Stereotactic Radiosurgery for Cavernous Sinus Tumors

Introduction

The cavernous sinus (CS) is a dural venous sinus that extends from the apex of the orbit to the apex of the petrous temporal bone and houses many critical neurovascular structures, including the carotid artery, the oculomotor nerve, the trochlear nerve, the first and second division of the trigeminal nerve, and the abducens nerve.

Neoplastic, inflammatory, infectious, granulomatous, and vascular pathology can occur within the CS. The differential for neoplastic CS lesions is broad and includes meningioma, pituitary adenoma, schwannoma, cavernous hemangioma, plexiform neurofibroma, malignant peripheral nerve sheath tumor, melanocytoma, chordoma, chondrosarcoma, nasopharyngeal carcinoma, juvenile angiofibroma, sphenoid sinus carcinoma, rhabdomyosarcoma, metastases, lymphoma, leukemia, post-transplantation lymphoproliferative disorder, epidermoid and dermoid cysts, and histiocytosis. Many of these tumors, such as nasopharyngeal carcinoma, arise in adjacent structures and extend into the CS. Inflammatory, infectious, and granulomatous lesions such as sarcoidosis and Tolosa-Hunt syndrome, as well as vascular lesions such as carotid-cavernous fistulas and aneurysms, also occur in the CS.

Clinical presentation, laboratory studies, and imaging are used to diagnosis CS lesions. MR imaging with routine T2, fluid-attenuated inversion recovery (FLAIR), pre-contrast T1-weighted imaging, and axial/coronal post-contrast T1-weighted images with fat saturation is needed to adequately assess a CS lesion. Imaging characteristics can be very helpful in differentiating the different tumor types within the CS. For example, meningiomas commonly arise from the lateral wall of the CS and often have a dural tail or narrow the internal carotid artery, while trigeminal schwannomas (TS) typically follow the course of the trigeminal nerve and often have a dumbbell shape due to narrowing at the porous trigeminus. Lesions that are indeterminate may require a biopsy for definitive diagnosis. The Mayo Clinic reported a series of 85 patients who underwent a biopsy for an indeterminate CS or Meckel’s cave lesion. While the surgical approach varied, a diagnosis was obtained in 90% of cases. Frontotemporal craniotomy, endoscopic endonasal biopsy, and retrosigmoid craniotomy had a diagnostic success rate of 88%, 95%, and 83%, respectively, while percutaneous transforamen ovale biopsy obtained a diagnosis in only 50% of cases. The diagnosis was metastatic in 32%, primary (majority meningioma) in 25%, hematologic in 13%, and fungal in 5%.

Because of the critical neurovasculature contained within the CS, surgical resection can be challenging and pose an increased risk of damage to these critical structures. Stereotactic radiosurgery (SRS) is an excellent alternative treatment that can offer high rates of local control with less morbidity. Herein, the most common pathologies within the CS that may benefit from radiosurgery are discussed.

Meningiomas

Meningiomas, a common dural based tumor, have a propensity for sites of dural reflection like the CS. In a large radiosurgical series including 1045 meningiomas treated with SRS, 29.3% were located in the CS. Gross total resection (GTR) of a meningioma and its dural attachments is generally considered the ideal treatment for meningiomas; however, GTR of CS meningiomas pose an increased risk of vascular and neurological morbidity. Therefore, radiation therapy is often used as a suitable alternative for the management of these challenging tumors.

Due to the lack of level 1 evidence regarding the role of surgery versus radiation for CS meningiomas, Sughrue et al. performed a meta-analysis to compare the results of published series of patients undergoing a GTR, a subtotal resection without adjuvant radiation therapy, or SRS alone. In the combined surgical series, 218 patients underwent a GTR with an 11.8% risk of recurrence at a mean follow-up of 51 months, while 217 patients underwent a subtotal resection with no adjuvant therapy with an 11.1% risk of recurrence at a mean follow-up of 59 months. The risk of recurrence was statistically higher in the surgical patients compared with the 1309 SRS patients, for whom the risk of recurrence was 3.2% at 44-month mean follow-up ( P < .01). Similarly, the risk of cranial neuropathy was markedly higher in the surgical patients compared with the SRS patients—59.6% versus 25.7%, respectively. These aggregate data suggest SRS is preferred over surgical resection for most CS meningiomas; however, surgery may still have a role when a histological diagnosis is required or for debulking tumor with mass effect upon critical structures, such as the optic nerves, optic chiasm, adjacent temporal lobe, or brainstem.

Several studies confirm the efficacy and safety of SRS for CS meningiomas. The University of Pittsburgh recently updated its series of 200 patients who underwent Gamma Knife (GK) radiosurgery, including 120 (60%) patients for primary management, 46 (23%) patients for residual tumor, and 34 (17%) patients for recurrent tumors. The median tumor target volume was 7.5 cm ³ (range 0.1 to 37.3 cm ³ ), and the median prescription dose to the tumor margin was 13 Gy (range 10 to 20). After a median follow-up of 101 months, tumor control was maintained in 170 (85%) patients, with 61% experiencing a regression of the tumor. The progression-free survival at 5 and 10 years was 92% and 84%, respectively. Patients who underwent radiosurgery for progressive tumor after prior surgery had a higher risk of recurrence (29.4%). At the time of treatment, 170 (85%) patients had a least one or more cranial nerve (CN) deficits; 57% of these patients presented for primary treatment (no prior surgery). After treatment, 26% of patients experienced an improvement in CN function, more commonly in patients undergoing primary treatment. Fifteen (7.5%) patients developed permanent CN dysfunction, with no evidence of tumor progression at a median of 9 months after radiosurgery. The risk of permanent CN dysfunction was higher in patients with a tumor volume ≥10 cm ³ . Three (1.5%) patients developed pituitary dysfunction after SRS, but no radiation-induced tumors were identified. Table 92.1 provides a comprehensive review of recent studies involving SRS for CS meningiomas.

Several factors may influence clinical outcome in these patients. Park et al. showed local tumor control was better in patients who underwent radiosurgery alone as definitive treatment compared with patients who underwent radiosurgery for progressive tumor following surgery. This difference may be due to several confounding factors, including the inherent aggressiveness of a progressive tumor requiring radiation after surgery, as well as the difficulty delineating a target volume after surgery. Other studies have not shown a correlation between prior surgery and progression-free survival. In addition, Skeie et al. demonstrated that large tumors and suboptimal radiation coverage (due to sparing of adjacent critical structures) both negatively influenced tumor control.

Dose plays an important role in achieving tumor control. Over the past several decades, studies have shown lower radiation doses are sufficient for controlling meningiomas. One study compared an initial cohort of patients receiving a mean dose of 14.1 Gy to a later cohort of patients receiving a mean dose of 12.5 Gy. There was no difference in tumor control between the groups, with 10-year local control above 80% in both groups, suggesting the lower dose is sufficient for tumor control. Another study by Cohen-Inbar et al. did suggest a difference when comparing cohorts treated with a margin dose of less than 16 Gy to those with a margin dose of ≥16 Gy. The 10-year progression-free survival was 96% and 82% for a margin dose of ≥16 Gy and less than 16 Gy, respectively. The current dose recommendation for SRS to CS meningiomas by the International Stereotactic Radiosurgical Society is 11 to 16 Gy.

Many patients present with CN dysfunction, and improvement in neurological outcome, in addition to tumor control, is a goal of treatment. Studies suggest CN function is more likely to improve following definitive SRS for primary treatment of CS meningiomas as compared with those patients who also undergo surgery as part of their management. ^{, , ,} This finding is most likely due to pre-treatment CN deficits secondary to surgery rather than tumor, which are less likely to improve following radiosurgery. Spiegelmann et al. demonstrated CN dysfunction is more likely to improve if treatment occurs within 1 year of initial symptom presentation. In this series, 49% of patients with symptoms less than 1 year had improvement in CN dysfunction after radiosurgery compared to 19% of patients with symptom duration greater than 1 year. These data suggest treatment sooner rather than later is advantageous for symptomatic patients.

New or worsening CN dysfunction is the most commonly reported toxicity after radiosurgery for CS meningiomas, although improvement in pre-treatment CN function can occur as described previously. Table 92.1 shows the rate of new or worsening CN neuropathy in several contemporary studies. Trigeminal nerve dysfunction causing facial pain or numbness accounts for the majority followed by diplopia and/or visual field deficits. Pituitary dysfunction is reported in several radiosurgical series in 1% to 2% of the treatment population and is dependent upon the dose to the pituitary gland and, therefore, the size and location of the tumor. ^, Several series have demonstrated that the rate of radiation-induced toxicity increases with large (≥10cm ³ ) tumor size. ^, In one series, the rate of radiation-induced complications was 12% and 5% for tumors ≥10cm ³ and less than 10cm ³ , respectively.

After SRS, most meningiomas will either regress (22% to 74%) or remain stable (24% to 62%) with time on serial scans. ^{, , ,} Cohen-Inbar et al. reviewed volumetric patient-specific data at 3-year follow-up and found a strong linear correlation between volumetric measurements at 3 and 10 years, suggesting tumor control at 3 years may reliably predict long-term, 10-year tumor control.

Fractionated radiation therapy remains the standard of care following a subtotal resection of an atypical (WHO grade II) or anaplastic (WHO grade III) meningioma, and the role of radiation is controversial, though often recommended, following a GTR of more histologically aggressive meningiomas. The role of SRS in WHO grade II and III meningiomas is less clear. It is often used with variable results in patients who develop a small recurrence after a GTR, have small-volume residual disease, or have growth of residual disease after surgery. In a series from Mayo Clinic, 71 tumors (55 WHO grade II meningiomas and 16 WHO grade III meningiomas) were treated with SRS. After a median follow-up of 38 months, 1-year and 5-year local control were 85% and 45%, respectively, suggesting a need for alternative treatments (including fractionated radiotherapy) in this patient population.

Schwannomas

TS are the most common schwannoma to occur within the CS. TS have a typical dumbbell-shape due to constriction of the tumor at the porous trigeminus as it grows along the nerve in the prepontine cistern, Meckel’s cavity, and CS. Symptoms of facial paresthesias, facial pain, and occasionally other cranial neuropathies can occur. Due to the rarity of TS, optimal management of these tumors is unclear. Surgical resection has variable success in achieving a complete resection and can be associated with increased morbidity. Surgery is often reserved for patients with large tumors or significant facial pain. SRS is a less invasive alternative to surgical resection.

There are several series reporting single institution outcomes with radiosurgery for TS. Hasegawa et al. reported on 53 patients treated with SRS for TS, the largest study to date, with a follow-up of 98 months. This study reported 5- and 10-year progression-free survival (PFS) rates of 90% and 82%, respectively. Notably, tumors with compression of the brainstem and deviation of the fourth ventricle had worse local control. Excluding this subgroup of patients, 5- and 10-year PFS was 95% and 90%, respectively. In a similar publication, Kano et al. published the results of 33 patients who underwent SRS. After an average follow-up of 6 years, the 5- and 10-year PFS rate was 82%. Tumor volume ≥ 8.0 cm ³ and dumbbell-shaped tumors were associated with lower rates of tumor control following treatment. The prescription dose for TS ranges from 12 to 15 Gy. Fig. 92.1 shows a treatment plan for a patient with a right trigeminal schwannoma treated with SRS on a linear accelerator, with subsequent follow-up imaging showing a positive treatment response.

In addition to durable tumor control, clinical outcomes with radiosurgery are also favorable. In several studies that monitored trigeminal nerve function following SRS, improvement in facial numbness/paresthesias was documented in 36% to 46% of patients, and improvement in facial pain was noted in 58% to 82% of patients. ^{, ,} The risk of adverse radiation effects, primarily trigeminal neuropathy, was 6% to 16% in contemporary SRS studies.