INTRODUCTION

Meningiomas are benign tumors of the central nervous system with an annual incidence of 2.3 to 4 in 100,000.^1,2 The term was initially coined by Harvey Cushing in 1922 to describe a tumor originating from the meningeal covering of the brain and spinal cord.³ Intracranial meningiomas account for approximately 90% of these tumors and outnumber spinal meningiomas by 10-fold. Brain tumor series indicate meningiomas to be 20% of all intracranial tumors (second only behind primary glial neoplasms), and autopsy studies estimate an incidence of 30%.⁴ These neoplasms have also been described in the orbit, paranasal sinuses, the skin and subcutis, lung, mediastinum, and adrenal glands.⁵ Meningiomas originate from arachnoid cap cells, are often found in association with arachnoid villi at the dural venous sinus and veins, and commonly attach to the dural covering of the brain and spinal cord.⁴

The World Health Organization (WHO) classifies meningiomas into grade I (benign), grade II (atypical), and grade III (anaplastic), which account for 80%, 5% to 20%, and 1% to 2% of all meningiomas respectively.^5,6 In large clinical series there is a strong association between outcome and grade. Patients with WHO grade I meningiomas have a greater than 80% chance of being progression free at 10 years, while only 40% to 60% of WHO grade II patients are progression free at 10 years. The median recurrence-free rate of patients with anaplastic meningioma is 2 years.⁷

Meningiomas often express hormonal receptors, which may explain the increased prevalence of meningiomas in females where the overall ratio is 2:1 in the brain and up to 10:1 in the spine.⁸ The best established of these receptors is the progesterone receptor, which is found in greater than two thirds of these tumors. In addition, greater than 30% of meningiomas also express estrogen receptors and approximately 40% express androgen receptors. Moreover, patients with meningiomas have been found to have a higher incidence of breast cancer and breast cancer patients have a higher incidence of meningiomas.⁹ The most frequent genetic alteration seen in meningiomas is the loss of the neurofibromatosis 2 gene (NF2) on chromosome 22q which encodes a tumor suppressor called merlin (also known as schwannomin). Merlin is related to molecules of the Protein 4.1 superfamily that are involved with cell growth and regulation.^5,8

Meningiomas are slow growing and are often incidentally found on imaging. If symptomatic, they can present with headaches, seizures, mental status changes, or focal neurologic deficits depending on the location of the lesion. On magnetic resonance imaging (MRI), meningiomas classically present as an extra-axial mass that is hypointense on T1-weighted images, hyperintense on T2-weighted imaging, and uniformly contrast enhance with gadolinium. Other characteristic findings include an enhancing dural tail, cerebrospinal fluid/vascular cleft, and hyperostosis of nearby bone. However, 10% to 15% of meningiomas exhibit atypical MRI findings that can mimic brain metastases or primary glial tumors.¹⁰

TREATMENT: OVERVIEW

Gross total resection has traditionally been considered the gold standard treatment for meningiomas. Radiation historically was reserved for poor surgical candidates, subtotal resection, unresectable tumors, or high-grade and recurrent tumors. Chemotherapy has not been shown to be effective. A recent phase II study evaluating temozolomide for treatment of recurrent meningioma showed no benefit.¹¹ Targeted therapies against the hormonal receptors have been disappointing,¹² but other adjuvant therapies such as hydroxyurea and other targeted therapies are currently under evaluation.¹³

Simpson was the first to publish the prognostic importance of complete surgical resection and created a scale linking extent of resection and its relationship to risk of recurrence.¹⁴ Since then, many authors have confirmed this link. However, even after an apparent gross total resection, three large series with extended follow-up report 7% to 12%, 20% to 25%, and 24% to 32% rates of recurrence at 5, 10, and 15 years, respectively.^15–17 Four series with extended follow-up of patients with subtotal resection alone reported 37% to 47%, 55% to 63%, and greater than 70% recurrence rate after an interval of 5, 10, and more than 15 years, respectively. In addition to high recurrence rates, complete excision is associated with a range of potential risks depending on the location of the tumor. Surgical resection of cavernous sinus, foramen magnum, and petroclival meningiomas is accompanied by great risk of cranial nerve dysfunction and cerebrospinal fluid leakage, while surgery for tumors that invade or occlude the posterior sagittal sinus and torcula not infrequently lead to venous infarction. Especially for meningiomas in such locations, the development of modern radiosurgical techniques has provided an adjunct, or in some cases, an alternative therapy to surgery.

RADIATION FOR MENINGIOMAS

Goldsmith and colleagues¹⁸ demonstrated that local control rates of subtotally resected meningiomas treated with conventional external beam radiation could rival that of gross-total resection. In this study, the 5-year progression free survival (PFS) in the post-1980 era when three-dimensional computed tomography (CT) planning and conformal radiation therapy became available was 98% compared with 77% in patients treated before 1980. With doses greater than 52 Gray (Gy), the 10-year PFS rates was 93% versus 65% in patients treated with less than 52 Gy. These results with traditional radiation therapy provide an important basis for evaluating ever newer radiation delivery techniques and in particular the application of stereotactic radiosurgery (SRS).

Whether conventional linear accelerator, Gamma Knife® (GKS), or CyberKnife®-based, SRS has been increasingly utilized instead of external-beam radiotherapy for managing meningiomas. The tumors being treated with SRS are generally recurrent, partially resected, surgically inaccessible, or in patients deemed poor surgical candidates or who refuse surgery. Several authors have hypothesized that a therapeutic gain may be achieved by treating slowly proliferating tumors, such as meningiomas, with larger-sized fractions, a treatment that is possible only with radiosurgical techniques.^2,19,20 Since the early 1990s many reports have described the efficacy of adjunctive or primary GKS radiosurgery. These studies demonstrate 5-year local control rates between 86% and 99%, tumor regression rates between 28% and 70%, and symptom improvement in 8% to 65%; meanwhile, complications occur in 2.5% to 13% of all cases.²¹ Further analysis of a subgroup of these patients by Pollock and colleagues²² looked at 7-year progression-free survival after a median follow-up of 64 months. There was no significant difference between patients treated with GKS alone (95%) or patients who underwent Simpson grade 1 resection (96%); Simpson grade 1 resection represents gross total resection of all tumor and involved dura and bone. Meanwhile, the Pollock study also showed that patients treated with GKS alone had increased tumor control at 7 years when compared to patients with Simpson grade 2 (82%) and grade 3 to 4 (34%) resections. Finally, when comparing the outcomes between standard radiation therapy and radiosurgery, Metellus and colleagues demonstrated excellent local tumor control with both, but clearly superior tumor shrinkage with radiosurgery.²³ After more than two decades of experience, SRS has become a widely accepted tool for managing a broad spectrum of meningiomas in different locations.

The remainder of this chapter is devoted to describing CyberKnife radiosurgery and its role in managing meningiomas. A particular emphasis is placed on some of the unique advantages of the CyberKnife. Preliminary results using CyberKnife radiosurgery for the treatment of certain types of meningioma are also presented.

CYBERKNIFE RADIOSURGERY

Stereotactic radiosurgery is a radiation technique delivered in one to five sessions and defined by a high degree of spatial accuracy and rapid radiation dose fall off at the periphery of the target lesions. Since Lars Leksell first conceived of radiosurgery in the 1950s,²⁴ this procedure has advanced with technologic improvements in instrumentation, computing, and imaging. For three decades, stereotactic targeting that utilized rigid skeletal fixation was at the center of all radiosurgical systems, and especially the gold standard GKS. However, in the 1990s, a novel frameless radiosurgery system that uses the patient’s bony anatomy for image guidance was developed and termed the CyberKnife radiosurgical system. After more than a decade of improvements, this technology is now widely commercially available.

The CyberKnife is a frameless, image-guided, robotic radiosurgical instrument. Therapeutic radiation is emitted from a compact 6 MV linear accelerator (LINAC) mounted on a robotic manipulator that is able to bring in beams from greater than 1200 directions (Fig. 55-1). This allows for non-isocentric radiation planning that can optimize both dose conformality and dose homogeneity. The need for rigid fixation is circumvented by image guidance. For intracranial targets, patients are relatively immobilized with an Aquaplast mask. Flat panel X-ray detectors are mounted on either side of the treatment table and obtain orthogonal X-ray images in real time during the treatment. These images are automatically referenced to digitally reconstructed radiographs (DRRs) created during treatment planning; computer algorithms establish the patient’s position by analyzing adjacent bony anatomy and/or implanted fiducials. The image guidance software compares spatial differences in three translational and three rotational axes and minute adjustments are then made with the treatment couch and robotic manipulator. This process is continuously updated throughout the radiosurgical treatment to maintain accuracy.²⁵

FIGURE 55-1 Schematic of the CyberKnife Radiosurgical System.

Unlike radiosurgical instruments that require rigid skeletal fixation, the CyberKnife can be used to ablate extracranial targets. For spinal targets, the X-Sight image guided targeting system utilizes nonrigid deformation modeling and hierarchical mesh tracking to maintain the spatial accuracy of target location within 0.6 mm. The X-Sight system has been shown to be accurate with system error ranging from 0.52 mm to 0.61 mm depending on testing methods and conditions.²⁶ This image-guided frameless system also has sufficient flexibility to perform radiosurgery in either a single day or over multiple days. Such fractionation or multi-session delivery of the radiation dose can be especially important when treating lesions near radiation sensitive structures such as the spinal cord. As detailed in the text that follows, the CyberKnife affords users unique advantages over other radiosurgical methods when treating certain types of intracranial and spinal meningiomas.

SRS TOXICITY

Radiosurgical ablation of meningiomas is well tolerated, with symptomatic complication rates ranging from 2.5% to 14%.^27–29 As radiosurgical techniques have improved and doses to normal anatomy have been reduced, the risk of postradiosurgical complications has decreased. Patients treated since 1991 experience less toxicity (5.3% vs. 22.9%), largely because radiosurgery prescription doses were gradually decreased from a median marginal dose of 17 Gy (range 10–20 Gy) between 1987 and 1991 to 14 Gy (range 8.9 Gy–20 Gy) between 1991 and 2000.³⁰

Peritumoral edema after SRS has been reported in 16% of cases and is related to tumor location.³¹ In a study of 179 patients with meningiomas, 9.3% of the 140 patients with MRI follow-up had symptomatic imaging changes that were manifested by headaches in most, but also seizures (4/33 patients) and other neurologic deficits (3/33 patients). In their analysis, only tumor location was found to be a significant predictor for peritumoral edema. Those tumors arising from the convexity (18%), parasagittal region (40%), and falx (6.7%) developed more edema than skull-base lesions (1.3%).²⁸ Ganz and colleagues also found that a sagittal sinus or midline location or a marginal dose greater than 18 Gy increased the risk of edema as well.³² Other studies have not found this relationship to dose. Kondziolka and colleagues reported a 16% actuarial 5-year rate of symptomatic edema in patients with parasagittal meningiomas treated with SRS; this complication rate was unrelated to dose but was influenced by the presence of pretreatment neurologic deficit.³¹ After conservative management, all of Kondziolka and colleagues’ patients had symptom resolution after a median of 15 months.

Cranial nerve damage, particularly optic neuropathy, is a well documented complication of all skull-base irradiation including SRS. Tischler and colleagues studied the risk of injury to cranial nerves II to VI in 62 patients (42/62 meningiomas) treated with radiosurgical doses between 10 and 40 Gy near the cavernous sinus. With a median follow-up of 19 months, 12 patients developed cranial nerve injury 3 to 41 months after treatment. Injury to cranial nerves III to VI was not related to dose. But in the four patients with optic nerve injury, there was increased risk in patients receiving greater than 8 Gy to any portion of the optic apparatus as compared to those receiving less than 8 Gy (24% vs. 0%, P = 0.009).³³ Another study reported that patients treated with larger doses (10–15 Gy) had 26.7% incidence of optic neuropathy.³⁴