Radiosurgery of Benign Intracranial Tumors

Published on 26/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 1399 times

CHAPTER 259 Radiosurgery of Benign Intracranial Tumors

Jean Régis, Marc Lévêque, Frédéric Castinetti, Pierre-Hugues Roche

Radiosurgery has become one of the primary modalities used for the management of intracranial tumors.¹^–³ The extensive radiosurgical experience that has developed worldwide over the years has allowed refinement of technique and a better understanding of radiobiologic concepts.⁴^–⁷ Additionally, improvement in patient selection, advances in magnetic resonance imaging (MRI) technology, and access to powerful computer-based navigation systems have led to very significant improvements in clinical radiosurgical results. Globally, radiosurgery has revolutionized neurosurgery by providing neurosurgeons with a novel approach that allows safe and efficient treatment of small, deeply seated lesions that would normally be associated with a high risk for functional deterioration with microsurgery. The remarkable safety-efficacy ratio of radiosurgery is especially valuable for patients with benign intracranial lesions that may not have an impact on life expectancy.

History

After its creation by Leksell in the 1950s and introduction of the Gamma Knife in the 1970s, radiosurgery was initially dedicated to functional neurosurgery.^8,⁹ However, with the discovery of new drugs effective in the treatment of functional disorders and with its proven efficacy in treating the nidus of arteriovascular malformations (AVMs), radiosurgery was used mainly for the treatment of AVMs. Even though the first vestibular schwannoma (VS) was treated in 1968 by Leksell and Noren,¹⁰ it was not until the 1980s that sufficiently accurate neuroradiologic targeting devices were available to allow the development of Gamma Knife surgery (GKS) as a modern, high-accuracy, image-guided surgical option for skull base surgery. In the 1980s, VSs and meningiomas became prominent indications for radiosurgery.

Since 1960, roughly 400,000 patients have been treated worldwide with GKS. Thus, GKS is no longer an experimental method, and sufficient evidence is available for us to draw a precise picture of the potential role of radiosurgery in the neurosurgical management of benign intracranial tumors.³

Definition

Radiosurgery is a neurosurgical procedure that uses convergent narrow ionizing beams, delivered in a single session with stereotactic accuracy and precision, to destroy or modify a biologically predefined target without injuring critical surrounding structures.¹¹ Radiotherapy was already being used for the treatment of intracranial tumors when Leksell developed the concept of radiosurgery in the 1950s. Although ionizing radiation is used for both conventional radiotherapy and radiosurgery, it is the physical properties inherent in each technique that make the radiobiologic effects of these two methods dramatically different. Thus, the clinical effects, radiologic effects, indications, risks, complications, and requirements for use will be very much different for both techniques.

Theoretically, radiotherapy is poorly adapted to the treatment of brain lesions because of the ordinarily high level of radioresistance of these tumors and the high sensitivity to radiation of precious neural structures. Radiotherapy, when delivered by fractionation, attempts to attenuate this effect through biologic selectivity. To adhere to the philosophy of treating intracranial lesions while sparing and avoiding any form of physical aggression involving normal neurological structures, neurosurgeons developed GKS, a topologic selectivity instrument that delivers the vast majority of the energy to the target and very little to the surrounding structures.¹¹

Radiobiology

Radiobiology is teaching us that with a single dose, the radiobiologic effect on the targeted structure is much more important than when the same total dose is spread over several fractions. Calculation of the biologic equivalent dose of radiosurgery to determine what dose should be used instead of fractionated treatment (e.g., 2 Gy per fraction) for the same radiobiologic effect ¹¹ requires that the nature of the targeted tissue be taken into account (linear quadratic formula). Such calculation indicates that there is a strong radiobiologic benefit of using radiosurgery instead of radiotherapy when treating benign intracranial tumors (slow-reacting tissue). This is also a demonstration that the radiobiology of radiosurgery and radiotherapy is extremely different.¹¹ Craniopharyngiomas are among the tumors traditionally requiring high radiotherapy doses because of their well-known radioresistance, in contrast to their high radiosensitivity to radiosurgery, which allows safe and effective radiosurgical treatment of small residual craniopharyngiomas even when they are very close to the optic pathways.

Topologic Differential Effect

The topologic differential effect is the basic principle for safety in radiosurgery, in which a high dose is delivered in a single session, as opposed to radiotherapy, in which an attempt is made to spare normal tissue by determining the biologic differential effect related to fractionation. Thus, accuracy and precision in radiosurgery are turning out to be particularly important. Especially high spatial accuracy is required for both imaging and dose planning, particularly when the tumor is located close to highly functional and fragile surrounding structures. The Gamma Knife with its fixed sources is providing us with an excellent falloff dose. However, to ensure highly accurate and precise radiosurgical treatment, precision and accuracy of the radiosurgical instrument are crucial, but not sufficient. Each step of the procedure needs to be performed precisely to guarantee such safety. Consequently, quality control will play a major role in establishing safety of the procedure. As a purely image-guided surgery with no immediate possible control of the effect of the procedure, radiosurgery requires very strict quality control of the whole procedure, especially with regard to the imaging aspect of it. Stereotactic MRI performed after application of the frame is the more informative imaging modality for benign tumors. Precise definition of the target boundaries and the relative topography of critical structures in the vicinity will usually require, in addition to a three-dimensional T1-weighted sequence, other specific sequences that can provide the neurosurgeon with complementary qualitative information. These sequences are selected on the basis of the topography of the lesion, the past therapeutic history of the patient (fat saturation [FATSAT] in previous surgery), and the patient’s functional status. The weakness of MRI technology is its potential non-negligible distortion because of inhomogeneity and nonlinearity of the field and gradients or magnetic susceptibility effects. Thus, we recommend systematic use of a second modality. CT scan is less informative but has the advantage of high resolution and a much lower risk for distortion than is the case with MRI.

Long-Term Complications

Long-term complications must be scrutinized because of the extensive use of stereotactic radiosurgery (SRS) for many benign tumors in young patients over the past 2 decades. The potential long-term carcinogenic risk associated with SRS was not evaluated until recently. The definition of radioinduced tumors is based on the following criteria proposed by Cahan and colleagues: the tumor must occur in a previously irradiated field after a long interval from the time of irradiation and must be pathologically different from the primary tumor and not present at the time of irradiation.¹² In addition, the patient must not have a genetic predisposition for the tumor. A low dose of radiation, such as 1 Gy, has been associated with second tumor formation at a relative risk of 1.57 to 8.75. This relative risk increases to 18.4 for an interval of time between 20 and 25 years. The radiation-associated incidence of tumor is linked to different factors such as age and individual genetic susceptibility. At this time, three radiation-associated gliomas and five malignant VSs have been reported in the literature. Moreover, these second tumors do not meet all the criteria of Cahan and coworkers.¹³ Long-term follow-up ranging from 5 to 30 years is needed to observe the crude incidence of radiation-induced tumors. The relative risk is estimated at less than 1 per 1000 and must be reported to each patient before any radiosurgical procedure.¹³

Numerous intracranial lesions are potentially good indications for radiosurgery, especially when they are small, well defined, and difficult to remove safely microsurgically; the main indications for SRS of these lesions are reviewed in this chapter.

Vestibular Schwannomas

Gamma Knife SRS was first used in 1969 by Leksell to treat VS.¹⁰ Lunsford’s group has established the optimal treatment parameters for control of VSs in conjunction with preservation of hearing and facial nerve function. Lunsford and associates have confirmed the importance of Noren’s policy of administering the “lowest irradiation doses that are therapeutically effective.”¹⁴ They demonstrated the impact of technical advances on the improvement in clinical results.¹⁵ Flickinger and colleagues recently reviewed their series of 313 patients with previously untreated unilateral VS who underwent SRS (a marginal dose of 12 to 13 Gy) between 1991 and 2001.¹⁶ The actuarial 6-year resection-free tumor control rate was 98.6%. The 6-year actuarial rates for preservation of facial nerve function, trigeminal nerve function, and hearing were 100%, 95.6 ± 1.8%, and 78.6 ± 5.1%, respectively. Using outcomes analysis and retrospective review, Lunsford’s team suggested an advantage (in terms of functional outcome) of SRS over microsurgery in the management of VS.¹⁷

In the past 3 decades, microsurgery and SRS have become well-established management options for VS. The Marseille SRS experience includes 2000 patients, with more than 1000 patients having follow-up longer than 3 years.¹⁸ A long-term tumor control rate of 97%, an incidence of transient facial palsy of lower than 1%, and a probability of preservation of functional hearing of between 50% and 95% were achieved in this large series of patients treated by state-of-the-art SRS (Fig. 259-1). Between 1973 and 2004, a total of 2577 VSs were surgically resected or treated with SRS by the Marseille group. Surgical resections were performed from the translabyrinthine, middle fossa, and retrosigmoid/suboccipital approaches. Approximately 1500 patients have been treated with Gamma Knife SRS, and all were evaluated prospectively. All patients with VS treated by SRS at Marseille receive a 12-Gy dose to the tumor margin. Three major technical advances have clearly influenced this practice. The availability of high-resolution stereotactic MRI,¹⁴ workstations for selecting the dose and treatment plan (Gamma Plan),¹⁹ and installation of the robotic Automatic Positioning System (APS) have allowed our group to achieve more conformal and selective dose planning. The average number of isocenters used in 1992 was less than 5 but increased to more than 15 in 2003 with the APS system. Consequently, if we consider the first 100 patients to represent the learning curve, four treatment periods can be defined. The rate of transient facial palsy and hemifacial spasm in patients with VS treated by our group was 3% and 3% during the first phase (June 1992 to December 1994, 100 patients), 1.4% and 2.8% during the second period (December 1994 to July 1997, 212 patients), 0.55% and 0.83% during the third period (July 1997 to May 2000, 360 patients), and 0% and 0% during the last period (May 2000 to January 2002, 258 patients), respectively.¹⁸

FIGURE 259-1 Gamma Knife radiosurgery for a Koos stage II vestibular schwannoma. Typical planning with the Perfexion Gamma Knife is illustrated. Note avoidance of the cochlea to limit the risk of hearing loss.

Comparison to Microsurgery

In a comparison of 110 VSs resected surgically and 97 treated by SRS, a lower rate of facial palsy and a higher probability of preservation of functional hearing were both achieved after SRS.²⁰ All patients had Koos stage II-III tumors with a minimum follow-up of 4 years. (Koos grading is as follows²¹: stage I, small intracanalicular tumor; stage II, small tumor with protrusion into the cerebellopontine angle and no contact with the brainstem; stage III, tumor occupying the cerebellopontine cistern with no brainstem displacement; and stage IV, large tumor with brainstem and cranial nerve displacement.) The three other studies in which the safety and efficacy of microsurgery and radiosurgery were compared are consistent with these results.^17,^18,²²

Efficacy of Radiosurgery

To better define the accuracy and efficacy of SRS, the imaging-determined morphologic changes of 1000 tumors treated between July 1992 and January 2002 ²³ by the otoneurosurgical group of the Timone Hospital in Marseille were evaluated systematically. Evaluation with MRI before and after SRS (intervals of 6 months and 1, 2, 3, 5, 7, and 10 years) was reviewed. Systematic measurements have been performed on all tumors treated. Preoperatively, 129 patients had progressive tumors. At the time of SRS, the median tumor volume was 732 mm³ (mean, 1346; range, 20 to 14,405). According to the Koos topographic classification, among the 996 patients with complete follow-up, there were 80 stage I, 538 stage II, 322 stage III, and 56 stage IV tumors. Loss of central enhancement was noted on MRI at 6 months and/or 1 year postoperatively in 45.5% of patients. In 64% of these patients, loss of central contrast enhancement occurred. A significant increase in tumor size was recorded in 15% of the patients (Figs. 259-2 and 259-3). In 3% of the patients, progression led to a second procedure, either resection or repeated SRS. Failure is defined as continuous tumor progression 3 years after SRS.¹⁸ Tumor control was achieved in 97% of the patients. Because the natural history of VS includes growth of 2 mm/yr, these results confirm the efficacy of SRS.

FIGURE 259-2 Overall evolution of the volume of treated vestibular schwannomas. GK, Gamma Knife radiosurgery.

FIGURE 259-3 Outcome of Gamma Knife radiosurgery for vestibular schwannoma. In a 42-year-old woman with tinnitus of the right ear and Gardner-Robertson class 2 residual hearing on the right side, a Koos stage III right vestibular schwannoma was discovered in June 1997. On the day of Gamma Knife surgery (September 10, 1997), magnetic resonance imaging (MRI) demonstrated a dramatic increase in the tumor, which became a Koos stage IV tumor that was causing a slight mass effect on the brainstem (A). A decision was made with the patient to proceed with treatment despite enlargement of the tumor. Six months after MRI, loss of the central contrast enhancement is seen (B). After a transient asymptomatic increase in the tumor, subsequent images (C) demonstrated a dramatic reduction (from 5 cm³ to less than 1 cm³) in tumor volume (D). On the last MRI in February 2008, the tumor has shrunk, and there is no longer any mass effect on the brainstem. The patient has experienced no complications, side effects, or discomfort since radiosurgery and has retained functional hearing. Between the first audiometry session in April 1997 and treatment (5 months), the patient has experienced an average loss of 10 dB. Since radiosurgery and after 11 years of follow-up, the average loss is just 15 dB (E). The probability of preserving normal facial function after radical microsurgical removal should have been poor in this woman with a large lesion, and the probability of preserving functional hearing should have been close to zero after microsurgery. GKS, Gamma Knife surgery.

Hearing Preservation

In the SRS experience of the Marseille group, 175 patients with VS and functional preoperative hearing (Gardner-Robertson stage 1 or 2) were initially treated with SRS and have a follow-up of longer than 3 years.²³ Hearing was preserved in 60% of all patients after SRS. Univariate and multivariate analysis has revealed parameters that influence the probability of preservation of functional hearing at 3 years, including limited hearing loss (Gardner-Robertson stage 1), the presence of tinnitus, younger age, and small lesion size. The rate of preservation of functional hearing at 3 years was 77.8% in patients with stage 1 hearing, 80% in patients with tinnitus as a first symptom, and 95% when the patient had both stage 1 hearing and tinnitus. In these patients, the probability of preservation of functional hearing at 5 years was 84%.²⁴

Facial Nerve Preservation

Facial nerve palsy is very rare (<1%) in the series from the three centers with the largest experience (Marseille, Pittsburgh, and Rhode Island).¹⁸ Schwannomas originating from the facial nerve itself are more inclined to interfere with facial motor function. Because of the paucity of facial palsy after SRS for VS, the use of SRS may be reasonable in this difficult group of patients.²⁵ Among the 1000 schwannomas of the cerebellopontine angle treated by SRS in Marseille between July 1992 and March 2003, 9 were diagnosed as originating from cranial nerve VII. The criterion for diagnosis was involvement of the second or third portion of cranial nerve VII (seven patients) or intraoperative demonstration during a previous surgical resection (two patients). Facial palsy occurring within 18 months of radiosurgery for VS was determined in patients with more than 2 years of follow-up (eight patients). Four patients had previous facial paresis. Normal facial motor function was observed in two patients before SRS (House-Brackmann grade 2 in six patients and grade 3 in one patient). Follow-up ranged from 2 to 7 years in all patients. Worsened facial palsy did not occur in any patient, whereas two patients had improvement in their preoperative facial palsy. Our results and those from the main series in the literature confirm the important role of SRS in the management of VS and preservation of facial nerve function (Table 259-1).²⁶^–³⁶

TABLE 259-1 Treatment of Vestibular Schwannomas by Radiosurgery in the Literature

The best candidates for radiosurgery are young patients with small and medium VSs and few symptoms. Patients with Koos stage II and III tumors are good candidates as well. In addition, patients with intracanalicular, cystic, previously resected, and Koos stage IV tumors may be candidates under certain conditions. Originally, until 1999, Koos stage I tumors were considered for radiosurgery at numerous institutions only in patients with tumor progression. Retrospective analysis of the tumor’s growth rate and preservation of functional hearing and requests for radiosurgery by patients have led to modification of this practice. According to the Marseille group, patients treated with SRS have a higher probability of preservation of functional hearing.³⁷ Consequently, patients with a stage I lesion and functional hearing may be considered for radiosurgery in the absence of tumor progression.

Large Vestibular Schwannomas: Combined Microsurgical and Radiosurgical Treatment

Conservative treatment of large vestibular schwannomas (LVSs) exposes patients to severe and permanent functional impairment, as well as life-threatening complications. However, this option can be considered for selected elderly people with limited life expectancy.

Radiosurgical treatment of LVSs is associated with less probability of tumor control and increases the risk for radioinduced complications and hydrocephalus.

Microsurgical resection is therefore generally recommended for LVSs. Continuous technical refinements in microsurgery had led to a dramatic reduction in mortality and severe morbidity in the past 30 years. Nevertheless, identification, complete dissection, and preservation of the facial nerve are challenging in many of these large tumors when radical resection is attempted. In the hands of the best skull base teams, such treatment offers a 60% to 70% rate of normal to nearly normal facial nerve function at last follow-up after surgery for LVSs, thus indicating that there is still a significant rate of permanent facial nerve deficit.

To reduce the incidence of this unacceptable complication, several centers have adopted a combined approach in which optimal subtotal microsurgical resection is followed by radiosurgical treatment of the residual tumor.³⁸^–⁴⁰ In our own otoneurosurgical group, a series of 51 patients (16 males, 35 females) harboring LVSs were managed between January 2003 and January 2008 with an intentional combined approach. The initial symptoms were hearing deterioration, imbalance, ataxia, and hydrocephalus. Our treatment strategy was to perform optimal resection of the tumor with preservation of facial nerve integrity. Radiologically, 51 patients had Koos stage IV LVSs.²¹ The median tumor diameter in the cerebellopontine angle was 30 mm (range, 25 to 35 mm). Microsurgically, a widened translabyrinthine approach was used in 38 patients, whereas 13 underwent a retrosigmoid transmeatal approach. Postoperative evaluation of patient status and facial nerve function was determined at 3 and 6 months and 1, 2, 3, 5, 7, and 9 years after surgery, including assessment of the intermedius nerve (the smaller root of cranial nerve VII). Three months after surgery, gadolinium-enhanced MRI of the posterior fossa was performed to assess the size of the tumor remnant.

The need for radiosurgical treatment of the remnant tumor was determined postoperatively and depended on the size of the tumor remnant and facial nerve status. In patients with grade I and II postoperative facial motion, radiosurgery was performed in the 3- to 6-month period after surgery (Fig. 259-4). However, GKS was postponed to 9 to 12 months if facial nerve motion was worse than House-Brackmann grade 2 because recovery could reasonably be expected. Total removal was achieved in 9 patients, nearly total removal (>95% tumor resection) in 7 patients, subtotal removal (95% to 80% tumor resection) in 34 patients, and partial removal (<80% tumor resection) in 1 patient. Postoperative facial nerve preservation (House-Brackmann grade 1 or 2) was achieved in 66% of the patients immediately and in 82.2% at last follow-up, as shown in Table 259-1. The functional status of the intermedius nerve was able to be assessed in 39 patients at longer than 6 months after surgery. The nerve was normal in 20 patients, partially impaired in 13, and totally deficient in 6. Twenty-four patients underwent GKS at a 4- to 13-month interval after subtotal resection of their tumor. Thus far, no complications have been observed after radiosurgery, even with particular attention directed to facial and the intermedius nerve status, which remained unchanged. The median radiologic and clinical follow-up after microsurgery was 24 months (range, 8 to 68 months). Permanent tumor control was achieved in all patients with no need to reoperate or offer another GKS procedure.

FIGURE 259-4 A and B, Combined approach to the treatment of large vestibular schwannomas. C, Microsurgical resection was followed by radiosurgery of the residual tumor.

Resection of LVSs is accompanied by a significant rate of complications, but leakage of cerebrospinal fluid, meningitis, posterior fossa hematomas, and general complications do not depend on the extent of resection. Conversely, an attempt at radical resection requires extensive dissection of the facial nerve in its entire course from the brainstem to the fundus because the nerve is usually shifted to the anterior pole of the tumor capsule. This dissection is at particular risk into the internal auditory canal and at the level of the meatus. Therefore, one may presume that avoiding dissection of the nerve at this point will reduce the risk for injury. Because lowering the risk for damage to the facial nerve by all means was one of our major goals, we decided to be less radical and intentionally left some tumor in each instance when resection was risky. As shown in this study, safe resection could be achieved in 17.6% of our population, but in other patients it was necessary to stop the resection; this possibility was mentioned to the patients before surgery.

Justification of the necessity for additional radiosurgery in the event of a tumor remnant is an important issue. In actuality, we have few data concerning the fate of conservatively treated residual tumor. Before the modern era of neuroimaging, reports indicated a 40% rate of clinical regrowth after incomplete surgery.⁴¹ W. F. House had to perform reoperations on 10 of 22 patients 4 to 7 years after initial partial resection, with increased morbidity.⁴² More recently, Silverstein reported a 20% rate of regrowth in 15 patients older than 65 years who initially underwent subtotal resection.⁴³ In studies in which computed tomography (CT) or MRI was available, rates of regrowth and reoperation after incomplete removal were 40% and 20%, respectively.⁴⁴ Therefore, we consider it warrented to offer adjunctive GKS when a clear target is identified on postoperative MRI. This proactive strategy is also justified by the necessity of reoperating on patients with a high risk for damage to the facial nerve.⁴⁵

Meningiomas

We consider radical microsurgical removal as the reference for the treatment of intracranial meningiomas (Simpson grade 0). In all situations in which this goal seems achievable, there is limited indication for radiosurgery. When complete safe resection of a large meningioma appears unachievable because of its topography and close relationship to critical structures, radiosurgery may intentionally be considered as the second stage of combined treatment after subtotal removal.

First-line radiosurgery is recommended for meningiomas in the cavernous sinus when they are small enough and sufficiently far from the optic pathways.⁴⁶ When the lesion is too big or too close to the optic pathway, a combined approach (resection of the portion of the lesion out of the cavernous sinus) or conventional radiotherapy is advocated (Fig. 259-5).^47,⁴⁸

FIGURE 259-5 A and B, Large cavernous sinus meningiomas treated by a combined approach. C, Microsurgical resection was followed by radiosurgery of the residual tumor.

Petroclival meningiomas may be an excellent indication for radiosurgery, especially when they are growing, causing few clinical signs, and small enough that they can be treated with lower morbidity than possible with microsurgery.⁴⁹

Parasagittal meningiomas have been demonstrated to correlate with the highest rate of brain edema, especially in patients previously operated on or with a neurological deficit at the time of radiosurgery.⁵⁰

Rational for Radiosurgery

Meningiomas account for about 20% of intracranial tumors, with an incidence of 1 to 6 per 100,000 persons. Surgical resection is historically the treatment of reference. In many instances, meningiomas cannot be completely excised as a result of unacceptable risks of morbidity because of their crucial neurovascular environment. Analysis of the modern literature indicates that radical microsurgical resection of skull base meningiomas is achieved in 50% to 60% of patients, with permanent morbidity rates reaching 20%, a mortality rate of 5% to 15%,^51,⁵² and a recurrence rate of 10% at 10 years after surgery.⁵³ Additionally, because 90% to 95% of diagnosed meningiomas are benign (World Health Organization [WHO] grade I), preservation of quality of life together with control of tumor growth should be the major goals of treatment of this select group of meningiomas. In this scenario, SRS has become an effective alternative to microsurgery.

Based on several models (quadratic linear model, alpha-beta ratio), radiobiologic data indicate that meningioma tissue responds to single-session and high-dose radiation therapy. Meningiomas are usually sharply delineated and encapsulated and rarely infiltrate the brain tissue. Clear identification of meningiomas on CT and MRI allows precise definition of target volumes and steep radiation falloff with sparing of adjacent normal tissue.

Patient Selection

Tumor size, tumor growth, and the patient’s age and neurological status are the key parameters to consider in determining which therapeutic option to choose. As a general rule, patients with large tumors associated with neurological deficits are candidates for primary microsurgical resection. Conversely, elderly patients with an incidentally discovered meningioma are preferentially treated conservatively with serial observation. Apart from these two typical situations, radiosurgery can be discussed as the first treatment option on an individual basis. Even in patients with size-compatible meningiomas, the neurovascular environment of the tumor is carefully assessed on the radiologic work-up before treatment, with special attention being directed to the optic pathway and brainstem. Previous treatments (radiation therapy, microsurgery) and histologic grading of the meningioma are also taken into consideration before radiosurgery.

A selective review of published series is outlined in Table 259-2.⁵⁴^–⁶⁰ The target can be situated anywhere, but skull base locations are predominant. In a recent series published by the Pittsburgh team,⁶¹ tumor was localized to the middle fossa in 351 patients, posterior fossa in 307, convexity in 126, anterior fossa in 88, parasagittal region in 113, and other areas in 115 patients.

TABLE 259-2 Main Series of Long-Term Tumor Control after Gamma Knife or Linear Accelerator Treatment of Meningiomas

Technical Considerations

Because radiosurgery involves the delivery of high-energy radiation in a single session, it is necessary to ensure selective and usually complete coverage of the target with the treatment. Therefore, excellent high-resolution contrast-enhanced axial MRI planning is needed to provide selective and conformal treatment while avoiding the delivery of radiation to surrounding tissues. Fat suppression images are used for tumors adjacent to the orbit or infratemporal fossa or after a previous skull base approach. Image fusion with a bone-window CT scan may be helpful for meningiomas in cranial base locations.

The optimal radiosurgery dose for meningioma is still under debate. Selection of doses depends on tissue response, tumor volume, and the tolerance of neighboring structures. Doses of less than 12 Gy have been reported to be a significant factor in failure to control the growth of meningiomas.⁶² According to Kondziolka and associates, marginal doses greater than 15 Gy do not provide better tumor control.⁶³ Chin and colleagues recommended 12 to 14 Gy for tumors larger than 3 cm, 16 Gy for tumors 1 to 3 cm, and 18 Gy for tumors smaller than 1 cm.⁶⁴ Tishler and coworkers investigated the tolerance of the second through sixth cranial nerves to radiosurgery.⁶⁵ They found a higher incidence of damage to the optic pathway in patients receiving more than 8 Gy to any part of the optic apparatus. Leber and coworkers indicated that no signs of radiation-induced optic neuropathy were observed when the dose to the visual pathway was less than 10 Gy whereas this complication occurred in 26.7% of their patients when the dose varied from 10 to less than 15 Gy and in 77.8% when the dose was 15 Gy or higher.⁶⁶ In contrast, Morita and colleagues showed that it was possible to safely deliver 12 to 16 Gy to short segments of the visual pathway.^66a Tishler and associates indicated that the oculomotor nerves in the cavernous sinus could tolerate doses greater than 20 Gy whereas the trigeminal nerve was potentially at risk with doses beyond 19 Gy.⁶⁵ Several studies have recommended that the lateral wall of the cavernous sinus, pituitary gland and stalk, hypothalamus, and brainstem not receive more than 15 Gy.

Results

The two main goals of radiosurgery for meningiomas are long-term tumor control and preservation of function. Evaluation of these results takes into account specificities related to the histology, clinical evaluation, and response of the tumor to treatment.

Histopathology

For patients treated primarily, histologic proof of success is not available. In several locations such as the cavernous sinus or jugular foramen, it may be difficult to distinguish meningioma from schwannoma, plasma cell granuloma, hemangioma, or paraganglioma. A series published in 2003 reported on the treatment of 219 imaging-diagnosed meningiomas with a mean follow-up of 29 months.⁶⁷ Progression of tumor after GKS occurred in 7 patients, 2 of whom were found to have different tumors. We fully agree with the recommendation that biopsy should be performed before treatment in patients with atypical imaging features or unusual clinical findings. In patients with the typical radiologic features of meningioma, however, the risk of misdiagnosis seems too low to justify biopsy, but such an approach is probably not representative.

Clinical Evaluation

Without treatment, the natural history of skull base meningiomas consists of spontaneous recovery and fluctuations in symptoms such as cranial nerve deficits. Thus, clinical improvement may be mistakenly attributed to radiosurgery without adequate long-term follow-up.

Tumor Response

Long-term follow-up is required for evaluation of the response of the tumor to treatment in view of the slow growth of many meningiomas. Yano and coauthors recently showed in a cohort of patients treated conservatively that lack of hyperintense tumor signal on T2-weighted sequences and the presence of calcifications were predictive of slow growth whereas perifocal edema was associated with the potential for more rapid growth.^67a

Tumor Control

Most series indicate that the tumor is controlled in 86.7%⁶⁸ to 100%^69,⁷⁰ of patients. In a recent study of a large group of patients, Kollovà and associates reported a 5-year actuarial tumor control rate of 97.9%,⁶² but studies providing long-term follow-up are sparse (see Table 259-2).

The results of stabilization and regression of tumor size after GKS are diverse. Several series have reported a decrease in tumor size in 60% of patients and tumor stability in 40%,^63,^68,^70,⁷¹ whereas others indicate that in only 13% to 16% of patients does the tumor decrease in size.⁷² In this later report, the mean reduction in size of benign meningiomas was 16.1%, and the mean response rate was highest in the cavernous sinus location. The heterogeneity of results may be due to different ways of evaluating tumor behavior. Stafford and colleagues defined tumor regression as a decrease in tumor size of greater than 2 mm,⁷³ but this criterion is not universally adopted, and many teams believe that any decrease in tumor volume indicates regression of the tumor. Huge variations in length of follow-up may also explain these heterogeneous results. Moreover, Zachenhofer and associates showed that patients could be “late responders,” with tumors starting to shrink more than 4 years after GKS.⁵⁹ Several studies have reviewed the factors associated with failure. In the study reported by Dibiase and coauthors, female gender, a conformity index of 1.4 or greater, dural tail treatment, and gross tumor volume (GTV) of less than 10 cm³ were significantly linked with longer disease-free survival (DFS) in univariate analysis.^73a In multivariate analysis, GTV greater than 10 cm³ was the only factor associated with worse 5-year DFS (hazard ratio, 4.58).

Functional Outcome

Neurological improvement after SRS can be expected when the deficits are incomplete and of recent onset. Such improvement or recovery has been reported in 14% to 48% of patients in the literature. Although nonspecific symptoms such as headaches and vertigo may be eliminated after SRS,^59,⁷⁴ the most dramatic effect involves the cranial nerves. Trigeminal neuralgia improves in 13% to 91% of patients, third nerve deficits in 17% to 67%, and abducens nerve deficits in 21% to 71%. In our personal experience, abducens nerve deficits improved after SRS in 60% of patients with petroclival tumors and in 42% with parasellar meningiomas.^49,⁷⁵ Such recovery may or may not be linked to shrinkage of the tumor. The probability of recovery from optic pathway deficits and trigeminal hypoesthesia is low.

The toxicity associated with radiosurgery for meningiomas is mainly due to symptomatic edema or damage to cranial nerves. In the majority of patients, the morbidity is temporary and rarely disabling; however, permanent complications have been reported in 2.5% to 9% of patients. Flickinger and coworkers showed that in their series, the risk for adverse sequelae after radiosurgery decreased in patients treated after 1991, which corresponded to the routine use of stereotactic MRI for planning treatment and a decrease in the doses delivered.⁶⁷ Symptomatic peritumoral imaging changes developed in 4% of the patients reported by Kondziolka and coauthors at a mean of 8 months⁶¹ and in 9.3% of patients who were monitored by sequential MRI in another study.⁷⁶ In this later series, imaging changes developed at a mean of 7.8 months (range, 2.8 to 48.9 months) and were sustained for 13.5 months. The mechanisms for these changes are unclear, but they have been ascribed to vasogenic edema, with a potential role being played by vascular endothelial growth factor or radiation injury to the vasculature. Symptoms attributable to edema may consist of transient headaches, seizures, or other neurological deficits. Postradiation edema has been correlated with tumor location. Parasagittal,⁶³ parafalcine, and anterior fossa locations are at risk.⁷⁷ In the study by Chang and colleagues, imaging changes were evident in 5.1% of skull base meningiomas and in 50% of hemispheric locations.⁷⁶ Previous surgery, age older than 60 years, mean GTV larger than 10 cm³, maximum radiosurgical dose greater than 30 Gy, and dose to the tumor margin higher than 16 Gy have also been linked to an increased risk for edema.

Cranial nerve neuropathies are unusual. The occurrence of a new deficit depends on the doses delivered to the tissue, the length of the segment of nerve exposed, and the class of nerve. In the event of worsening, one should make sure that the tumor is not growing. We personally reported that treatment failure was observed in two of the three patients with oculomotor deficits that occurred after GKS for cavernous sinus meningiomas.⁷⁵ Brainstem radiolesions and anterior pituitary failure have also been reported. Vascular occlusion of the intracavernous internal carotid artery has been documented to occur in 1% to 2% of patients with cavernous sinus meningiomas.^73,⁷⁵ Seizures are reported in 1.3% to 5.8% of patients, with most seizures occurring between 1 and 16 weeks after GKS and usually being transient. Their cause is unknown, but they could be due to gliotic scarring or peritumoral edema.

Special Situations

Large Tumors

We have stressed that GTV is the key parameter in predicting tumor control after GKS, and therefore GKS is not recommended as a first stage for large meningiomas. Staged radiosurgery has been proposed,⁷⁸ but no data are available to recommend this option. Another proposition is to deliberately expose the entire volume of the tumor but spare a compartment that is not supposed to grow or not put the patient at risk in the event of growth. We personally think that extension to the paranasal sinuses or to the infratemporal fossa occurs with large meningiomas but does not systematically require treatment.

Another option is to perform microsurgery initially to reduce tumor volume, decompress critical structures, and allow safer performance of radiosurgery. A combined strategy requires excellent collaboration between treatment teams and careful evaluation of the postoperative images to identify the target for GKS.

The target of GKS is the portion of the meningioma in which excision would pose a risk for neurological symptoms.^69,^74,⁷⁹ Radiosurgery does have more side effects and a greater risk for failure when given as adjunctive treatment after incomplete resection than when performed as primary treatment. Whether radiosurgery should be routinely performed in the months after surgery or only in the case of tumor regrowth is still debatable.

Histology

Although the 5-year tumor control rate is uniformly greater than 90% for WHO grade I meningiomas, the results are less favorable for aggressive meningiomas. Kondziolka and coauthors reported tumor control rates of 50% and 17% for grade II and grade III meningiomas, respectively.⁶¹ In these cases, particularly in patients with grade III meningioma, an aggressive management strategy combining extensive microsurgery with fractionated radiation therapy should be considered. Malignant and atypical meningiomas are clearly more likely to fail to respond to radiosurgery.⁶¹ Hemangiopericytomas are frequently similar to meningiomas from a radiologic point of view but are usually much more sensitive to radiosurgery.^80,⁸¹

Pituitary Tumors

Pituitary adenomas are benign tumors that may cause clinical signs either by secreting hormones (acromegaly, Cushing’s disease, and prolactinomas) or by local mass effect, particularly chiasmatic compression. The therapeutic algorithm is thus different when the aim of treatment is to control hormone hypersecretion (e.g., in a secreting microadenoma) or tumor volume (e.g., in a large nonsecreting pituitary adenoma). In secreting pituitary adenomas, transsphenoidal surgery is the treatment of choice in most cases, except for prolactinomas; however, this surgical approach does not always allow definite remission, and efficacious adjunctive treatment is frequently needed. Antisecretory drugs are inconsistently effective and sometimes poorly tolerated, and conventional radiotherapy, though highly effective, is associated with high rates of panhypopituitarism.⁸²^–⁸⁵

Acromegaly

With growth hormone (GH)-secreting pituitary adenomas, the first-line treatment of acromegaly is transsphenoidal surgery (Fig. 259-6), but remission rates range from 44% to 74%, with variations being due to neurosurgeon experience, size of the adenoma (microadenoma or macroadenoma), and presurgical GH levels. In other cases, surgery is contraindicated and cure is impossible because of cavernous invasion. Currently published studies have reported a wide range of efficacy of SRS varying from 17% to 100%.⁸⁶^–⁹⁸ This broad range is probably due to various criteria for remission; only few studies to date have used the most recent criteria, and SRS is probably effective in about 40% to 50% of patients.⁹⁴^–⁹⁷ In our study just 17% of patients were in remission, but the adenomas in 23% of our patients were able to be controlled with somatostatin agonists, whereas they were not with the same treatment before SRS; withdrawal of this treatment would probably have increased our remission rate (because these patients were considered uncured at the end of the study).⁹⁴ We and others have shown similar efficacy whether SRS was used as primary treatment (without surgery) or as adjunctive treatment (after unsuccessful surgery), thus suggesting that SRS could be performed as an alternative first-line treatment in patients with contraindications to surgery.⁹⁴^–⁹⁶ One of the main drawbacks of the technique is the delay in remission, and with the mean time to remission varying from 12 to 60 months in the literature, adjunctive treatment (somatostatin agonists or pegvisomant, for instance) is required during this period to control excess secretion.^{94–97,99–101}

FIGURE 259-6 Residual pituitary adenoma treated by radiosurgery via a transnasal approach in an acromegalic patient. Typical planning with the Perfexion Gamma Knife is illustrated.

Factors predictive of remission vary with the study; the dose to the target, initial GH and insulin-like growth factor type I (IGF-I) levels, and target volume could be valuable predictive factors, but their impact on the final result was not systematically reported in the literature. The question of the radioprotective effect of somatostatin agonists given at the time of radiosurgery remains a matter of controversy, and the initial hypothesis that somatostatin agonists could reduce the proliferation rate of the adenoma ⁹⁸ and then decrease the efficacy of radiosurgery was not confirmed by two of the three recent studies on the topic and was not evaluated in one.⁹⁴^–⁹⁷ However, we share other investigators’ viewpoint that such treatment should be stopped shortly before SRS because the initial GH and IGF-I levels could be factors predictive of remission.

Cushing’s Disease

With adrenocorticotropic hormone (ACTH)-secreting pituitary adenomas (Cushing’s disease), first-line treatment is transsphenoidal surgery. The remission rate ranges from 50% to 80% and varies with the type of adenoma and the experience of the neurosurgeon ¹⁰²^–¹⁰⁶; recent studies in the literature have reported an elevated risk for recurrence (>25%) with prolonged postsurgical follow-up (>5 years).¹⁰⁷

Only a few studies have reported the results of treatment of Cushing’s disease with SRS, and remission rate has ranged from 10% to 83%.^89,¹⁰⁸^–¹¹³ However, the higher rates of remission were observed in radiosurgical series conducted in the pre-MRI era and involved elevated rates of induced hypopituitarism (two thirds of treated patients had least one new anterior pituitary deficiency at last follow-up).¹⁰⁸ The remission rate with SRS is probably closer to 40% to 50%, as observed in our study.¹¹³ As seen with acromegaly, SRS seemed to be equally effective as an adjunctive or primary treatment. The main drawback of the technique is again the delay until remission, estimated to be 24 to 36 months, with efficacious medical treatment being required during this period to control signs of excess cortisol,^100,^101,¹¹³ which may prove challenging.

Factors predictive of remission varied with the study; however, dose and target volume seem to be valuable predictive factors. Interestingly, we found that the group of patients treated with ketoconazole at the time of radiosurgery had a lower rate of remission than did the group that was not treated,¹¹³ but the physiopathologic explanation for this mechanism remains unclear, even though in vitro data show a decrease in ACTH with ketoconazole infusion and thus suggest a central effect of the drug (in addition to its well-known peripheral adrenal effects).¹¹⁴ None of the studies published to date have reported an effect of pre-SRS hormonal levels on the final result of the procedure.

Prolactinomas

With prolactin-secreting pituitary adenomas, first-line treatment depends on the type of adenoma: dopamine agonists can always be proposed, surgery less frequently in the case of macroadenoma. Medical or surgical remission was reported in about 90% of patients with microprolactinomas, and surgical remission rates do not exceed 50% in those with macroprolactinomas. However, dopamine agonists are sometimes not tolerated, and after unsuccessful surgery or in patients with a contraindication to surgery, an adjunctive treatment may be proposed.⁸²

The rate of remission after SRS ranges from 20% to 80% (mean, 50%) after a mean follow-up of about 30 months.^87,^89,^90,^93,^110,¹¹⁵^–¹²⁰ Factors predictive of remission include the dose to the target and the target volume. No study published to date has reported that a low pre-SRS prolactin level has any effect on the final result of the procedure. However, because of the high efficacy of medical and surgical treatments, published studies were always based on small numbers of patients. Only one series reported SRS treatment (128 patients), but the low rate of remission (15%) could suggest that the large number of patients in this series did not represent the ideal indication for SRS because they were systematically treated primarily with SRS.¹¹⁸ It is important to stress that as in acromegaly, a few studies have reported a presumably radioprotective effect of dopamine agonists; patients treated at the time of radiosurgery were less frequently cured than were those who were not treated with dopamine agonists. Although this might be due to selection bias, withdrawal of dopamine agonists before SRS may thus be advocated.^117,¹²¹

Antitumoral Effects

The antitumoral effects of SRS are excellent. Decreases in the volume of pituitary adenomas were reported to range from 70% to 100%,^100,^101,¹²² with results varying with the dose to the tumor and cavernous sinus invasion.¹²² One study described a transient increase in the days after SRS, probably because of inflammatory edema of the target.⁹⁷ These results are of importance, especially in patients with nonsecreting adenomas, in which the main objective of the procedure is to control tumor size. A recent study reported the outcome of 62 patients at a median follow-up of 64 months in whom SRS was used to treat nonfunctioning pituitary adenomas; only 2 patients exhibited regrowth of the tumor after the radiosurgical procedure, whereas tumor size decreased in 60% and remained unchanged in 37%. The fact that all patients in this study had previously been treated surgically confirms that SRS could be a valuable adjuvant treatment in those with incomplete surgical resection of nonfunctioning pituitary adenomas.¹²³

Hypopituitarism

The risk for hypopituitarism ranges from 0% to 66% and varies with the dose (dose to the stalk in one study), accuracy in defining the target (one study reported hypopituitarism in 66% of patients in the pre-MRI era of SRS), and pre-SRS radiotherapy or radiosurgery (which increases the risk for hypopituitarism).^94,⁹⁹^–¹⁰¹ ^,114 ^,124 However, most published studies were based on short-term follow-up, and thus prolonged studies will be necessary to confirm this relatively low incidence of hypopituitarism. The risk for hypopituitarism increases with the length of time after radiosurgery, and increased rates of remission are usually associated with increased rates of hypopituitarism. The need for a precise target is evident because total sellar radiosurgery will induce panhypopituitarism in the majority of cases.

Optic Nerve Neuropathy

Optic nerve neuropathy is estimated to occur in less than 2% of patients. The risk increases if the target-to-chiasm distance is less than 5 mm and if the dose to the chiasm is greater than 8 to 10 Gy.^65,^95,^125,¹²⁶ It is thus currently not recommended that GKS be performed if the target is too close to the optic chiasm.

Other Potential Adverse Effects

Additional potential adverse effects include transient headaches in the days after the procedure.^99,¹⁰¹ Other severe side effects, including radiation-induced tumors and cognitive dysfunction, have never been reported. However, the mean follow-up of currently published studies is too short to draw any firm conclusions on these adverse effects, which have previously been reported after conventional radiotherapy. The safety of SRS is not absolutely defined, but it will be determined with longer term studies.

Conventional Radiotherapy

The efficacy of conventional radiotherapy in controlling hormone hypersecretion is estimated to be about 50% to 90%, whatever the type of secretion.¹²⁷^–¹³² However, conventional radiotherapy has the following two drawbacks: the time until remission is long, 5 to 10 years, with effective medical treatment therefore being required during this period, and there is an elevated risk for side effects, including hypopituitarism (in more than 80% of patients),¹²⁷^–¹³² optic neuritis, radiation-induced cerebral tumors, cerebral infarction, and cognitive dysfunction.^90,¹³³^–¹³⁶ These latter side effects occurred after a mean of 10 to 20 years and have not been reported with the use of SRS (however, studies based on SRS had a shorter follow-up time, and thus prolonged studies will be necessary to confirm the absence of such side effects). In nonfunctioning pituitary adenomas, the results of conventional radiotherapy are comparable to those of SRS, with unchanged or decreased tumor volume occurring in the majority of patients.

Comparison between radiosurgery and radiotherapy is difficult because the indications for each procedure are theoretically different. The indications for SRS are limited by the need for a well-delimited, small target volume sufficiently far from the optic chiasm (>5 mm); the dose prescription to the target must take into account the secreting characteristics of the adenoma (a lower dose is needed when the objective is only to control tumor volume, whereas a higher dose is needed in the case of secreting adenomas) and the requirement to keep the dose delivered to the chiasm low (risk of optic neuropathy if greater than 8-10 Gy). Conversely, conventional radiotherapy is indicated for larger tumors.

Summary

For all types of adenoma, SRS should be performed in patients with a small target volume sufficiently far from the optic chiasm to avoid the chance of radiation-induced optic neuropathy. For secreting pituitary adenomas, SRS represents an alternative treatment after unsuccessful surgery and in patients resistant to or intolerant of medical treatment. Although still a matter of debate, it seems logical to withdraw medical therapies before SRS, not only because of a potential risk of radioprotective effects of the drugs but perhaps more importantly to be able to compare hormone levels before and after treatment while not undergoing medical therapy. One should keep in mind that the mean time to remission is approximately 24 to 36 months, so effective medical treatment is required during this period. Follow-up evaluations should be performed after withdrawal of medical treatment to be able to correctly evaluate the efficacy of the procedure. In patients with nonsecreting pituitary adenomas, SRS may be performed for adenomas recurring after incomplete surgical resection. SRS thus seems to be an effective procedure that achieves control of excess hormone secretion in 40% to 50% of patients and control of tumor volume in all patients. Because the risk for hypopituitarism is lower with SRS than with conventional radiotherapy, SRS seems to be a valuable adjunctive treatment after unsuccessful surgery, as well as an alternative treatment in patients who tolerate medical therapy poorly or as a first-line treatment in patients with contraindications to surgical treatment of a small tumor.

Other Benign Tumors

Nonvestibular Schwannomas

Schwannomas, particularly those arising from the trigeminal nerve or from the foramen jugulare, are now considered to be potentially very good indications for radiosurgery when their size is suitable for this approach (Fig. 259-7). When these tumors are too large for radiosurgery as a primary treatment, especially in young patients with few or no clinical signs, combined approaches (partial microsurgical resection followed by radiosurgery on the remnant) must be discussed in all cases.

FIGURE 259-7 Radiosurgery for a foramen jugulare schwannoma. Typical planning with the Perfexion Gamma Knife is shown.

Craniopharyngiomas

Craniopharyngiomas must ideally be removed radically when diagnosed. After so-called radical removal, the recurrence rate has been reported to be between 15% and 38% in the recent literature.¹³⁷ However, aggressive management can result in severe complications, including morbid obesity, growth retardation, diabetes insipidus, sexual problems, and deficiency of GH, TH, and ACTH. In our experience with 53 patients with a minimum follow-up of 9 years, the recurrence rate was 27.9%. Because of the unacceptably high complication rate and lack of total prevention of recurrence after radical tumor resection, there has been a growing advocacy for less invasive tumor resection with adjuvant therapy.¹³⁸

However, radiotherapy results in serious morbidity (radionecrosis, optic neuritis, malignancies, and disturbances in cognition) in 6% to 18% of patients.^139,¹⁴⁰ With a median follow-up of 17 years, Regine and colleagues demonstrated radiation-related complications in 58% of children and 46% of adults, with an increased risk for endocrinologic morbidity and vascular complications when the maximum dose exceeds 61 Gy.¹⁴¹ Even conventional radiotherapy at a maximum dose of 54 to 55.8 Gy has been demonstrated to result in cognitive worsening in a significant percentage of patients.¹⁴² Globally, predictors of deterioration in long-term quality of life are visual field defects, repeated surgery, and radiotherapy.¹⁴³ In this study, general and physical fatigue, energy, physical condition, and physical mobility were all significantly adversely affected compared to controls (142 patients).