHL, implicating increasing age as a possible risk factor.15,22,24,33 Oneseries16 reported a statistically significant increase in the incidence of post-RTSNHL above the age of 50. Pretreatment hearing loss, which is more common withadvancing age, may be a more important risk factor for post-RT hearing loss.22,24

Adjuvant Chemotherapy

A number of reports12,16 have associated chemoradiation with the development ofSNHL, yet others15,29 have reported no relationship. The ototoxic effect of cisplatin,one of the most commonly used chemotherapeutic agents, is well known.34,35

Bhandare et al628

Cisplatin ototoxicity is thought to be greater if given after RT, compared with beforeRT.15,36 It is unclear whether cisplatin and RT ototoxicities are truly synergistic.34

Atrophy of the stria vascularis and loss of inner and outer hair cells with reduced spiralganglion cells, have been reported in patients receiving cisplatin, RT, or the twocombined.12

RADIATION THERAPY FOR VESTIBULAR SCHWANNOMA

The options for delivery of radiation therapy for VS include SRS, FSRT, intensity-modulated radiotherapy (IMRT), three-dimensional conformal radiotherapy (3DCRT)and, rarely, conventional RT. SRS uses multiple, convergent, nonparallel, noncoplanerradiation beams to deliver a high dose of irradiation to a small target volume whiledelivering low doses to the surrounding normal tissues. This can be accomplishedwith either a gamma knife (GK-SRS) or a linear accelerator. Depending on the choiceof prescription isodose line covering the target, up to 50% or more variation may beseen between the peripheral dose and the maximum dose.

FSRT offers the advantages of SRS and conventional RT. Like SRS, FSRT allows theconformation of the irradiation dose precisely to the target volume. In addition, itallows the fractionated delivery of the dose, which is thought to result in differentialsparing of cranial nerves relative to the tumor.1,2

VS typically presents with auditory or vestibular dysfunction. This dysfunction isthought to result from any one, or a combination, of the direct compression of theeighth cranial nerve, labyrinthine vasculature, or the development of poorly under-stood changes within the inner ear fluids. Symptoms may include tinnitus, ataxia,dizziness, unsteadiness, vertigo, a sensation of fullness in the ear and gradual hearingloss.37 Most patients present with hearing loss (98%), tinnitus (70%), and disequilib-rium (67%).38 Thus, it can be difficult to distinguish the auditory and vestibular effectsof RT for VS from those caused by the natural progression of VS.

As hearing is the most readily measured parameter of auditory and vestibular func-tion, hearing preservation serves as the best indicator of RT impact on the auditorysystems, particularly for VS. Hearing preservation rates, according to the GardnerRobertson criteria, ranging from 50% to 70%,39–42 and local control rates rangingfrom 95% to 100% have been reported with GK-SRS39,43,44 and LA-SRS45–47 forVS. Immediate SNHL post-SRS for VS has been rarely reported.48 SNHL within3 months of SRS may result from neural edema or demyelination. Most SNHL hasbeen reported at 3 to 24 months after SRS.49 Foote and colleagues38 reporteda median time to the onset of SNHL after GK-SRS among subjects with sporadicVS to be 18 months (range, 8.9–30 months). In subjects with neurofibromatosis type2 (NF-2)-associated VS treated with SRS, Linskey and colleagues50 observedprogressive SNHL with a median time of onset to be 4 months (range, 3–12 months).Subach and colleagues51 reported a median time of onset of SNHL at 4 months(range, 3–15 months) with none documented before 3 months after SRS.

Many RT and tumor parameters have been found to affect auditory outcomes in thetreatment of VS. These parameters include involved nerve length, total dose, fraction-ation, and tumor type.

Irradiated Nerve Length

Niranjan and colleagues49 estimated that, for an intracranial VS, the length of theeighth nerve irradiated in a GK-SRS varied between 4 to 12 mm, with many tumors(5 of 15) occupying only part of the internal auditory canal (4–7 mm) and more tumors(10 of 15 tumors) extending the entire length of the canal (8–12 mm). The authors


concluded that neither the position of the tumor in the canal (lateral versus medial) northe length of the nerve irradiated correlated with hearing preservation (P 5 0.1).Although tumor diameter was significantly related to hearing preservation in theunivariate analysis (P 5 0.025), only margin dose was significant in the multivariateanalysis (P 5 0.0009).

Total Radiation Dose

Paek and colleagues52 reported that subjects who retained useful hearing after SRShad a mean dose delivered to the cochlea of 6.9 Gy, compared with 11 Gy in subjectswho exhibited significant hearing loss. Niranjan and colleagues49 found that SRS doseextending beyond the canalicular tumor volume was the most important factorresponsible for cochlear nerve injury. Serviceable hearing was preserved in 100% ofthe subjects who received a dose of 14 Gy or less, which dropped to 20% in subjectswho received more than 14 Gy. Massager and colleagues53 suggested that intracana-licular tumor volume (<100 cm3 versus R100 cm3) and intracanalicular integrateddose (dose � volume) were determinants of hearing loss and postulated that ‘‘wors-ening of hearing after GK-SRS could be attributed to cochlear injury inside the internalauditory canal caused by enlargement of the intracanalicular part of the VS during theinflammatory edema phase after SRS through an increase of intracanalicular pres-sure.’’ Rowe and colleagues54 reported hearing results after treatment using GK-SRS for NF-2. These authors reported three groups of subjects: group one treatedwith a mean marginal dose of 25 Gy, group two with a mean dose of 17.8 Gy and groupthree with a mean dose of 13.4 Gy. The authors concluded that the dose reductionfrom group one to group three increased hearing preservation (P 5 0.05). Based onthe available data, the general trend has been to reduce SRS dose from 18 to20 Gy initially, to 12 to 16 Gy.

Fractionation

Combs and colleagues55 reported outcomes of 106 subjects with VS treated withFSRT to a total dose of 57.6 Gy delivered by standard fractionation. The 5-year localcontrol rate was 93% with serviceable hearing preservation in 94%. In a recentreport56 of 26 subjects treated with SRS to a median total dose of 13 Gy (range 11–20 Gy), local control of 91% and serviceable hearing preservation of 55% wasreported. Onset of SNHL after FSRT has been reported mostly in the first 6 monthsand less commonly beyond 12 months after treatment.55,57–59 Andrews andcolleagues60 found that the probability of retaining serviceable hearing was signifi-cantly higher after FSRT versus GK-SRS (81% versus 33%; P 5 0.0228). Comparedwith GK-SRS, FSRT resulted in a significantly lower pure tone average (P 5 0.0120)and significantly higher mean speech discrimination scores (P 5 0.0466). Their highrate of early hearing preservation at pretreatment levels after FSRT is in agreementwith an earlier report by Williams.61 A greater chance of hearing preservation maybe associated with FSRT using standard-dose fractionation not exceeding a conven-tional 2 Gy dose per fraction.61–63 Further prospective studies comparing postirradia-tion hearing preservation are needed to evaluate the potential advantages of FSRTcompared with GK-SRS.

Compared with standard fractionation, hypofractionation delivers the total doseover a shorter period of time using a smaller number of higher dose fractions, butcompared with single high-dose fractions as used in SRS, it delivers a lower doseper fraction. Most hypofractionation schedules deliver total doses ranging from 21to 30 Gy in 3 to 10 days with a dose per fraction ranging from 3 to 7 Gy. Althoughhypofractionation reduces fraction size compared with that delivered in SRS, late

Bhandare et al630

irradiation effects, such as hearing loss, are strongly dependent on fraction size andrises as fraction size increases above the standard once-daily fraction size of 1.8 to2.0 Gy. Experience with hypofractionated RT in regards to hearing preservationremains very limited. Meijer and colleagues57 treated 12 subjects with SRS to a totaldose of 10 to 12 Gy and 25 with a hypofractionated SRT regimen of either five fractionsof 4 Gy or five fractions of 5 Gy. They found no significant differences in the rate of localcontrol and hearing loss. It is still not clear if there is an advantage in using hypofrac-tionation compared with SRS for hearing preservation. For further informationregarding hypofractionation hearing preservation rates, see the Stanford UniversityExperience Section in the CyberKnife Radiosurgery for Vestibular SchwannomaChapter.

Neurofibromatosis Type 2

Unlike sporadic VS that tends to displace the auditory portion of the eighth cranialnerve, VS associated with NF-2 often form grapelike clusters,64 engulf the facial andcochlear nerves,65,66 and invade the cochlea and other parts of the temporalbone.65 Accordingly, there are lower rates of tumor control and hearing preservationwith SRS or FSRT for VS associated with NF-2.60,66,67

VERTIGO

There is little data that addresses the impact of SRS or FSRT on the vestibular system.Petit and colleagues68 reported 17 subjects with imbalance before GK-SRS for VS.The imbalance was unchanged in 13 of the 17 subjects after treatment, while foursubjects had improvement of their imbalance. New cases of vertigo were notobserved. Niranjan and colleagues49 found that 3 of 11 subjects continued to haveepisodic vertigo who presented with vertigo after treatment with GK-SRS for VS. Itoand colleagues69 reported the results of caloric function testing after GK-SRS forVS. Thirteen of 46 subjects had intact caloric function before treatment. Nine of the13 subjects developed canal paresis after GK-SRS. The median onset of loss ofcaloric function was 8 months. The authors did not report the subjects’ symptom ofdizziness in their manuscript.

SUMMARY

RT to the region of the temporal bone and surrounding structures may result in audi-tory and vestibular dysfunction both peripherally and centrally. Auditory and vestibulardysfunction following RT for VS may be limited by limiting the total dose of cranialnerve VIII irradiation and by fractionation.

REFERENCES

1. Buatti JM, Friedman WA, Meeks SL, et al. The radiobiology of radiosurgery andstereotactic radiotherapy. Med Dosim 1998;23:201–7.

2. Safdari H, Fuentes JM, Dubois JB, et al. Radiation necrosis of the brain: time ofonset and incidence related to total dose and fractionation of radiation. Neurora-diology 1985;27:44–7. PM:3974866.

3. Masurovsky EB, Bunge MB, Bunge RP. Cytological studies of organotypiccultures of rat dorsal root ganglia following X-irradiation in vitro. I. Changes inneurons and satellite cells. J Cell Biol 1967;32:467–96. PM:10976234.

4. Masurovsky EB, Bunge MB, Bunge RP. Cytological studies of organotypiccultures of rat dorsal root ganglia following X-irradiation in vitro. II. Changes in


Schwann cells, myelin sheaths, and nerve fibers. J Cell Biol 1967;32:497–518.PM:10976235.

5. Linskey ME, Johnstone PA. Radiation tolerance of normal temporal bone struc-tures: implications for gamma knife stereotactic radiosurgery. Int J Radiat OncolBiol Phys 2003;57:196–200. PM:12909233.

6. Pollock BE, Lunsford LD, Kondziolka D, et al. Outcome analysis of acousticneuroma management: a comparison of microsurgery and stereotactic radiosur-gery. Neurosurgery 1995;36:215–29.

7. Gamble JE, Peterson EA, Chandler JR. Radiation effects on the inner ear. ArchOtolaryngol 1968;88:156–61. PM:5301997.

8. Bohne BA, Marks JE, Glasgow GP. Delayed effects of ionizing radiation on theear. Laryngoscope 1985;95:818–28. PM:4010422.

9. Ocho S, Iwasaki S, Umemura K, et al. A new model for investigating hair celldegeneration in the guinea pig following damage of the stria vascularis usinga photochemical reaction. Eur Arch Otorhinolaryngol 2000;257:182–7. PM:10867831.

10. Winther FO. X-ray irradiation of the inner ear of the guinea pig. Early degenerativechanges in the vestibular sensory epithelia. Acta Otolaryngol 1969;68:514–25.PM:4907323.

11. Nadol JB Jr. Comparative anatomy of the cochlea and auditory nerve inmammals. Hear Res 1988;34:253–66. PM:3049492.

12. Hoistad DL, Ondrey FG, Mutlu C, et al. Histopathology of human temporal boneafter cis-platinum, radiation, or both. Otolaryngol Head Neck Surg 1998;118:825–32. PM:9627244.

13. Felix H. Anatomical differences in the peripheral auditory system of mammalsand man. A mini review. Adv Otorhinolaryngol 2002;59:1–10. PM:11885648.

14. Leach W. Irradiation of the ear. J Laryngol Otol 1965;79:870–80. PM:5830715.15. Kwong DL, Wei WI, Sham JS, et al. Sensorineural hearing loss in patients treated

for nasopharyngeal carcinoma: a prospective study of the effect of radiation andcisplatin treatment. Int J Radiat Oncol Biol Phys 1996;36:281–9. PM:8892450.

16. Bhandare N, Antonelli PJ, Morris CG, et al. Ototoxicity after radiotherapy for headand neck tumors. Int J Radiat Oncol Biol Phys 2007;67:469–79. PM:17236969.

17. Poen JC, Golby AJ, Forster KM, et al. Fractionated stereotactic radiosurgery andpreservation of hearing in patients with vestibular schwannoma: a preliminaryreport. Neurosurgery 1999;45:1299–307.

18. Borsanyi SJ, Blanchard CL. Ionizing radiation and the ear. JAMA 1962;181:958–61. PM:13871468.

19. Jereczek-Fossa BA, Zarowski A, Milani F, et al. Radiotherapy-induced ear toxicity.Cancer Treat Rev 2003;29:417–30. PM:12972360.

20. Raaijmakers E, Engelen AM. Is sensorineural hearing loss a possible side effectof nasopharyngeal and parotid irradiation? A systematic review of the literature.Radiother Oncol 2002;65:1–7. PM:12413668.

21. Bhide SA, Harrington KJ, Nutting CM. Otological toxicity after postoperativeradiotherapy for parotid tumours. Clin Oncol (R Coll Radiol) 2007;19:77–82.PM:17305258.

22. Ho WK, Wei WI, Kwong DL, et al. Long-term sensorineural hearing deficitfollowing radiotherapy in patients suffering from nasopharyngeal carcinoma:a prospective study. Head Neck 1999;21:547–53. PM:10449671.

23. Wang LF, Kuo WR, Ho KY, et al. A long-term study on hearing status in patientswith nasopharyngeal carcinoma after radiotherapy. Otol Neurotol 2004;25:168–73. PM:15021778.

Bhandare et al632

24. Honore HB, Bentzen SM, Moller K, et al. Sensori-neural hearing loss after radio-therapy for nasopharyngeal carcinoma: individualized risk estimation. RadiotherOncol 2002;65:9–16. PM:12413669.

25. Grau C, Overgaard J. Postirradiation sensorineural hearing loss: a common butignored late radiation complication. Int J Radiat Oncol Biol Phys 1996;36:515–7. PM:8892478.

26. Singh IP, Slevin NJ. Late audiovestibular consequences of radical radiotherapy tothe parotid. Clin Oncol (R Coll Radiol) 1991;3:217–9. PM:1657112.

27. Chao WY, Tseng HZ, Tsai ST. Caloric response and postural control in patientswith nasopharyngeal carcinoma after radiotherapy. Clin Otolaryngol Allied Sci1998;23:439–41. PM:9800080.

28. Gabriele P, Orecchia R, Magnano M, et al. Vestibular apparatus disorders afterexternal radiation therapy for head and neck cancers. Radiother Oncol 1992;25:25–30. PM:1410586.

29. Johannesen TB, Rasmussen K, Winther FO, et al. Late radiation effects onhearing, vestibular function, and taste in brain tumor patients. Int J Radiat OncolBiol Phys 2002;53:86–90. PM:12007945.

30. Young YH, Ko JY, Sheen TS. Postirradiation vertigo in nasopharyngeal carcinomasurvivors. Otol Neurotol 2004;25:366–70. PM:15129119.

31. Grau C, Moller K, Overgaard M, et al. Sensori-neural hearing loss in patientstreated with irradiation for nasopharyngeal carcinoma. Int J Radiat Oncol BiolPhys 1991;21:723–8. PM:1869465.

32. Pan C, Eisbruch A. Ototoxicity after radiotherapy of head-and-neck cancer: theperils of retrospective dose-response estimations: in regard to Bhandare et al.(Int J Radiat Oncol Biol Phys 2007;67:469–479). Int J Radiat Oncol Biol Phys2007;68:1582.

33. Moretti JA. Sensori-neural hearing loss following radiotherapy to the naso-pharynx. Laryngoscope 1976;86:598–602. PM:1263730.

34. The Department of Veterans Affairs Laryngeal Cancer Study Group. Inductionchemotherapy plus radiation compared with surgery plus radiation in patientswith advanced laryngeal cancer. N Engl J Med 1991;324:1685–90.

35. Skinner R, Pearson AD, Amineddine HA, et al. Ototoxicity of cisplatinum in chil-dren and adolescents. Br J Cancer 1990;61:927–31. PM:2372498.

36. Walker DA, Pillow J, Waters KD, et al. Enhanced cis-platinum ototoxicity in chil-dren with brain tumours who have received simultaneous or prior cranial irradia-tion. Med Pediatr Oncol 1989;17:48–52. PM:2913475.

37. Mendenhall WM, Friedman WA, Amdur RJ, et al. Management of acousticschwannoma. Am J Otolaryngol 2004;25:38–47.

38. Foote RL, Coffey RJ, Swanson JW, et al. Stereotactic radiosurgery using thegamma knife for acoustic neuromas. Int J Radiat Oncol Biol Phys 1995;32:1153–60.

39. Noren G. Long-term complications following gamma knife radiosurgery of vestib-ular schwannomas. Stereotact Funct Neurosurg 1998;70(Suppl 1):65–73 PM:9782237.

40. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry 1983;46:797–803. PM:6352865.

41. Thomassin JM, Epron JP, Regis J, et al. Preservation of hearing in acousticneuromas treated by gamma knife surgery. Stereotact Funct Neurosurg 1998;70(Suppl 1):74–9. PM:9782238.

42. Traquina DN, Guttenberg I, Sasaki CT. Delayed diagnosis and treatment ofacoustic neuroma. Laryngoscope 1989;99:814–8. PM:2755290.


43. Kondziolka D, Lunsford LD, McLaughlin MR, et al. Long-term outcomes afterradiosurgery for acoustic neuromas. N Engl J Med 1998;339:1426–33.

44. Ogunrinde OK, Lunsford DL, Kondziolka DS, et al. Cranial nerve preservationafter stereotactic radiosurgery of intracanalicular acoustic tumors. StereotactFunct Neurosurg 1995;64(Suppl 1):87–97. PM:8584844.

45. Flickinger JC, Kondziolka D, Niranjan A, et al. Results of acoustic neuroma radio-surgery: an analysis of 5 years’ experience using current methods. J Neurosurg2001;94:1–6. PM:11147876.

46. Flickinger JC, Kondziolka D, Niranjan A, et al. Acoustic neuroma radiosurgerywith marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys 2004;60:225–30. PM:15337560.

47. Iwai Y, Yamanaka K, Shiotani M, et al. Radiosurgery for acoustic neuromas:results of low-dose treatment. Neurosurgery 2003;53:282–7 PM:12925242.

48. Pollack AG, Marymont MH, Kalapurakal JA, et al. Acute neurological complica-tions following gamma knife surgery for vestibular schwannoma. Case report.J Neurosurg 2005;103:546–51. PM:16235688.

49. Niranjan A, Lunsford LD, Flickinger JC, et al. Dose reduction improves hearingpreservation rates after intracanalicular acoustic tumor radiosurgery. Neurosur-gery 1999;45:753–62. PM:10515468.

50. Linskey ME, Lunsford LD, Flickinger JC. Tumor control after stereotactic radiosur-gery in neurofibromatosis patients with bilateral acoustic tumors. Neurosurgery1992;31:829–38. PM:1436407.

51. Subach BR, Kondziolka D, Lunsford LD, et al. Stereotactic radiosurgery in themanagement of acoustic neuromas associated with neurofibromatosis Type 2.J Neurosurg 1999;90:815–22. PM:10223445.

52. Paek SH, Chung HT, Jeong SS, et al. Hearing preservation after gamma knifestereotactic radiosurgery of vestibular schwannoma. Cancer 2005;104:580–90.PM:15952200.

53. Massager N, Nissim O, Delbrouck C, et al. Role of intracanalicular volumetric anddosimetric parameters on hearing preservation after vestibular schwannomaradiosurgery. Int J Radiat Oncol Biol Phys 2006;64:1331–40. PM:16458446.

54. Rowe JG, Radatz MW, Walton L, et al. Clinical experience with gamma knifestereotactic radiosurgery in the management of vestibular schwannomassecondary to type 2 neurofibromatosis. J Neurol Neurosurg Psychiatry 2003;74:1288–93. PM:12933938.

55. Combs SE, Volk S, Schulz-Ertner D, et al. Management of acoustic neuromas withfractionated stereotactic radiotherapy (FSRT): long-term results in 106 patientstreated in a single institution. Int J Radiat Oncol Biol Phys 2005;63:75–81. PM:16111574.

56. Combs SE, Thilmann C, Debus J, et al. Long-term outcome of stereotactic radio-surgery (SRS) in patients with acoustic neuromas. Int J Radiat Oncol Biol Phys2006;64:1341–7. PM:16464537.

57. Meijer OW, Vandertop WP, Baayen JC, et al. Single-fraction vs. Is radiotherapy investibular schwannoma: early results from a single centre. Clin Oncol (R CollRadiol) 2007;19:517–22. PM:17400433.

58. Horan G, Whitfield GA, Burton KE, et al. Fractionated conformal radiotherapy investibular schwannoma: early results from a single centre. Clin Oncol (R CollRadiol) 2007;19(7):517–22.

59. Sakamoto T, Shirato H, Takeichi N, et al. Annual rate of hearing loss falls after frac-tionated stereotactic irradiation for vestibular schwannoma. Radiother Oncol2001;60:45–8.

Bhandare et al634

60. Andrews DW, Suarez O, Goldman HW, et al. Stereotactic radiosurgery andfractionated stereotactic radiotherapy for the treatment of acoustic schwanno-mas: comparative observations of 125 patients treated at one institution. Int J Ra-diat Oncol Biol Phys 2001;50:1265–78.

61. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas. ActaNeurochir (Wien) 2002;144(12):1249–54; discussion 1254.

62. Andrews DW, Silverman CL, Glass J, et al. Preservation of cranial nerve functionafter treatment of acoustic neurinomas with fractionated stereotactic radio-therapy. Preliminary observations in 26 patients. Stereotact Funct Neurosurg1995;64:165–82. PM:8817804.

63. Hoban PW, Jones LC, Clark BG. Modeling late effects in hypofractionated stereo-tactic radiotherapy. Int J Radiat Oncol Biol Phys 1999;43:199–210. PM:9989527.

64. Jaaskelainen J, Paetau A, Pyykko I, et al. Interface between the facial nerve andlarge acoustic neurinomas. Immunohistochemical study of the cleavage plane inNF2 and non-NF2 cases. J Neurosurg 1994;80:541–7. PM:8113868.

65. Linthicum FH Jr, Brackmann DE. Bilateral acoustic tumors. A diagnostic andsurgical challenge. Arch Otolaryngol 1980;106:729–33. PM:7436847.

66. Hirsch A, Noren G. Audiological findings after stereotactic radiosurgery inacoustic neurinomas. Acta Otolaryngol 1988;106:244–51. PM:3051887.

67. Flickinger JC, Lunsford LD, Linskey ME, et al. Gamma knife radiosurgery foracoustic tumors: multivariate analysis of four year results. Radiother Oncol1993;27:91–8. PM:8356233.

68. Petit JH, Hudes RS, Chen TT, et al. Reduced-dose radiosurgery for vestibularschwannomas. Neurosurgery 2001;49:1299–307.

69. Ito K, Kurita H, Sugasawa K, et al. Analyses of neuro-otological complicationsafter radiosurgery for acoustic neurinomas. Int J Radiat Oncol Biol Phys 1997;39:983–8. PM:9392535.

Gamma KnifeRadiosurgeryfor VestibularSchwannoma

Robert A. Battista, MDa,b,c,*

KEYWORDS

� Gamma knife radiosurgery � Vestibular schwannoma� Neurofibromatosis type 2 � Cystic vestibular schwannomar� Tumor control

In 1951, Dr. Lars Leksell from Sweden conceived of what is now known as gammaknife radiosurgery (GKRS). The Leksell Gamma Knife (Elekta, Norcross, Georgia)uses a focused array of 201 intersecting beams of cobalt-60 gamma radiation to treatlesions within the brain. The gamma knife is so named because of its accuracy in deliv-ering an exact field of radiation to the target. The combined intensity of radiation at thefocus (or isocenter) is extremely high whereas the intensity only a short distance fromthe isocenter is very low. This enables a high dose of radiation to be delivered to thelesion while sparing the adjacent healthy brain tissue.

Leksell first used the gamma knife in 1969 to treat a patient who had a vestibularschwannoma.1 Gamma knife radiosurgery has grown considerably in the UnitedStates since the first gamma knife unit was installed in North America at the Universityof Pittsburgh in 1987. Currently, there are more than 50 gamma knife centers in theUnited States and 249 centers worldwide. Based on a worldwide survey in December2006 by the Leksell Gamma Knife Society, 397,672 lesions have been treated withGKRS since the introduction of the gamma knife (202 of 249 gamma knife centersreporting). Slightly more than 9% (36,180) of the lesions treated were vestibularschwannomas.2 A similar survey in December 2004 by the Leksell Gamma KnifeSociety reported 28,306 cases of patients who had vestibular schwannoma hadbeen treated worldwide (181 out of 213 centers reporting).3

a Department of Otolaryngology, Northwestern University Medical School, Chicago, IL 60611,USAb The Ear Institute of Chicago, LLC, 11 Salt Creek Lane, Suite 101, Hinsdale, IL 60521, USAc Illinois Gamma Knife Center, Alexian Neurosciences Institute, Alexian Brothers MedicalCenter, Elk Grove Village, IL 60007, USA* Corresponding author. The Ear Institute of Chicago, 11 Salt Creek Lane, Suite 101, Hinsdale,IL 60521.E-mail address: [email protected]




Battista636

Based on these data, it seems that there is a steady, growing demand for GKRS astreatment of vestibular schwannomas. This may be due, in part, to improved patientoutcomes and an expanding body of knowledge of long-term tumor control. Prior to1992 high doses (typically 16 Gy or greater to the tumor margin) resulted in significantrates of facial weakness (21%) and facial numbness (27%) after GKRS.4 Since then,dose reduction (typically 12–13 Gy to the tumor margin), the introduction of MRI scans,and more sophisticated dose planning software and equipment have improvedoutcomes considerably. This article outlines the technique of GKRS and discussesthe current results of the use of this technique to treat vestibular schwannomas usingmarginal doses of 12 to 13 Gy or less.

STEREOTAXY

The aim of the stereotactic technique as it pertains to the gamma knife is to relate thelocation of structures or lesions within the boundaries of a head frame to a 3-D Carte-sian axis system. This is done initially by fixing a rigid titanium head frame to the head(Fig. 1). The borders of the frame then constitute the Cartesian axes, whereas thecranium serves as a platform to support the frame. The way in which a point withinthe frame is defined in terms of the Cartesian axes is depicted in Fig. 2. By convention,the X axis runs from side to side, the Y axis runs from behind forwards, and the Z axisruns from above downwards. The common zero point is above and behind the rightear. It is the frame and not the head that is used to localize. This is helpful becausethere may be times when it is necessary to place the frame eccentrically or to rotatethe frame with respect to the head.

Reference points relative to the titanium head frame must be visible on radiographicimaging. This is accomplished by means of plastic plates that are attached to eachside of the head frame. Embedded in the walls of the plastic plates are three metalrods or strips (CT), or three hollow channels filled with a copper-containing liquid(MRI). These rods or channels are radiopaque, creating fiducial lines. Of the three

Fig. 1. Leksell titanium head frame. (Courtesy of Elekta Inc., Norcross, GA; with permission.)

Fig. 2. Cartesian axis system as referenced to the head frame.

Gamma Knife Radiosurgery 637

rods or channels, two are parallel and vertical. The third rod or channel is oriented ata 45� angle to the other two. This creates three fiducial lines in the configuration ofa backwards ‘‘N’’. In this way, on each image slice, three sets of dots are identifiable(Fig. 3). As one moves through an imaging set, only the middle dot moves sequentially.Determination of the separation of the two fixed dots in relation to the moving middledot makes calculation of the Z coordinate for each slice a simple geometric exercise.Thus, a stereotactic space is created.

THE GAMMA KNIFEMACHINE

There have been several generations of the Leksell Gamma Knife machine. The LeksellGamma Knife 4B and 4C models are the machines used most commonly at present.The basic construction of the model 4 is described here.

The Leksell Gamma Knife 4C unit (Fig. 4) consists of a thick-cast, hemispheric steelshell containing a dome-shaped core of 201 cylindric cobalt-60 sources, each 1 mm indiameter and 20 mm in length. These sources are radially aligned toward a commonfocal point. The steel shell is equipped with two thick mechanical doors through whichthe head of the patient is introduced. Each cobalt-60 source undergoes a primarycollimation within the casting. A second collimation is performed through a helmet

Fig. 3. Schematic of fiducial lines created by the MRI or CT box attachment to the Leksellhead frame. Two separate ‘‘slice’’ examples are shown.

Fig. 4. (A) Leksell Gamma Knife 4C unit, front view. (B) Schematic of Leksell Gamma Knife 4Cunit, side view. (Courtesy of Elekta Inc., Norcross, GA; with permission.)

Battista638

in which the patient’s head is positioned. The collimator helmet bears 201 collimatorsthat are designed to align with the corresponding primary collimator. The collimatorhelmet is fixed to a movable bed on which the patient lies. With the patient’s headin the Leksell head frame, the head frame is likewise fixed to the bed. The frame isplaced inside the concavity of the collimator helmet so that the targeted lesion is atthe center of the helmet. When the primary collimator and the collimator helmet arealigned, it is possible for the gamma rays to converge on the target.

Four collimator helmet sizes are available corresponding to different aperture sizes(4, 8, 14, and 18 mm). Helmet sizes are chosen based on the prescribed treatment plan(discussed later). More than one size of helmet may be used throughout the course ofa treatment.

The only movements in the Leksell Gamma Knife system are that of the bed, whichdraws patients in and out, and that of the doors at the mouth of the machine, whichopen prior to movement of the bed. When the treatment is over, the bed automaticallymoves out and the doors close.

INDICATIONS/TREATMENT GOALS

The goals of any form of treatment for vestibular schwannoma should be long-termtumor control, preservation of cranial nerve function, and maintenance of a highquality of life. The aim of any form of treatment for vestibular schwannoma is to prevent


the functional consequences due to growth of the untreated tumor rather thanimprovement of preoperative symptoms. GKRS attempts to achieve this goal by stop-ping tumor growth. With this in mind, the indications for GKRS have expanded over theyears as more centers have gained experience with this technique. Common indica-tions for GKRS for vestibular schwannoma are listed in Table 1. It is recommendedthat patients who have tumors larger than 3 cm in diameter not undergo GKRSbecause of the high probability of radiation damage that may occur to the surroundingbrainstem. In addition, patients who have signs of intracranial hypertension or brain-stem compression should not undergo GKRS. These patients need rapid decompres-sion of their tumor and for this reason require microsurgical removal.

TECHNIQUEPre-gamma Knife Evaluation

Prior to undergoing GKRS, patients are evaluated with a thorough neurotologic historyand physical examination, high-resolution MRI with gadolinium, and audiometric eval-uation to include pure-tone air/bone and word recognition score testing. Hearing isgraded with the American Academy of Otolaryngology–Head and Neck Surgery1995 guidelines5 or the Gardner-Robertson6 classification system (Table 2). Facialnerve function is assessed most commonly using the House-Brackmann (H-B)grading system.7

After the initial evaluation, gamma knife treatment consists of four main steps: (1)placement of a head frame; (2) an imaging study; (3) treatment/dose planning; and(4) radiation delivery.

Several specialists are typically involved in the gamma knife treatment process.These specialties include a radiation oncologist, a physicist, and a neurosurgeon orneurotologist. Most commonly, the neurosurgeon or neurotologist places the headframe while three specialists—a radiation oncologist, physicist, and neurosurgeonor neurotologist—all are involved in the treatment planning.

Placement of Head Frame/Imaging Study

The gamma knife procedure begins with a patient’s head undergoing rigid fixation inan MRI-compatible Leksell stereotactic head frame. Fixation is performed with varioussize pins in a bifrontal and bicoronal fashion after application of a local anesthetic. Anoral sedative often is given prior to placement of the head frame. The frame is placed inan attempt to locate the tumor as close to the center of the frame as possible. Thisensures proper positioning of the head frame and the targeted lesion in or near thecenter of the collimator helmet.

Plastic plates with radiopaque lines for fiducial marking are attached to the headframe. High-resolution MRI then are obtained after double-dose gadolinium is given.Double-dose gadolinium is given to maximize contrast between tumor andsurrounding tissue. Volume acquisition studies require 1- to 1.5-mm axial slice thick-nesses that subsequently are reformatted in coronal and sagittal projections. At the

Table 1Indications for gamma knife radiosurgery

1. Small- to medium-sized tumor (tumor <3 cm in maximum diameter)

2. Elderly/medically infirm patient

3. Recurrent/residual tumor

4. Patient choice

Table 2Gardner-Robertson hearing classification

Class Pure ToneAverage (dB) Speech Discrimination (%)1—Good 0–30 70–100

2—Serviceable 31–50 50–69

3—Nonserviceable 50–90 5–49

4—Poor 91 maximum 1–4

5—None Nontestable 0

If the pure-tone average and speech discrimination scores do not qualify for the same class, theclass appropriate for the poorer of the two scores is used.

Data from Gardner G, Robertson JH. Hearing preservation in unilateral acoustic neuromasurgery. Ann Otol Rhinol Laryngol 1988;97(1):55–66.

Battista640

author’s treating facility, T1 and T2 constructive interference in steady state magneticresonance sequences are obtained. The T1 images define the tumor margins whereasthe constructive interference in steady state sequences help to define cranial nerves.

Treatment Planning/Radiation Delivery

The imaging study is downloaded into the treatment software in order to begin treat-ment planning. The goal of GKRS treatment planning is to provide the maximum, safeamount of radiation to a tumor without damaging surrounding structures. Typicalparameters considered by a gamma knife team are the number and position of isocen-ters, their collimator sizes, their relative weights, and the prescription isodose. Theplan of radiation delivery is designed to closely adhere, or conform, to the shape ofthe tumor with a sharp falloff of radiation at the edge of the tumor. A high degree ofconformity of the prescription isodose to the target volume, while ensuring that tumoris completely encompassed, is one of the prerequisites for complication-free lesioncontrol with good tumor control.8 These parameters and goals are described morefully later.

The center of the radiosurgery beam is called the isocenter. Treatment planningconsists of placing these beams, or shots, over the target volume. The isocenter radi-ation distribution tends to be circular or elliptic in shape. A typical radiosurgical treat-ment can be conceived of as trying to fill an irregular-shaped volume (the tumor) with‘‘balls’’ of different sizes. Thus, the Leksell Gamma Knife has 4-, 8-, 14-, and 18-mm‘‘balls’’ (corresponding to the different-sized collimator helmets) with which dose planscan be created. The goal is to maximally fill the treatment volume with these shots.Because most tumors are not truly spherical, the dose from a single beam of radiationwill not match the shape of the tumor. By overlapping multiple isocenters of varyingsizes, the dose distribution can be fashioned to produce a conformal treatmentplan. The ‘‘weight’’ of each isocenter can be adjusted, if necessary, to further conformto the tumor volume.

The amount of radiation given is called the prescribed isodose. The distribution of theisodose around the tumor can be demonstrated visually through circumferential linesthat extend outward from the target. These lines are the isodose curves (or isodoselines) of radiation distribution (Fig. 5). The isodose lines are expressed in percentages.Assuming the center, or near center, of the target receives 100% radiation, the isodoselines are constructed with a 10% decrement: 90%, 80%, 70%, and so forth.

The prescription isodose is defined most commonly as the amount of radiationdelivered to the 50% isodose line. Most gamma knife plans prescribe the 50% isodoseto the margin of the tumor. This is because there is a steep falloff in radiation dosage at

Fig. 5. Leksell gamma Plan for a right-sided vestibular schwannoma. 13 Gy is prescribed tothe 50% isodose line (yellow line).


the edge of the 50% isodose line, which ensures radiation sparing of surroundingstructures. Gamma knife prescription isodoses typically are 12 to 13 Gy to the marginof the tumor. This also may be expressed as 12 to 13 Gy prescribed to the 50%isodose line.

Factors taken into account when determining a prescription isodose include the sizeand location of the tumor, the hearing level, and the tumor status (primary versusrecurrent/residual). In general, 12 Gy is given when the tumor of medium size or thehearing is serviceable. When the tumor is small, the hearing is nonserviceable, andthe tumor is recurrent/residual, 13 Gy is typically considered.

Finally, the gamma knife planning software is capable of mapping the location of themaximal dose, or hot spot, in the stereotactic space. A gamma knife team attempts tokeep the hot spot well away from the facial nerve and, if necessary, the cochlear nerve.

Using the Leksell Gamma Knife 4C unit, the actual time of radiation delivery variesfrom approximately 45 to 120 minutes. At the conclusion of the procedure, the headframe is removed and patients receive intravenous steroids (typically 40 mg methyl-prednisolone). Patients are observed for a short period and are discharged thesame day unless immediate complications (rare) develop, such as vomiting, dizzinessor headache.

FOLLOW-UP

Patients are followed with serial contrast-enhanced MRI studies, which typically areobtained at 6 months, 12 months, 2 years, 4 years, 8 years, and 16 years. Initial radio-graphic follow-up often demonstrates loss of central enhancement along with slighttumor enlargement and capsular thickening. Audiometric testing is performed at thetime of their MRI follow-up for patients who have preserved hearing. Patients typicallyreturn to their pretreatment level of functioning or employment within 3 to 10 days posttreatment.9

RESULTS/COMPLICATIONSTumor Control

Currently there is no consensus as to the definition of tumor control after radiosurgery.One nonstandardized means to describe tumor control is tumor shrinkage or no tumor

Battista642

growth, also known as radiologic control. A widely accepted definition of tumor growthis that of Flickinger and colleagues,10 who describe tumor growth as a 1-mm increasein tumor diameter in any two directions or 2-mm in one direction. Another definition ofgrowth is a greater than 10% increase in tumor volume11 whereas another is anincrease of 100% of the initial volume at 2 years post treatment.12

Clinical, rather than radiologic, tumor control is another means of defining theoutcome of GKRS. Clinical tumor control means, ‘‘the absence of the need for furthermicrosurgical or radiosurgical intervention.’’13 Patients who require microsurgical orradiosurgical treatment after radiosurgery are patients who have documentedprogression of tumor growth or worsening symptoms, such as ataxia.

Tumor expansion after vestibular schwannoma radiosurgery does not necessarilymean a lack of tumor control, however. Transient tumor swelling occurs in as manyas 80% of cases after GKRS.9,14,15 The swelling typically peaks between 6 to 7months16,17 but may continue for 1 to 3 years after treatment.18–20 In one study, themean increase in tumor volume was 120% of the initial tumor volume.16,19,20 Pollock20

and Myrseth and colleagues21 have noted that some schwannomas may enlarge afterGKRS and continue to remain stable in size. In a study by Pollock,20 8 of 28 (29%) oftumors enlarged after GKRS and remained stable in size for a maximum follow-upperiod of 8.2 years (median 56 months).

Transient tumor swelling, or growth that plateaus after radiosurgery, may lead tounnecessary microsurgery or revision radiosurgery.19,20 In order to differentiate truetumor growth from tumor expansion, the tumor size must be mapped with serialMRI scans. Hasegawa and colleagues19 also have suggested that magnetic reso-nance spectroscopy, single photon emission tomography, or positron emissiontomography may help determine if tumors are actively growing or showing signs oftemporary expansion.

Tumor regression is a favorable prognostic sign of tumor control. Tumors that shrinkrarely show signs of growth.22 In addition, the chance for tumor growth is extremelylow for stable tumors 4 years post GKRS.13,18,23 Continuous monitoring after 4 yearsis necessary, however, because there has been at least one report of a patient devel-oping an expanding cyst at the tumor margin 60 months after GKRS.18

Some factors associated with continued tumor growth for unilateral vestibularschwannoma after gamma knife treatment have been imaging/targeting error, tumorbiology, and large tumor volume and cystic tumor (the latter two factors are discussedlater). Spatial distortion is known to occur with MRI.24,25 These distortions are mostpronounced at soft tissue-bone interfaces, such as that in the internal auditory canal.As a means to accommodate this distortion and to prevent imaging/targeting error,some centers combine CT and MRI.16,26 Another possible cause for MRI error isthat gadolinium may not enhance the lateral most extent of the intracanalicular portionof a vestibular schwannoma.27

There has been no difference, however, in tumor control rates between patientsreceiving current doses of 12 to 13 Gy to the tumor margin and those patients whohave received higher doses.10,13

Table 3 provides a summary of tumor control rates from various, worldwide treat-ment centers. Inclusion criteria for the table were the following: minimum 100 patientsreported, mean marginal dose of 13.5 Gy or less, and an average of 3 years of follow-up or longer. In a study by Myrseth and colleagues,21 tumor size was reported as themaximal tumor diameter, including the intracanalicular portion (17.5% were 0–10 mm;66% were 11–20 mm; and 16.5% were 21–30 mm).

Based on studies (see Table 3), radiologic control ranged from 89% to 96%,whereas clinical control (no further radiation treatment or microsurgery) ranged from

Table 3Control rates from various treatment centers for patients underwent gamma knife radiosurgery for vestibular schwannoma

Author,Year N

PreviousMicrosurgery(N)

Mean TumorVolume (Range)

Mean MarginalDose (Range)

Mean Follow-up(Range)

NoGrowth(%)

Regression(%)

Growth(%)

UnderwentFurtherMicrosurgery orRadiosurgery(%)

Prasad,200028

153 57 2.8 cm3 (0.02–18.3 cm3) 13.3 Gy (9–20 Gy) 38 mo (1–10 yrs) 17 75 8 NR

Litvack,200329

134 18 NR 12 Gy (�0.6 Gy) 36 mo (12–72 mo) 61 46 4 2

Chung,200530

195 76 4.1 cm3 13 Gy (median)(11–18 Gy)

36 mo (2–109 mo) 35 58 7 3

Lundsford,200513

157 NR 2.5 cm3 12–13 Gy 120–180 mo 25 73 NR 2

Myrseth,200521

102 5 See text 12 Gy (10–20 Gy) 71 mo (12–170 mo) 40 49 11 5

Hasegawa,200518

301 72 <15 cm3: 277 R15 cm3: 24 13.2 Gy 91 mo median 31 62 7 9

Hempel,200631

116 NR 1.6 cm3 (0.1–9.9 cm3) 13 Gy (median)(10–14.5 Gy)

98 mo (63–129 mo) 40 56 4 3

As a reference, a tumor volume of 15 cm3 corresponds approximately to a mean intracranial diameter of 3 cm.Abbreviation: NR, not reported.

Gam

ma

Kn

ifeR

ad

iosu

rgery

643

Battista644

91% to 98%. Tumor growth was variously defined as greater than 2 mm in meanpretreatment tumor diameter,18,29 greater than 110% of the original tumor volume,30

a 1-mm increase in tumor diameter in any two directions or 2-mm in one direction,13

and no growth or tumor regression.21 Two studies did not clearly define tumorgrowth.28,31

Two results in the table should be clarified. First, the definition of radiologic tumorcontrol is strict in the study by Myrseth and colleagues21 (no growth or tumor regres-sion). Four of 102 patients in this study had a growth plateau. If these four patientswere taken into account, the radiologic control rate would be 93%. Secondly, in thestudy by Hasegawa and colleagues18 the clinical control rate of 91% may be skewed.In this study, six patients (2%) underwent microsurgery after GKRS at a separate insti-tution. These patients had unchanged tumor volume or slight tumor expansion withoutneurologic changes. It is possible that microsurgery may have been unnecessary ifthere was a longer period of observation.

Repeat GKRS for continued tumor growth has been reported.18,32 In one report, sixpatients underwent repeat GKRS between 24 and 70 months after the first gammaknife treatment.18 The mean marginal dose was 11 Gy. Three of the six patientsachieved tumor shrinkage without complication and the other three required microsur-gery for continued growth. In another report, eight patients underwent repeat GKRS.32

Indications for retreatment were documented tumor growth for at least 3 years afterinitial GKRS. The median marginal tumor dose at retreatment was 12 Gy (range,10–12 Gy). Follow-up after repeat GKRS was from 26 to 121 months, with six patientsdemonstrating regression and two patients stabilization of tumor growth at last follow-up. There were no new neurologic deficits.

Acute Side Effects/Complications

The vast majority of acute side effects after GKRS are mild and brief.33,34 These side-effects may include headache, nausea, vomiting, marked fatigue, and pin-site relatedproblems.

There have been a few reports of acute facial paralysis occurring between 1 and7 days after GKRS.35–38 In all cases, the facial paralysis improved to at least a H-Blevel 2 within 6 to 24 months after onset. Oral or intravenous steroids were given ineach case, but it is unknown if this treatment improved facial function. A case ofpermanent, profound sensorineural hearing loss also has been reported shortly afterGKRS. The hearing loss was the result of acute intracochlear hemorrhage occurringwithin 24 hours of GKRS.35

Hearing Results

Preservation of functional hearing (Gardner-Robertson class 1 or 2) has been reportedas ranging from 33% to 79%, with observation periods of 2 to 15 years of follow-up(Table 4). There also are sporadic reports of permanent hearing improvement afterGKRS.9,12,13,28,29 In those patients whose hearing worsens, hearing loss typicallydoes not develop until 3 to 12 months after gamma knife treatment. Once hearingloss develops, the hearing deterioration may continue up to the eighth year posttreatment.28 In general, hearing preservation rates worsen as tumor size increases.29

The cause for hearing deterioration after GKRS is not precisely known. Some theo-ries of hearing deterioration include direct radiation damage to the auditory system,40

decreased blood flow to the auditory system due to hyalinization of blood vessels, andtumor enlargement in the internal auditory canal. Paek and colleagues40 measured theradiation doses delivered to the cochlea, cochlear nerve, and cochlear nucleus forpatients treated with gamma knife for vestibular schwannoma. The investigators

Table 4Hearing preservation rates after gamma knife radiosurgery

Author,YearMeanMarginalDose (Range)

Mean Follow-up(Range)

PreoperativeWith FunctionalHearing

PostoperativeWith FunctionalHearing

Preservation(%)

Prasad,200028

13.3 Gy(9–20 Gy)

51 mo 36 21 58

Andrews,200139

12 Gy 30 mo 69 23 33

Rowe,200222

14.6 Gy(13–15 Gy)

35 mo 49 37 76

Regis,20029

(12–14 Gy) 48 mo 48 24 50

Litvack,200329

12 Gy(�0.6 Gy)

26 mo(12–60 mo)

47 29 62

Chung,200530

13 Gy(median)(12–17 Gy)

31 mo median(6–74 mo)

20 12 60

Lunsford,200513

(12–13 Gy) 36–180 mo 267 210 79

Hasegawa,200518

%13 Gy 91 mo median 74 50 68

Functional hearing defined as Gardner-Robertson class 1 or 2.


concluded that the maximum dose delivered to the cochlear nucleus was the mostsignificant prognostic factor of hearing deterioration.

Facial Paralysis

New-onset, permanent facial neuropathy (defined as a decline in H-B facial nervegrade) is rare after GKRS using the current marginal doses of 12 to 13 Gy. Rates ofnew-onset, permanent facial neuropathy have been reported as ranging from 0% to0.5%.9–11,21,22,30,31,41 Transient facial paralysis has been reported with rates ofapproximately 1% to 1.5%.9–11,21,22,30,31,41 Transient facial paralysis may last up to6 months. If facial paralysis develops, it usually develops within the first year aftertreatment but has been reported as late as 62 months after GKRS.30

Trigeminal Neuralgia/Neuropathy

New-onset, temporary or permanent trigeminal nerve symptoms, ranging from numb-ness to trigeminal neuralgia, are uncommon. Rates for trigeminal symptoms rangefrom 0.2% to 4%.9–11,13,28,30 There also are a few reports of improvement of pretreat-ment trigeminal symptoms.9,22,29,30 Similar to facial nerve dysfunction, trigeminalsymptoms have been reported to develop between 5 and 48 months posttreatment.10,28

Hemifacial Spasm

Approximately 2% to 4% of patients develop hemifacial spasm after GKRS.9,20,42 Inmost cases, the symptoms improve with time or respond to carbamazepine. Hemifa-cial spasm has been reported to occur as early as 1 month after radiosurgery42

whereas several reports note that spasm typically develops between 1 and 2 years

Battista646

after radiosurgery.9,20 Pollock20 has postulated that hemifacial spasm is the result oftumor expansion with irritation of the facial nerve or a delayed vascular insult.

Dizziness

The symptom of dizziness is rarely discussed in the current gamma knife literature.Two articles that specifically address this condition report new-onset dizziness afterGKRS in 16 of 120 patients (13%)31 and in 27 of 104 patients (26%),9 respectively.If dizziness develops, it often is transient but may last for several weeks.9,31 TheUniversity of Pittsburgh report on 829 cases states that their group has no data tosupport the notion that dizziness before treatment is made worse by GKRS.13

As a means to treat patients who have dizziness before GKRS, the author’s grouphas begun to use intratympanic gentamicin prior to and, as necessary, after GKRS.This treatment is based on the finding of Brantberg and colleagues43 who successfullytreated a patient who had dizziness and a vestibular schwannoma and who refusedmicrosurgery and radiosurgery.

Hydrocephalus

Hydrocephalus in the absence of progressive tumor growth has been reported asoccurring in 0% to 7% of patients.9,12,13,18,21,28 The median time to development ofhydrocephalus is approximately 1 year. Hydrocephalus is believed the result of tumornecrosis, with proteinaceous debris blocking cerebrospinal fluid flow. The develop-ment of hydrocephalus is more common after treatment of larger (>25 mm diameter)tumors. The hydrocephalus may resolve spontaneously44 but most often requires ven-ticuloperitoneal shunting.

SPECIAL SITUATIONSLarge Tumors

GKRS is not an ideal treatment for patients who have large vestibular schwannomas(mean tumor diameter greater than 3 cm). One reason is that there is an increased riskfor complications relative to smaller tumors. Significant tumor edema invariably occursafter GKRS for tumors larger than 3 cm in diameter.45 The edema has the immediateeffect of potentially causing severe headache, vomiting, and dizziness. Brainstemcompression may develop leading to ataxia and obstructive hydrocephalus. Asa means to avoid these complications, it has been recommended that the doseprescribed be decreased when treating large tumors.23 Marginal doses lower than12 Gy, however, have been associated with increased tumor recurrence.46

A second reason that GKRS is not ideal for large tumors is that large tumors demon-strate reduced rates of tumor control compared to smaller tumors even with marginaldoses of 12 Gy or higher. In one study, the 10-year progression-free survival rates forvestibular schwannoma GKRS was found statistically better for small- to-medium-sized tumors (<15 cm3 in total volume) versus larger tumors (>15 cm3) (95% versus57% progression-free survival, respectively).18 These results indicate that largetumors (>15 cm3) should not undergo GKRS as initial treatment. A more favorableoption for treatment of large vestibular schwannomas is complete microsurgical exci-sion and, when necessary, adjuvant GKRS for residual tumor.

Residual/Recurrent Vestibular Schwannoma After Microsurgery

GKRS has been used to treat residual and recurrent vestibular schwannoma aftermicrosurgery.45,47–49 In general, reported rates of tumor control are similar to thoseof primary GKRS for unilateral tumors (Table 5).

Table 5Summary of reports for gamma knife radiosurgery for residual or recurrent vestibular schwannoma after microsurgery

Author N Residual RecurrentMedian TimeAfterMicrosurgery (Range)

Median TumorVolume

Median MarginalDose (Range)

MedianTime ofFollow-up (Range) Facial Paralysisa Tumor Controlb

Pollock et al,199848

78 52 26 57 mo (12–144 mo) 2.8 cm3 15 Gy (12–20 Gy) 37 mo (12–101 mo) 23% 92%

Unger et al,200249

50 34 16 39 mo (6–324 mo) 3.4 cm3 13 Gy 75 mo (42–114 mo) 0% 96%

Yang et al,200745

61 — — 5.8 mo (0.3–96 mo) 3.65 cm3 12.5 Gy mean(9–14 Gy)

53.7 mo (24–102 mo) 0% 98%

a Facial paralysis: new or worsened paralysis of patients who have pre-GKRS with H-B grade I–III.b Tumor control: absence of the need for further microsurgery or radiosurgery.

Gam

ma

Kn

ifeR

ad

iosu

rgery

647

Battista648

In a study by Pollock and colleagues,48 76 patients (78 tumors) were treated withGHRS for residual/recurrent vestibular schwannoma after microsurgery. Documentedtumor growth was the indication for treatment in 86% of cases. The indication fortreatment in the remaining 14% of cases was not reported. Of the 47 patients whohad H-B grade I–III facial function prior to GKRS, 11 patients developed new orincreased facial weakness at a median of 7 months (range 3–11 months) afterGKRS. When these results were stratified based on a marginal dose of 12 to14 Gy,only 3 of 22 (14%) patients developed new or worsened facial paralysis. Twelvepercent of patients developed new, permanent trigeminal symptoms. All three patientslost hearing and had Gardner-Robertson class 1 or 2 hearing prior to GKRS. Onepatient required a ventriculoperitoneal shunt for hydrocephalus. There were sixpatients who required revision microsurgery after radiosurgery because of increasedataxia (two patients) or documented tumor growth (four patients). Failure of GKRS wasdocumented between 12 and 45 months after treatment.

In a study by Unger and colleagues,49 the indications for GKRS for residual/recur-rent vestibular schwannoma after microsurgery were not reported. There was tempo-rary facial paralysis and transient trigeminal dysfunction in five and four cases,respectively. There was no change in pre-existing hearing status in any patient. Onepatient required a ventriculoperitoneal shunt for hydrocephalus. After gamma knife,two tumors required revision microsurgery at 4 and 18 months post radiosurgerydue to progressive tumor enlargement.

In the report by Yang and colleagues,45 the indications for GKRS also were notreported. There were no reported changes in hearing or trigeminal function as a resultof GKRS. At last follow-up, four patients had tumor enlargement. One of these fourpatients required revision microsurgery whereas the other patients had continuedtumor observation due to lack of symptoms.

Based on the results of these studies, it seems that GKRS is an acceptable thera-peutic option for small- to medium-sized residual/recurrent vestibular schwannomaafter microsurgery. Adjuvant GKRS for residual vestibular schwannoma should bereserved for cases of documented, progressive tumor growth, because it is knownthat residual tumors may not grow.50,51

Cystic Tumors

Vestibular schwannoma cysts may be classified as peritumoral, when the cyst lies onthe outside of the tumor, and intratumoral when inside. Relative to solid vestibularschwannoma, vestibular schwannoma with intratumoral cysts is associated morecommonly with sudden expansion,52–54 shorter symptom duration,52,54,55 atypicalinitial symptoms (such as dysgeusia, vertigo, facial pain and unsteadiness),55 andan increased rate of preoperative facial palsy.52,55,56

Treatment results are mixed regarding the use of the gamma knife to treat intratumoralcystic vestibular schwannomas. In one study, six cases of vestibular schwannoma withmacrocystic components (cyst component of at least one third of the entire tumorvolume) demonstrated significant enlargement after treatment with GKRS.57 The sixcases were from a series of 74 cases of vestibular schwannoma treated with thegamma knife. Cyst enlargement developed between 1 and 8 months after GKRS.Cyst enlargement was associated with new or worsened neurologic symptoms in allcases. Three cases required microsurgery for treatment, whereas two casesdecreased spontaneously and one case remained stable. Neurologic symptomsresolved in the two cases that decreased spontaneously.

In another study by Hasegawa and colleagues19 there was no correlation foundbetween the presence of pretreatment intratumoral cysts and the chance for


post-treatment tumor expansion. Cyst formation after GKRS, however, was correlatedsignificantly with tumor expansion associated with neurologic changes.

Neurofibromatosis Type 2

Approximately 5% of patients who have a vestibular schwannoma have neurofi-bromatosis type 2 (NF-2). NF-2 vestibular schwannomas are histologically distinctfrom sporadic vestibular schwannomas.58 Only a few groups have reported theirresults using GKRS for NF-2 vestibular schwannoma. Table 6 summarizes theseresults.

Kida and colleagues59 reported the results of 20 patients who had NF-2 who under-went GKRS to manage 20 vestibular schwannomas. Seventy percent of these patientshad prior microsurgery. The group reported 100% tumor control (50% no growth and50% regression) after a mean follow-up period of 33.6 months. Transient radiation-induced facial paresis, hemifacial spasm, and ataxia occurred in two patients each,but each symptom resolved completely.

Roche and colleagues60 reported the outcome of 27 patients who had NF-2 inwhom GKRS was used to manage 37 vestibular schwannomas. There were twotemporary facial nerve deficits and one permanent. The group reported that 26(74%) of the tumors were controlled by GKRS (40% no growth and 30% regression)at last follow-up. Three tumors required microsurgical removal for continued growth.The treatment required for the remaining eight cases was not reported.

Rowe and colleagues61 reported a large series of NF-2 patients who had vestibularschwannoma managed using GKRS. They irradiated 122 tumors in 96 patientsbetween 1986 and 2000. Twenty percent of the cases were treated with one ormore prior microsurgical resections. The majority of patients were treated for docu-mented tumor growth. For those cases without tumor growth, other indications forGKRS included a rapid deterioration in hearing, tumor near 3 cm in size, or plannedadjuvant therapy for significant residual tumor after microsurgery. The serviceablehearing preservation rate was 22%. When hearing results were examined usingcurrent marginal radiation doses (10–16 Gy, mean 13 Gy), 38% of the tumors hadhearing preservation at 3 years after radiosurgery. In this lower-dose group, facialand trigeminal function worsened in 8% and 2% of the patients, respectively. Therewas a statistical reduction in cochlear, facial, and trigeminal dysfunction with reduced

Table 6Results of gamma knife radiosurgery for patients who had neurofibromatosis type 2

SeriesPatients/Tumors

AverageFollow-up(Range)

AverageTumor Size

AverageMarginalDose(Range)

ServiceableHearingPreservationa

LocalControlRateb

Kida et al,200059

20 patients/20 tumors

34 mo(18–84 mo)

24.4 mm 13 Gy(10–15 Gy)

33.3% 100%

Roche et al,200060


62 mo(27–123 mo)

4.0 cm3 13 Gy(10–18 Gy)

57% 74%

Rowe et al,200361


50 mo(4–154 mo)

5.3 cm3 15 Gy(10–25 Gy)

22% 79%

Mathieuet al, 200762


64 mo(4–196 mo)

5.7 cm3 14 Gy(11–20 Gy)

40% 88%

a Serviceable hearing defined as Gardner-Robertson class 1 or 2.b Local control rate defined as no further microsurgery or radiation after GKRS.

Battista650

radiation dosage. Tumor volume was found to be the only factor predictive of localcontrol, with worse control associated with larger tumors.

Mathieu and colleagues62 reported the results of 60 patients and 72 tumors treatedbetween 1987 and 2005. Twenty-one tumors in 17 patients had at least one microsur-gical resection before GKRS. The indications for GKRS were progressive tumorgrowth, recurrence of tumor after gross total resection, continued growth after frac-tionated radiation therapy, or hearing deterioration. Overall, the serviceable hearingpreservation rate was 40%. When hearing results were examined using marginal radi-ation doses less than 14 Gy, the serviceable hearing preservation rate was 53%. Atlatest evaluation, 39% of tumors regressed in size, 44% were stable in size, and17% (12 tumors) increased in size. Nine of these 12 enlarging tumors underwentmicrosurgical resection, further GKRS, or both. The treatment required for the remain-ing three failures was not reported. As with the study by Rowe and colleagues,61 therewas a statistical reduction in control rate with increasing tumor size.

Based on these results, GKRS seems less efficacious in treating vestibular schwan-noma secondary to NF-2 when compared with sporadic tumors. This finding is mostlikely due to the differences in tumor biology of NF-2 tumors and the tendency of thesetumors to invade adjacent cranial nerves.58 Despite the differences in resultscompared to sporadic tumors, GKRS does seem to have a role to play in treatingNF-2 vestibular schwannoma. As more data become available using GKRS forNF-2, the management strategies for these tumors may become more clearly defined.

FURTHER CONSIDERATIONS

The Leksell Gamma Knife 4C system and earlier models are limited to treatment ofcranial lesions. The latest gamma knife machine, first put into practice in 2006, theLeksell Gamma Knife Perfexion, has the capability to treat lesions of the cervical spineand other areas of the head and neck. The Perfexion unit is based on a single, inte-grated permanent collimator system (Fig. 6). The collimators are partitioned into eightsegments around the circumference of the device, each containing an independentlymoveable sector. These eight sectors, each containing 24 cobalt-60 sources, areoperated by individual servo drives. Depending on the sector position, the collimatorsize of each sector can be individually varied between 4, 8, and 16 mm. In this way, theneed for changing the collimator helmet during treatment is no longer necessary,which can result in a reduction in treatment time.

In addition to improvements in radiation delivery, MRI and CT technology certainlywill continue to evolve. Improvements in radiologic techniques will translate directly toimprovements in accuracy and predictability of 3-D stereotactic radiation.

Fig. 6. Schematic drawing of side view of Leksell Gamma Knife Perfexion showing one unithousing 4-, 8- and 16-mm collimators. (Courtesy of Elekta Inc., Norcross, GA; with permission.)


SUMMARY

GKRS is now an established means of treatment of small- to medium-sized vestibularschwannomas. The procedure has a low morbidity when used appropriately. Tumorcontrol, however, needs to be better defined. The results of GKRS treatment ofresidual/recurrent vestibular schwannoma after microsurgery suggest that GKRS isa viable treatment option for residual/recurrent vestibular schwannoma. Some centersare now reporting their results of revision GKRS for vestibular schwannoma; the long-term results of retreatment are, as yet, unknown. When radiosurgery is offered astreatment, the prospect of radiosurgical failure, and the results of subsequent treat-ment for radiosurgical failure, must be discussed with patients.

REFERENCES

1. Leksell L. A note on the treatment of acoustic tumors. Acta Chir Scand 1971;137:763–5.

2. Society Leksell Gamma Knife. World wide indications treated. Available at:http://www.elekta.com/healthcare_us_treatment_statistics.php; 2006. AccessedJanuary 20, 2008.

3. Society Leksell Gamma Knife. World wide indications treated. 2004.4. Flickinger JC, Kondziolka D, Pollock BE, et al. Evolution in technique for vestibular

schwannoma radiosurgery and effect on outcome. Int J Radiat Oncol Biol Phys1996;36(2):275–80.

5. American Academy of Otolaryngology–Head and Neck Surgery Foundation, INC.Committee on Hearing and Equilibrium guidelines for the evaluation of hearingpreservation in acoustic neuroma (vestibular schwannoma). American Academyof Otolaryngology–Head and Neck Surgery Foundation, INC. Otolaryngol HeadNeck Surg 1995;113(3):179–80.

6. Gardner G, Robertson JH. Hearing preservation in unilateral acoustic neuromasurgery. Ann Otol Rhinol Laryngol 1988;97(1):55–66.

7. House JW, Brackmann DE. Facial nerve grading system. Otolaryngol Head NeckSurg 1985;93(2):146–7.

8. Lomax NJ, Scheib SG. Quantifying the degree of conformity in radiosurgery treat-ment planning. Int J Radiat Oncol Biol Phys 2003;55(5):1409–19.

9. Regis J, Pellet W, Delsanti C, et al. Functional outcome after gamma knife surgeryor microsurgery for vestibular schwannomas. J Neurosurg 2002;97(5):1091–100.

10. Flickinger JC, Kondziolka D, Niranjan A, et al. Acoustic neuroma radiosurgerywith marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys 2004;60(1):225–30.

11. Wowra B, Muacevic A, Jess-Hempen A, et al. Outpatient gamma knife surgery forvestibular schwannoma: definition of the therapeutic profile based on a 10-yearexperience. J Neurosurg 2005;102(Suppl):114–8.

12. Pellet W, Regis J, Roche PH, et al. Relative indications for radiosurgery andmicrosurgery for acoustic schwannoma. Adv Tech Stand Neurosurg 2003;28:227–82 [discussion: 82–4].


14. Nakamura H, Jokura H, Takahashi K, et al. Serial follow-up MR imaging aftergamma knife radiosurgery for vestibular schwannoma. AJNR Am J Neuroradiol2000;21(8):1540–6.

http://www.elekta.com/healthcare_us_treatment_statistics.php

Battista652

15. Linksey ME, Lundsford LD, Flickinger JC. Neuroimaging of acoustic nerve sheathtumors after stereotaxic radiosurgery. AJNR Am J Neuroradiol 1991;12:1165–75.

16. Muacevic A, Jess-Hempen A, Tonn JC, et al. Results of outpatient gamma kniferadiosurgery for primary therapy of acoustic neuromas. Acta Neurochir (Wien)2004;91:75–8.

17. Yu CP, Cheung JY, Leung S, et al. Sequential volume mapping for confirmation ofnegative growth in vestibular schwannomas treated by gamma knife radiosur-gery. J Neurosurg 2000;93(Suppl 3):82–9.

18. Hasegawa T, Fujitani S, Katsumata S, et al. Stereotactic radiosurgery for vestib-ular schwannomas: analysis of 317 patients followed more than 5 years. Neuro-surgery 2005;57(2):257–65 [discussion: 257–65].

19. Hasegawa T, Kida Y, Yoshimoto M, et al. Evaluation of tumor expansion afterstereotactic radiosurgery in patients harboring vestibular schwannomas. Neuro-surgery 2006;58(6):1119–28 [discussion: 1119–28].

20. Pollock BE. Management of vestibular schwannomas that enlarge after stereo-tactic radiosurgery: treatment recommendations based on a 15 year experience.Neurosurgery 2006;58(2):241–8 [discussion: 241–8].

21. Myrseth E, Moller P, Pedersen PH, et al. Vestibular schwannomas: clinical resultsand quality of life after microsurgery or gamma knife radiosurgery. Neurosurgery2005;56(5):927–35 [discussion: 35].

22. Rowe JG, Radatz M, Walton L, et al. Gamma knife radiosurgery for unilateralacoustic neuromas. J Neurol Neurosurg Psychiatry 2003;74:1536–42.

23. Inoue HK. Low-dose radiosurgery for large vestibular schwannomas: long-termresults of functional preservation. J Neurosurg 2005;102(Suppl):111–3.

24. Borden JA, Tsai JS, Mahajan A. Effect of subpixel magnetic resonance imagingshifts on radiosurgical dosimetry for vestibular schwannoma. J Neurosurg2002;97(Suppl 5):445–9.

25. Yu C, Apuzzo ML, Zee C, et al. A phantom study of the geometric accuracy ofcomputed tomographic and magnetic resonance imaging stereotactic localiza-tion with the Leksell Stereotactic System. Neurosurgery 2001;48:1092–8.

26. Poetker DM, Jursinic PA, Runge-Samuelson CL, et al. Distortion of magneticresonance images used in gamma knife radiosurgery treatment planning:implications for acoustic neuroma outcomes. Otol Neurotol 2005;26(6):1220–8.

27. Selesnick SH, Rebol J, Heier LA, et al. Internal auditory canal involvement ofacoustic neuromas: surgical correlates to magnetic resonance imaging findings.Otol Neurotol 2001;22(6):912–6.

28. Prasad D, Steiner M, Steiner L. Gamma surgery for vestibular schwannoma.J Neurosurg 2000;92(5):745–59.

29. Litvack ZN, Noren G, Chougule PB, et al. Preservation of functional hearing aftergamma knife surgery for vestibular schwannoma. Neurosurg Focus 2003;14(5):1–5.

30. Chung WY, Liu KD, Shiau CY, et al. Gamma knife surgery for vestibular schwan-noma: 10-year experience of 195 cases. J Neurosurg 2005;102(Suppl):87–96.

31. Hempel JM, Hempel E, Wowra B, et al. Functional outcome after gamma knifetreatment in vestibular schwannoma. Eur Arch Otorhinolaryngol 2006;263(8):714–8.

32. Yomo S, Arkha Y, Delsanti C, et al. Repeat gamma knife surgery for regrowth ofvestibular schwannomas. Neurosurgery 2009;64(1):48–55.

33. St. George EJ, Kudhail J, Perks J, et al. Acute symptoms after gamma knife radio-surgery. J Neurosurg 2002;97(Suppl 5):631–4.


34. Werner-Wasik M, Rudoler S, Preston PE, et al. Immediated side effects of stereo-tactic radiosurgery. Int J Radiat Oncol Biol Phys 1999;43:299–304.

35. Franco-Vidal V, Songu M, Blanchet H, et al. Intracochlear hemorrhage aftergamma knife radiosurgery. Otol Neurotol 2007;28(2):240–4.

36. Ogunrinde OK, Lunsford LD, Flickinger JC, et al. Cranial nerve preservation afterstereotactic radiosurgery for small acoustic tumors. Arch Neurol 1995;52(1):73–9.

37. Pollack AG, Marymont MH, Kalapurakal JA, et al. Acute neurological complica-tions following gamma knife surgery for vestibular schwannoma. Case report.J Neurosurg 2005;103(3):546–51.

38. Tago M, Terahara A, Nakagawa K, et al. Immediate neurological deteriorationafter gamma knife radiosurgery for acoustic neuroma. Case report. J Neurosurg2000;93(Suppl 3):78–81.

39. Andrews DW, Suarez O, Goldman HW, et al. Stereotactic radiosurgery and frac-tionated stereotactic radiotherapy for the treatment of acoustic schwannomas:comparative observations of 125 patients treated at one institution. Int J RadiatOncol Biol Phys 2001;50(5):1265–78.

40. Paek SH, Chung HT, Jeong SS, et al. Hearing preservation after gamma knifestereotactic radiosurgery of vestibular schwannoma. Cancer 2005;104(3):580–90.

41. Hasegawa T, Kida Y, Kobayashi T, et al. Long-term outcomes in patients withvestibular schwannomas treated using gamma knife surgery: 10-year follow up.J Neurosurg 2005;102(1):10–6.

42. Noren G. Long-term complications following gamma knife radiosurgery of vestib-ular schwannomas. Stereotact Funct Neurosurg 1998;70(Suppl 1):65–73.

43. Brantberg K, Bergenius J, Tribukait A. Gentamicin treatment in peripheral vestib-ular disorders other than Meniere’s disease. ORL J Otorhinolaryngol Relat Spec1996;58(5):277–9.

44. Selch MT, Pedrose A, Lee SP. Stereotactic radiosurgery for the treatment ofacoustic neuromas. J Neurosurg 2004;101(Suppl 3):362–70.

45. Yang SY, Kim DG, Chung HT, et al. Evaluation of tumor response after gammaknife radiosurgery for residual vestibular schwannomas based on MRI morpho-logical features. J Neurol Neurosurg Psychiatry 2007;79:431–6.

46. Foote RL, Coffey RJ, Swanson JW, et al. Stereotactic radiosurgery using thegamma knife for acoustic neuromas. Int J Radiat Oncol Biol Phys 1995;32(4):1153–60.

47. Park CK, Jung HW, Kim JE, et al. Therapeutic strategy for large vestibularschwannomas. J Neurooncol 2006;77(2):167–71.

48. Pollock BE, Lunsford LD, Flickinger JC, et al. Vestibular schwannoma manage-ment. Part I. Failed microsurgery and the role of delayed stereotactic radiosur-gery. J Neurosurg 1998;89(6):944–8.

49. Unger F, Walch C, Papaefthymiou G, et al. Radiosurgery of residual and recurrentvestibular schwannomas. Acta Neurochir (Wien) 2002;144(7):671–6 [discussion:6–7].

50. Lownie SP, Drake CG. Radical intracapsular removal of acoustic neurinomas.J Neurosurg 1991;74:422–5.

51. Wazen J, Silverstein H, Norrell H. Preoperative and postoperative growth rates inacoustic neuromas documented with CT scanning. Otolaryngol Head Neck Surg1985;93:151–5.

52. Charabi S, Tos M, Thomsen J, et al. Cystic vestibular schwannoma—clinical andexperimental studies. Acta Otolaryngol 2000;543:11–3.

53. Lanser MJ, Jackler RK, Pitts LH. Intratumoral hemorrhage and cyst expansion asa cause of acute neurological deterioration in acoustic neuroma patients. In: Tos M,

Battista654

Thomsen J, editors. Acoustic Neuroma: Proceedings of the First International Confer-enceonAcoustic Tumors, Copenhagen, Denmark, August 25–29, 1991. Amsterdam:Kugler Publications; 1992. p. 229–34.

54. Benech F, Perez R, Fontanella MM, et al. Cystic versus solid vestibular schwan-nomas: a series of 80 grade III-IV patients. Neurosurg Rev 2005;28(3):209–13.

55. Kameyama S, Tanaka R, Kawaguchi T, et al. Cystic acoustic neurinomas: studiesof 14 cases. Acta Neurochir (Wien) 1996;138(6):695–9.

56. Matthies C, Samii M, Krebs S. Management of vestibular schwannomas (acousticneuromas): radiological features in 202 cases–their value for diagnosis and theirpredictive importance. Neurosurgery 1997;40(3):469–81 [discussion: 81–2].

57. Pendl G, Ganz JC, Kitz K, et al. Acoustic neurinomas with macrocysts treatedwith gamma knife radiosurgery. Stereotact Funct Neurosurg 1996;66(Suppl 1):103–11.

58. Sobel RA. Vestibular (acoustic) schwannomas: histologic features in neurofibro-matosis 2 and in unilateral cases. J Neuropathol Exp Neurol 1993;52(2):106–13.

59. Kida Y, Kobayashi T, Tanaka T, et al. Radiosurgery for bilateral neurinomas asso-ciated with neurofibromatosis type 2. Surg Neurol 2000;53(4):383–9 [discussion:389–90].

60. Roche PH, Regis J, Pellet W, et al. [Neurofibromatosis type 2. Preliminary resultsof gamma knife radiosurgery of vestibular schwannomas]. Neurochirurgie 2000;46(4):339–53 [French] [discussion: 54].

61. Rowe JG, Radatz MW, Walton L, et al. Clinical experience with gamma knifestereotactic radiosurgery in the management of vestibular schwannomassecondary to type 2 neurofibromatosis. J Neurol Neurosurg Psychiatry 2003;74(9):1288–93.

62. Mathieu D, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for vestib-ular schwannomas in patients with neurofibromatosis type 2: an analysis of tumorcontrol, complications, and hearing preservation rates. Neurosurgery 2007;60(3):460–8 [discussion: 468–70].

StereotacticRadiotherapyfor VestibularSchwannoma

Patrick Sweeney, MDa,*, SantoshYajnik, MDb, William Hartsell, MDc,George Bovis, MDa, JagannathVenkatesan, MSa

KEYWORDS

� Vestibular schwannomas � Schwannoma� Stereotactic radiotherapy � Radiosurgery

Vestibular schwannomas (acoustic neuromas) are typically benign tumors that arisefrom the Schwann’s cells of the eighth (VIII) cranial nerves. Treatment options includeobservation, surgical resection, and radiation therapy, which can be delivered ina conformal or stereotactic manner. The choice of treatment is multifactorial andincludes patient and physician preference, age, tumor size, and symptoms.

Stereotactic techniques for treating vestibular schwannoma include single fractionstereotactic radiosurgery (SRS) and fractionated stereotactic radiotherapy (SRT).Evaluation of management trends from 10 years ago predicted an explosive growthof SRS for vestibular schwannomas in the next century,1 and this growth has beenrealized. A similar growth potential also can be surmised for SRT, but there is noconsensus as yet on the optimal technique, dose, or fraction size. This articledescribes the basic concepts of stereotactic radiotherapy and the radiobiologic ratio-nale for this treatment and reports the clinical results.

STEREOTACTIC RADIOSURGERY VERSUS RADIOTHERAPYStereotactic Radiosurgery

The definition of SRS is defined variably but generally is considered to be a single highdose of radiation delivered with rigid immobilization to an intracranial target to achieve

a Illinois Gamma Knife Center, Alexian Neurosciences Institute, Alexian Brothers MedicalCenter, 800 Biesterfield Road, Elk Grove Village, IL 60007, USAb Advocate Illinois Masonic Medical Center, Department of Radiation Oncology, 836 WestWellington Avenue, Chicago, IL 60657, USAc Advocate Good Samaritan Hospital, Department of Radiation Oncology, 3815 HighlandAvenue, Downers Grove, IL 60515, USA* Corresponding author.E-mail address: [email protected] (P. Sweeney).




Sweeney et al656

a biologic result.2 As a rule, this treatment implies the use of a halo or frame that is af-fixed rigidly to the patient ’s skull to allow for precise immobilization and to provide thereference for defining the target in space relative to a Cartesian coordinate system.The treatment can be delivered with a dedicated radiosurgery machine such as a Lek-sell Gamma Knife (Elekta, Norcross, Georgia) or a modified linear accelerator, of whichthere are various commercially available products. The invasive nature of this fixation,usually by pins that are screwed into the skull, makes this technique feasible only forsingle-fraction treatment. Further, the fact that a high dose is delivered in one fractionlimits this treatment to small targets, usually less than 3 cm. SRS has been very effec-tive in treating various intracranial targets, including vestibular schwannomas, menin-giomas, vascular lesions, trigeminal neuralgia, and primary and metastatic braintumors.

Stereotactic Radiotherapy

SRT is a technique that attempts to combine the biologic advantages of fractionation(ie, delivering multiple small doses of radiation) with the physical advantages of theprecision immobilization seen in SRS. This technique is performed on a linear accel-erator with a relocatable stereotactic frame such as with the Novalis (BrainLab Incor-porated, Westchester, Illinois) or Radionics (Burlington, Massachusetts) systems orwith frameless systems that use optical tracking such as Cyberknife (Accuray Incorpo-rated, Sunnyvale, California). Much of published experience with SRT has been withthe Gill-Thomas-Cosman relocatable head ring, an example of which is shown inFig. 1. Treatments can be delivered with multiple noncoplanar arcs or with fixed fields.Multiple noncoplanar arcs imply that a finely collimated linear accelerator delivers theradiation dose while rotating around the patient, whereas with fixed fields, the beam isfixed but delivered through multiple points that converge on the target lesion. As a rule,the precision for SRT is less than that of SRS, with a relocatable accuracy of approx-imately 2 mm,3 and often safety margins of 2 to 5 mm are added to the planningvolume to account for this uncertainty.

In recent years, the development of intensity-modulated radiation therapy (IMRT)and image-guided radiation therapy (IGRT) has refined SRT further. IMRT is the abilityto modulate the intensity of the radiation beam through the treatment volume to allowfor maximum sparing of normal tissues with improved dosimetry to the target. IGRT isthe use of daily imaging of the target volume before treatment delivery to account forpatient movement to assure target accuracy. One could argue that these techniques

Fig. 1. The Gill-Thomas-Cosman relocatable head ring. (Courtesy of Integra LifeSciencesCorporation, Plainsboro, NJ; with permission).

Stereotactic Radiotherapy for Vestibular Schwannoma 657

have replaced SRT in that they are imaging- and not coordinate (or frame)-based.However defined, these techniques are used widely in many radiation oncologydepartments and have allowed for treatments with SRT-like precision to be performedon many of the same intracranial targets as SRS. One posited advantage of SRT rela-tive to SRS is that because of the ability to fractionate the treatment, targets largerthan 3 cm/or those involving or adjacent to critical structures such as the brainstemor optic apparatus can be treated with SRT without undue risk of complications. Thesesize and dose constraints are much more limiting in single-fraction SRS. Further, thesedevelopments are being applied to extracranial targets, including tumors in the lungs,pancreas, liver, prostate, and spine.

RADIOBIOLOGIC RATIONALE FOR STEREOTACTIC RADIOTHERAPY

SRS involves delivery of a single dose of radiation while SRT involves administeringa more protracted and fractionated course of treatment. For both SRS and SRT, themegavoltage radiation delivered during treatment causes free radical-mediateddamage to the DNA of the target. When treating fast-growing tumors such as squa-mous cell carcinomas of the head and neck or uterine cervix, both fraction size andoverall treatment time help to determine the tumor control probability.4 Such fast-growing tumors can experience accelerated repopulation, whereby surviving cancercells, during a protracted course of radiotherapy, begin to divide and multiply morerapidly. The potential for accelerated repopulation necessitates that the treatment ofsuch fast-growing tumors be completed in a timely manner. If there are unplannedbreaks during a course of radiotherapy, this may lead to a decrease in tumor control.For slow-growing tumors such as vestibular schwannomas, on the other hand, theoverall treatment time has little influence on tumor control probability. The dailydose of the radiation therapy that is delivered is considered the principle factor indetermining the late effects of radiation therapy.4 Therefore, a radiobiological conceptfavoring SRT over SRS for treating vestibular schwannoma would be that fractionationwould not be expected to adversely impact tumor control probability, while it mayreduce long-term damage to surrounding normal tissues by allowing nearby criticalstructures to repair sublethal radiation-induced damage.

Little is known about the precise radiobiological mechanisms that mediate theeffects of ionizing radiation on vestibular schwannoma. An interesting study wasdone at the University of Pittsburgh in which athymic mice were transplanted withhuman xenografts of vestibular schwannoma tissue.5 This study sought to describethe mechanism of SRS-induced tumor control. Various SRS regimens ranging from10 Gy to 40 Gy were delivered in a single fraction, and mice subsequently were sacri-ficed and the tissue studied at 2 weeks, 1 month, and 2 months after SRS. An averagetumor volume reduction of about 16.4% was noted at 3 months in the 10 Gy arm.Histologic examination showed a higher incidence of hemosiderin deposits andvascular mural hyalinization in SRS xenografts versus controls, suggesting botha cellular and vascular effect from radiosurgery.

As will be shown in the next section, both SRS and SRT lead to high tumor controlprobability rates, with good preservation of hearing and limited facial nerve dysfunc-tion. Understanding of the radiobiological mechanisms governing tumor control andnormal tissue damage from radiation, however, is in its nascency. Although it is notknown what the optimal dose and optimal fractionation regimen are for treating vestib-ular schwannoma, tumor control probability with SRS and SRT is excellent, and therisk of long-term morbidity from these treatments remains low.

Sweeney et al658

CLINICAL RESULTS

In reviewing the data on SRT for vestibular schwannoma, one must compare outcomemeasures with the mature results obtained for SRS. These measures include tumorcontrol, hearing preservation, and the development of cranial neuropathies. Chopraand colleagues6 recently reported the results from the University of Pittsburgh ofpatients treated with Gamma Knife radiosurgery with marginal tumor doses of 12Gy to 13 Gy. The authors reported a 10-year actuarial resection-free control of98%, with a cranial nerve complication rate of less than 5%. Although hearing preser-vation is a more difficult parameter to quantify, they reported that among thosepatients who had testable hearing, the hearing preservation rate was 78% to keepthe same Gardner-Robertson level and 97% to preserve any testable level of hearing.Similarly, Friedman and colleagues,7 in their series of 390 patients treated with single-fraction linear accelerator radiosurgery with a marginal dose of 12.5 Gy, describeda 5-year actuarial control rate of 90%, with a cranial nerve complication rate of lessthan 5%. Thus, when analyzing the results of SRT, it is important that the outcomesapproach those of these and other mature SRS series.

In addition, it is helpful to distinguish between SRT series that use conventional frac-tionation and those that use hypofractionation. Conventional fractionation SRT isbased on longstanding experience in the treatment of cranial (and extracranial) tumorswith nonstereotactic techniques and generally implies 5 to 6 weeks of daily treatmentwith total doses of between 45 Gy and 58 Gy. Conversely, hypofractioned SRT is bydefinition the use of larger but fewer doses of radiation delivered over a shorter period.Hypofractionation is used to gain some of the advantages of fractionation in terms ofsparing normal tissues, with the convenience of a more limited treatment time. Hypo-fractionation schedules typically range from three to six fractions. Unfortunately, theoptimal dose/fractionation schedule for this approach is unknown in terms of tumorcontrol and complication potential. Hypofractionation is more akin to single-fractionSRS than conventional SRT and, therefore, would require more precision than conven-tional fractionation SRT. Conventional fractionated SRT, on the other hand, is moreanalogous to conventional radiation therapy techniques.

A summary of the reported series for conventionally fractionated SRT is shown inTable 1. In the studies referenced in Table 1 treatment consisted of conventional frac-tionation schedules between 45 Gy and 57.6 Gy delivered over 5 to 7 weeks. These dataare remarkable for being fairly consistent with local control rates of greater than 86%.Local control generally is defined as tumor stability or regression. Included in this groupis the series with the longest follow-up of 80 months reported by Maire and colleagues.8

Maire and colleagues described an actuarial 15-year local control rate of 86%, which isquite good given the large median tumor size of 3.1 cm, but also underscores the possi-bility of late recurrences and the need for life-long follow-up, especially in youngerpatients. It is interesting that this series of patients was treated without stereotacticimmobilization but only with mask fixation. In other words, the target was definedusing mask fiducials on a day-to day basis as opposed to stereotactically definingthe isocenter as with SRT. Thus, although stereotactic treatment has become thepredominant treatment approach for vestibular schwannomas, conventional externalbeam treatment remains an effective modality. The remainder of the series withstereotactic immobilization show local control rates of greater than 90% but at shorterfollow-up intervals.9–17

The reported complications of the studies in Table 1 were low for both facial andtrigeminal neuropathies and were similar to what has been reported for SRS.6,7 Otherreported complications included the development of hydrocephalus requiring shunt

Table1Acoustic neuromas: conventional stereotactic radiotherapy

First Author/Institution(Reference)

NumberofPatients

Dose(Gy)

FractionSize (Gy) Technique

Tumor Size(Range) cm3

LocalControl(%)

MedianFollow-up (mo)

ComplicationsHearingPres (%)

CranialNerveVII

CranialNerveV Other

Maire/Bordeaux,France8

45 51 1.8 Mask–fixed

3.1*(1.1–5.5)

86% 80 0% 0% 4- hydrocephalous;1- mpnst

77.7%

Andrews/T. Jefferson,Philadelphia9

56 50 2 GTC–arcs 2.78 97% 26.5 2% 7% 81%

Sawamura/Hokkaido,Japan10

101 48 2 Mask–arcs& fixed

1.9*(0.3–4.0)

91% 45 1% 4% 11%hydrocephalousrequiring shunt

71%

Selch/University ofCalifornia LosAngeles11

48 54 1.8 mask- arcs& fixed

2.2*(0.6–4.0)

100% 36 2.1% 2.2% 93%

Combs/HeidelbergGermany12

106 57.6 1.8 mask-mmlcfixed

3.9(2.7–30.7)

93% 48.5 2.3% 3.4% 94.5%

Chan/MassachusettsGeneral13

70 54 1.8 GTC–arcs 2.4*(0.05–21.1)

98% 45.3 1% 4% 84%

Horan/Addenbrookes,Cambridge14

95 50 1.66 Mask/GTC–arc& fixed

2.0*(1.0–4.0)

97% 18.6 3.2% 0% 100%

Koh/PrincessMargaret15

60 50 2 GTC– arcs 4.9(0.3–49)

96% 31.9 0% 0% GBM 77.3%

Thomas/BritishColumbia16

34 45 1.8 Mask–arcs& fixed

2.25(0.3–17.93)

96% 36.5 7% 0% 2- hydrocephalous 63%

McClelland III/Minnesota17

20 54 1.8 GTC–arcs 2.1*(1.1–3.4)

100% 22 0% 10% 100%

Abbreviations: Gy, gray; mmlc, mini-multi-leaf collimator; GTC, gill-thomas-cosman frame; Local Ctrl, local control; Median f/u (mos), median follow-up in months;CN, cranial nerve; hydroceph, hydrocephalus; GBM, glioblastoma multiforme; NF2, neurofibromatosis, type II; no diff, no difference in outcome; hearing pres (%),proportion of patients with preservation of hearing after treatment.

* maximal diameter in cm instead of volume measurements.

Stere

ota

cticR

ad

ioth

era

py

for

Vestib

ula

rSch

wan

no

ma

659

Table 2Hypofractionated stereotactic radiotherapy

First Author/Institution(Reference)

Number ofPatients

Dose(Gy)

FractionSize(Gy) Technique

Tumor Size(Range) cm3

LocalControl(%)

MedianFollow-up (mo)

Complications

HearingPres (%)

CranialNerveVII

CranialNerveV Other

Lederman/Staten Island,New York19

38 20 4/5 6.9(0.1–32)

100% 28 0% 0%

Kalapurakal/Philadelphia20 19 30 5/6 GTC-arcs 3.5*(2.3–4.9)

100% 54

Williams/Hopkins, Baltimore21 111 25 5 Mask 1.4 100% 21.6 0% 0% 72%14 30 3 8.1 100% 21.6 0% 0%

Meijer/VU, Amsterdam22 80 20/25 4/5 GTC–arcs 2.5*(0.8–3.8)

100% 33 3% 2% 61%

Chang/Stanford23 61 21/18 7/6 Cyberknife 1.8*(0.5–3.2)

98% 48 0% 0% 74%

Abbreviations: GTC, Gill-Thomas-Cosman frame; hearing pres (%), proportion of patients with preservation of hearing after treatment.* maximal diameter in cm instead of volume measurements.

Sween

ey

et

al

660


placement8,10,16 and the rare development of second malignancies.8,15 With respectto hearing preservation, all the studies demonstrate that greater than two thirds ofpatients can preserve useful hearing, although there is trend for hearing preservationrates to decline with longer follow-up. The hearing preservation rates for SRT aresimilar to those reported for SRS.6,7 An exception to this is the report of the groupfrom Thomas Jefferson University (Philadelphia, Pennsylvania), which comparedGamma Knife radiosurgery patients treated at the authors’ institution with thosetreated with SRT.9 These authors reported a statistically significant improvement inhearing preservation in those treated with SRT compared with those treated withGamma Knife.9 This study was limited, however, by a small number of patients whounderwent prospective testing and a short follow-up interval. Recent data fromThomas Jefferson University by Werner-Wasik, and colleagues,18 with a median followup of 69 months, has suggested that lowering the total radiation dose from 50.4 Gy to46.8 Gy with conventional fractionation may be associated with higher hearing pres-ervation rates without changing local control.

Table 2 shows a summary of the series treating vestibular schwannoma patientswith hypofractionated SRT.19–23 The first report in the literature from Lederman andcolleagues19 in 1997 was with four or five weekly fractions to a total dose of 20 Gy.They described 100% tumor control with no permanent cranial neuropathies. Similarlygood results have been reported from the Stanford group.23 The Stanford group’sinitial experience with hypofractionation for vestibular schwannoma was with the rigidfixation of an SRS frame with three fractions spaced 8 hours apart to a total dose of 21Gy.24 With this technique, the Stanford group reported 97% tumor control with a 77%hearing preservation rate at a median follow-up interval of 24 months. In recent years,Stanford has been using the frameless Cyberknife system to deliver SRT. With thistechnique, they are treating with three 6 Gy fractions delivered on consecutive daysto a total dose of 18 Gy. With a minimum follow-up of 36 months, they have reportedsimilar rates of local control, complication risk, and hearing preservation as their earlierexperience.23 The latest results from Stanford using the Cyberknife are reported in thearticle by Sakamoto and colleagues, elsewhere in this issue. Thus it would appear thathypofractionated SRT, like conventionally fractionated SRT, leads to similar outcomesin vestibular schwannoma patients as single-fraction SRS.

SUMMARY

The optimal dose and fractionation schedule for vestibular schwannoma SRT remainto be determined. It is clear that the tumor control rates for both conventionally frac-tionated and hypofractionated SRT are similar to SRS. In addition, the risk of cranialneuropathies is low for SRT and SRS. The posited advantage of fractionation foracoustic schwannomas over single-fraction radiosurgery is preservation of hearing.Most current data, however, do not show an advantage of hearing preservationwhen using either conventionally or hypofractionated SRT when compared withSRS. The exception to this are the data from Thomas Jefferson University that appearto show that conventionally fractionated SRT leads to better hearing preservationwhen compared with single-fraction SRS.9 This study, however, is limited by a smallnumber of patients and a short follow-up interval. In addition, data from Stanford usingthe Cyberknife have demonstrated a 74% hearing preservation rate at 2 years offollow-up (see the article concerning Cyberknife radiotherapy for vestibular schwan-noma, by Sakamoto and colleagues, elsewhere in this issue).

Flickinger 25 has noted that because of the high success rate in controlling vestibularschwannomas with SRS and SRT, it is unlikely that reliable radiobiologic parameters

Sweeney et al662

will be extracted from dose–response data. He notes that tumor control rates were notchanged as the doses of SRS were lowered from 16 Gy to 20 Gy to the current level of11 Gy to 13 Gy. Morbidity has decreased, however, as dosages were reduced. It islikely that a similar lowering of total dose and fraction sizes for SRT will continue tolead to excellent tumor control. The impact on cranial neuropathies and hearing pres-ervation will require lengthy follow-up and large case series.

As a practical matter, fractionated SRT, as opposed to SRS, would appear to bemore suitable for larger lesions or those that are adherent to, or compressing, thebrainstem. Although the optimal dose and fraction size are not known, it wouldseem that conventional SRT would be most useful in those radiotherapy departmentswithout the ability to image patients on a daily basis before treatment (IGRT), thusrequiring larger safety margins and leading to a larger volume of normal tissue inthe radiated volume. The tolerance of the normal brain to radiation, although notunderstood completely, is related partly to the volume irradiated and to fractionsize.26 Partial inclusion of the brainstem in the target volume receiving a full dosewith conventional fractionation should be tolerated well. Hypofractionated SRT prob-ably should be used in those departments with IGRT capability where prospectiveimaging of the target allows for smaller margins of normal tissue and thus the minimi-zation of brainstem in the treated volume. Small vestibular tumors can be treatedeffectively with any of these stereotactic techniques.

REFERENCES

1. Pollock BE, Lunsford DL, Noren G. Vestibular schwannoma management in thenext century: a radiosurgcal perspective. Neurosurgery 1998;43(3):475–81.

2. Verhey LJ, Smith V. The physics of radiosurgery. Semin Radiat Oncol 1995;5(3):175–91.

3. Kooy HM, Dunbar SF, Tarbell NJ, et al. Adaptation and verification of the relocat-able Gill-Thomas-Cosman frame in stereotactic radiotherapy. Int J Radiat OncolBiol Phys 1994;30(3):685–91.

4. Hall E, Giaccia A. Time, dose and fractionation in radiotherapy. In: Hall E,Giaccia, editors. Radiobiology for the radiologist. 5th edition. Philidelphia:Lippincott Wilkins & Williams; 2000. p. 397–418.

5. Linskey ME, Martinez AJ, Kondziolka D, et al. The radiobiology of human acousticschwannoma xenografts after stereotactic radiosurgery evaluated in the subrenalcapsule of athymic mice. J Neurosurg 1993;78(4):645–53.

6. Chopra R, Kondziolka D, Niranjan A, et al. Long-term follow-up of acousticschwannoma radiosurgery with marginal doses of 12 to 13 Gy. Int J Radiat OncolBiol Phys 2007;68(3):845–51.

7. Friedman WA, Bradshaw P, Myers A, et al. Linear accelerator radiosurgery forvestibular schwannomas. J Neurosurg 2006;105(5):657–61.

8. Maire JP, Huchet A, Milbeo Y, et al. Twenty years ’ experience in the treatment ofacoustic neuromas with fractionated radiotherapy: a review of 45 cases. Int JRadiat Oncol Biol Phys 2006;66(1):170–8.

9. Andrews DW, Suarez O, Goldman HW, et al. Stereotactic radiosurgery and frac-tionated stereotactic radiotherapy for the treatment of acoustic schwannomas:comparative observation of 125 patients treated at one institution. Int J RadiatOncol Biol Phys 2001;50(5):1265–78.

10. Sawmura Y, Shirato H, Sakamoto T, et al. Management of vestibular schwannomaby fractionated stereotactic radiotherapy and associated cerebrospinal fluidmalabsorption. J Neurosurg 2003;99(4):685–92.


11. Selch MT, Pedroso A, Lee SP, et al. Stereotactic radiotherapy for the treatment ofacoustic neuromas. J Neurosurg 2004;101(Suppl 3):362–72.

12. Combs SE, Volk S, Schulz-Ertner D, et al. Management of acoustic neuromas withfractionated stereotactic radiotherapy (FSRT): long-term results in 106 patientstreated at a single institution. Int J Radiat Oncol Biol Phys 2005;63(1):75–81.

13. Chan A, Black PM, Ojemann RG, et al. Stereotactic radiotherapy for vestibularschwannomas: favorable outcome with minimal toxicity. Neurosurgery 2005;57(1):60–70.

14. Horan G, Whitfield GA, Burton KE, et al. Fractionated conformal radiotherapy investibular schwannoma: early results from a single center. Clin Oncol 2007;19(7):517–22.

15. Koh ES, Millar BA, Menard C, et al. Fractionated stereotactic radiotherapy foracoustic neuroma: single institution experience at princess Margaret hospital.Cancer 2007;109(6):1203–10.

16. Thomas C, Di Maio S, Ma R, et al. Hearing preservation following fractionatedstereotactic radiotherapy for vestibular schwannomas: prognostic implicationsof cochlear dose. J Neurosurg 2007;107(5):917–26.

17. McClelland S III, Gerbi BJ, Higgins PD, et al. Safety and efficacy of fractionatedstereotactic radiotherapy for acoustic neuromas. J Neurooncol 2008;86(2):191–4.

18. Werner-Wasik W, Curran WJ, Evans J, et al. Improved hearing preservation withdeescalated dose of fractionated stereotactic radiotherapy (SRT) in patients withacoustic schwannomas: 50.4 Gy vs. 48.6 Gy. Int J Radiat Oncol Biol Phys 2006;66(3):S241.

19. Lederman G, Lowry J, Wertheim S, et al. Acoustic neuroma: potential benefits offractionated stereotactic radiosurgery. Stereotact Funct Neurosurg 1997;69:175–82.

20. Kalapurakal JA, Silverman CL, Akhtar N, et al. Improved trigeminal and facialnerve tolerance following fractionated stereotactic radiotherapy for large acousticneuromas. Br J Radiol 1999;72(864):1202–7.

21. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas. IntJ Radiat Oncol Biol Phys 2002;54(2):500–4.

22. Meijer OWM, Vandertop WP, Baayen JC, et al. Single-fraction vs. fractionatedlinac-based radiosurgery for vestibular schwannomas: a single-institution study.Int J Radiat Oncol Biol Phys 2003;56(5):1390–6.

23. Chang SD, Gibbs IC, Sakamoto GT, et al. Staged stereotactic irradiation foracoustic neuroma. Neurosurgery 2005;56(6):1254–63.

24. Poen JC, Golby AJ, Forster KM, et al. Fractionated stereotactic radiosurgery andpreservation of hearing in patients with vestibular schwannoma: a preliminaryreport. Neurosurgery 1999;45(6):1299–307.

25. Flickinger JC. What is the optimal dose and fractionation for stereotactic irradia-tion of acoustic neuromas? Int J Radiat Oncol Biol Phys 2002;54(2):311–2.

26. Marks JE, Baglan RJ, Prassad SC, et al. Cerebral radionecrosis: incidence andrisk in relation to dose, time, fractionation, and volume. Int J Radiat Oncol BiolPhys 1981;7(2):243–52.

CyberknifeRadiotherapy forVestibularSchwannoma

GordonT. Sakamoto, MDa,*, Nikolas Blevins, MDb, Iris C. Gibbs, MDc

KEYWORDS

� Stereotactic radiosurgery � Vestibular schwannomas� Cyberknife � Fractionation

Most vestibular schwannomas are benign and slow-growing. Based on that fact,conservative management with serial imaging is a viable alternative.1 For patientswho undergo treatment because of tumor growth, progressive symptoms, or personalpreference, options include serial observation, microsurgical resection, fractionatedstereotactic radiotherapy, and stereotactic radiosurgery. There remains some debateas to how best to manage these tumors. Historically, vestibular schwannomas havebeen treated with microsurgical resection. Loss of hearing and facial nerve injury,however, are not uncommon microsurgical complications. Fractionated radiotherapywas used initially as an adjunct to microsurgery in patients who had undergonesubtotal resection, and demonstrated the overall effectiveness of radiation for treatingvestibular schwannomas.2 Over the last few decades, stereotactic radiosurgery hasemerged as a safe and effective treatment modality for vestibular schwannomas.3–11

Owing mainly to the unknown risks of secondary tumors and possible late effects ofstereotactic radiation, stereotactic radiosurgery initially was reserved for patientswho were poor microsurgical candidates, or as an adjuvant treatment for residual orrecurrent tumor. The long-term data of stereotactic radiosurgery now support the effi-cacy of this treatment modality. Historical reviews also suggest that secondary tumorsare rare after radiosurgery.12 For these reasons, stereotactic radiosurgery hasemerged as the preferred treatment modality, as it offers lower morbidity than surgicalresection and excellent long-term tumor control.4,7,13,14

Traditionally, stereotactic radiosurgery required the use of rigid immobilization usinga stereotactic frame to achieve treatment precision and accuracy. The Cyberknife(Accuray, Incorporated, Sunnyvale, California), which was introduced in 1994, does

a Department of Neurosurgery, Stanford University School of Medicine, Stanford, CA, USAb Department of Otolaryngology, Stanford University School of Medicine, Stanford, CA, USAc Radiation Oncology, Stanford University School of Medicine, Stanford, CA, USA* Corresponding author.E-mail address: [email protected] (G.T. Sakamoto).




Sakamoto et al666

not require the use of rigid immobilization. Although the Cyberknife does not requirethe use of a stereotactic frame, its accuracy is comparable to frame-based radiosur-gery.15 Additionally, because the Cyberknife does not employ a rigid frame, it is prac-tical to fractionate or stage Cyberknife treatments over several days to decrease therisk of radiation injury to adjacent critical structures, such as the brainstem or cochlea.

THE CYBERKNIFE

Stereotactic radiosurgery is a radiation technique that can deliver therapeutic doses ofradiation with submillimeter accuracy and a rapid dose fall-off at the periphery of thetarget. Leksell first conceived the concept of radiosurgery in the 1950s16 and createdthe Gamma Knife, which used multiple cobalt sources to treat a target as defined onpretreatment imaging data. Since then, stereotactic radiosurgery has advanced withtechnologic improvements in computing and imaging. For three decades, stereotactictargeting that required rigid skeletal fixation to provide registration between pretreat-ment plan and target anatomy was at the center of all radiosurgical systems, espe-cially the gold standard Gamma Knife. In the 1990s, John Adler, however, createdthe CyberKnife, a robotic frameless radiotherapy system that uses real-time acquisi-tion of the patient’s bony anatomy for image-guidance.17

The Cyberknife is a frameless, image-guided, robotic radiotherapy system. Thera-peutic radiation is emitted from a compact 6 MV linear accelerator mounted ona robotic arm (Fig. 1). The arm is programmed to move the linear accelerator sequen-tially through a predetermined series of locations. At each location, radiation is deliv-ered to the target volume. The trajectory and dose delivered at each location arecalculated so that their cumulative effect optimizes the coverage of the target volumewhile minimizing the exposure of adjacent tissues. The Cyberknife can generatebeams from more than 1200 directions. This allows for nonisocentric radiation plan-ning that can optimize dose conformality and homogeneity. The need for rigid fixationis circumvented by real-time image guidance. For intracranial targets, patients arerelatively immobilized with an Aquaplast mask (WFR/Aquaplast Company, Wyckoff,New Jersey) (Fig. 2). Flat- panel radiograph detectors are mounted on either side of

Fig.1. The Cyberknife is a linear accelerator mounted on a robotic arm that employs imageguidance to eliminate the need for a stereotactic frame.

Fig. 2. A custom-fit Aquaplast mask is shown attached to the treatment table.

Cyberknife Radiotherapy for Vestibular Schwannoma 667

the treatment table and obtain orthogonal radiograph images in real time during thetreatment. These images are referenced to digitally reconstructed radiographs(DRRs) that are created from CT datasets obtained during treatment planning. Thepatient’s position is verified by his or her own bony anatomy. The image guidance soft-ware compares differences in three translational and three rotational axes, and adjust-ments are made with the treatment couch and robotic arm. This process is updatedthroughout the radiotherapy treatment to maintain accuracy.15

RADIOSURGERY FOR VESTIBULAR SCHWANNOMA

Leksell pioneered the use of stereotactic radiosurgery for treating vestibular schwan-nomas in 1969. He later published his initial experience with vestibular schwannomaradiosurgery using the Gamma Knife in 1971.18 Since then, stereotactic radiosurgeryhas proven to be a safe and effective treatment option for managing vestibularschwannomas. Numerous radiosurgical series exist in the literature detailing excellenttumor control rates ranging from 92% to 100% in the first several years after treatmentusing the Gamma Knife.4,7–9,13,19,20 Additionally, many studies have confirmed excel-lent longer-term control rates at 5 and 10 years after treatment.7,13,19,20

Over the last decade, attention has been directed to improving the hearing preser-vation rates following radiosurgery and reducing other treatment-related morbidities.Initial radiosurgery series reported hearing preservation rates that ranged from 51% to60%7–9 with significant rates of facial weakness and numbness.7 Improvements inconformal radiation treatment delivery and use of lower marginal prescription doses,however, have improved steadily the 3- to 5-year hearing preservation rate to between68% and 77% while also reducing rates of facial weakness and numbness.4,10,11,13,19

Despite improved Gamma Knife radiosurgical techniques and lower marginal doses,a recent report has shown a somewhat disappointing 10-year actuarial hearing pres-ervation rate of 44.5%, with hearing loss developing as much as 6 years aftertreatment.13

Sakamoto et al668

Studies have demonstrated that the total radiation dose to the cochlea is a criticalfactor in hearing preservation.21,22 Fractionation of the total dose, or staging, however,also may play a fundamental role. Staging radiation treatments long has beenproposed as a means to reduce the risk of injury to adjacent critical structures suchas the brainstem, cranial nerves, and cochlea.23–28 Fractionation of the prescribeddose takes advantage of radiobiologic principles to reduce toxicity and maintain tumorcontrol.28,29 The authors’ initial experience with staged frame-based radiotherapyusing 21 Gy in three fractions and a 12-hour interfraction interval showed a 77%hearing preservation rate at 2 years.26 Experience, however, with dose staging usingan interfraction interval of 24 hours was limited, as frame-based radiotherapy tech-niques would require that the patient remain hospitalized while wearing a stereotacticframe continuously over the course of several days.

Based on the proof of principle established in the authors’ staged framed-basedstudy and the introduction of the Cyberknife, it became possible to accurately deliverhighly conformal radiation to vestibular schwannomas and fractionate the prescribeddose in an attempt to spare the adjacent brainstem and cochlea. The treatment of thefirst vestibular schwannoma with the Cyberknife occurred in 1999. Since then, over350 vestibular schwannomas have been treated at Stanford using the Cyberknife em-ploying a standard 3-day staging protocol.

PATIENT SELECTION

A team that consists of a neurosurgeon or otolaryngologist and a radiation oncologistexperienced in the treatment of vestibular schwannomas evaluates all patients treatedwith Cyberknife radiosurgery. All treatment options, including conservative manage-ment and microsurgical resection, are considered by the team.

Cyberknife radiosurgery is offered to patients with unilateral or bilateral vestibularschwannomas that are less than 3 cm in diameter within the cerebellopontine angle.The Cyberknife also can be used to treat residual vestibular schwannomas aftera planned subtotal microsurgical resection or those that recur despite apparent grosstotal removal. Patients demonstrating tumor growth following radiotherapy or radio-surgical treatment also can be considered for retreatment. Retreatment is consideredno sooner than 1 year after radiotherapy treatment. Patients who are not microsurgicalcandidates because of advanced age or other risk factors or patients who do not wishto undergo microsurgical treatment may be candidates for Cyberknife treatment.Additionally, patients who have neurofibromatosis type II (NF2) are also candidatesfor Cyberknife treatment, although as a group this population does not seem torespond as well to treatment as do patients who have sporadic unilateral tumors.30

PRETREATMENT EVALUATION

A pretreatment gadolinium-enhanced MRI is obtained in all patients within the 3months before treatment. The tumor is measured in three orthogonal dimensions.The intracanalicular portion (if any) is included in the maximal transverse diameterwhen calculating measurements.

Pretreatment audiograms are obtained within 3 months before treatment to docu-ment baseline hearing unless anacusis has been documented previously. The wordrecognition score and pure tone average in decibels are recorded. Hearing is gradedon the Gardner-Robertson (G-R) scale.31

Clinical evaluation of the patient’s neurologic status is performed before treatment.Special attention is given to the fifth, seventh, and eighth cranial nerve examination,and any baseline deficits or palsies are noted. Trigeminal nerve function is graded


on a semiquantitative scale as normal sensation, decreased sensation, or no sensa-tion. Facial nerve function is graded on the House-Brackmann (H-B) scale.32

TREATMENT PLANNING AND RADIOSURGERY DELIVERY

All patients initially are fitted with a custom made Aquaplast mask and thin foam head-rest to ensure consistent positioning from the acquisition of the imaging studiesthrough the radiotherapy treatment. While in the mask, a thin slice (1.25 mm) high-resolution (CT) scan of the entire head is obtained with a GE Light Speed 8i Scanner(Milwaukee, Wisconsin) after the intravenous administration of 125 mL of Omnipaquecontrast (iohexol, 350 mg I/mL; Nycomed, Incorporated, Princeton, New Jersey). Theacquired images then are transferred to the Cyberknife treatment planning worksta-tion. When the tumor is identified readily on the CT, with morphology and dimensionsconsistent with those of the pretreatment MRI, target planning can be done using CTimaging alone. In the authors’ experience, nearly all vestibular schwannomas can bedelineated accurately with the high-resolution, thin-slice contrast CT imaging alone.

In the rare instance that CT imaging is not optimal, or if a patient cannot tolerateiodinated contrast used during acquisition of the CT, a thin slice (2 mm) gadolinium-enhanced T1 weighted MRI is obtained. This MRI then is fused to the CT using theCyberknife treatment planning software to create a composite image for tumor local-ization. This method helps define the extent of smaller tumors and can help furtherdefine intracanalicular lesions. Additionally, most digital imaging and communicationin medicine (DICOM) format imaging studies can be imported and fused to the stan-dard CT image. The high-resolution, thin-slice CT scan is always necessary to providethe skeletal anatomy used for real-time patient tracking during treatment.

The treating surgeon then manually defines the tumor volumes and critical struc-tures on the axial images (Fig. 3). Although the axial images normally are used fortreatment planning, it is possible to delineate structures on the either the coronal orsagittal images. It is usually necessary to adjust the windows and levels of the treat-ment planning CT dataset to optimize clarity of the bone of the internal auditory canalto improve definition of the tumor margins. In addition, the cochlea is defined better

Fig. 3. Right-sided intracanalicular vestibular schwannoma outlined as the radiotherapytarget with the cochlea and brainstem outline as critical structures.

Sakamoto et al670

when CT window levels are set for bone visualization. It should be emphasized thattarget definition can be a complex process in recurrent or postoperative lesions,and in rare circumstances, it may require the input of an experienced neuroradiologist.

Once the tumor and critical structures are delineated, nonisocentric, inverse plan-ning using the Cyberknife treatment planning software helps to achieve a highlyconformal radiotherapy dose that minimizes dose to the adjacent critical structures(Fig. 4). Nonisocentric planning allows the beams of radiation to originate frommany different points in space to create an even dose distribution to nonsphericallesions. This differs from isocentric planning, in which one or more spherical dosedistributions are used to cover the tumor shape. The Cyberknife can do both isocentricand nonisocentric planning; however, for irregularly shaped tumors, such as vestibularschwannomas, nonisocentric planning is used. Inverse planning allows the physicianto input specific treatment criteria and dose tolerances. Then the treatment planningsoftware selects the beams and beam weights to compute a radiotherapy plan thatmeets the treatment criteria. This is in contrast to forward planning, in which theuser manually selects the beams and beam weights to be used. The treatment planis evaluated by the treating surgeon and the radiation oncologist. The treatmentplan is evaluated and selected based on an analysis of the volumetric dose and thedose–volume histograms of the target volume and the adjacent critical structures.The number of paths and beams used for each patient varies and is determined bythe selected individual treatment plan.

Fig. 4. Treatment plan for a left-sided vestibular schwannoma involving the left internalauditory canal and cerebellopontine angle cistern treated with 18 Gy delivered in threeequal fractions. The dose distribution is contoured to minimize irradiation of the cochleaand brainstem. The tumor volume is depicted as well as the 50% and 25% isodose linesin three planes.


Once the treatment plan has been created and approved by the treating physicians,the patient is brought back and placed supine on the treatment couch. The Aquaplastmask then is put on the patient and affixed to the treatment couch to ensure consistentpositioning throughout the treatment. The treatment couch is adjusted as necessary toalign the patient with the previously generated DRRs. Once the patient is aligned prop-erly, treatment begins. Each fraction lasts between 30 and 45 minutes. Following eachradiotherapy fraction, patients are treated with an oral dose of 4 mg of dexamethasonefor prophylaxis of acute radiation toxicity.

POST-TREATMENT FOLLOW-UP

For the first 2 years after radiosurgery, a thin slice (2 to 3 mm) gadolinium-enhancedMRI is obtained every 6 months. MR images are obtained annually after the first 2years. Tumor size is measured in three orthogonal dimensions on each follow-upscan and compared with the pretreatment measurements. Tumors then are catego-rized as stable, smaller, or larger than the pretreatment dimensions.

For the first 2 years after radiosurgery, audiograms are obtained every 6 months.After the first 2 years, pure-tone audiograms and word recognition scores are obtainedannually or if a hearing change is reported. Whenever possible, patients obtain audio-grams at the same center to minimize discrepancies related to technique. Directedneurologic examination and testing of the fifth, seventh, and eighth cranial nerves isperformed every 6 months for the first 2 years. Neurologic testing occurs annuallythereafter.

STANFORD UNIVERSITY CYBERKNIFE EXPERIENCE

Between 1999 and 2001, 61 patients who had unilateral vestibular schwannomasunderwent tri-fractionated Cyberknife radiosurgery at Stanford University using theprotocol outlined previously.33 Mean patient age was 54 years (range 27 to 79 years),and 31 (51%) of the vestibular schwannomas were located on the right side, while 30were located on the left side. Eight patients had a prior microsurgical resection on thetreated side. These eight patients either demonstrated residual tumor on postopera-tive scans or a new recurrence on follow-up scans. None of the patients in this seriescarried the diagnosis of NF2.

The mean pretreatment maximal tumor dimension was 18.5 mm (range 5 to 32 mm).The initial treatment dose was 21 Gy over 3 days for the first 14 patients. This dose wasbased on the authors’ prior experience with frame-based radiosurgery and selectedbased on the radiobiological equivalence to a single 14 Gy dose.26 Based on excellenttumor control and evidence that a reduced dose could maintain tumor control rateswhile improving hearing preservation,10,34 the dose was lowered to 18 Gy deliveredin three equal fractions delivered on consecutive days (18 Gy in three fractions isapproximately equal to 11.5 Gy in one single treatment session). The treatmentdose was prescribed to the 70% to 80% isodose contour line at the periphery ofthe tumor. In every case, the total dose was given in equal fractions over the courseof 3 consecutive days. After each fraction was delivered, 4 mg of dexamethasonewas administered orally.

All patients received an MRI, audiogram, and clinical follow up every 6 months forthe first 2 years. After the first 2 years, the patients were followed with annual MRIs,audiograms, and clinic visits. Mean clinical and radiological follow-up for this serieswas 48 months (range 36 to 62) months.

Radiologically, 29 of 61 tumors decreased in size, and 31 tumors had no change insize, producing a tumor control rate of 98% (Fig. 5). The loss of central enhancement

Fig. 5. (A) Axial T1 weighted gadolinium-enhanced MRI of a right-sided vestibular schwan-noma before Cyberknife radiosurgery. (B) Follow-up axial T1 weighted gadolinium-enhanced MRI taken 6 years after Cyberknife radiosurgery demonstrating no evidence oftumor progression.

Sakamoto et al672

within the tumor was noted routinely between 6 and 12 months after radiosurgery. Onepatient was noted to have an increase in tumor size 4 years after treatment. The tumordimensions all increased by 50%. At 3 years after treatment, the tumor was minimallylarger than pretreatment dimensions. The patient did not develop new neurologicsymptoms with tumor growth. This patient ultimately underwent a microsurgicalresection.

Of the 61 patients in this series, 13 patients had no hearing (G-R grade 5) beforetreatment. These patients were not followed with serial audiograms. The remaining48 patients all had some degree of hearing (G-R grade 1 to 3) before treatment. Ofthe 48 patients with hearing before treatment, 43 (90%) retained hearing (G-R grade1 to 3) at last follow-up. Thirty-five patients had G-R grade 1 to 2 hearing before treat-ment, and 26 (74%) of these patients maintained this level of hearing at last follow-up.Two patients’ hearing actually improved after treatment with the Cyberknife. Onepatient’s hearing improved from a G-R grade 2 to a grade 1, and the other patientimproved from a G-R grade 3 to a grade 1.

No patient treated with the Cyberknife developed a new facial weakness. Onepatient had a H-B grade 3 facial weakness before treatment that remained stable afterradiosurgery. Two patients had transient facial twitching within the first 12 monthsafter radiosurgery. One patient’s twitching resolved in 3 months, and the otherresolved in 5 months.

Trigeminal function was assessed at each follow up using a semiquantitative scale.Trigeminal dysfunction usually is seen in association with larger tumors. Smallertumors and intracanalicular tumors are relatively distant from the trigeminal nerve. Inthis series, no patients experienced trigeminal nerve deficits after radiosurgery,regardless of tumor size.

One patient who had undergone previous microsurgical resection developed symp-toms from brainstem edema within the first 12 months after radiotherapy. Five monthsafter radiotherapy, the patient experienced left lower extremity sensory loss. Imagingdemonstrated a T2 signal change along the lateral brainstem. The patient’s symptomsresolved over a period of 3 months and the imaging abnormalities fully resolved onsubsequent studies.

In summary, this series demonstrated excellent short tumor control rates with goodhearing preservation and few cranial nerve deficits. A recent Gamma Knife study


suggests that cranial nerve deficits after radiosurgery can occur up to 5 years aftertreatment.35 Additionally, another recent Gamma Knife study reported that the hearingpreservation rate continues to decline even 6 years after treatment.13 Therefore, addi-tional studies with longer follow-up after Cyberknife radiosurgery are underway todetermine if fractionation of the radiotherapy dose can improve upon previously re-ported Gamma Knife results.

SUMMARY

Vestibular schwannomas are benign tumors, which are amenable to both microsur-gical and radiotherapy or radiosurgical treatments. For patients with small- tomedium-sized tumors, who are poor microsurgical candidates or who do not wishto undertake the risks associated with microsurgical resection, Cyberknife stereo-tactic radiotherapy offers excellent tumor control and hearing preservation withminimal facial and trigeminal morbidity. Further studies with longer follow-up areneeded and underway. Based on the Stanford University experience, fractionatedCyberknife radiotherapy appears to be a safe and effective primary treatment forvestibular schwannomas.

REFERENCES

1. Smouha EE, Yoo M, Mohr K, et al. Conservative management of acousticneuroma: a meta-analysis and proposed treatment algorithm. Laryngoscope2005;115:450–4.

2. Wallner KE, Sheline GE, Pitts LH, et al. Efficacy of irradiation for incompletelyexcised acoustic neurilemomas. J Neurosurg 1987;67:858–63.

3. Bertalanffy A, Dietrich W, Aichholzer M, et al. Gamma knife radiosurgery ofacoustic neurinomas. Acta Neurochir (Wien) 2001;143:689–95.

4. Flickinger JC, Kondziolka D, Niranjan A, et al. Results of acoustic neuroma radio-surgery: an analysis of 5 years’ experience using current methods. J Neurosurg2001;94:1–6.

5. Flickinger JC, Lunsford LD, Coffey RJ, et al. Radiosurgery of acoustic neurino-mas. Cancer 1991;67:345–53.

6. Flickinger JC, Lunsford LD, Linskey ME, et al. Gamma knife radiosurgery foracoustic tumors: multivariate analysis of four-year results. Radiother Oncol1993;27:91–8.

7. Kondziolka D, Lunsford LD, McLaughlin MR, et al. Long-term outcomes after ra-diosurgery for acoustic neuromas. N Engl J Med 1998;339:1426–33.

8. Linskey ME, Lunsford LD, Flickinger JC. Radiosurgery for acoustic neurinomas:early experience. Neurosurgery 1990;26:736–44 [discussion: 44–5].

9. Lunsford LD, Linskey ME. Stereotactic radiosurgery in the treatment of patientswith acoustic tumors. Otolaryngol Clin North Am 1992;25:471–91.

10. Niranjan A, Lunsford LD, Flickinger JC, et al. Dose reduction improves hearingpreservation rates after intracanalicular acoustic tumor radiosurgery. Neurosur-gery 1999;45:753–62 [discussion: 62–5].

11. Spiegelmann R, Gofman J, Alezra D, et al. Radiosurgery for acoustic neurinomas(vestibular schwannomas). Isr Med Assoc J 1999;1:8–13.

12. Rowe J, Grainger A, Walton L, et al. Risk of malignancy after gamma knife stereo-tactic radiosurgery. Neurosurgery 2007;60:60–5 [discussion: 65–6].

13. Chopra R, Kondziolka D, Niranjan A, et al. Long-term follow-up of acousticschwannoma radiosurgery with marginal tumor doses of 12 to 13 Gy. Int J RadiatOncol Biol Phys 2007;68:845–51.

Sakamoto et al674

14. Pollock BE, Lunsford LD, Kondziolka D, et al. Outcome analysis of acousticneuroma management: a comparison of microsurgery and stereotactic radiosur-gery. Neurosurgery 1995;36:215–24 [discussion: 24–9].

15. Chang SD, Main W, Martin DP, et al. An analysis of the accuracy of the Cyber-knife: a robotic frameless stereotactic radiosurgical system. Neurosurgery2003;52:140–6 [discussion: 46–7].

16. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand1951;102:316–9.

17. Adler JR Jr, Chang SD, Murphy MJ, et al. The Cyberknife: a frameless roboticsystem for radiosurgery. Stereotact Funct Neurosurg 1997;69:124–8.

18. Leksell L. A note on the treatment of acoustic tumours. Acta Chir Scand 1971;137:763–5.

19. Hasegawa T, Fujitani S, Katsumata S, et al. Stereotactic radiosurgery for vestib-ular schwannomas: analysis of 317 patients followed more than 5 years. Neuro-surgery 2005;57:257–65 [discussion: 57–5].

20. Hasegawa T, Kida Y, Kobayashi T, et al. Long-term outcomes in patients withvestibular schwannomas treated using gamma knife surgery: 10-year follow up.J Neurosurg 2005;102:10–6.

21. Massager N, Nissim O, Delbrouck C, et al. Irradiation of cochlear structuresduring vestibular schwannoma radiosurgery and associated hearing outcome.J Neurosurg 2007;107:733–9.

22. Thomas C, Di Maio S, Ma R, et al. Hearing preservation following fractionatedstereotactic radiotherapy for vestibular schwannomas: prognostic implicationsof cochlear dose. J Neurosurg 2007;107:917–26.

23. Flickinger JC. An integrated logistic formula for prediction of complications fromradiosurgery. Int J Radiat Oncol Biol Phys 1989;17:879–85.

24. Lederman G, Lowry J, Wertheim S, et al. Acoustic neuroma: potential benefits offractionated stereotactic radiosurgery. Stereotact Funct Neurosurg 1997;69:175–82.

25. Lo YC, Ling CC, Larson DA. The effect of setup uncertainties on the radiobiolog-ical advantage of fractionation in stereotaxic radiotherapy. Int J Radiat Oncol BiolPhys 1996;34:1113–9.

26. Poen JC, Golby AJ, Forster KM, et al. Fractionated stereotactic radiosurgery andpreservation of hearing in patients with vestibular schwannoma: a preliminaryreport. Neurosurgery 1999;45:1299–305 [discussion: 305–07].

27. Varlotto JM, Shrieve DC, Alexander E 3rd, et al. Fractionated stereotactic radio-therapy for the treatment of acoustic neuromas: preliminary results. Int J RadiatOncol Biol Phys 1996;36:141–5.

28. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas. Int JRadiat Oncol Biol Phys 2002;54:500–4.

29. Andrews DW, Suarez O, Goldman HW, et al. Stereotactic radiosurgery and frac-tionated stereotactic radiotherapy for the treatment of acoustic schwannomas:comparative observations of 125 patients treated at one institution. Int J RadiatOncol Biol Phys 2001;50:1265–78.

30. Mathieu D, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for vestib-ular schwannomas in patients with neurofibromatosis type 2: an analysis of tumorcontrol, complications, and hearing preservation rates. Neurosurgery 2007;60:460–8 [discussion: 68–70].

31. Gardner G, Robertson JH. Hearing preservation in unilateral acoustic neuromasurgery. Ann Otol Rhinol Laryngol 1988;97:55–66.


32. House JW, Brackmann DE. Facial nerve grading system. Otolaryngol Head NeckSurg 1985;93:146–7.

33. Chang SD, Gibbs IC, Sakamoto GT, et al. Staged stereotactic irradiation foracoustic neuroma. Neurosurgery 2005;56:1254–61 [discussion: 61–3].

34. Spiegelmann R, Lidar Z, Gofman J, et al. Linear accelerator radiosurgery forvestibular schwannoma. J Neurosurg 2001;94:7–13.


StereotacticRadiosurgeryand StereotacticRadiotherapy in theTreatment of SkullBase Meningiomas

JohnM. McGregor, MD*, Atom Sarkar, MD, PhD

KEYWORDS

� Stererotactic � Radiosurgery � Fractionated � Radiotherapy� Skull base � Meningioma � Gamma Knife � Particle therapy� LINAC

Meningiomas are the most common of the nonglial brain tumors. They account forapproximately 20% of primary brain tumors. They tend to be slow growing, tendto be benign, and often can reach substantial sizes before becoming symptomatic.Complete surgical resection of intracranial meningiomas, including the dura of originand infiltrated bone, remains the treatment of choice when able to be accomplishedsafely. Extent of resection has been shown to be a predictor of recurrence.1,2 Addi-tional prognostic factors for recurrence after surgical resection include the WorldHealth Organization (WHO) pathologic grade and the tumor cell proliferationrate.3,4 Location of a meningioma within the cranial vault may make a completesurgical resection unlikely, and tumors arising from the dura of the skull base areparticularly challenging in this regard. Advances in radiation therapy, includingstereotactic techniques, expand the options for treatment available in these situa-tions. They may be used as adjuncts to surgery or as alternative modalities in thetreatment of these complex tumors. The therapy for these patients can be individu-alized depending on several factors, including clinical presentation, patient comor-bidities, tumor progression, tumor location, and involvement of adjacentneurovascular structures.

Department of Neurological Surgery, The Ohio State University College of Medicine, 410 West10th Avenue, Columbus, OH 43212, USA* Corresponding author.E-mail address: [email protected] (J.M. McGregor).




McGregor & Sarkar678

SKULL BASEMENINGIOMAS

One of the most difficult areas in which to achieve complete resection is the basalskull. This region includes tumors that arise from the dura of the clivus and petrousbone; dura of the cavernous sinus and tentorium; sphenoid bone, including sella tur-cica; and the olfactory groove; and the optic nerve sheath. Approximately 20% ofintracranial meningiomas reside at these regions of the skull base.5,6 In these locationsthey tend to present clinically with symptoms due to the proximity of adjacent vascularand cranial nerve structures or due to mass effect on intracranial contents.

Treatments of meningiomas of the skull base often have to be tailored to the localenvironment in which the tumor arises and the surrounding structures that find them-selves involved in tumor. Because of the location of the tumor, treatment options mustbe weighed against the sensitivity of normal adjacent structures and by the morbidityassociated with the treatment approaches required and the natural history of thetumor itself. Complexities in the clinical decision making regarding treatment for skullbase meningiomas center around the risk for morbidity and from intervention andexpectant management. These tumors tend to be benign and slow growing althoughthey can be aggressive clinically and histologically.

The natural history and the tolerance of surrounding neural tissues must be consid-ered when treatments are discussed. Bindal and colleagues addressed the issue ofnatural history of skull base meningiomas. They reviewed 40 patients who were diag-nosed with skull base meningiomas based on radiographic studies and who were fol-lowed without treatment over a mean of nearly 7 years. Twenty-seven percentprogressed neurologically, all with worsening cranial nerve deficit. No new strokesor brainstem symptoms developed. On serial follow-up imaging studies, 18% showedsome degree of growth. The 2-, 5-, and 10-year radiologic evaluations showed theirtumors were progression free on serial imaging at rates of 97%, 80%, and 42%,respectively. Forty percent of their patients eventually underwent some form of radio-surgical or surgical treatment.7 These observations need to be considered whenrecommendations for treatment are made. They also need to be considered whentrying to assess the results of various treatment modalities. Still, with the threefoldadvances of imaging capabilities, microsurgical techniques, and focused radiationdelivery systems, the options to individualize therapy, taking into account tumor-and patient-specific variables, have never been as great.

SURGERY FOR SKULL BASEMENINGIOMAS

Microsurgery for meningiomas of the skull base remains a viable and in many cases anoptimal treatment of this condition. The abilities of a surgeon to affect completeremoval, obtain tissue diagnosis, and affect an immediate decompression of compro-mised neural structures with minimal new neurologic sequelae are important points inthe decision-making process. Yet the ability to accomplish those goals surgically mustbe assessed on a case-by-case basis. Every tumor and every patient has unique char-acteristics that affect a decision for surgery.

Microsurgery of the skull base has been shown effective, yet there are risks for adja-cent structures and there are recurrences. Current microsurgical techniques haveadvanced the abilities of surgeons to perform aggressive resections in the skullbase area with minimal morbidity and mortality and prolongation of progression-freesurvivals. Complete surgical resection in all cases of skull base meningiomas withminimal morbidity remains, however, elusive. In addition, several reports evaluatingpathologic specimens of skull base meningiomas have demonstrated that tumor infil-tration can be seen into neurovascular structures adjacent to tumor, suggesting that

Skull Based Meningiomas 679

there are limits to which microsurgery can achieve complete resection.8–10 De Jesusand colleagues,11 in the analysis of their series of 119 patients who had cavernoussinus meningiomas at a mean of 2.5 years follow-up, identified a recurrence rate of10% in patients initially believed to have complete resection and a progression rateof 15% in patients who had initial subtotal resection. DeMonte and colleagues12

observed, in their 41 patients followed over a mean 45 months, 7% mortality, 7%morbidity, 10% new cranial nerve deficit, and 10% recurrence rate for patients whohad complete resection. O’Sullivan and colleagues reviewed 89 patients, overa mean of 2 years, who had cavernous meningiomas of varying histologic grade.They reported a 12% incidence of new cranial nerve deficit with surgery. They wereable to attain a complete resection in 8 patients who had WHO grade I tumors whohad no recurrence during the follow-up period. The progression rate for the patientswho had subtotal resection was 6%.13

Surgery for resection of skull base meningiomas of the petroclival region involvesa particularly difficult area. Natarajan and colleagues reported on the long-term resultsof 150 patients evaluated at a mean of 102 months after microsurgery for skull basemeningiomas in this area. They reported new neurologic deficits in 20% of patients.Gross total resection was accomplished in 40% of patients. Recurrence developedin 4% of patients who had complete resection and 5% of patients who had incompleteresection. They reviewed the literature on surgery in this area and reported a range oftotal or gross total resection rates of between 40% and 79% with recurrence/progres-sion rates reported as 0 to 36%. Follow-up times for the other reports are compara-tively shorter, however (14–73 months).14

RADIATION THERAPY FOR SKULL BASEMENINGIOMA

Historically, radiation therapy has been shown a reasonable adjunct to surgical resec-tion, providing enhanced local control of tumor growth for benign meningiomas.Barbaro and colleagues15 in 1987 reviewed their series, which included all typesand grades of meningiomas, and showed an improvement in progression-free survivalat 6.5 years with external-beam fractionated radiotherapy after subtotal resectionwhen compared with subtotal resection alone (68% versus 40%).

In the ensuing years, advancements in imaging and improvements in radiationdelivery methods have enhanced abilities to conform the dose of radiation evermore precisely to tumor, have allowed for more intense tumor dosing, and haveallowed for reduction in radiation to nontumor structures. These concepts havebeen particularly advantageous for tumors that arise at the skull base. The proximityof important radiosensitive neurovascular structures, brainstem, and pituitary glandto tumor margins requires that radiation to skull base lesions be delivered in a waythat provides therapeutic benefit while maintaining enhanced margins of safety foradjacent tissues. Further improvements in conformality have been achieved with theintroduction of focused radiation techniques, including stereotactic radiosurgery(SRS) and stereotactic radiotherapy (SRT); each may be delivered by gamma photons,x-ray photons, or charged particle beam therapy.

Stereotactic therapy is the term used to identify treatment modalities that use high-dose radiation, delivered from multiple beam locations, to precise targets registeredwithin a defined 3-D space. Stereotactic localization uses high-quality imaging to iden-tify tumor, tumor margins, critical neurovascular structures, and identifiable fiducialpoints, which then are digitally processed to register a patient and anatomy intoa defined 3-D space. Stereotactic therapy includes a mechanism of radiation accuracyassurance during treatment. Repetitive image localization during treatment or fixation


of a patient and target throughout the treatment are used as methods to confirm theaccuracy of the radiation dose delivered.

Radiation Sources

The radiation sources for SRS or SRT that are used most commonly in clinical practiceare the gamma photons generated from the decay of cobalt-60 or the x-ray photonsgenerated from a linear accelerator (linac). Cobalt-60 decays with the release of twogamma photons. The isotope is housed in such a way as to allow the release of onlythose photons emitted in a predetermined direction. A series of cobalt sources are ar-ranged to deliver multiple gamma photons to a single convergent point. Targeting thenis accomplished by adjusting a patient’s position so as to place the tumor into the pointof convergence. The isotope has a half-life of 5.28 years. The sources typically are re-placed every 5 to 8 years because the treatment times increase as the sources age.

Linear Accelerators

Linear accelerators can generate x-ray photons by the collision of accelerated elec-trons and a target metal. The resulting photon beam exits the generator followinga linear path. The beam then can be directed by moving the accelerator into the properposition. Targeting is accomplished by positioning the accelerator to allow the beamto pass through the target site from multiple entrance trajectories. These radiationbeams, gamma radiation and x irradiation, are characterized by having high enoughenergy to penetrate the cranial tissues allowing for treatment of the intracranialcontents. Additionally, they deposit radiation over the entire length of the beam,including both target tissue and nontarget tissue proximal and distal to the target.By arranging multiple portals for the delivery of these radiation beams and havingthem all converge on the exact same target point, the target can accumulate signifi-cant radiation exposure while the dose delivered to nontarget tissues is minimized.A steep gradient of radiation dosage is generated with a substantial falloff of thedose measured at distances adjacent to and distal to the target. In addition to varyingthe location of beam delivery, current delivery systems allow gamma photon beamsand x-ray photon beams to be varied in their intensity along each particular deliverypath, providing variable dosing per exposure and greater opportunities for dose con-touring. These techniques add to the ability to minimize the dose to nontarget tissueswhile maintaining the planned target dose accumulation.16

Particle Therapy

Particle therapy refers generally to beams of heavier charged particles used as radia-tion sources. These particles include protons (hydrogen nuclei), helium nuclei, andcarbon nuclei. Using a cyclotron, these beam sources are generated by the strippingof the underlying atoms of their electrons and accelerating the charged nuclei to a vari-able exiting energy level. The particle beam then is released in the intended directionof the target. As with photon-based therapies, particle therapies vary the direction andthe intensity of the beam entering the patient to improve conformality of the dose tothe tumor. Because the particles have mass and charge, they can be focused magnet-ically. Most importantly, particle beams release their energy essentially completely ata single interaction point at a predetermined depth within the target tissue. This char-acteristic is known as the Bragg peak effect. By varying the intensity of the beam as itleaves the cyclotron, the distance the particle beam travels in tissue before discharg-ing its energy can be varied. Because so little energy is delivered beyond the length ofthe Bragg peak, the radiation exposure is decreased to the surrounding nontarget


tissues. This allows for increased conformality of the radiation dose and providesa theoretically improved radiobiologic profile.6,17

FOCUSED RADIATION AND FRACTIONATION

SRS, then, is a radiation delivery method that uses multiple radiation beams focused ina 3-D manner within a defined target space (stereotactically) to a single target. Typi-cally, the term, radiosurgery, refers to treatment performed in a single fraction,although up to five fractions may be considered radiosurgery according to convention.The radiation may be generated from fixed sources, such as the gamma knife unit. Inthis unit, 201 fixed, cobalt-60, gamma ray–emitting sources housed in cast iron arepreset in a spherical array focused to the unit’s isocenter. A series of tungsten helmetcollimators are used to further focus the individual sources. Patients are immobilized ina rigid titanium stereotactic frame applied to the skull in four-point pin fixation. Thepins secure the frame with tension applied to the outer table of the skull. A fiducialbox is attached to the frame, and the patient then undergoes stereotactic MRI orCT scanning or both for localization of the target and adjacent structures. Other rele-vant images (positron emission tomography, functional MRI, or diffusion tensorimaging) may be merged to the stereotactic images as necessary. Once targets areidentified and digitized and a treatment plan optimized, the patient is attached viathe rigid frame to the gamma knife unit in such a manner as to place the patient’starget in the machine’s isocenter. The table then moves the patient into the focal pointof the sources and the radiation for that isocenter is delivered. Depending on thecomplexity of the plan, the patient may require several isocenter exposures tocomplete the treatment plan. Once the treatment plan is completed, the head frameis removed.

The radiation output for SRS also may be generated as x rays from a nonfixed linac.In this instance electrons are accelerated and then collide with a metal target gener-ating a beam of x-ray photons. The photon source then is rotated in multiple arcsaround the patient lying on the treatment table, positioning the x-ray beam in multipleorientations focused to a fixed isocenter. This method generates a spherical distribu-tion at the isocenter. Tertiary collimators can be added to the beam to adjust the doseat each position, allowing for a nonuniform, or shaped, isocenter. For 3-D localization,the patient may be immobilized by a stereotactic rigid frame applied to the skull,implanted with attached fiducials, or registered with a frameless system in whichrepeated x-ray imaging confirms patient position throughout the treatment. A high-resolution, stereotactic MRI or CT is performed with an attached fiducial referenceframe. The images are used for target localization and treatment planning. Dependingon the complexity of the tumor and the plan, there may be a single isocenter used totreat the tumor or multiple isocenters. In the treatment paradigms that use rigid fixa-tion, the radiation dose delivered can be contoured to the margin of the tumor itself.No additional treatment beyond the margin is needed.

STEREOTACTIC RADIOTHERAPY

SRT is a variation of the stereotactic focused radiation that combines the benefits ofstereotactic localization and the advantages of radiation fractionation. It allows forfocused radiation to be delivered over several sessions instead of a single session.This method uses the precision of the stereotactic localization in 3-D space ofa patient, target, nontarget structures and delivery methods that include multiple-beam entry points targeting the isocenter in a repeatable, accurate manner. Accuracyis assured over multiple treatment sessions using frameless head holders, face masks,


or bite blocks; with repetitive imaging during treatment; or by the use of skull mountedimplanted anchors to provide repeatable accurate positioning at each treatmentsession.18 Because of the ability of normal tissue to affect repair between fractions,SRT is considered beneficial in the treatment of larger lesions that likely would exposesurrounding central nervous system tissues to a higher dose if treated in a single frac-tion and in the treatment of those tumors adjacent to critical radiation sensitive struc-tures. Patients undergo stereotactic MRI or CT imaging with or without additionalimage merging, followed by 3-D registration and dose planning. Treatment then is per-formed over a series of repeat sessions, in a conventional fashion with 25 to 30 frac-tions (fractionated SRT) or over a much fewer number of fractions (hypofractionatedSRT [HSRT]).5

TISSUE TOLERANCES

Microsurgical resection currently remains the gold standard in the treatment of skullbase meningiomas to which all new therapies are compared. The use of radiosurgeryin these treatments has evolved over time as improvements in imaging, tumor identi-fication, tolerance of the cranial structures, and minimal effective dosing schemeswere identified. Radiosurgery initially was reserved for those lesions that were recur-rent, more malignant, or identified in patients who had associated medical conditionsthat precluded surgery. The ability to incorporate MRI into dose planning has allowedimproved accuracy by better identifying tumor margins and cranial nerves. The initialmarginal doses used for the treatment of skull base meningiomas were derived fromeffective doses used in other central nervous system malignancies. Over time, theidentification of tolerances of the cranial nerve structures has led to modification ofdoses with good results. Current therapeutic dosing for meningiomas of the skullbase has allowed for reasonable therapeutic benefits and minimal cranial nervecomplications.

OPTIC NERVE CONSIDERATIONS

In planning anterior skull base radiosurgery, the critical structure to be addressed isthe optic apparatus. Its tolerance to single-fraction radiosurgery has been shown tobe less than 10 Gy,19 although some investigators have observed radiation-inducedoptic neuropathy at doses greater than 9 Gy20 and greater than 8 Gy.21 Leber andcolleagues, in reviewing their series of 50 patients treated with gamma knife radiosur-gery, identified no radiation-induced optic neuropathy when the dose to the opticapparatus was kept below 10 Gy. They reported, however, a 26% incidence whenthe optic nerve received doses of between 10 and 15 Gy and a 77% incidence atdoses greater than 15 Gy after a mean follow-up of 40 months.19 Kondziolka andcolleagues22 recommend keeping the dose to the optic nerve at less than 8 Gy andhave observed radiation-induced optic neuropathy in only 1 of 78 patients (1.3%)from 2001 to 2008. Alternatively, Morita and colleagues,23 in their series of 88 patientswho had skull base meningiomas, observed over a median follow-up period of 35months, reported no incidence of optic nerve deterioration in which only shortsegments (9–12 mm) of the apparatus were exposed to doses of 12 to 16 Gy. Theydid not include tumors that had distorted or elevated the optic nerves and chiasm.Stafford and colleagues24 reported a 1.1% incidence of significant radiation-inducedoptic neuropathy in their series of 218 patients treated for mixed anterior skull basetumor types who received 12 Gy or less and were followed for a median of 40 months.


CRANIAL NERVE CONSIDERATIONS

The cranial nerves of the cavernous sinus seem to tolerate relatively higher doses thanthe optic apparatus. Morita and colleagues23 observed trigeminal nerve dysfunction in13% of their 88 patients who had Meckel’s cave doses of greater than 19 Gy but inonly 7% of their patients who received less than 19 Gy. They reported only one patientwho had abducens palsy and one who had an oculomotor deficit among their 88patients. Leber and colleagues,19 in their series of 50 patients, observed no newtrigeminal nerve deficits (dose range of 6–20 Gy) and no new deficits in cranial nerve3, 4, or 6 (dose range 5–30 Gy).

PITUITARYGLAND CONSIDERATIONS

The pituitary gland function is also at risk from radiosurgery to the skull base; however,although the incidence reported is much lower than that seen in the treatment of pitu-itary adenomas, the actual safe dose to the hypophysis is not known. Vladyka andcolleagues studied pituitary dose tolerance in their series of 63 patients who hadadenoma and were evaluated at a median of 5 years after gamma knife radiosurgery.They observed a 60% risk for gonadotropic dysfunction at a mean dose to thehypophysis of greater than 17 Gy, a 15% risk at a mean dose of less than 17 Gy,and no hypofunction at a mean dose of 15 Gy or less. They observed an 85% riskfor hypothyroidism when the mean dose was greater than 17 Gy, a 15% risk if themean dose was less than 17 Gy, and no dysfunction if the mean dose was 15 Gy orless. They observed corticotropic hypofunction in 85% of patients receiving a meandose of greater than 20 Gy, a 10% risk if the dose was less than 20 Gy, and noobserved corticotropic hypofunction if the dose was less than 18 Gy.25

TREATMENT RESULTS FOR SKULL BASEMENINGIOMASStereotactic Radiosurgery Results

Initially used to treat postoperative recurrences and medically unsuitable surgicalcandidates, radiosurgery has evolved in its application as a treatment modality of valuein the care of patients who have these difficult lesions. Currently, radiation is considereduseful as primary treatment and as an adjunct to surgical therapy. Several investigatorshave reported their rates of tumor control and progression after radiosurgery for skullbase meningiomas and, recently, long-term data have been reviewed. Kreil andcolleagues reviewed the results of their 200 patients who had benign skull base menin-giomas evenly divided between postsurgical radiosurgery and primary radiosugerytreatments to a median marginal dose of 12 Gy (range 10–20), with a median follow-up period of 95 months. They reported a 10-year progression-free survival of 97%.They observed a neurologic deterioration in 4.5% of patients.26 Zachenhofer andcolleagues evaluated their series of 36 patients who had cranial base meningiomasand were treated with gamma knife radiosurgery with or without previous surgery. Theirfollow-up period was a mean of 103 months. They prescribed a mean marginal dose of17 Gy. They reported a long-term tumor control rate of 94% (31 of 33 patients) with sizereduction in 70%. One patient developed late pituitary dysfunction, and one patientdeveloped progressive cranial nerve palsy. In addition, they reviewed the literatureon results of gamma knife radiosurgery on skull base meningiomas with meanfollow-up of 15 to 82 months. They reported tumor control rates of 82% to 100%and new neurologic deficits in 0 to 27%.4 Iwai and colleagues27 evaluated their 108patients after lower-dose gamma knife radiosurgery at a mean of 86 months after giving


a median marginal dose of 12 Gy (range 8–12 Gy). They reported a 10-year progressionfree survival of 86% and adverse neurologic sequelae in 6%.

Comparable studies using linac-based radiosurgery units to treat skull base menin-giomas generally indicate similar tumor control rates of 92% to 100% over medianfollow-up periods of 23 to 48 months using median marginal doses of 14.6 to 18Gy. More recent studies identified neurologic dysfunction post treatments of 4.7%to 24%.26,28,29

Stereotactic Radiotherapy Results

Fractionated SRT techniques have been applied with success to meningiomas of theskull base. Lo and colleagues published a retrospective comparison of SRS and frac-tionated SRT for the treatment of intracranial meningiomas, which included 30 skullbase lesions in 53 patients. They reported similar control rates of 92% for each groupat 3 years, using median tumor marginal doses of 14 Gy in single fraction and a mediandose of 54 Gy at the 88% isodose given over a median of 30 fractions. They suggestthat larger tumors and those closer to sensitive neural structures, such as the opticapparatus, would benefit from fractionation.30 Hamm and colleagues reviewed their224 patients who had skull base meningiomas alone and underwent radiosurgery,fractionated SRT, or HSRT. Those patients who had the largest tumor volumes orthose in the closest proximity to the optic apparatus (<2 mm) underwent the moreconventional fractionated SRT with a median dose of 55.8 Gy at the 90% to 95%isodose (1.8–2.0 Gy per fraction). HSRT consisted of doses of 4 to 5 Gy per fraction.Single-fraction radiosurgery used a marginal dose of 12.8 to 18 Gy. Their medianfollow-up was 3 years. They reported a local control rate of 96.9% and a 4.4% neuro-logic deterioration but no differences between the three types of focused radiation. Intheir review of published studies, they observed similar control rates for gamma kniferadiosurgery, linac radiosurgery, and SRT (92%–100% overall) at the various centersover the median follow-up periods (19–82 months), and similar mean deteriorationrates were 6%–7% for the three treatment modalities.29

Gorman and colleagues examined their results from HSRT for 38 patients who hadskull base meningiomas who had a median follow-up of 47 months. They treated witha median dose of 37.5 Gy to the 80% isodose in 15 fractions. They reported no tumorprogression on imaging in this early follow-up period and 11% new cranial nervedysfunction.5 Milker-Zabel and colleagues published their long-term follow-up results(median 78 months) for their 57 patients who had cavernous sinus meningiomastreated with linac-based fractionated SRT using a median dose of 57.6 Gy to the90% isodose, 1.8 Gy per fraction, with a median of 32 fractions. Their 5- and 10-year local tumor control rates were 100%, with a 7% new cranial nerve dysfunction,and no pituitary dysfunction.31

Fractionated SRT has been shown safe and effective in the treatment of skull basemeningiomas with low rates of neurotoxicity comparable to those of radiosurgery.Although those patients whose tumors were treated with single-fraction radiosurgerytended to have smaller tumors that were located farther from sensitive structures,such as the brainstem and optic apparatus, compared with those patients treatedwith fractionated SRT, the tumor control rates and neurologic toxicities werecomparable.

Proton Beam Results

Although fewer total data have been accumulated and published using protons thaneither LINAC based x-ray or Cobalt-60 based Gamma radiation, several investigatorshave described their results in treating meningiomas of the skull base with proton beam


fractionated SRT. Weber and colleagues reviewed their series of 16 patients who hadmeningiomas that were pathologically confirmed or highly suggestive based onimaging characteristics. Thirteen of these were skull base lesions. All were treatedwith proton beam therapy. The investigators targeted the tumor, dural tail, and a 4-to 6-mm margin. The median tumor volume was 17.5 cm3. The median doseprescribed was 56 cobalt gray equivalents (CGE 5 proton Gy � 1.1). The medianfollow-up time was 34.1 months. Their 3-year local control rate was 91.7% witha 19% overall late neurologic complication rate consisting of one case each of opticneuropathy, retinopathy, and radiation-induced cortical injury.32

Vernimmen and colleagues reviewed their 23 patients who underwent proton beamtherapy for skull base meningiomas. Five were treated with SRT (>15 fractions) and 18patients were treated with a hypofractionated therapy (HSRT) (3 fractions). Their meantarget volumes were 15.6 cm3 and the mean single-fraction dose equivalent (sfe) in theHSRT group was 20.3 sfe/CGE. In the SRT group, the dose ranged from 54 CGE in 27fractions to 61.6 CGE in 16 fractions. Vernimmen and colleagues demonstrated for theHSRT group a radiologic response rate of 88% and 29% regressed partially orcompletely. Clinically, 89% improved or remained the same. The SRT group all re-mained stable on MRI and stable or improved clinically. The mean follow-up timewas 40 months in the two groups combined. They reported 11% late side effects con-sisting of an eighth nerve palsy and one new case of temporal lobe epilepsy.33

SUMMARY

Skull base meningiomas remain a difficult clinical condition with several options formanagement. Surgery provides the gold standard for therapy when complete resec-tion can be accomplished with minimal neurologic risk. Focused radiation techniqueshave become more important as an adjunct to surgical therapy and as a primary treat-ment in selected individual cases. SRS in a single fraction or SRT in full- or hypofrac-tionated delivery systems offers the opportunity for tumor control with reasonable riskprofiles. The source of radiation may be provided by gamma photons, x-ray photons,or charged nuclei particles, such as protons, each with clinical and practical advan-tages and disadvantages; however, the complexity of the anatomy of the skull basedoes not eliminate the risk for surrounding neurovascular structures with any of thesemodalities. Additionally, the natural history of these tumors needs to be consideredwhen the decision to move forward with treatment is made. There has been a movetoward more individualized recommendations regarding the application of therapyfor these conditions.

Focused radiation techniques and microsurgery have evolved as complementarytools in the treatment of meningiomas of the skull base. Patients should be consideredappropriate for microsurgery in those clinical situations in which complete surgicalresection can be achieved and a minimum of acceptable morbidity is likely. Patientsshould be considered for SRT or SRS in situations in which surgery is not desired,is not deemed safe, or is unlikely to accomplish complete resection without significantmorbidity. Additionally, surgery and focused radiation techniques can be used ina complementary fashion. When tumor is purposefully left behind at surgery or tumorprogresses on follow-up after surgery, SRT or SRS may afford better tumor control.Certain clinical conditions also arise where anatomic reduction of tumor from criticalstructures, such as the brain, brainstem, or the optic apparatus, is required forsymptom relief, and as such a reduction may improve the ability to subsequentlydeliver optimal focused radiation to the residual tumor with the least risk for neurologiccomplications.


Finally, although complete surgical resection continues to be the result with the bestlong-term outcome, in situations where this approach is not possible or the risks forsurrounding central nervous system tissues and vessels are too great, surgery canbe reasonably modified to the point where a surgeon resects as much as is consideredacceptable, preserving neurovascular structures as best as possible, and addressingresidual tumor with focused radiation techniques. Heros has an excellent review ofthese issues in which he describes his current surgical approaches to cranial menin-giomas at various locations throughout the skull base and his application of radiosur-gery in practice.34

Focused stereotactic radiation techniques continued to evolve. They currently canmake a significant contribution to the care of the patients who have skull base menin-giomas. The particular care and direction of treatment of these patients can be individ-ualized based on accurate anatomy identified on ever-improving imaging techniquesand on expected biologic behavior given a combination of clinical conditions, patho-logic identification, and sequential image analysis. There are limits to accurate deci-sion making as these are slow growing tumors with a potentially long natural history,and the lengths of follow-up analyses published on the various modalities are relativelyshort. The best long-term outcomes for patients will be obtained when all the optionsavailable for each patient are considered, the risks of the various treatment modalitiesare addressed, and the likely outcomes are reviewed in light of the natural history ofthese tumors. Although the treatment of these tumors remains an exceptional clinicaltherapeutic challenge, focused radiation techniques, such as SRS and SRT, offerconsiderable help in the application of treatment options for individual patients.

REFERENCES

1. Simpson D. The recurrence of intracranial meningiomas after surgical treatment.J Neurol Neurosurg Psychiatr 1957;20(1):22–39.

2. Adegbite AB, Khan MI, Paine KW, et al. The recurrence of intracranial meningi-omas after surgical treatment. J Neurosurg 1983;1:51–6.

3. Commins DL, Atkinson RD, Burnett ME. Review of meningioma histopathology.Neurosurg Focus 2007;23(4):E3 [review].

4. Zachenhofer I, Wolfsberger S, Aichholzer M, et al. Gamma-knife radiosurgery forcranial base meningiomas: experience of tumor control, clinical course, andmorbidity in a follow-up of more than 8 years. Neurosurgery 2006;58(1):28–36[discussion: 28–36].

5. Gorman L, Ruben J, Myers R, et al. Role of hypofractionated stereotactic radio-therapy in treatment of skull base meningiomas. J Clin Neurosci 2008;15(8):856–62 [Epub 2008 Jun 12].

6. Elia AE, Shih HA, Loeffler JS. Stereotactic radiation treatment for benign meningi-omas. Neurosurg Focus 2007;23(4):E5 [review].

7. Bindal R, Goodman JM, Kawasaki A, et al. The natural history of untreated skullbase meningiomas. Surg Neurol 2003;59(2):87–92.

8. Shaffrey M, Dolenc V, Lanzino G, et al. Invasion of the internal carotid artery bycavernous sinus meningiomas. Surg Neurol 1999;53:167–71.

9. Sen C, Hague K. Meningiomas involving the cavernous sinus: histological factorsaffecting the degree of resection. J Neurosurg 1997;87(4):535–43.

10. Larson J, van Loveren H, Balko M, et al. Evidence of meningioma infiltration intocranial nerves: clinical implications for cavernous sinus meningiomas. J Neuro-surg 1995;83(4):596–9.


11. De Jesus O, Sekhar LN, Parikh HK, et al. Long-term follow-up of patients withmeningiomas involving the cavernous sinus: recurrence, progression, and qualityof life. Neurosurgery 1996;39(5):915–9 [discussion: 919–20].

12. DeMonte F, Smith HK, al-Mefty O. Outcome of aggressive removal of cavernoussinus meningiomas. J Neurosurg 1994;81(2):245–51.

13. O’Sullivan MG, van Loveren HR, Tew JM Jr. The surgical resectability of meningi-omas of the cavernous sinus. Neurosurgery 1997;40(2):238–44 [discussion:245–47].

14. Natarajan SK, Sekhar LN, Schessel D, et al. Petroclival meningiomas: multimodal-ity treatment and outcomes at long-term follow-up. Neurosurgery 2007;60(6):965–79.

15. Barbaro N, Gutin P, Wilson C, et al. Radiation therapy in the treatment of partiallyresected meningiomas. Neurosurgery 1987;20(4):525–8.

16. Kondziolka D, Lunsford LD, Flickinger JC. The application of stereotactic radio-surgery to disorders of the brain. Neurosurgery 2008;62(Suppl 2):707–19[discussion: 719–20] [review].

17. Chen CC, Chapman P, Petit J, et al. Proton radiosurgery in neurosurgery. Neuro-surg Focus 2007;23(6):E5 [review].

18. Ammirati M, Bernardo A, Ramsinghani N, et al. Stereotactic radiotherapy ofcentral nervous system and head and neck lesions, using a conformal inten-sity-modulated radiotherapy system (Peacock system). Skull Base 2001;11(2):109–19.

19. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the visual pathwaysand cranial nerves of the cavernous sinus to stereotactic radiosurgery. J Neuro-surg 1998;88(1):43–50.

20. Duma CM, Lunsford LD, Kondziolka D, et al. Stereotactic radiosurgery ofcavernous sinus meningiomas as an addition or alternative to microsurgery.Neurosurgery 1993;32(5):699–704 [discussion: 704–5].

21. Tishler RB, Loeffler JS, Lunsford LD, et al. Tolerance of cranial nerves of thecavernous sinus to radiosurgery. Int J Radiat Oncol Biol Phys 1993;27(2):215–21.

22. Kondziolka D, Flickinger JC, Lunsford LD. The principles of skull base radiosur-gery. Neurosurg Focus 2008;24(5):E11 [review].

23. Morita A, Coffey RJ, Foote RL, et al. Risk of injury to cranial nerves after gammaknife radiosurgery for skull base meningiomas: experience in 88 patients. J Neu-rosurg 1999;90(1):42–9.

24. Stafford SL, Pollock BE, Leavitt JA, et al. A study on the radiation tolerance of theoptic nerves and chiasm after stereotactic radiosurgery. Int J Radiat Oncol BiolPhys 2003;55(5):1177–81.

25. Vladyka V, Lisc�ak R, Novotny J Jr, et al. Radiation tolerance of functioning pitui-tary tissue in gamma knife surgery for pituitary adenomas. Neurosurgery 2003;52(2):309–16 [discussion: 316–17].

26. Kreil W, Luggin J, Fuchs I, et al. Long term experience of gamma knife radiosur-gery for benign skull base meningiomas. J Neurol Neurosurg Psychiatr 2005;76(10):1425–30.

27. Iwai Y, Yamanaka K, Ikeda H. Gamma knife radiosurgery for skull base menin-gioma: long-term results of low-dose treatment. J Neurosurg 2008;109(5):804–10.

28. Deinsberger R, Tidstrand J, Sabitzer H, et al. LINAC radiosurgery in skull basemeningiomas. Minim Invasive Neurosurg 2004;47(6):333–8.


29. Hamm K, Henzel M, Gross MW, et al. Radiosurgery/stereotactic radiotherapy inthe therapeutical concept for skull base meningiomas. Zentralbl Neurochir2008;69(1):14–21.

30. Lo SS, Cho KH, Hall WA, et al. Single dose versus fractionated stereotactic radio-therapy for meningiomas. Can J Neurol Sci 2002;29(3):240–8.

31. Milker-Zabel S, Zabel-du Bois A, Huber P, et al. Fractionated stereotactic radiationtherapy in the management of benign cavernous sinus meningiomas: long-termexperience and review of the literature. Strahlenther Onkol 2006;182(11):635–40[review].

32. Weber DC, Lomax AJ, Rutz HP, et al. Swiss Proton Users Group. Spot-scanningproton radiation therapy for recurrent, residual or untreated intracranial meningi-omas. Radiother Oncol 2004;71(3):251–8.

33. Vernimmen FJ, Harris JK, Wilson JA, et al. Stereotactic proton beam therapy ofskull base meningiomas. Int J Radiat Oncol Biol Phys 2001;49(1):99–105.

34. Heros RC. Effect of modern radiation techniques on the surgery of nonmalignantintracranial tumors. Neurosurg Focus 2008;24(5):1–5.

Radiosurgeryfor GlomusJugulare Tumors

Jonathan P. Miller, MDa, MarounT. Semaan, MDb,Robert J. Maciunas, MD, MPHa,c, Douglas B. Einstein, MD, PhDc,Cliff A. Megerian, MDa,b,*

KEYWORDS

� Glomus jugulare tumor � Gamma knife radiosurgery� CyberKnife radiosurgery � LINAC radiosurgery� Minimally invasive glomas jugulare surgery

Paragangliomas (glomus tumors) are slow-growing, usually benign, highly vasculartumors of paraganglionic tissue. During embryogenesis, paraganglionic tissue isderived from the migration of neural crest cells in close association with the sympatheticnervous system. Within the head and neck these cell rests are predominantly distrib-uted throughout the middle ear in close association with the Jacobson nerve (branchof cranial nerve [CN] IX) and the Arnold nerve (branch of CN X), the jugular foramen, va-gus nerve, and carotid body. Tumors arising from paraganglionic tissue can be dividedinto adrenal paraganglioma (pheochromocytoma) and extra-adrenal paraganglioma.The term ‘‘branchiomeric paraganglioma’’ is used to describe extra-adrenal tumorsarising from head and neck paraganglionic tissue.1 The presence of paraganglionictissue associated with the Jacobson nerve was recognized as early as 1840.2 Guildcoined the term ‘‘glomic tissue’’ to describe the vascularized ganglionic tissue alongthe adventitia of the jugular bulb and promontory.3 In 1945, Rosenwasser4 successfullyremoved a middle ear paraganglioma that he called a carotid body of the middle ear.Although frequently used, the term glomus is a misnomer and denotes the previouslybelieved origin from specialized pericytes within arteriovenous complexes (glomus).Carotid and aortic bodies are the only two paraganglia known to function as chemore-ceptors. Hence, the term chemodectoma is improperly used to designate paraganglio-mas. The most common head and neck paraganglioma is the carotid body tumor.

a Department of Neurological Surgery, Case Western Reserve University and UniversityHospitals of Cleveland, 11100 Euclid Avenue, Cleveland, OH 44106, USAb Department of Otolaryngology–Head and Neck Surgery, Case Western Reserve University andUniversity Hospitals of Cleveland, 11100 Euclid Avenue, Cleveland, OH 44106, USAc Department of Radiation Oncology, Case Western Reserve University and University Hospitalsof Cleveland, 11100 Euclid Avenue, Cleveland, OH 44106, USA* Corresponding author. Department of Neurological Surgery, Case Western Reserve Universityand University Hospitals of Cleveland, 11100 Euclid Avenue, Cleveland, OH 44106.E-mail address: [email protected] (C.A. Megerian).




Miller et al690

Within the temporal bone two types of paragangliomas exist: glomus tympanicum andglomus jugulare or jugulotympanic paraganglioma (JTP). Glomus tympanicum arisesfrom rests of paraganglionic tissue associated with Jacobson and Arnold nerves.Glomus jugulare is believed to arise from similar paraganglionic rests within the adven-titia of the jugular bulb, intimately associated with CNs IX, X, and XI. The incidence ofparagangliomas is estimated at 0.012%.5 No racial predilection has been noted andthey commonly occur in the fourth and fifth decades of life.6 Multicentricity is seen in3% to 10%.7,8 Familial paraganglioma, inherited as an autosomal dominant disorderwith genetic imprinting through paternal transmission, occurs in 1/30,000 head andneck tumors.9 In contrast to the sporadic type, individuals affected by this rare conditiondevelop multiple paragangliomas, often bilateral and at an earlier age. Multicentricity isseen in 30% of cases of familial paragangliomas. The most common association isa carotid body tumor and ipsilateral glomus tympanicum.8 Recent genetic studiesfound linkage to two distinct chromosomal loci, 11q13.1 (PGL 2) and 11q22.3-q23(PGL 1).9–11 Germline mutations in the genes encoding for the three subunits of the mito-chondrial complex II or succinate dehydrogenase (SDHD) (an essential component ofKrebs cycle), SDHB, SDHC, and SDHD, have been recently found.12–14 Some reportssuggested that the same pattern of mutations may be found in sporadic cases and otherhave disputed a common genetic alteration.14,15

Paragangliomas spread through pathways of least resistance: air cell tracts,vascular channels, naturally occurring fissures, and foramina. Different patterns ofintracranial spread, ‘‘dangerous triangles,’’ have been described.16 Paragangliomascan travel through the peritubal air cells into the petrous apex, petrous carotid artery,and middle cranial fossa, or through the hypotympanic air cell tract between thejugular bulb and carotid artery into the posterior fossa.

Malignant paragangliomas are rare and reported in 5% of cases.17 The diagnosis ofmalignancy is based on the confirmed presence of distant metastasis. Cellular criteriaand invasiveness has not been established as a prerequisite for making the diagnosisof malignant paragangliomas. Paragangliomas may originate in the temporal bone andinvade into the cerebellopontine angle (CPA).18,19

PATHOLOGY

Paragangliomas contain two cell types: the chief cells and sustentacular cells. Thechief cells possess secretory granules that contain catecholamines. They are deriva-tives of neural crest cells and belong to the diffuse neuroendocrine system (DNES).20

Cells that are members of the DNES are capable of secreting neurotransmitters andhave similar cell receptors.

Despite the detection of catecholamine precursors in most paragangliomas, only1% to 3% of head and neck paragangliomas excrete norepinephrine. Unlike adrenalparagangliomas (pheochromocytoma), extra-adrenal paragangliomas rarely produceepinephrine, because the rate-converting enzyme phenylethanolamine-N-methyltransferase is absent.20,21

On light microscopy, chief cells form clusters (zellballen) embedded with supportcells (sustentacular cells) within an abundant vascular stroma. Mitosis and capsularinvasion have been described in the benign variant and are not considered determi-nant of malignant behavior. Unmyelinated nerve fibers may be seen.6

CLINICALMANIFESTATIONS

Paragangliomas may be sporadic or part of an inherited syndrome with an autosomaldominant mode of transmission with genetic imprinting. The hereditary form is


characterized by a higher incidence of multicentricity and associated tumors.22 Thegenetic imprinting through paternal transmission seen in these tumors is reflectedby the absence of phenotypic expression in offspring of females carrying the mutatedgene, whereas inheriting a mutated copy from the father results in phenotypic expres-sion with high penetrance.10,11

Early-stage paragangliomas present with symptoms related to involvement of themiddle ear cleft. Pulsatile tinnitus and conductive hearing loss are seen in 98% and63%, respectively.23 Glomus tympanicum tends to spread through pathways of leastresistance along the peritubal air cells, intrapetrous carotid artery, and petrous apex.Glomus jugulare presents with pulsatile tinnitus and cranial neuropathy. Theseneoplasms tend to spread trough the hypotympanic air cells tract, around the jugularbulb, inferior petrosal sinus, and carotid artery into the jugular foramen and posteriorfossa. The jugular foramen syndrome may be seen in 50% of tumors.18 Involvement ofCN IX has been reported in 4% to 43% of cases, CN X in 5% to 57%, CN XI in 4% to43%, and CN XII in 7% to 43%.6,18,24,25 Glomus tympanicum is seen as a retrotym-panic red mass on the promontory. Glomus jugulare presents as a middle ear masswhen it erodes into the floor of the hypotympanum, and as an aural polyp if erosionof the tympanic membrane has occurred. Invasion of the middle ear results in conduc-tive hearing loss. Of interest, paragangliomas encroach onto the ossicles but do notcause ossicular erosion.26 Pulsatile tinnitus is an indicator of the tumor hypervascular-ity. Brown sign (tumor blanching with positive pressure using pneumatoscopy) orAquino sign (cessation of pulsations with compression of the ipsilateral carotid artery)may be seen. As tumor invades into deeper structures, additional lower cranialneuropathies, sensorineural hearing loss, vertigo, and pain may ensue. Extension ofthe tumor through the facial recess and retrofacial air cells may result in encasementof the facial nerve. Facial nerve involvement is present in 21% to 33%.25–27 In largetumors, Horner syndrome, facial hypesthesia, and diplopia may be seen due to exten-sion into the carotid artery and intradural or extradural involvement of CN VI.25,26 Inone study,18 posterior fossa involvement was seen in 50% of cases with jugularforamen syndrome and in 75% of cases with CN XII neuropathy.

It is prudent not to biopsy a vascular middle ear mass, because this may result inprofuse bleeding or exsanguination if not properly controlled. An aberrant carotidartery or high-riding, dehiscent jugular bulb may present as a reddish or bluishmass in the hypotympanum and possibly masquerade as a glomus tumor.

In cases in which multicentricity is seen, an appropriate workup should includescreening for other adrenal and extra-adrenal tumors and for familial type paraganglio-mas. A 24-hour urinary vanillylmandelic acid (VMA), plasma catecholamines, andurinary beta-metanephrine and normetanephrine may be obtained as part of thebiochemical screening workup.28 History suggestive of labile hypertension, attacksof headache, anxiety, sweating, and flushing may suggest a catecholamine-producingtumor. Because of the rarity of functional head and neck paragangliomas, elevatedplasma catecholamines should prompt the search for a pheochromocytoma, and peri-operative alpha blockade may be indicated to avoid catecholaminergic crisis. Geneticscreening is not widely available as a screening tool and remains confined to theresearch laboratory. A gadolinium-enhanced MRI of the head and neck is the goldstandard modality to screen for multicentric paragangliomas.

DIAGNOSIS

The role of neuroradiology in determining the origin, extent, and nature of the tumor isof paramount importance. Its role is not limited to the characterization of the lesion

Miller et al692

itself, but is essential in differentiating these lesions from vascular anomalies ortemporal bone malignant neoplasms and in screening for other contralateral or ipsilat-eral lesions in cases of familial paragangliomas.

On high-resolution CT of the temporal bone, glomus tympanicum in its early phasesappears as a well-circumscribed soft tissue mass localized on the promontory. Thedifferential diagnosis of a soft tissue density confined to the promontory includescongenital cholesteatoma, persistent stapedial artery, and aberrant carotid artery.29

Radiographically, glomus jugulare tumors are associated with an irregular erosiveenlargement of the jugular plate (floor of the hypotympanum) and the jugulocarotidspine, a pattern that is described as ‘‘moth-eaten.’’ Depending on their origin, theseneoplasms extend through the skull base to involve the jugular foramen in its neuraland vascular compartment and eventually progress intracranially through intra or ex-tradural pathways or extracranially through cervical extension.30

MR weighted images are superior to evaluate tumor vascularity, extension alongneural foramina, and multicentricity.31 In addition, MR venography is helpful inensuring intraluminal patency of the jugular vein and retrograde involvement of thevenous system and the status of the contralateral venous system. On T1-weightedimages, paragangliomas appear hypointense with a speckled appearance. On gado-linium-enhanced T1 images, early and pronounced enhancement is seen witnessingthe hypervascular nature of the neoplasm. On T2-weighted images, paragangliomasare hyperintense, and when larger then 2 cm the serpentine flow void pattern isdescribed as a ‘‘salt and pepper’’ appearance.31,32

Paragangliomas exhibit an intense blush or a ‘‘bag of worm’’ appearance on angi-ography. MR angiography may substitute for four-vessel angiography as a tool to eval-uate the vascularity of skull base tumors. Small vascular anomalies and blood feedersare better seen on conventional angiography, however.33 Four-vessel angiography isindicated if preoperative embolization is planned. The latter may decrease intraoper-ative blood loss and operative time.34

TREATMENTClassification Schemes

Different classification schemes have been described. The scheme advised by Fischdivides these lesions into four categories:35 type A, tumors limited to the middle ear;type B, tumors limited to the tympanomastoid compartment; type C, tumorsinvolving the infralabyrinthine air cells tract and intrapetrous carotid canal and ex-tending into the petrous apex; and type D, tumors with intracranial extension. Thescheme described by Jackson and colleagues36 divides JTPs into four types: typeI, small tumors involving the jugular bulb, middle ear, and mastoid; type II, tumorsextending under the internal auditory canal that may have an intracranial extension;type III, tumors extending into the petrous apex with or without intracranial extension;and type IV, tumors extending beyond the petrous apex into the clivus or infratem-poral fossa. The House ear group adopted the classification devised by AntonioDe La Cruz.23 JTPs are considered tympanic when tumors are entirely confined tothe mesotympanum, tympanomastoid when they extend beyond the limits of themesotympanum without eroding the jugular plate, jugular bulb when the tumorsare confined to the jugular foramen without involvement of the carotid artery orextension intracranially, carotid artery when the tumor involves the intrapetrouscarotid artery, and transdural when the tumor extends intracranially. These classifica-tions help guide the surgeon in the selection of the most appropriate approach toeradicate the disease.


Operative Management

Tympanic tumors may be completely excised by a transcanal approach. To ensureadequate exposure, the tympanic annulus is elevated circumferentially and leftattached to the manubrium. The tumor is excised using cup forceps and hemostasisis obtained with hemostatic sealant or absorbable gelatin foam (Gelfoam). Tympano-mastoid tumors are removed by a classic mastoidectomy with an extended facialrecess approach. This approach can at times obviate the need to elevate a tympano-meatal flap. For jugular bulb tumors the mastoid-neck approach is used. Only afterpreoperative embolization has been achieved is a mastoidectomy performed. Themastoid tip is taken down along with the insertion of the sternocleidomastoid muscle.The posterior belly of the digastric is dissected free and reflected anteriorly to exposethe great vessels. The jugular vein is ligated in the neck. The proximal sigmoid sinus ispacked extraluminally and the anterior wall of the segment involved by the tumor isexcised preserving the posterior dural surface of the vein intact. Bleeding from theinferior petrosal sinus is controlled by packing and the remainder of the tumor isremoved. Some centers perform preoperative embolization of the inferior petrosalsinus to minimize intraoperative bleeding and facilitate neural microdissection.37

Facial nerve rerouting can often be avoided if a fallopian bridge technique is used toaccess the tumor.24 This technique involves the removal of the retrofacial and infrala-byrinthine air cell tracts. Anatomic factors, such as an anteriorly displaced jugularbulb, may limit exposure. For carotid artery tumors the infratemporal fossaapproaches described by Fisch38 provide adequate access. Detailing the surgicaltechnique is beyond the scope of this discussion. Transdural involvement is ad-dressed by different posterior fossa approaches. Paragangliomas involving the CPAare often extensive, and their treatment involves a combined extradural and intraduralapproach, which may or may not be a staged procedure.39 The transpetrous approachmay be performed for complete tumor excision, with or without facial nerve rerouting,with or without a transcochlear approach, and with additional craniotomies designedas dictated by tumor extension.40 For tumors extending into the CPA without violatingthe otic capsule, a transsigmoid, retrolabyrinthine approach may provide adequateexposure of the upper compartment of the CPA. For lesions with inferior extension,a retrosigmoid approach, which may be combined with a far-lateral approach, maybe needed. Lesions extending to the middle cranial fossa with preserved hearingmay be removed using the subtemporal-retrolabyrinthine approach.41 These exten-sive operations are often associated with significant morbidity and increased opera-tive time.

CONVENTIONAL RADIOTHERAPY

Conventional fractionated external beam radiotherapy represents an alternative tosurgery for paragangliomas and has been used as primary treatment in patients forwhom surgery is contraindicated because of advanced age or other comorbidities.Radiotherapy has also been used as an adjuvant to surgery for patients who havelarge or unresectable tumors and as salvage treatment for residual disease.42,43

Glomus tumors are relatively radioresistant and tend to persist after radiotherapy,although control of tumor growth has been demonstrated.44–46 The effect of radiationon paraganglioma tissue is not well understood. Postirradiation histopathologicstudies demonstrate variable amounts of perivascular fibrosis45,47 and increase instromal connective tissue,48,49 but there is minimal effect on tumor cells themselvesor catecholamine secretion after radiation.46 Nevertheless, long-term control of tumorgrowth can be obtained using radiotherapy.

Miller et al694

Tumor control seems to be related to dose, with frequent recurrence reported atlower radiation doses.50–56 At higher doses, multiple studies have shown tumorcontrol comparable to that obtained by surgery, with only rare progression,54,57–61

and similar or even lower rate of complication.62 One retrospective review of studiesfrom 1965 to 1988 demonstrated long-term control rates after surgery alone, surgeryand radiotherapy, and radiotherapy alone to be 86%, 90%, and 93%, respectively,with lowest complication rates in the radiotherapy group.63 Another review thatincluded 582 patients from 24 series comparing surgery and radiotherapy demon-strated recurrence in 7% versus 8% and death in 2.5% versus 6%.64

Unfortunately, radiotherapy is associated with risk for several significant long-termcomplications. To obtain an adequate dose of radiation to the tumor, normal tissue ofthe upper neck and skull base must be included in the radiation field. The temporalbone itself is exquisitely radiosensitive and is susceptible to osteoradionecro-sis.59,61,65 Other reported complications of radiotherapy for glomus tumors includeradiation-induced otitis, mastoiditis, altered taste sensation, alopecia, mucositis,dermatitis, cranial nerve palsy (such as facial weakness or hearing loss), brainnecrosis, radiation-induced secondary malignancy, metastasis, cerebrospinal fluidleakage, and insufficiency of pituitary or hypothalamic function.54,59,61,66–71 For thisreason, radiotherapy is not commonly used to treat glomus tumors. Elderly patientsor patients for whom extensive surgery is contraindicated may benefit from a limitedsurgical procedure with adjuvant radiotherapy.72

RADIOSURGERY

Over the past several years, radiosurgical techniques have emerged as a promisingalternative to other therapeutic strategies for treatment of glomus jugulare tumors.Developed by Lars Leksell in 1951, radiosurgery is a refinement of traditional radio-therapy in which a large number of intersecting beams of radiation are focused directlyon the tumor.73,74 This method allows for a steep drop-off in radiation dose around thetumor margin so that a large dose of radiation can be provided to the tumor while mini-mizing radiation exposure to adjacent tissues. Although glomus tumors are relativelyradioresistant, they are otherwise ideal candidates for radiosurgical treatment,because they are well-demarcated on MRI, rarely invade the brain, are usually fairlysmall, and lie close to vital structures.75 Because the volume of irradiated tissue adja-cent to the tumor is small, the rate of complications is also significantly lower than forconventional radiotherapy.75,76 Several studies have documented stability or reduc-tion in tumor size after radiosurgery without new neurologic deficits.76–89 There arecurrently three techniques available for radiosurgery: gamma knife, linear accelerator(LINAC), and CyberKnife. A schematic of each of these is shown in Fig. 1, and photo-graphs of a gamma knife and a CyberKnife machine are shown in Fig. 2.

GAMMA KNIFE RADIOSURGERY

The most widely studied technique for radiosurgical treatment of glomus tumors isgamma knife, in which convergent beams of gamma radiation from 201 separatecobalt-60 sources are focused on a point into which the patient is moved.74 Beforetreatment, an MRI is obtained while the patient is wearing a stereotactic head frame,and the tumor is then carefully analyzed by a neurosurgeon and radiation oncologistusing a computer to identify a treatment plan that will effectively treat the tumor andavoid adjacent tissue. The patient is placed into the machine so that the tumor remainsat the focal point of radiation for the precise amount of time to receive the appropriatedose of radiation. Positioning of the patient within the machine can be then be slightly

Fig. 1. Three types of radiosurgery. All use the intersection of many beams of radiation toproduce a high dose to the tumor with low dose to adjacent tissues. (A) Gamma knife radio-surgery. A total of 201 separate cobalt-60 sources are arranged in a hemisphere, each behinda column of lead that produces beams that converge. The tumor is placed at the location ofthe convergence to provide radiation. (B) Conventional LINAC radiosurgery. Convergenthigh-voltage x-ray beams are produced by a gantry that rotates and pivots around thetumor. (C) CyberKnife radiosurgery. A LINAC is mounted on a robotic arm and aims x-raybeams at the tumor from different directions.


altered and the process repeated several times to provide treatment to an irregularvolume. The entire process takes only a few hours and can be performed in an outpa-tient setting.

Gamma knife radiosurgery to treat glomus jugulare tumors was first reported in1995 in a case series of skull base tumors.77 Two glomus jugulare tumors wereincluded, and there was no radiographic or clinical progression 19 months after treat-ment. Two years later, Foote and colleagues76 reported a study of 9 patients designedto evaluate complications of radiosurgery for glomus tumors; they noted control oftumor growth in 8 of 9 and no apparent complications. A subsequent study fromthe same authors added 16 patients for a total of 25 and demonstrated no radio-graphic progression and only one complication (vertigo) after 25 months.78 Multipleother studies have shown high rates of tumor control with rare and transient compli-cations. In 1999, Eustacchio and colleagues79 reported 10 patients who had glomusjugulare, among whom 7 were receiving primary treatment, and noted 40% had

Fig. 2. Photographs of radiosurgical devices. (A) Gamma knife. The patient’s head is immo-bilized in a stereotactic head frame that is attached to the metal hemisphere, and the entiretable moves into the machine so that the tumor is precisely placed at the intersection of thebeams of radiation. (B) CyberKnife. The patient lies on the table and two orthogonal x-rays(visible at the top of the image) continuously check patient position and adjust the planaccordingly. The robotic arm moves the LINAC device around the patient, allowing for radi-ation to be delivered to the tumor from many directions, sparing adjacent tissue.

Miller et al696

decrease in tumor size after median of 37.6 months with no evidence of clinical orradiographic progression in any patient. In a subsequent publication, 9 more patientswere added for a total of 19 (10 treated primarily), with similar results.80 Meanwhile,Liscak and colleagues81,82 reported a multicenter experience including 66 glomus jug-ulare tumors, of whom 30 were primarily treated, and all patients had stabilization ordecrease in tumor size with decreased vascularity seen on angiography at median24 months. Jordan and colleagues83 reported 8 patients who were ineligible forsurgery who all underwent gamma knife as primary treatment with no radiographicprogression over 27 months and only one complication (vertigo). Saringer andcolleagues75 reported 13 patients treated with gamma knife who had no tumor growthat 5 years and transient cranial nerve complications in 2 patients.

More recently, Sheehan and colleagues84 reported 8 patients who had recurrent,residual, or unresectable glomus tumors; all 8 had no clinical or radiographic progres-sion over median 32 months, and tumor shrinkage occurred in 3 of these patients.Gerosa and colleagues85 treated 20 patients and noted shrinkage in nearly half.Among patients in this study, symptoms remained stable in 13, improved in 5, andworsened (hearing loss) in 2. A radiologic volumetric analysis of 15 patients whohad glomus tumors treated with gamma knife radiosurgery discovered transientincrease in the size of tumors in 7 but sustained enlargement in only 4, with 8decreasing in size and 5 unchanged.86 In summary, gamma knife radiosurgery hasbeen shown to be effective to control growth in the vast majority of cases.

LINEAR ACCELERATOR AND CYBERKNIFE

LINAC radiosurgery is performed using a particle accelerator mounted on a gantry thatis able to rotate around the patient’s head. CyberKnife is a special type of LINAC inwhich a compact linear accelerator is mounted on a robotic arm that moves freelyaround the patient to irradiate the tumor from multiple directions. Conventional LINACradiosurgery requires placement of a frame before imaging and planning, but Cyber-Knife radiosurgery is performed without a frame, because there are x-ray devices that


repeatedly check patient position and reregister orientation in case of patientmovement.

In contrast to gamma knife, there are only a few reports of LINAC or CyberKnifeused for glomus jugulare tumors, but the results seem similar to those obtained usinggamma knife. Maarouf and colleagues87 reported 14 patients with glomus jugularetumors treated with LINAC; none had progression, and 8 reported symptomaticimprovement. Lim and colleagues88 reported 13 patients with 16 tumors treatedwith LINAC (5 patients) or CyberKnife (8 patients). Dosage ranged from 18 to 25 Gyusing one to three fractions, with no evidence of growth in any patient througha mean follow-up period of 44 months. Subsequently, Lim and colleagues89 reported21 tumors treated with LINAC or CyberKnife with a mean follow-up of 60 months,including 8 patients followed for at least 10 years. Two patients had transient hearingloss and tongue weakness, but no patient at last follow-up had any signs of radio-graphic or clinical progression.

COMBINED SURGERY-RADIOSURGERY

Radiosurgery has repeatedly been shown to be a safe and effective alternative tosurgery and radiotherapy. Most published series indicate that radiosurgery oftendoes not lead to improvement of symptoms, however, especially tinnitus, otalgia,and hearing loss. Liscak and colleagues82 reported improvement in tinnitus in only12 of 53 patients and improvement in hearing loss in 1 of 40 patients. Eusatacchioand colleagues80 treated 10 patients who had tinnitus, and only 4 reported improve-ment. Gerosa and colleagues85 noted improvement in tinnitus in 2 patients out of 4.Varma and colleagues86 treated 17 patients and noted tinnitus improved in 4, otalgiain 2, and hearing in 1; 7 patients had unchanged symptoms. Results from 9 recentstudies with respect to tumor control and clinical outcome are shown in Table 1. Alto-gether, of 175 patients reported to have been treated with gamma knife radiosurgeryalone, the vast majority (93%) had no evidence of tumor growth, but fewer than half(42%) had any improvement in symptoms. Radiosurgery also does not allow fora tissue diagnosis, and it is remotely possible that what radiographically appears tobe a glomus tumor is in fact another type of tumor that requires more aggressivetreatment.90

To effectively deal with a symptomatic glomus jugulare tumor without the risk oftotal surgical excision, it is possible to combine a limited tailored surgical resectionwith radiosurgery. Originally reported by Willen and colleagues91 in 2005, this para-digm combines the advantages of both techniques to produce a better outcomethan is obtainable by either alone. Use of planned postoperative radiosurgery allowsthe surgical procedure to be completed in an outpatient setting without dissectionof the entire jugular bulb where neurologic injury often occurs. Evacuation of tumorfrom the middle ear, mastoid, and dome of the jugular bulb produces immediateimprovement of symptoms caused by tumor in that area, and it displaces radiosurgicaltargeting away from the highly radiosensitive tympanic portion of the temporal bone.Both are relatively conservative methods that are easily tolerated by elderly patients inpoor health. Finally, this paradigm allows for tissue diagnosis, which can be importantto exclude other masses in the region of the jugular foramen that might masquerade asglomus jugulare tumors.

The approach for the tailored surgical resection consists of transcanal resection ofmesotympanic tumor or posterior auricular/transmastoid resection of middle ear ormastoid tumor using the extended facial recess approach. Tumor is dissected fromthe undersurface of the tympanic membrane and then dissected from the ossicular

Table 1Summaryofmajor series inwhich gammaknife radiosurgery alonewas used to treat glomus jugularetumors

Radiographic Response Clinical Outcome

SeriesNo.Patients Larger Stable Smaller Unknown Better Stable Worse Unknown

Eustacchio et al,199979

13 0 6 4 3 5 5 0 3

Liscak et al,199982

52 0 31 21 0 15 34 3 0

Jordan et al,200083

8 0 3 4 1 7 1 0 0

Saringer et al,200175

13 0 10 3 0 6 7 0 0

Eustacchio et al,200280

19 1 11 7 0 10 8 1 0

Foote et al,200278

25 0 17 8 0 15 9 1 0

Sheehan et al,200584

8 0 3 4 1 3 5 0 0

Gerosa et al,200685

20 8 11 1 0 5 13 2 0

Varma et al,200686

17 4 5 8 0 7 8 2 0

Total 175 13 97 60 5 73 90 9 3

Miller et al698

chain. At this point no attempt is made to remove tumor from the jugular bulb or toreroute the facial nerve. The tumor is dissected from the facial nerve as far as themastoid tip using facial nerve monitoring. A diagram of the surgical site after exposureand resection of the tumor is shown in Fig. 3. The surgical procedure is relatively

Fig. 3. Transmastoid posterior auricular approach to glomus jugulare tumor after exposure(A) and removal (B) of tumor. The tumor is resected from the middle ear and hypotympa-num without altering the course of the facial nerve. Tumor in the main portion of jugularbulb is left for radiosurgical treatment after removal of tumor in the hypotympanum anddome of the bulb. EFR, extended facial recess; FN, facial nerve; GJ, glomus jugulare tumor;I, incus; ISJ, incudostapedial joint; M, malleus; RFC, retrofacial air cells; SCC, semicircularcanals; SS, sigmoid sinus.


straightforward and can be performed as an outpatient. After a few weeks, gammaknife radiosurgery is performed. The frame is placed low and shifted toward theside of the tumor to place the tumor as close as possible to the center of the frame.The plan is then developed in such a way as to minimize radiation exposure to theinternal auditory canal and cochlea. An example of a representative gamma knifeplan is shown in Fig. 4.

To date, five patients have been treated using this paradigm. Patient characteristicsand treatment results after median 29 months of follow-up (range, 11–40 months) areshown in Table 2. All patients had immediate improvement or resolution of pulsatiletinnitus or ear pain, and one patient had complete resolution of both symptoms.Four patients had hearing preoperatively, and all four had improved hearing after treat-ment. Audiogram results for these patients are shown in Fig. 5. One patient who hadpreoperative facial nerve palsy experienced improvement from House grade IV tograde II, and one patient who had baseline CN XI weakness did not experience anychange postoperatively. The only treatment complication was transient facial palsyin one patient that appeared approximately 2 weeks after surgery, but it fully resolvedbefore radiosurgery and did not recur. Postoperative MR imaging revealed stabletumor size in two patients and a decrease by at least 20% in net tumor volume in threepatients. There were no immediate or delayed complications due to radiosurgery.

This novel paradigm combining outpatient tailored microsurgical and radiosurgicaltreatment of glomus jugulare tumors in the symptomatic patient appears to be safe,improving tinnitus, otalgia, and hearing loss while controlling tumor growth. Extendedfollow-up reveals that control of tumor size and clinical symptoms persists for yearsafter treatment (J.P. Miller, unpublished data, 2009).

Fig. 4. Axial contrast-enhanced T1-weighted MR image demonstrating radiosurgical treat-ment plan for glomus tumor in axial (A), coronal (B), and sagittal (C) section. 30% (9 Gy),40% (12 Gy), and 50% (15 Gy) isodose curves are noted.

Table 2Baseline characteristics, treatment parameters, and neurologic response for patients treated with combined surgery/radiosurgery

Age/sex Stage (Fisch)SurgicalApproach

Time toGKRS (mo)

Dose(50%/100%isodose)

Tumor/MatrixVolume(cm3) Conformality

Follow-Up (mo)

ClinicalPresentation

ClinicalOutcome TumorResponse

61/F D1 Transmastoid 3.2 15 Gy/30 Gy 4.0/6.8 1.7 11 Dead earOtalgia

Hearingunchanged

Otalgiaimproved

No change

73/F C3 Transcanal 2.3 15 Gy/30 Gy 2.7/5.6 2.1 40 Pulsatiletinnitus

Tinnitusimproved

No change

73/F D1 Transmastoid 3.6 15 Gy/30 Gy 6.8/10.3 1.5 36 Pulsatiletinnitus

Otalgia

Tinnitusresolved

Otalgiaresolved

Reduced size

78/F D1 Transmastoid 4.9 15 Gy/30 Gy 2.5/4.0 1.6 21 Pulsatiletinnitus

OtalgiaGrade IV

facial palsy

Tinnitusimproved

Otalgiaresolved

Grade IIfacial palsy

Reduced size

63/F C3 Transmastoid 4.0 15 Gy/30 Gy 4.7/7.6 1.6 35 Pulsatiletinnitus

OtalgiaCranial

nerveXI palsy

Tinnitusimproved

Otalgiaimproved

No changein CN XI

Reduced size

Abbreviation: GKRS, Gamma knife radiosurgery.

Mille

ret

al

700

Fig. 5. Audiogram results for four patients who had hearing at presentation. Each patientwas noted to have improvement in hearing threshold (pure-tone average).


SUMMARY

Glomus jugulare tumors represent a treatment challenge because of their locationnear many vital structures. Traditional surgical resection can be performed but issometimes associated with significant morbidity, and radiotherapy is fraught withcomplications. Radiosurgery represents a promising alternative to surgery or radio-therapy, because it allows for control of tumor growth without significant side effects.It can also be combined with a limited surgical resection to produce immediate allevi-ation of symptoms without the risk for an extensive gross total resection. Althoughlong-term data are not yet available, results have so far been promising, and radiosur-gical techniques will likely become even more important in the treatment of glomusjugulare tumors in the future.

REFERENCES

1. Glenner G, Grimley P. Tumors of the extra-adrenal paraganglion system(including chemoreceptors). Atlas of tumor pathology. 2nd series, fascicle 9.Washington, DC: Armed Forces Institute of Pathology; 1974.

2. Valentin G. Ueber eine gangliose Anschwellung in der Jacobsonchen Anasto-mose des Menschen. Arch Anat Physiol Wissensch Medicin 1840;89:287–90.

3. Guild SR. The glomus jugulare, a nonchromaffin paraganglion, in man. Ann OtolRhinol Laryngol 1953;62(4):1045–71.

4. Rosenwasser H. Carotid body tumor of the middle ear and mastoid. Arch Otolar-yngol 1945;41:64–7.

Miller et al702

5. Lack EE, Cubilla AL, Woodruff JM, et al. Paragangliomas of the head and neckregion: a clinical study of 69 patients. Cancer 1977;39(2):397–409.

6. Alford BR, Guilford FR. A comprehensive study of tumors of the glomus jugulare.Laryngoscope 1962;72:765–805.

7. Zacks SI. Chemodectomas occurring concurrently in the neck (carotid body),temporal bone (glomus jugulare) and retroperitoneum; report of a case with histo-chemical observations. Am J Pathol 1958;34(2):293–309.

8. Spector GJ, Ciralsky R, Maisel RH, et al. Multiple glomus tumors in the head andneck. Laryngoscope 1975;85(6):1066–75.

9. Mariman EC, van Beersum SE, Cremers CW, et al. Fine mapping of a puta-tively imprinted gene for familial non-chromaffin paragangliomas to chromo-some 11q13.1: evidence for genetic heterogeneity. Hum Genet 1995;95(1):56–62.

10. Heutink P, van der Mey AG, Sandkuijl LA, et al. A gene subject to genomicimprinting and responsible for hereditary paragangliomas maps to chromosome11q23-qter. Hum Mol Genet 1992;1(1):7–10.

11. Milunsky J, DeStefano AL, Huang XL, et al. Familial paragangliomas: linkage tochromosome 11q23 and clinical implications. Am J Med Genet 1997;72(1):66–70.

12. Milunsky JM, Maher TA, Michels VV, et al. Novel mutations and the emergence ofa common mutation in the SDHD gene causing familial paraganglioma. AmJ Med Genet 2001;100(4):311–4.

13. Astuti D, Hart-Holden N, Latif F, et al. Genetic analysis of mitochondrial complex IIsubunits SDHD, SDHB and SDHC in paraganglioma and phaeochromocytomasusceptibility. Clin Endocrinol (Oxf) 2003;59(6):728–33.

14. Mhatre AN, Li Y, Feng L, et al. SDHB, SDHC, and SDHD mutation screen insporadic and familial head and neck paragangliomas. Clin Genet 2004;66(5):461–6.

15. Bikhazi PH, Messina L, Mhatre AN, et al. Molecular pathogenesis in sporadichead and neck paraganglioma. Laryngoscope 2000;110(8):1346–8.

16. Spector GJ, Sobol S, Thawley SE, et al. Panel discussion: glomus jugulare tumorsof the temporal bone. Patterns of invasion in the temporal bone. Laryngoscope1979;89(10 Pt 1):1628–39.

17. Manolidis S, Shohet JA, Jackson CG, et al. Malignant glomus tumors. Laryngo-scope 1999;109(1):30–4.

18. Spector GJ, Ciralsky RH, Ogura JH. Glomus tumors in the head and neck: III.Analysis of clinical manifestations. Ann Otol Rhinol Laryngol 1975;84(1 Pt 1):73–9.

19. Kinney SE. Glomus jugulare tumor surgery with intracranial extension. OtolaryngolHead Neck Surg 1980;88(5):531–5.

20. Gulya AJ. The glomus tumor and its biology. Laryngoscope 1993;103(11 Pt 2Suppl 60):7–15.

21. Myssiorek D. Head and neck paragangliomas: an overview. Otolaryngol ClinNorth Am 2001;34(5):829–36.

22. McCaffrey TV, Myssiorek D, Marrinan M. Familial paragangliomas of the head andneck. Arch Otolaryngol Head Neck Surg 1994;120(11):1211–6.

23. Green JD Jr, Brackmann DE, Nguyen CD, et al. Surgical management of previ-ously untreated glomus jugulare tumors. Laryngoscope 1994;104(8 Pt 1):917–21.

24. Pensak ML, Jackler RK. Removal of jugular foramen tumors: the fallopian bridgetechnique. Otolaryngol Head Neck Surg 1997;117(6):586–91.

25. Spector GJ, Gado M, Ciralsky R, et al. Neurologic implications of glomus tumorsin the head and neck. Laryngoscope 1975;85(8):1387–95.


26. Brown LA. Glomus jugulare tumor of the middle ear; clinical aspects. Laryngo-scope 1953;63(4):281–92.

27. Jackson CG, Harris PF, Glasscock ME III, et al. Diagnosis and management ofparagangliomas of the skull base. Am J Surg 1990;159(4):389–93.

28. Schwaber MK, Glasscock ME, Nissen AJ, et al. Diagnosis and management ofcatecholamine secreting glomus tumors. Laryngoscope 1984;94(8):1008–15.

29. Lo WW, Solti-Bohman LG. High-resolution CTof the jugular foramen: anatomy andvascular variants and anomalies. Radiology 1984;150(3):743–7.

30. Lo WW, Solti-Bohman LG, Lambert PR. High-resolution CT in the evaluation ofglomus tumors of the temporal bone. Radiology 1984;150(3):737–42.

31. Olsen WL, Dillon WP, Kelly WM, et al. MR imaging of paragangliomas. AJR AmJ Roentgenol 1987;148(1):201–4.

32. Weber AL, McKenna MJ. Radiologic evaluation of the jugular foramen. Anatomy,vascular variants, anomalies, and tumors. Neuroimaging Clin N Am 1994;4(3):579–98.

33. Rodgers GK, Applegate L, De la Cruz A, et al. Magnetic resonance angiography:analysis of vascular lesions of the temporal bone and skull base. Am J Otol 1993;14(1):56–62.

34. Murphy TP, Brackmann DE. Effects of preoperative embolization on glomus jug-ulare tumors. Laryngoscope 1989;99(12):1244–7.

35. Jenkins HA, Fische U. Glomus tumors of the temporal region. Technique ofsurgical resection. Arch Otolaryngol 1982;107(4):209–14.

36. Jackson CG, Glasscock ME III, Nissen AJ, et al. Glomus tumor surgery: theapproach, results, and problems. Otolaryngol Clin North Am 1982;15(4):897–917.

37. Lustig LR, Jackler RK. The variable relationship between the lower cranial nervesand jugular foramen tumors: implications for neural preservation. Am J Otol 1996;17(4):658–68.

38. Fisch U. Infratemporal fossa approach for extensive tumors of the temporal boneand base of the skull. In: Silverstein H, Norell H, editors. Neurological surgery ofthe ear. Birmingham (AL): Aesculapius; 1977. p. 34–53.

39. Sanna M, Jain Y, De Donato G, et al. Management of jugular paragangliomas: theGruppo Otologico experience. Otol Neurotol 2004;25(5):797–804.

40. Blevins NH, Jackler RK, Kaplan MJ, et al. Combined transpetrosal-subtemporalcraniotomy for clival tumors with extension into the posterior fossa. Laryngoscope1995;105(9 Pt 1):975–82.

41. Daspit CP, Spetzler RF, Pappas CT. Combined approach for lesions involving thecerebellopontine angle and skull base: experience with 20 cases—preliminaryreport. Otolaryngol Head Neck Surg 1991;105(6):788–96.

42. Pemberton LS, Swindell R, Sykes AJ. Radical radiotherapy alone for glomus jug-ulare and tympanicum tumours. Oncol Rep 2005;14(6):1631–3.

43. Krych AJ, Foote RL, Brown PD, et al. Long-term results of irradiation for paragan-glioma. Int J Radiat Oncol Biol Phys 2006;65(4):1063–6.

44. Spector GJ, Compagno J, Perez CA, et al. Glomus jugulare tumors: effects ofradiotherapy. Cancer 1975;35(5):1316–21.

45. Hawthorne MR, Makek MS, Harris JP, et al. The histopathological and clinicalfeatures of irradiated and nonirradiated temporal paragangliomas. Laryngoscope1988;98(3):325–31.

46. Schwaber MK, Gussack GS, Kirkpatrick W. The role of radiation therapy in themanagement of catecholamine-secreting glomus tumors. Otolaryngol HeadNeck Surg 1988;98(2):150–4.

Miller et al704

47. Brackmann DE, House WF, Terry R, et al. Glomus jugulare tumors: effect of irra-diation. Trans Am Acad Ophthalmol Otolaryngol 1972;76(6):1423–31.

48. Spector GJ, Maisel RH, Ogura JH. Glomus tumors in the middle ear. I. An anal-ysis of 46 patients. Laryngoscope 1973;83(10):1652–72.

49. Spector GJ, Maisel RH, Ogura JH. Glomus jugulare tumors. II. A clinicopatho-logic analysis of the effects of radiotherapy. Ann Otol Rhinol Laryngol 1974;83(1):26–32.

50. Spector GJ, Fierstein J, Ogura JH. A comparison of therapeutic modalities ofglomus tumors in the temporal bone. Laryngoscope 1976;86(5):690–6.

51. Gibbin KP, Henk JM. Glomus jugulare tumours in South Wales—a twenty-yearreview. Clin Radiol 1978;29(6):607–9.

52. Kim JA, Elkon D, Lim ML, et al. Optimum dose of radiotherapy for chemodecto-mas of the middle ear. Int J Radiat Oncol Biol Phys 1980;6(7):815–9.

53. Simko TG, Griffin TW, Gerdes AJ, et al. The role of radiation therapy in the treat-ment of glomus jugulare tumors. Cancer 1978;42(1):104–6.

54. Sharma PD, Johnson AP, Whitton AC. Radiotherapy for jugulo-tympanic paragan-gliomas (glomus jugulare tumours). J Laryngol Otol 1984;98(6):621–9.

55. Powell S, Peters N, Harmer C. Chemodectoma of the head and neck: results oftreatment in 84 patients. Int J Radiat Oncol Biol Phys 1992;22(5):919–24.

56. Mukherji SK, Kasper ME, Tart RP, et al. Irradiated paragangliomas of the headand neck: CT and MR appearance. AJNR Am J Neuroradiol 1994;15(2):357–63.

57. Larner JM, Hahn SS, Spaulding CA, et al. Glomus jugulare tumors. Long-termcontrol by radiation therapy. Cancer 1992;69(7):1813–7.

58. Boyle JO, Shimm DS, Coulthard SW. Radiation therapy for paragangliomas of thetemporal bone. Laryngoscope 1990;100(8):896–901.

59. Hinerman RW, Mendenhall WM, Amdur RJ, et al. Definitive radiotherapy in themanagement of chemodectomas arising in the temporal bone, carotid body,and glomus vagale. Head Neck 2001;23(5):363–71.

60. Elshaikh MA, Mahmoud-Ahmed AS, Kinney SE, et al. Recurrent head-and-neckchemodectomas: a comparison of surgical and radiotherapeutic results. Int JRadiat Oncol Biol Phys 2002;52(4):953–6.

61. Hu K, Persky MS. The multidisciplinary management of paragangliomas of thehead and neck, Part 2. Oncology (Williston Park) 2003;17(8):1143–53.

62. Wang ML, Hussey DH, Doornbos JF, et al. Chemodectoma of the temporal bone:a comparison of surgical and radiotherapeutic results. Int J Radiat Oncol BiolPhys 1988;14(4):643–8.

63. Springate SC, Haraf D, Weichselbaum RR. Temporal bone chemodectomas—comparing surgery and radiation therapy. Oncology (Williston Park) 1991;5(4):131–7.

64. Carrasco V, Rosenman J. Radiation therapy of glomus jugulare tumors. Laryngo-scope 1993;103(Suppl 60):23–7.

65. Wang CC, Doppke K. Osteoradionecrosis of the temporal bone—consideration ofnominal standard dose. Int J Radiat Oncol Biol Phys 1976;1(9–10):881–3.

66. Cole DJ. Glomus jugulare tumours seen in Oxford 1960–1984. Clin Radiol 1988;39(1):83–6.

67. Springate SC, Weichselbaum RR. Radiation or surgery for chemodectoma of thetemporal bone: a review of local control and complications. Head Neck 1990;12:303–7.

68. Lalwani AK, Jackler RK, Gutin PH. Lethal fibrosarcoma complicating radiationtherapy for benign glomus jugulare tumor. Am J Otol 1993;14(4):398–402.


69. Pluta RM, Ram Z, Patronas NJ, et al. Long-term effects of radiation therapy fora catecholamine-producing glomus jugulare tumor. Case report. J Neurosurg1994;80:1091–4.

70. Gabriel EM, Sampson JH, Dodd LG, et al. Glomus jugulare tumor metastatic tothe sacrum after high-dose radiation therapy: case report. Neurosurgery 1995;37:1001–5.

71. Mumber MP, Greven KM. Control of advanced chemodectomas of the head andneck with irradiation. Am J Clin Oncol 1995;18(5):389–91.

72. Maniglia AJ, Sprecher RC, Megerian CA, et al. Inferior mastoidectomy-hypotym-panic approach for surgical removal of glomus jugulare tumors: an anatomicaland radiologic study emphasizing distances between critical structures. Laryn-goscope 1992;102(4):407–14.

73. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand1951;102(4):316–9.

74. Lunsford LD, Flickinger J, Lindner G, et al. Stereotactic radiosurgery of the brainusing the first United States 201 cobalt-60 source gamma knife. Neurosurgery1989;24(2):151–9.

75. Saringer W, Khayal H, Ertl A, et al. Efficiency of gamma knife radiosurgery in thetreatment of glomus jugulare tumors. Minim Invasive Neurosurg 2001;44(3):141–6.

76. Foote RL, Coffey RJ, Gorman DA, et al. Stereotactic radiosurgery for glomus jug-ulare tumors: a preliminary report. Int J Radiat Oncol Biol Phys 1997;38(3):491–5.

77. Kida Y, Kobayashi T, Tanaka T, et al. [A new strategy for the treatment of jugularforamen tumors using radiosurgery] [article in Japanese]. No Shinkei Geka 1995;23(8):671–5.

78. Foote RL, Pollock BE, Gorman DA, et al. Glomus jugulare tumor: tumor controland complications after stereotactic radiosurgery. Head Neck 2002;24(4):332–8.

79. Eustacchio S, Leber K, Trummer M, et al. Gamma knife radiosurgery for glomusjugulare tumours. Acta Neurochir (Wien) 1999;141(8):811–8.

80. Eustacchio S, Trummer M, Unger F, et al. The role of gamma knife radiosurgery inthe management of glomus jugular tumours. Acta Neurochir Suppl 2002;84:91–7.

81. Liscak R, Vladyka V, Simonova G, et al. Leksell gamma knife radiosurgery of thetumor glomus jugulare and tympanicum. Stereotact Funct Neurosurg 1998;70((Suppl 1):152–60.

82. Liscak R, Vladyka V, Wowra B, et al. Gamma knife radiosurgery of the glomus jug-ulare tumour - early multicentre experience. Acta Neurochir (Wien) 1999;141(11):1141–6.

83. Jordan JA, Roland PS, McManus C, et al. Stereotactic radiosurgery for glomusjugulare tumors. Laryngoscope 2000;110(1):35–8.

84. Sheehan J, Kondziolka D, Flickinger J, et al. Gamma knife surgery for glomus jug-ulare tumors: an intermediate report on efficacy and safety. J Neurosurg 2005;102(Suppl):241–6.

85. Gerosa M, Visca A, Rizzo P, et al. Glomus jugulare tumors: the option of gammaknife radiosurgery. Neurosurgery 2006;59(3):561–9.

86. Varma A, Nathoo N, Neyman G, et al. Gamma knife radiosurgery for glomus jug-ulare tumors: volumetric analysis in 17 patients. Neurosurgery 2006;59(5):1030–6.

87. Maarouf M, Voges J, Landwehr P, et al. Stereotactic linear accelerater-basedradiosurgery for the treatment of patients with glomus jugulare tumors. Cancer2003;97(4):1093–8.

Miller et al706

88. Lim M, Gibbs IC, Adler JR Jr, et al. Efficacy and safety of stereotactic radiosur-gery for glomus jugulare tumors. Neurosurg Focus 2004;17(2):68–72.

89. Lim M, Bower R, Nangiana JS, et al. Radiosurgery for glomus jugulare tumors.Technol Cancer Res Treat 2007;6(5):419–24.

90. Megerian CA, McKenna MJ, Nadol JB Jr. Non-paraganglioma jugular foramenlesions masquerading as glomus jugulare tumors. Am J Otol 1995;16(1):94–8.

91. Willen SN, Einstein DB, Maciunas RJ, et al. Treatment of glomus jugulare tumorsin patients with advanced age: planned limited surgical resection followed bystaged gamma knife radiosurgery: a preliminary report. Otol Neurotol 2005;26(6):1229–34.

MicrosurgeryAfter Radiosurgeryor Radiotherapyfor VestibularSchwannomas

William H. Slattery III, MD

KEYWORDS

� Radiation therapy � Tumor changes � Surgical excision cases� House clinic surgical excision data � Tumor pathology� Malignant transformation surgery

Vestibular schwannomas (VS) are benign neoplasms of the eighth cranial nerves.Vestibular schwannomas usually present with symptoms associated with the eighthcranial nerve including hearing loss, tinnitus, or dizziness. Historically, microsurgeryhas been the mainstay of therapy. Complete tumor removal has been the goal of treat-ment. Since the early 1990s, radiosurgery and radiotherapy have developed as alter-natives to microsurgical excision of vestibular schwannomas in the United States. Thegoals of microsurgery and radiation therapy are very different. The goal of microsur-gical excision is complete tumor removal with preservation of the cranial nerves.The goal of radiosurgery or radiotherapy is tumor control or prevention of tumorgrowth, with preservation of cranial nerve function.

There are several modalities by which radiation may be used to treat vestibularschwannomas. Treatment types may include a single-session treatment, used withgamma knife radiosurgery, or fractionated therapy that is performed over severalsessions. Fractionated radiotherapy is more common with the linear particle acceler-ator (LINAC) system such as Cyberknife (Accuray, Sunnyvale, California). The use ofa head frame or facial mask may be used for stereotactic computer localization ofthe tumor to guide radiation localization to the tumor. All of these therapies havea significant tumor control rate as defined by lack of tumor growth. Approximately2% to 10% of tumors continue to grow, however, and cause symptoms despite treat-ment with radiosurgery or radiotherapy.

Clinical Studies, House Ear Institute, House Ear Clinic, University of Southern California, 2100West Third Street, Los Angeles, CA 90057, USAE-mail address: [email protected]




Slattery708

ADVANTAGES OF RADIATION THERAPY

The indications for radiation therapy for vestibular schwannomas have evolved overthe years. The initial indications for radiation therapy for vestibular schwannomasincluded elderly patients, patients who had significant medical conditions precludingmicrosurgery, or patients who refused microsurgical removal. Recently, the indica-tions for radiation therapy have expanded to include all patients who wish to have ra-diosurgery or radiotherapy for their vestibular schwannoma. Although somephysicians will treat all vestibular schwannomas with radiation therapy, others recom-mend that there should be documented tumor growth before radiosurgery or radio-therapy. Because the goal of radiation therapy is to stop tumor growth, thesephysicians believe that the tumor should demonstrate growth before having the riskassociated with radiation therapy. Many vestibular schwannomas do not grow and,therefore, a watch-and-wait policy can be recommended before undergoing radiationtherapy.

The main advantage of radiation therapy is that it is less invasive compared withmicrosurgery. Radiosurgery or radiotherapy is considered less invasive, as the treat-ment usually is performed as an outpatient procedure, or requires only a one-nighthospital stay. The patient may return back to work much faster following this type oftreatment when compared with microsurgery. Radiosurgery or radiotherapy oftenresults in preservation of facial nerve function, although facial nerve paralysis canoccur. The risks of significant immediate adverse effects are low. Radiation therapyoffers the ability to preserve hearing in tumors whose size normally would be consid-ered a contraindication for a hearing preservation type of microsurgery. As anexample, tumors larger than 2 cm in size have a poor prognosis for hearing preserva-tion with microsurgery. The possibility of preserving hearing can be considered withradiation for tumors larger than 2 cm. Hearing preservation after microsurgery andradiation therapy will be discussed later in this article.

DISADVANTAGES OF RADIATION THERAPY

The main disadvantage of radiation therapy for vestibular schwannoma is that thetumor is still present after treatment and, therefore, the long-term control rate isunknown. As long as the tumor is still present there is always the possibility for tumorgrowth. Patients treated with radiation therapy must undergo routine MRI scans for theremainder of their lives. It is known that tumors swell initially after radiation therapyand, therefore, a size limit of 2.5 to 3.0 cm is recommended as the maximum treatmentsize. Tumors larger than 2.5 to 3.0 cm can swell, causing the patient to have significantintracranial complications such as brainstem compression requiring medical or micro-surgical intervention. Therefore, most centers will not treat patients who have tumorslarger than 2.5 to 3.0 cm in size. Radiation treatment also includes a risk to the cranialnerve’s seventh and eighth complexes. This risk includes partial or complete loss ofhearing. Most studies have demonstrated that hearing may be preserved in the imme-diate months following radiation treatment; however, there is a definite decrease in therate of hearing preservation the longer the patient is followed. Facial nerve complica-tions can include temporary or permanent paralysis.

Facial nerve spasm is also a symptom that is associated with postradiation treat-ment. Patients can have significant hemifacial spasm that can be quite disabling. Insome cases, the hemifacial spasm is severe enough to warrant microsurgical excisionof the tumor. Botox therapy also may be used to treat the hemifacial spasm. New-onset or worsening preexisting dizziness or ataxia may develop following radiationtreatment. Vestibular rehabilitation with physical therapy may help in some cases,

Microsurgery After Radiosurgery 709

but in other cases, the dizziness or ataxia can be so severe that the patient is at signif-icant risk of falls. Two percent to 10% of patients treated with radiation therapy havetumors that continue to grow and that may require microsurgical treatment. Microsur-gical treatment may consist of insertion of a ventricular peritoneal shunt for hydro-cephalus or microsurgical excision of the tumor.

The long-term effects of radiation therapy for vestibular schwannomas are notknown. The earliest reported series from within the United States used a higherdose of radiation to treat vestibular schwannomas than is used currently. This higherrate of radiation was associated with a higher rate of radiation-induced complications;therefore the total dose to the tumor has been reduced over the past 10 years. Thelong-term effect of lowering the treatment dose with respect to tumor control will beassessed only with time. There are very few reports on the long-term effects after15 years from stereotactic radiation treatment. Radiation poses the risk of possiblemalignant transformation of the tumor or the risk of inducing additional tumor forma-tion within the radiation field. This risk of malignancy associated with stereotactic radi-ation has been estimated to be approximately 1%. More information regardingmalignancy risk is available elsewhere in this issue. Other radiation treatment modal-ities have been associated with an increase incidence of tumor formation 15 to 30years after treatment. The risk of subsequent tumor formation is thought to be higherfor younger patients, as they have a longer life expectancy following radiation treat-ment and therefore have more time to develop additional tumors. Meningioma forma-tion and vestibular schwannoma have been associated in patients who haveexperienced external beam radiation therapy 15 to 30 years before the presentationand diagnosis of these tumors.

TUMOR CHANGES AFTER RADIATION

There are consistent changes that occur following radiation therapy for vestibularschwannomas. These changes include an increase in size of the tumor 6 to 9 monthsafter treatment and loss of central MRI enhancement. Pollock and colleagues1 foundin 208 patients that the median time for tumor enlargement was 9 months. The medianvolume was increased by 75%. A loss of MRI central enhancement was noted in 93%of tumors at the time of tumor expansion. The increase in size of the tumor ispresumed to be caused by swelling of tumor tissue. The tumors typically increase insize 6 to 12 months following treatment and then decrease in size over the 12 to 24months following treatment. Reduction of MRI signal intensity and the time it takesfor the tumor to reach maximum swelling following treatment have been associatedwith the total tumor dose. The higher tumor dose is associated with a more rapidswelling and more central MRI changes.

Pollock1 describes three types of growth patterns that may occur following stereo-tactic radiosurgery for vestibular schwannomas. Type 1 is seen most commonly. Thevestibular schwannoma usually will enlarge by several millimeters within 9 to 12months after stereotactic radiotherapy. This enlargement is followed by a volumereduction that reverts back to the initial tumor size. The type 2 pattern includes tumorsthat enlarge and remain larger than the size before radiosurgery. This occurs inapproximately 30% of cases. Although these tumors may be larger, they do not causeany additional symptoms. The type 3 pattern consists of vestibular schwannomas thatprogressively grow in serial imaging. These are usually the tumors that require treat-ment when symptoms change or develop.

Long-term follow-up of type 1 tumors reveals that they are very stable. The type 2tumors require long-term follow-up, because these tumors may continue to grow

Slattery710

and fall into the type 3 category. It is possible that the increase in size and stabilizationof type 2 tumors may represent the natural history of the benign tumor that has spon-taneous periods of growth arrest.

Tumor size may increase after radiation therapy because of three factors: (1) solidexpansion of the tumor, (2) tumor necrosis, or (3) tumor cyst formation.

Tumor cyst formation may result from increased extracellular proteins that leak fromthe blood vessel walls in the tumor. This can result in a multiseptated cyst associatedwith the tumor. Cyst formation following radiation therapy is not uncommon. Thistheory of cyst formation is supported by the fact that cerebrospinal fluid (CSF) proteinlevels are higher in patients who have tumor enlargement requiring ventriculo–perito-neal (VP) shunts after radiation therapy of their vestibular schwannomas.

In the early experience of stereotactic radiation therapy for vestibular schwanno-mas, the swelling found after treatment was thought to represent tumor growth, indi-cating the radiation did not work; therefore microsurgery was undertaken for thegrowth. This resulted in some patients undergoing microsurgery, which was probablyunnecessary. In addition, in the early use of stereotactic radiation therapy for vestib-ular schwannomas, larger tumors were treated. Unfortunately, many of the patientswith these large tumors had normal swelling that occurred following radiation therapy,and they developed hydrocephalus or brainstem compression. This has led to sizeconstraints for patients who are candidates for radiation therapy.

SURGICAL INDICATIONS FOR VESTIBULAR SCHWANNOMA REMOVALAFTER RADIATION THERAPY

Microsurgery following radiation therapy is relatively uncommon, as most cases ofstereotactic radiation therapy for vestibular schwannomas have good short-termtumor control rates. As a result of the good short-term tumor control rate, there arefew large series of microsurgery following radiation cases. The degree of increasedmicrosurgical technical difficulties between nonradiated and radiated tumors hasbeen discussed, although conclusions regarding the relative difficulty of tumor dissec-tion following radiation therapy have not been settled. Surgical treatment after radia-tion therapy may include insertion of VP shunt, excision of tumor-associated cyst,subtotal tumor resection, or total tumor resection.

It is advisable to avoid surgery during the first 2 years after radiation therapyunless there are significant complications associated with tumor growth or radia-tion. The rationale for avoiding surgery during this time is twofold. First, the cranialnerves are more susceptible to damage from surgical manipulation during the firstyear. Second, differentiation of long-term tumor growth from expected tumorswelling during this first 2 years is difficult. Radiation-induced cranial neuropathyis known to occur during the first year after routine radiation therapy. Cranialneuropathy associated paralysis is unusual after the first year. Tumor swellingusually occurs during the first year after radiation therapy, and this effect mayextend into the second year. Therefore, it is recommended to observe the tumorfor 2 years after radiation therapy before considering a surgical intervention unlessserve tumor-related symptoms develop. These severe tumor-related symptoms areusually brainstem compression or hydrocephalus.

REVIEWOF PUBLISHED CASES OF SURGICAL EXCISION

Most centers have only one or two cases of microsurgery following stereotactic radi-ation therapy for vestibular schwannomas. Many of these cases have not been


reported in the literature. There are several small case series that report on the micro-surgical results following stereotactic radiation therapy for vestibular schwannomas,and these are reviewed here.

In 1998, Pollock and colleagues2 described 13 patients who underwent microsur-gery at a median time of 27 months after they had undergone radiosurgery. Thisrepresents a 2.9% failure rate for their series of gamma knife radiosurgery proce-dures. Tumor enlargement was the indication in five patients who had stable symp-toms; tumor enlargement with new or increased symptoms occurred in five patients,and increased symptoms without evidence of tumor growth occurred in three. Grosstotal resection was achieved in seven patients, and a near gross total resection wasachieved in four patients. The authors found the microsurgery was more difficult in 8of 13 patients, easier in 1, and no different in 4, relative to microsurgery without priorradiation treatment. All patients had lost hearing as a result of their tumors. Threepatients required VP shunts for progressive symptoms of hydrocephalus despitethe tumor removal. Three patients’ postoperative course required persistent care-givers. One patient in the series died of metastatic tumor progression followingradiosurgery.

Kwon and colleagues3 reported the results of Korea’s gamma knife unit in 1999.They reported a 91% control rate of their vestibular schwannomas followed over 6months with a mean of 55 months. They had seven patients whose tumor increasedsignificantly in size during the follow-up. Three of these were observed as the patientshad no symptoms. Four patients underwent microsurgical excision, as the tumor wassolid and increased in size. One patient developed brainstem compression, anddespite microsurgery was left in a persistent vegetative state. Three patients devel-oped multiseptated cysts that caused brainstem compression. Loss of central MRIenhancement was seen in all three of these tumors despite the cysts. The cystswere associated with tumors that were larger than 3.0 cm in size before radiation.The time for microsurgical intervention following radiation was 5, 10, 34, and 64months. The histopathology of the tumors was typical for routine vestibular schwan-nomas. The authors stated the facial nerve was preserved in all cases; however,they noted 50% of the patients had some degree of postoperative facial nerve paral-ysis. The authors stated, ‘‘all tumors were relatively easy to remove.’’ They furtherstated the tumor was avascular and that they had a shorter operating time removingthe previously radiated tumors. They additionally stated the facial nerve was not anydifferent to dissect than other nonradiated cases. The authors failed to explain whythe facial nerve outcome in these postradiated cases, if it was so easy to remove,was so significantly poor in 50% of the cases.

In 2000, Battista4 sent a survey to the American Neurotology Society requestingsurgeons experienced with treating vestibular schwannomas to describe theirexperience in treating patients microsurgically who previously had undergonestereotactic radiotherapy. The survey response was 36% and described 12patients who had tumor resection following stereotactic radiation therapy. Themean time to microsurgery was 35 months after radiotherapy (range 3 to 72months). Seven of the 12 tumors were larger than 3.0 cm at the time of microsur-gery, and four had had prior microsurgical resection before radiation therapy. Nineof 12 surgeons believed the microsurgery was more difficult than microsurgerywithout prior radiation therapy, as no plane was found between the tumor andthe facial nerve. This survey study concluded that microsurgery was more difficultfollowing radiation therapy.

In 2003, Lee5 reported on the Johns Hopkins (Baltimore, Maryland) experience offour patients who had undergone previous gamma knife or fractionated stereotactic

Slattery712

radiotherapy (FRT). Tumor resection was performed by means of the retrosigmoidapproach. Indications for treatment for the four cases included

Facial spasm without tumor growth at 12 months following radiotherapyTumor growth with left-sided tongue and oral numbness 2 years after FRTProgressive facial spasms 2 years after FRTFacial spasm with tumor growth

Tumor growth varied from 2 to 5 mm from the time of radiation therapy until micro-surgical resection. The time between radiation and microsurgery was an average of1.6 years, (range, 1 to 2 years). Intraoperatively, capsular fibrosis was seen with scar-ring of the tumor capsule in all cases. This made microsurgical dissection of the facialnerve and other cranial nerves more difficult. Anatomic facial nerve preservation wasachieved in all cases, but it is thought to be more difficult to dissect along the facialnerve compared with cases without prior radiation therapy. One patient developeda House-Brackmann (H-B) grade 6 facial paralysis, while the remaining three patientsvaried from H-B grade 1 to 2. After microsurgery, the facial spasm resolved in the firstcase, but the outcome of facial spasm was not reported in cases three and four. Histo-logically, elongated bipolar cells were seen in all cases. The tumors were moderatelycellular, but no real findings of radiation were noted.

Iwai and colleagues6 published their results from Japan in 2007 of 265 patientstreated over a 10-year period (1994 to 2004) with gamma knife therapy. They hadseven patients who failed to respond to radiation therapy and who required microsur-gical removal. Five were performed at Osaka City Hospital in Osaka, Japan, and twowere reported from other institutions. There were five patients who had unilateraltumors and one patient who had NF2. The size of the tumors at radiation therapywas an average diameter of 27.5 mm (19.4 to 45.9 mm). The average radiation dosewas 11 Gy (10 to 12.5 Gy). The indications for microsurgery were cerebellar ataxiawith increased intracranial pressure and worsening neurologic status. The meaninterval between microsurgery and radiation therapy was 28 months (4 to 74 months).The size of the tumor at microsurgery averaged 30.6 mm, with a range of 28.5 to 58.5mm. All microsurgical procedures were done by means of the retrosigmoid approach.The authors did not open the internal auditory canal and left tumor on the brainstem;thus, a subtotal or partial resection was performed in all cases. Findings at the time ofmicrosurgery included thickening of the arachnoid with increased bleeding relative tomicrosurgery without prior radiation therapy. Histopathology findings included normaltumor cells with foamy macrophages.

In 2008, Shuto and colleagues7 reported their postradiation microsurgical experi-ence in 12 patients from a total of 559 patients who underwent gamma knife therapy.These 12 patients had microsurgical removal of the tumor. The time between gammaknife and microsurgery was 29 months (range, 6.6 to 120 months). Four patients hadundergone previous microsurgery. The median gamma knife dose was 12.3 Gy. Theindications for microsurgical excision included tumor enlargement without deteriora-tion of symptoms in three patients. Tumor enlargement with new symptoms wasseen in nine patients. All patients had lost hearing by the time they underwent micro-surgical excision. The microsurgery was performed by means of the suboccipitalapproach, and tumor was removed subtotally except around the internal auditorycanal, which was left in place. Adhesion to the brainstem was very severe in sevenpatients, and identification of the facial nerve was easy in five patients and difficultin seven. The surgeons found dissection of the facial nerve difficult because of thesevere adhesions or lack of color change between the facial nerve and the tumor.Severe adhesions between the trigeminal nerve were observed in two patients. Facial


nerve function was unchanged in 7 of 12 patients, improved in 2, and deteriorated in 3.The brainstem compression symptoms were improved after tumor removal in allsymptomatic patients. It appeared from the cases presented that debulking the tumorwas the main objective in most of these cases.

The pathology demonstrated typical features of vestibular schwannomas such asbipolar cells and nuclear palisading. Hyalinization of the tumor vessels was found inmost cases. It was thought the hyalinization was the result of radiation. There wereno malignant changes found in any tumor specimens.

REVIEWOF HOUSE EAR CLINIC MICROSURGICAL EXCISION DATA

The initial House Ear Clinic experience with microsurgery following stereotactic radi-ation therapy for vestibular schwannomas was published in 1995.8 This report con-sisted of five cases that had undergone previous radiation therapy—three withgamma knife and two with proton beam. All five patients had lost all hearing preoper-atively and had significant facial weakness preoperatively. The symptoms initiatingmicrosurgery included:

Progressive facial weakness in one patient 12 years after gamma knife therapyBrainstem compression symptoms in two patients, 8 months and 2 years after

treatmentIncrease in tumor size in two patients, 2 and 2.5 years following treatment

All patients had an increase in tumor size during their time of follow-up after radia-tion before microsurgery. During microsurgery, the normal plane between the facialnerve and tumor was difficult to determine in three cases and completely lost in theremaining two. The histopathology examination demonstrated viable tumor cellsconsistent with schwannoma. The classic interweaving bundles of spindle-shapedcells with nuclei palisading was seen. Two tumors had thrombosed blood vessels inthe area of fibrotic necrosis. Viable tumor cells were seen in all five cases. As a resultof this initial study, the microsurgical approach for tumor removal at the House EarClinic has become much more conservative in regards to facial nerve dissection inpatients who previously have undergone radiation therapy.

The most recent published update of the House Ear Clinic experience with micro-surgical salvage after radiation therapy was published in 2005.9 This is the largest pub-lished series to date, with 44 patients having undergone microsurgery after previousirradiation. Six of the 44 patients had previous microsurgery and radiation therapybefore presentation to the House Ear Clinic and therefore were excluded from theanalysis. The remaining 38 patients treated with microsurgery following radiation treat-ment were compared with a matched random sample of 38 patients who underwentprimary microsurgery alone. The most common type of radiation therapy was gammaknife, which occurred in 22 patients. LINAC was the second most common type ofradiation, used in nine patients, and proton beam therapy was used in four patients.The mean time from irradiation to microsurgical treatment was 3.3 years (range 5.2months to 15.8 years). Tumor growth was the indication for microsurgery in mostpatients.

Patients who had undergone radiation therapy were more likely to have moderate-to-severe adhesions of the tumor to the facial nerve when compared with the nonirra-diated microsurgical control group. Brainstem adherence also was increasedsignificantly with the radiation group. Total tumor resection is the goal at the HouseEar Clinic for cases without prior irradiation. Over 97% of microsurgically treatedvestibular schwannoma patients achieved total tumor removal in those cases without

Slattery714

prior irradiation. In the group with previous irradiation, the number of patients with totaltumor removal has dropped to 79%. This decrease in total tumor removal in previouslyirradiated patients represents a change in treatment philosophy at the House EarClinic. This change in approach was because of the earlier microsurgical experiencewhere severe adhesions of the tumor to the facial nerve were found in the irradiatedpatients. The irradiated facial nerve did not recover from microsurgical trauma whencompared with the nonirradiated facial nerve. The facial nerve was lost in a greatergroup of previously irradiated microsurgical patients compared with the nonirradiatedgroup. There were no significant differences in operative complications other than theadherence to the facial nerve and facial nerve results. Histopathology in this largergroup of published patients again demonstrated signs of typical vestibularschwannomas.

Since the 2005 report, the author’s experience has increased to include 62 patientswith tumors undergoing microsurgery following radiotherapy. The incidence ofpatients who have growing tumors following irradiation is increasing. This probablyrepresents the larger number of patients who have opted for radiosurgery or radio-therapy as the initial treatment. The time from radiation to microsurgery averagesapproximately 3.1 years. The average size of the tumor is approximately 3.1 cm inlinear direction. Almost all procedures are performed through the translabyrinthineapproach, as the chance of preserving hearing in a patient who has undergone radi-ation therapy is very poor. In fact, hearing preservation from microsurgery in a previ-ously irradiated patient has not yet been reported. The author has become lessaggressive at obtaining complete tumor removal because of the significant adhesionsthat found between the facial nerve and the tumor. Approximately 87% of patients stillhave moderate-to-severe adhesion of the tumor to the facial nerve. This is similar tothe earlier House Ear Clinic study that demonstrated an increased risk with thisgroup.9 At present, there are not enough data to comment on the long-term outcomeof those patients who have undergone subtotal excision during revision microsurgeryafter irradiation.

PATHOLOGYOF TUMORS

The histopathology of tumors following radiation therapy has been very similar totumors that have not been radiated. Some authors who have investigated this havereported tumor fibrosis with loss of the peri-tumor arachnoid plane. There is thickeningof the arachnoid membrane with adjacent scarring. There are changes in the vascu-larity, with increased scarring of the anterior–inferior cerebellar artery to tumor.

MALIGNANT TRANSFORMATION SURGERY

Malignant vestibular schwannomas may occur sporadically or be induced by radia-tion.10–12 (More information is available elsewhere in this issue.) Malignant tumorsrequire more aggressive microsurgery than benign tumors. Usually the tumor planewith the facial nerve is lost, and sacrifice of the facial nerve is required for total tumorremoval. The planes between the other intracranial components can vary, as some arevery scarred, and others separate from the brainstem much more easily.

SUMMARY

The House Ear Clinic experience with microsurgery after irradiation has demonstratedthat the facial nerve is different once it has been radiated. An irradiated facial nerve’sregeneration potential is diminished, and the recovery from microsurgical trauma is


not as robust. It is recommended that patients who require microsurgical excisionfollowing radiosurgery or radiotherapy have a subtotal or near total resection per-formed as opposed to a gross total resection. A gross total resection may beperformed if the facial nerve planes are pristine. If the planes are not pristine, thennear-total resection should be performed with some tumor left on the facial nerve.The long-term results of these tumor remnants will require further follow-up.

REFERENCES

1. Pollock BE. Management of vestibular schwannomas that enlarge after stereo-tactic radiosurgery: treatment recommendations based on a 15-year experience.Neurosurgery 2006;58(2):241–8.

2. Pollock BE, Lunsford LD, Kondziolka D, et al. Vestibular schwannoma manage-ment: part II—failed radiosurgery and the role of delayed microsurgery. J Neuro-surg 1998;89:949–55.

3. Kwon Y, Khang SK, Kim CJ, et al. Radiologic and histopathologic changes aftergamma knife radiosurgery for acoustic schwannoma. Stereotact Funct Neurosurg1999;72(Suppl 1):2–10.

4. Battista RA, Wiet RJ. Stereotactic radiosurgery for acoustic neuromas: survey ofthe American Neurotology Society. Am J Otol 2000;21:371–81.

5. Lee DJ, Westra WH, Staecker H, et al. Clinical and histopathologic features ofrecurrent vestibular schwannomas (acoustic neuroma) after stereotactic radiosur-gery. Otol Neurotol 2003;24:650–60.

6. Iwai Y, Yamanaka K, Yamagata K, et al. Surgery after radiosurgery for acousticneuromas: Surgical strategy and histological findings. Neurosurgery 2007;60:75–82.

7. Shuto T, Inomori S, Matsunaga S, et al. Microsurgery for vestibular schwannomaafter gamma knife radiosurgery. Acta Neurochir (Wien) 2008;150(3):229–34.

8. Slattery WH III, Brackmann DE. Results of surgery following stereotactic irradia-tion for acoustic neuromas. Am J Otol 1995;16:315–21.

9. Friedman RA, Brackmann DE, Hitselberger WE, et al. Surgical salvage after failedirradiation for vestibular schwannoma. Laryngoscope 2005;115:1827–32.

10. Cahan WG, Woodard HQ, Higinbotham NL, et al. Sarcoma arising in irradiatedbone: report of eleven cases, 1948. Cancer 1998;82:8–34.

11. Hanabusa K, Morikawa A, Murata T, et al. Acoustic neuroma with malignant trans-formation: case report. J Neurosurg 2001;95:518–21.

12. Shin M, Ueki K, Kurita H, et al. Malignant transformation of a vestibular schwan-noma after gamma knife radiosurgery. Lancet 2002;360:309–10.

NeoplasticTransformationAfter Radiosurgeryor Radiotherapy :Risk and Realities

Ajay Niranjan, MBBS, MCha,c,*, Douglas Kondziolka, MDa,b,c,L. Dade Lunsford, MDa,b,c

KEYWORDS

� Malignant transformation � Radiation-induced tumor� Radiosurgery � Radiation therapy � Gamma Knife

The risk of a radiation-induced tumor is a controversial topic. Patients who undergoradiosurgery or fractionated radiation therapy may be at an increased risk for develop-ment of a second tumor. The role of concomitant environmental and genetic riskfactors is not known.1 Radiation-induced oncogenesis is a special a concern whenradiation is used to manage benign tumors where prolonged survival is expected.2

Typically, radiation-induced tumors occur several years after radiation exposure.The concept of radiation oncogenesis was established in 1948 by Cahan and

colleagues,3 who observed 11 patients who had sarcomas that developed after radio-therapy was administered for bone tumors and breast cancer. They outlined thefollowing criteria that must be met for a tumor to be designated as a radiation-inducedtumor:

A certain latency interval is required between delivery of the radiation and tumordevelopment.

The new tumor must arise in the irradiated region.The new tumor must be histologically distinct from the original irradiated tumor.There must be imaging evidence that the second tumor was not present at the time

of irradiation.The patient must not have a genetic predisposition for developing cancer.

a Department of Neurological Surgery, University of Pittsburgh, Pittsburgh, PA 15213, USAb Department of Radiation Oncology, University of Pittsburgh, Pittsburgh, PA 15213, USAc The Center for Image-Guided Neurosurgery, University of Pittsburgh Medical Center,Pittsburgh, PA 15213, USA* Corresponding author. Suite B-400, University of Pittsburgh Medical Center Presbyterian, 200Lothrop Street, Pittsburgh, PA 15213.E-mail address: [email protected] (A. Niranjan).




Niranjan et al718

SECOND TUMOR AFTER FRACTIONATED RADIATION THERAPY

Several large studies have documented an increased incidence of a second tumor inthe body after radiation therapy. Brenner and colleagues4 reported data from theNational Cancer Institute’s Surveillance Epidemiology and End Result (SEER)program, which contained information on 51,584 men who had prostate cancertreated by radiotherapy and 70,539 treated by surgery. The risk of a second tumorwas found to be higher by about 6% in the radiotherapy group. Boice and colleagues5

studied the risk of a second malignancy in various organs in 150,000 patients who hadcarcinoma of uterine cervix treated with either surgery or radiotherapy. The overallconclusion of this study was that an increased cancer risk was associated with radi-ation therapy compared with surgery. Bhatia and colleagues6 reported that 17 of 483women treated with radiation for Hodgkin’s disease later developed breast cancer.

There is evidence establishing radiation as a carcinogen in animal and humanstudies.3,7–13 Since Cahan’s publication, radiation-induced neoplasms affecting thecentral nervous system (CNS) after fractionated radiotherapy have been documented.The risk of developing a radiation-induced tumor after fractionated radiation therapy tothe CNS is estimated to be 1% to 3%.7,10,14 Most of these tumors are meningiomas,but gliomas, sarcomas, and schwannomas have been reported.7,8,10,12–16 The term‘‘radiation-induced,’’ however, seems inappropriate, because it implies that there isdefinitive evidence at a molecular level that radiation was the causative factor. Suchinformation has not been reported definitely.17 Perhaps ‘‘radiation-associated’’ maybe a better term.

Published data on the risks of oncogenesis after fractionated radiotherapy deliveredto the brain suggests that even very low doses (less than 2 Gy) have been associatedwith development of a second tumor. Between 1948 and 1960, 10,834 children inIsrael received scalp irradiation to induce alopecia as part of the management of tineacapitis.10 Mean doses to the brain in these children were estimated to be 1.5 Gy for theentire cohort. The relative risk of tumor formation at 30 years compared with thegeneral population was 18.8 for schwannomas, 9.5 for meningiomas, and 2.6 forgliomas. A clear-cut dose–response effect was observed, with the relative riskapproaching 20 after doses of approximately 2.5 Gy. The mean latency interval washistology-dependent. In this study, the mean latency interval was 21 years for menin-giomas, 15 years for schwannomas, and 14 years for gliomas.

In a study of radiation-associated second tumors, Sadetzki and colleagues18

described 253 patients who developed meningiomas after radiation for tinea capitis.The mean time from exposure to meningioma diagnosis was 36 years (range 12 to49 years). The authors found a higher incidence of multiple lesions, a younger ageat diagnosis, and a higher percentage of calvarial lesions in this group of patientscompared with those who developed meningiomas without previous exposure toionizing radiation. Dalton and colleagues19 reported on 1597 children who weretreated with or without 28 Gy of prophylactic whole-brain irradiation as part ofmanagement for acute lymphoblastic leukemia. Thirteen second tumors werereported in this group, with a median follow-up period of 7.6 years. Five secondtumors were in the CNS (four astrocytomas and one meningioma). The median latencyinterval for astrocytoma development was 9 years compared with 16.6 years formeningioma detection. Among the irradiated children, there was a 0.085 incidenceper 100 patient years of developing second tumors compared with 0 incidence forthose not receiving cranial irradiation.

Brada and colleagues20 reported the risk of second brain tumor formation in 334patients treated for pituitary tumors with surgery and fractionated small-field

Risk and Realities 719

irradiation therapy using 45 Gy. Five patients developed second tumors (two astrocy-tomas, two meningiomas, and one meningeal sarcoma) in long-term follow-up. Thelatency period for the second tumor was 6 to 21 years, with malignant tumors devel-oping earlier than benign tumors. The cumulative risk of developing second braintumors was 1.3% at 10 years and 1.9% at 20 years. The relative risk of a second braintumor was 9.38 compared with the normal population. In these patients, the irradiatedvolumes were typically small. On the basis of this experience, it seems that the relativerisk of second tumor formation is substantially less than that seen in patients treatedwith larger-volume radiation. Balasubramaniam and colleagues21 reported a radia-tion-associated glioblastoma multiforme (GBM) in the temporal lobe 5 years afterstereotactic fractionated radiation therapy (25 fractions) for a vestibular schwannoma.The area where the GBM developed had received between 1.45 to 6.94 Gy.

SECOND BRAIN TUMORS AFTER RADIOSURGERY

Physicians have speculated for years that second tumor formation would be a risk forpatients after radiosurgery. It also was thought, however, that the risk would be quitelow, for the following reasons:

The irradiated volume is very small compared with traditional radiotherapytechniques.

The high single doses of radiation given to the target volume during radiosurgerywould lead to cytotoxicity and not mutagenicity, which is required for tumorformation.

Volumes and doses of radiation along the entrance and exit pathways in radiosur-gery are so small that the likelihood of second tumors would be less.22

Nonetheless, there have been case reports of second tumor detection after radio-surgery. Currently, the number of patients worldwide who have received radiosurgeryis now likely in excess of 500,000. Reported cases of a second tumor after radiosur-gery can be grouped under the following three categories.

Radiosurgery-Associated Malignant Tumors

At present, five case reports (Table 1) of new malignant tumors after radiosurgerymeet Cahan’ criteria for radiation-induced neoplasm.23–27 Each of these cases issummarized. The new malignant tumors were discovered from 5.3 to 9 years afterradiosurgery. Each of the secondary tumors were high-grade gliomas.

Case report 1A 63-year-old woman underwent gamma knife radiosurgery for an imaging-definedmeningioma.27 She received 20 Gy to the tumor margin, with a maximal dose of 40Gy. Two years later, she underwent tumor resection because of persistent neurolog-ical symptoms and edema. The final pathological finding was a benign meningiomawith adjacent radiation necrosis. Seven years after radiosurgery and 5 years aftermicrosurgical resection, she was found to have an enhancing lesion in her left occipitallobe. Surgical resection revealed a GBM. It was estimated that the site at which theGBM originated had received 5 to 10 Gy at the time of radiosurgery 7 years earlier.

Case report 2A 57-year-old woman underwent gamma knife radiosurgery for a left-sided vestibularschwannoma.26 She was treated with 11 Gy to margin of an 8.6-cm3 tumor. Six-monthfollow-up MRI showed cystic degeneration of the lesion with mass effect on the brain-stem. The patient underwent microsurgical resection. Pathological study

Table1Radiosurgery-associated malignant tumors

Author,Year Patient Age/Sex Original Diagnosis Margin Dose Postradiosurgery Management Duration New TumorYu, 200027 63/F Meningioma 20 Gy Resection 2 years later 7 years GBM

Shamisa, 200126 57/F Vestibular Schwannoma 11 Gy Resection 6 months later 7.5 years GBM

Kaido, 200124 14/M Arteriovenous malformation 20 Gy None 6 years GBM

McIver, 200425 43.F Malignant Melanoma Metastases 15 Gy Whole-brain radiation therapy 5.25 years Anaplastic astrocytoma

Berman, 200723 34/F Arteriovenous malformation 15 Gy None 9 years GBM

Abbreviations: Duration, time between original treatment and diagnosis of malignant tumor; GBM, glioblastoma multiforme.

Nira

nja

net

al

720


demonstrated a benign schwannoma. Seven and one-half years later, she presentedwith progressive headaches, confusion, and left hemiparesis. She was found to havea large cystic, enhancing lesion in the left temporal lobe. Microsurgery was performed.Histology showed a GBM. The site at which the second tumor arose had received 4 Gy(14% of the maximum dose of 27.5 Gy) at the time of the radiosurgical treatment 7.5years earlier.

Case report 3A 14-year-old boy underwent gamma knife radiosurgery for a right parietal arteriove-nous malformation (AVM).24 He received 20 Gy to the tumor margin and 40 Gymaximum dose. Two years after radiosurgery, angiography confirmed that the AVMhad been obliterated. The patient developed headaches and vomiting 6 years after ra-diosurgery and was found to have a new lesion in the previous area of radiosurgery.Surgery was performed, and a GBM was discovered. The tumor seemed to arisewithin the full-dose region of his AVM radiosurgical treatment.

Case report 4A 43-year-old woman underwent radiosurgery for three brain metastases from malig-nant melanoma.25 Following radiosurgery, she underwent whole-brain radiotherapy(37.5 Gy). Twenty-two months later, a second radiosurgical procedure was performedfor a recurrent right temporal lobe metastasis. Five years and 4 months after initial ra-diosurgery, the patient was diagnosed with a cerebellar anaplastic astrocytoma. Thearea of cerebellum where the glioma developed had received a maximum dose of 7.7Gy and 1.5 Gy during the previous two radiosurgery procedures, respectively. Theadditional role of fractionated radiation therapy in the development of astrocytomais not clear.

Case report 5A 34-year-old woman underwent embolization followed by radiosurgery for a 4.5� 3.0cm � 3.0 cm AVM located in the pineal region.23 A margin dose of 15 Gy at 70%isodose line was delivered using linear accelerator linear accelerator (LINAC) radiosur-gery. The patient was lost to follow-up after treatment until she presented witha change in mental status, nausea, headaches, and a generalized seizure 9 years later.MRI demonstrated a 55 mm � 45 mm enhancing, heterogenous mass in the spleniumof the corpus callosum. Maximal debulking was performed. Pathological examinationdemonstrated an infiltrating glial neoplasm consistent with a GBM.

Radiosurgery-Associated Benign Tumors

Only four cases of new benign tumors after radiosurgery are reported (Table 2) in theliterature that meet Cahan’s criteria for radiation-induced neoplasms.17,28

Case report 1A 41-year-old man was diagnosed with acromegaly associated with a pituitary macro-adenoma.17 After trans-sphenoidal surgery, he underwent non-Bragg peak proton ra-diosurgery in Moscow, receiving 87 Gy by means of 25 separate 10 mm beams. Thebeams were centered on the sella, where the prescription dose was delivered. Sixteenyears after radiosurgery, he was evaluated for decreased vision, and repeat MRIconfirmed a Tuberculum sellae meningioma. Partial resection of a benign meningiomawas performed. The lesion was at the immediate periphery of the previous full-doseirradiated volume. The site at which the tumor arose was estimated to have received30% to 50% of the prescribed radiation dose.

Table 2Radiosurgery-associated benign tumors

Author,Year

PatientAge/Sex

OriginalDiagnosis Margin Dose Duration New Tumor

Loeffler200317

41/male Growth hormone-secretingpituitaryadenoma

87 Gy via 25separate beamsusing non-Braggpeak protonradiosurgery

16 years Tuberculum sellaemeningioma

Loeffler200317

53/male Growth hormone-secretingpituitaryadenoma

104 Gy via 12 fieldsusing Braggpeak protonradiosurgery

19 years Vestibularschwannoma

Sheehan200728

7/male Arteriovenousmalformation

20 Gy usinggamma knife

12 years Meningioma

Sheehan200728

12/female

Arteriovenousmalformation

25 Gy usinggamma knife

10 years Meningioma

Abbreviation: Duration, time between original treatment and diagnosis of malignant tumor.

Niranjan et al722

Case report 2A 53-year-old man underwent resection of a pituitary adenoma associated withacromegaly.17 Bragg peak proton radiosurgery followed. He received a peak doseof 104 Gy by means of 12 separate fields. Nineteen years later, MRI disclosed a leftvestibular schwannoma. The second tumor site had received 4.4 Gy during proton ra-diosurgery. A CT scan performed in 1979 showed no evidence of a soft tissue masswithin the posterior fossa or any widening of the internal auditory canal. Althoughthis case meets the criteria for a radiation-associated second tumor, one might arguethat if an MRI scan had been performed in 1979, a small intracanicular schwannomamight have been detected.

Case report 3A seven-year-old boy underwent gamma knife radiosurgery in May of 1990 usinga margin dose of 15 Gy (maximum dose of 30 Gy) to treat a right basal gangliaarteriovenous malformation.28 He underwent a second radiosurgery for persistentnidus in 1995 with a margin dose of 20 G (maximum, 40 Gy). Follow-up MRI in 2002showed a small dural-based tumor consistent with meningioma. This area received0.6 Gy and 0.25 Gy during first and second radiosurgery procedures.

Case report 4A 12-year-old girl underwent gamma knife radiosurgery in November of 1992 usinga margin dose of 25 Gy (maximum dose of 28 Gy) to treat a 1.2 cc right temporalarteriovenous malformation.28 She underwent a second radiosurgery for persistentnidus in 1995 with a margin dose of 20 G (maximum, 40 Gy). Follow-up MRI 10 yearsafter radiosurgery showed a mass consistent with meningioma in the previouslytreated area, which had received 25 Gy.

Radiosurgery-Associated Tumor Dedifferentiation

Spontaneous malignant transformation of tumors has been described in the literature,especially for glioma. There have been few case reports of malignant transformation


after radiosurgery, but such transformation is considered tumor dedifferentiation anddoes not meet Cahan’s criteria for a radiation-induced tumor. These tumors differ fromradiation-associated tumors in that the tumor was not induced by radiation but rathershowed evidence of malignant progression that involves the cellular evolution ofa benign lesion to malignancy (Table 3). Kubo and colleagues29 in their reportcontended that detailed histology prior to radiosurgery is essential, as biologicallyaggressive schwannomas are encountered in 0.14% of cases. These tumors exhibitlocal invasiveness, frequent recurrences, and systemic dissemination. If not identifiedbefore radiosurgery, an erroneous diagnosis of radiation-associated malignancy canbe made easily.

Of the 11 reported cases (see Table 3; Table 4) of tumor dedifferentiation, priorhistology was available only in five. On the other hand, even if the tumor exhibits unam-biguous histological proof of a benign pattern, its malignant progression after radio-surgery is not caused by irradiation necessarily. In fact, after incomplete surgicalresection, radiosurgery frequently is used only if residual neoplasm shows radiologi-cally confirmed regrowth. By itself, however, tumor growth after prior resection mayresult in malignant transformation of the initially benign tumor, either spontaneous,or even induced by microsurgery. Hanabusa and collegues30 reported on a patientwho had tumor recurrence 4 years after initial microsurgical resection for a vestibularschwannoma (pathologically proven from the first operation). The patient underwentradiosurgery for the recurrence. Six months later, a second microsurgical resectionwas performed. At reoperation, the tumor was found to be a malignant schwanno-ma.30 Whether the malignant transformation occurred because of radiosurgery orwas a result the natural history of malignant phenotype change is unclear. It iscommon for a surgeon to resect a benign meningioma, only to find at a second resec-tion for recurrence that the tumor is now atypical or even malignant.

The histological diagnosis of radiation-induced malignant transformation mayrepresent a significant challenge. In some reports, the aggressive locus was histolog-ically identifiable against a background of the benign tumor.31 In other reports, theentire recurrent tumor exhibited a malignant pattern.30,32,33 In the absence of specificclinical and histological criteria for identifying the irradiation-induced malignant trans-formation of the initially benign intracranial neoplasm, the pretreatment evaluation ofthe tumor proliferative potential seems to be extremely important.

RISK OF RADIOSURGERY-ASSOCIATED TUMORS

Prediction of the relative risk of radiation-associated second tumors in patients under-going radiosurgery is a challenging task. It is impossible to make an accurate assess-ment of the magnitude of radiosurgery-associated tumors because of lack ofsystematic long term follow-up of all cases and lack of knowledge of total numberof cases treated with radiosurgery. Approximately 80,000 patients in the United Stateswere treated with gamma knife radiosurgery for AVMs, benign tumors, and functionaldisorders up to 1996 (to allow 12 years of follow-up). It is this population that forms thelarge denominator of treated patients that will be needed to accurately estimate therisk of radiation-associated carcinogenesis.

A recent study of patients followed for up to 19 years after radiosurgery found noincreased risk of malignancies when compared with national mortality and cancerregistries.34 An accurate estimation of the probability that individual patients have ofdeveloping delayed tumors is the most critical issue left to be determined in radiosur-gery. Information from the studies on scalp irradiation for tinea capitis may be extrap-olated to provide an estimate of this risk.10,18 These articles showed that tissue

Table 3Radiosurgery-associated tumor dedifferentiation for sporadic vestibular schwannoma

Author, Year Patient Age/Sex CPAngle Tumor Histology Margin Dose Duration Final HistologyComey, 199831 50/male Sporadic No 14.4 Gy 5 years Triton tumor

Harada, 200340 7/male Sporadic VS RT 27 Gy 2 years Malignant schwannoma

Wilkinson, 200441 53/male Sporadic VS RT 7 years Malignant schwannoma

Hanabusa, 200130 51/female Sporadic VS 15 Gy 6 months Malignant schwannoma

Shin, 200232 26/female Sporadic VS 17 Gy 6 years Malignant schwannoma

Kubo et al, 200429 51/male Sporadic VS 14 Gy 8 months Malignant schwannoma

Abbreviations: Duration, time between original treatment and diagnosis of malignant tumor (all cases were treated with gamma knife radiosurgery unless other-wise noted); RT, fractionated radiation therapy; VS, vestibular schwannoma.

Nira

nja

net

al

724

Table 4Radiation therapy-associated tumor dedifferentiation for neurofibromatosis type 2 (NF2)

AuthorYearPatientAge/Sex

Cerebello-PontineAngleTumor Histology

MarginDose Duration

FinalHistology

Noren,199842

18/female NF2 No 20 Gy 5 years Triton tumor

Bari, 200243 28/female NF2 No 15 Gy 3.5 years Malignantschwannoma

Thomsen,200033

19/female NF2 No 12 Gy 6 years Meningosarcoma

McEvoy,200344

22/male NF2 No 15 Gy 2 years Not known

Ho, 200245 14/female NF2 No RT 18 Gy 7 months Death

All cases were treated with Gamma Knife radiosurgery unless otherwise noted.Abbreviations: Duration, time between original treatment and diagnosis of malignant tumor;

NF2, neurofibromatosis type 2.


receiving doses as low as 1 Gy are at increased risk for the development of secondtumors. Phantom studies based on the technique of irradiation used demonstratedthat tissues up to 2 cm from the scalp were exposed to this level of radiation. Thus,the volume of tissue at risk was approximately 900 cm3. By comparison, the totalvolume of adjacent tissue receiving 1 Gy or more during radiosurgery for a 1.5 cmvestibular schwannoma is approximately 120 cm3. Consequently, the tissue volumeat risk for later tumor formation after radiosurgery is less than 15% of the volumeirradiated for tinea capitis. Because the 30-year cumulative risk of radiation-inducedtumors after scalp irradiation was 0.8%, it can be predicted that the long-term riskafter radiosurgery should be approximately 1 in 1000.

Many argue that the numerator (five malignant and four benign radiation-associatedtumors reported) may not be reflective of the overall number of radiation-associatedtumors because of under-reporting or failure to recognize this complication. Still,the incidence must be relatively low when one considers a denominator of morethan 80,000 patients treated with radiosurgery worldwide up to 1996. Although itwill require 10 to 20 years of open-ended follow-up to determine the actual risk ofsecond tumor formation as a result of radiosurgery, enough information is availableto state that the relative risk is very low (1 in 1000 to 1 in 20,000 patients).

MECHANISM OF RADIATION-ASSOCIATED CARCINOGENESIS

The mechanism of radiation-associated carcinogenesis is multifaceted and not under-stood well. Radiation modulates various interdependent factors such as cell growth,apoptosis, mutations, repair, and genetic instability. Radiation-induced oncogenesisis thought to be secondary to sublethal nuclear damage of a tumor suppressorgene or a proto-oncogene, which is unable to be repaired. Subsequent acquiredenvironmental carcinogens then can produce oncogenesis. The time to acquire thesubsequent genetic defects is expressed as the latency between initial radiation expo-sure and carcinogenesis. This hypothesis is supported by the long-term follow-upstudy published by Rowe and colleagues,35 in which only one new primary intracranialtumor was reported among 4877 patients treated by radiosurgery. The age-matchedpopulation would have expected 2.47 cases.

The efficiency of tumor induction varies inversely with repair capacity, which in turndepends on the integrity of cell cycle checkpoints.36 The general form of the dose-

Niranjan et al726

response curve for radiation-associated second tumors is not clear, but several exper-iments on small animals suggest that the incidence increases with dose up to a maximumusually occurring between 3 and 10 Gy (delivered in a single dose), followed by a subse-quent decrease. Clinical evidence supports this biphasic relationship.37 There are alsostudies reporting that the highest incidence of radiation-associated second tumorsoccurs at field peripheries, where the dose is less than at the field center.36

There are no radiographic or histopathological features to differentiate betweenspontaneous and radiation-induced gliomas. It has been postulated that radiationmay result in genetic alterations that differ from those seen in spontaneous tumors.Genetic alterations that have been described include a three-base pair homozygousdeletion in exon 7 of the p53 gene,38 activating K-ras mutations,39 epidermal growthfactor receptor amplification, p16 deletions, pentaerythritol tetranitrate mutations,and p53 mutations.

There is also growing evidence that second tumors are more likely after combinedmodality treatment (radiation and chemotherapy), which is increasingly common. Therisk of radiation-associated cancer varies considerably with age at the time of irradi-ation. In some cases, younger patients may be especially vulnerable because ofa developmental window of opportunity for second tumor development.17

LESSONS LEARNED

Lessons learned include:

Radiation-associated tumors can occur within the full-dose region and in verylow-dose peripheral regions.

The risk is substantially lower with radiosurgery than that seen in patients treatedwith larger-volume radiation.

The latency period after radiosurgery is similar to what has been seen with fraction-ated therapy in the 6- to 20-year range, and malignant tumors occur earlier thantheir benign tumor counterparts.

Long-term follow-up should be obtained for all patients who undergo radiosurgeryfor benign brain lesions.

The current practice standards for radiosurgery should not be modified because ofthis very low risk of radiation-associated tumors.

FINALTHOUGHTS ON RADIOSURGERYOR RADIOTHERAPY FOR BENIGN TUMORS

In recent years there has been a dramatic increase in the number of patients selectingradiosurgery or radiotherapy as preferred management for benign tumors. Full disclo-sure of the risks and consequences of radiation and surgery is of paramount impor-tance in truly providing informed consent. All patients should be informed about therisk of radiation-associated tumors. They also should be aware that this risk remainssignificantly lower than that associated with the potential mortality after craniotomyand tumor resection (ie, from pulmonary embolus, myocardial infarction, meningitis,or some other adverse event). At the authors’ institution, all patients who undergoradiosurgery or radiation therapy with curative intent are informed about the risk fordeveloping a second tumor. Radiosurgery patients are told that this risk can occur5 to 30 years after the procedure. The authors estimate this risk to be low, less than1 in 1000, which is less than 10% of the risk for fractionated large-field irradiation.Currently, the authors have not seen such a case in their practice, which extends toover 9000 radiosurgery cases. Patients must weigh all of these factors as theymake educated decisions regarding their care.


REFERENCES

1. Hall EJ. Radiation carcinogenesis. Philadelphia: Lippincott Williams & Wilkins;2000.

2. Heros RC, Korosue K. Radiation treatment of cerebral arteriovenous malforma-tions. N Engl J Med 1990;323:127–9.

3. Cahan WG, Woodard HQ, Higinbotham NL, et al. Sarcoma arising in irradiatedbone: report of eleven cases. Cancer 1948;82:8–34.

4. Brenner DJ, Curtis RE, Hall EJ, et al. Second malignancies in prostate carcinomapatients after radiotherapy compared with surgery. Cancer 2000;88:398–406.

5. Boice JD Jr, Engholm G, Kleinerman RA, et al. Radiation dose and second cancerrisk in patients treated for cancer of the cervix. Radiat Res 1988;116:3–55.

6. Bhatia S, Robison LL, Oberlin O, et al. Breast cancer and other secondneoplasms after childhood Hodgkin’s disease. N Engl J Med 1996;334:745–51.

7. al-Mefty O, Kersh JE, Routh A, et al. The long-term side effects of radiationtherapy for benign brain tumors in adults. J Neurosurg 1990;73:502–12.

8. Liwnicz BH, Berger TS, Liwnicz RG, et al. Radiation-associated gliomas: a reportof four cases and analysis of postradiation tumors of the central nervous system.Neurosurgery 1985;17:436–45.

9. Modan B, Baidatz D, Mart H, et al. Radiation-induced head and neck tumours.Lancet 1974;1:277–9.

10. Ron E, Modan B, Boice JD Jr, et al. Tumors of the brain and nervous system afterradiotherapy in childhood. N Engl J Med 1988;319:1033–9.

11. Salvati M, Artico M, Caruso R, et al. A report on radiation-induced gliomas.Cancer 1991;67:392–7.

12. Shapiro S, Mealey J Jr, Sartorius C. Radiation-induced intracranial malignantgliomas. J Neurosurg 1989;71:77–82.

13. Simmons NE, Laws ER Jr. Glioma occurrence after sellar irradiation: case reportand review. Neurosurgery 1998;42:172–8.

14. Tsang RW, Laperriere NJ, Simpson WJ, et al. Glioma arising after radiationtherapy for pituitary adenoma. A report of four patients and estimation of risk.Cancer 1993;72:2227–33.

15. Sznajder L, Abrahams C, Parry DM, et al. Multiple schwannomas and meningi-omas associated with irradiation in childhood. Arch Intern Med 1996;156:1873–8.

16. Dweik A, Maheut-Lourmiere J, Lioret E, et al. Radiation-induced meningioma.Childs Nerv Syst 1995;11:661–3.

17. Loeffler JS, Niemierko A, Chapman PH. Second tumors after radiosurgery: tip ofthe iceberg or a bump in the road? Neurosurgery 2003;52:1436–40 [discussion:1440–432].

18. Sadetzki S, Flint-Richter P, Ben-Tal T, et al. Radiation-induced meningioma:a descriptive study of 253 cases. J Neurosurg 2002;97:1078–82.

19. Kimball Dalton VM, Gelber RD, Li F, et al. Second malignancies in patients treatedfor childhood acute lymphoblastic leukemia. J Clin Oncol 1998;16:2848–53.

20. Brada M, Ford D, Ashley S, et al. Risk of second brain tumour after conservativesurgery and radiotherapy for pituitary adenoma. BMJ 1992;304:1343–6.

21. Balasubramaniam A, Shannon P, Hodaie M, et al. Glioblastoma multiforme afterstereotactic radiotherapy for acoustic neuroma: case report and review of theliterature. Neuro Oncol 2007;9:447–53.

22. Larson DA, Flickinger JC, Loeffler JS. The radiobiology of radiosurgery. IntJ Radiat Oncol Biol Phys 1993;25:557–61.

Niranjan et al728

23. Berman EL, Eade TN, Brown D, et al. Radiation-induced tumor after stereotacticradiosurgery for an arteriovenous malformation: case report. Neurosurgery 2007;61:E1099 [discussion: E1099].

24. Kaido T, Hoshida T, Uranishi R, et al. Radiosurgery-induced brain tumor. Casereport. J Neurosurg 2001;95:710–3.

25. McIver JI, Pollock BE. Radiation-induced tumor after stereotactic radiosurgeryand whole-brain radiotherapy: case report and literature review. J Neurooncol2004;66:301–5.

26. Shamisa A, Bance M, Nag S, et al. Glioblastoma multiforme occurring in a patienttreated with gamma knife surgery. Case report and review of the literature.J Neurosurg 2001;94:816–21.

27. Yu JS, Yong WH, Wilson D, et al. Glioblastoma induction after radiosurgery formeningioma. Lancet 2000;356:1576–7.

28. Sheehan J, Yen CP, Steiner L. Gamma knife surgery-induced meningioma.Report of two cases and review of the literature. J Neurosurg 2006;105:325–9.

29. Kubo O, Chernov M, Izawa M, et al. Malignant progression of benign braintumors after gamma knife radiosurgery: is it really caused by irradiation? MinimInvasive Neurosurg 2005;48:334–9.

30. Hanabusa K, Morikawa A, Murata T, et al. Acoustic neuroma with malignanttransformation. Case report. J Neurosurg 2001;95:518–21.

31. Comey CH, McLaughlin MR, Jho HD, et al. Death from a malignant cerebellopon-tine angle triton tumor despite stereotactic radiosurgery. Case report. J Neuro-surg 1998;89:653–8.

32. Shin M, Ueki K, Kurita H, et al. Malignant transformation of a vestibular schwan-noma after gamma knife radiosurgery. Lancet 2002;360:309–10.

33. Thomsen J, Mirz F, Wetke R, et al. Intracranial sarcoma in a patient withneurofibromatosis type 2 treated with gamma knife radiosurgery for vestibularschwannoma. Am J Otol 2000;21:364–70.

34. Rowe J, Grainger A, Walton L, et al. Risk of malignancy after gamma knife stereo-tactic radiosurgery. Neurosurgery 2007;60:60–5 [discussion: 65–6].

35. Rowe J, Grainger A, Walton L, et al. Safety of radiosurgery applied to conditionswith abnormal tumor suppressor genes. Neurosurgery 2007;60:860–4 [discus-sion: 860–64].

36. Epstein R, Hanham I, Dale R. Radiotherapy-induced second cancers: are wedoing enough to protect young patients? Eur J Cancer 1997;33:526–30.

37. Boice JD Jr, Land CE, Shore RE, et al. Risk of breast cancer following low-doseradiation exposure. Radiology 1979;131:589–97.

38. Tada M, Sawamura Y, Abe H, et al. Homozygous p53 gene mutation in a radia-tion-induced glioblastoma 10 years after treatment for an intracranial germ celltumor: case report. Neurosurgery 1997;40:393–6.

39. Brustle O, Ohgaki H, Schmitt HP, et al. Primitive neuroectodermal tumors afterprophylactic central nervous system irradiation in children. Association with anactivated K-ras gene. Cancer 1992;69:2385–92.

40. Harada K, Nishizaki T, Adachi N, et al. Pediatric acoustic schwannoma showingrapid regrowth with high proliferative activity. Childs Nerv Syst 2000;16:134–7.

41. Wilkinson JS, Reid H, Armstrong GR. Malignant transformation of a recurrentvestibular schwannoma. J Clin Pathol 2004;57:109–10.

42. Noren G. Long-term complications following gamma knife radiosurgery of vestib-ular schwannomas. Stereotact Funct Neurosurg 1998;70(Suppl 1):65–73.

43. Bari ME, Forster DM, Kemeny AA, et al. Malignancy in a vestibular schwannoma.Report of a case with central neurofibromatosis, treated by both stereotactic


radiosurgery and surgical excision, with a review of the literature. Br J Neurosurg2002;16:284–9.

44. McEvoy AW, Kitchen ND. Rapid enlargement of a vestibular schwannomafollowing gamma knife treatment. Minim Invasive Neurosurg 2003;46:254–6.

45. Ho SY, Kveton JF. Rapid growth of acoustic neuromas after stereotacticradiotherapy in type 2 neurofibromatosis. Ear Nose Throat J 2002;81:831–3.

OCNA Sept 2009 Radiosurgery and Radiotherapy for Benign Skull Base Tumors

Documents

Transcript of OCNA Sept 2009 Radiosurgery and Radiotherapy for Benign Skull Base Tumors