Indications for SRS

Patients with newly presenting or recurrent benign intracranial neoplasms, determined based on imaging, biopsy, and history, may be given choices for surgical resection, SRS, and observation. Those patients who may benefit maximally from multi-session SRS often have the following characteristics:

• 1.
  

  Radiographic evidence of tumor progression after prior treatment

  
  
• 2.
  

  Symptomatic, untreated lesions with unfavorable locations for open surgery

  
  
• 3.
  

  Symptomatic lesions in patients who were poor surgical candidates because of advanced age or medical comorbidities

  
  
• 4.
  

  Patients choose to deny observation but refuse open surgery

Large benign tumors or tumors adjacent to sensitive critical neurologic structures, such as the optic nerve, may benefit from multi-session radiosurgery over single-session techniques.

Radiosurgical Technique

The CyberKnife Robotic Radiosurgical System (Accuray, Inc, Sunnyvale, California) was used to deliver the radiosurgical treatments. A high-resolution, thin-slice (1.25 mm) computed tomogram (CT) was obtained using a GE Light Speed 8i or 16i Scanner (Milwaukee, Wisconsin) after the administration of 125 mL of Omnipaque intravenous contrast (iohexol, 350 mg I/mL; Nycomed, Inc, GE Healthcare, Princeton, New Jersey). A stereotactic MRI scan was obtained and fused to the stereotactic CT scan to improve the target identification.

SRS is a joined effort among the neurosurgeon, radiation oncologist, and radiation physicist during the processes of tumor delineation, dose selection, and planning. The gross tumor volume of most benign intracranial neoplasms is defined as the residual or recurrent enhanced (if contrast enhancing) tumor seen on imaging. The radiosurgical dose is prescribed to cover the gross tumor volume with no additional margin. The treatment plan is created using the CyberKnife nonisocentric inverse treatment planning software. Once image-guided registration is performed, radiosurgery is administered to awake and usually unanesthetized patients. At the authors’ institution, patients have historically been given 4 mg of dexamethasone as well as antinausea medications as needed either before or after each treatment.

The quality of treatment plans is assessed by evaluating target coverage, dose heterogeneity, and conformity. Digitally reconstructed radiograms were computationally synthesized to allow near real-time patient tracking throughout radiosurgery. For multi-session treatments, the typical interfraction time interval was 24 hours.

Radiosurgical Dosage

SRS is a technique built largely on published center-specific experiences. Therefore, there is no strict dose selection for a specific pathological condition. Rather, it takes into account several patient- and center-specific factors. At Stanford University, the initial dose selection for multi-session radiosurgery was based on the following:

• 1.
  

  Extrapolation from the published single-fraction radiosurgery experience

  
  
• 2.
  

  A calculated biologic effective dose (BED) based on conventionally fractionated radiotherapy

  
  
• 3.
  

  A previous experience with fractionated frame-based radiosurgery for similar tumors

Afterward, the dose and fractionation decision for each patient depends on tumor volume, location, patient age, tumor histology, proximity to crucial neural structures, and history of irradiation.

A BED can be estimated from the following formula :

Multi-session SRS is routinely performed on an outpatient basis with an interfraction time of approximately 24 hours. When minimal toxicity is observed with multi-session radiosurgery, the prescribed radiobiologic dose can be gradually escalated. Dose escalation can be accomplished by either an increase in the dose per session or by a decrease in the number of sessions.

Meningiomas

Meningiomas are the most common nonglial primary intracranial tumors in adults. The main goal of management for these tumors is to minimize mass-induced symptoms. Small, asymptomatic lesions can have variable growth rates and can be observed with serial imaging. Easily accessible meningiomas can be easily treated with surgical resection. However, those in more eloquent locations, such as the skull base, and those that surround important neurovascular structures are less amenable to complete resection without severe morbidities. It has been estimated that 20% to 30% of meningioma cases are not amenable to complete surgical resection. For such reasons, SRS has become an attractive primary or adjuvant mode of treatment of meningiomas.

Local control rates following SRS of benign meningiomas were reported to range from 86% to 100% at 5 years and 83% to 95% at 10 years. Using historical control, Pollock and colleagues found that primary SRS for small- and medium-sized meningiomas had equivalent tumor control rates for SRS compared with Simpson grade I resection at 3 and 7 years. For less complete resections (Simpson grade II–IV), SRS provided better progression-free survival (PFS) at 3 and 7 years. The comparison of SRS and open surgery remains indirect because no such parallel studies exist and retrospective studies of this nature inherently carries substantial selection bias. However, the data thus far suggest that SRS as the primary treatment modality was somewhat comparable with open resection and without the invasiveness.

For skull-base meningiomas, which represent anatomic challenges for resection, Zachenhofer and colleagues reported no radiologic or symptomatic progression at late follow-up for the SRS-alone group and equivalent rates of neurologic improvement between the surgical and nonsurgical groups. In a series of 255 cases of skull-base meningiomas treated with GammaKnife SRS, Starke and colleagues reported actuarial PFS rates at 3, 5, and 10 years to be 99%, 96%, and 79%, respectively. Their data suggest that SRS can offer a good tumor control and neurologic preservation in patients with skull-base meningiomas, including those situated at the cerebellopontine angle, parasellar, or petroclival location.

Cavernous sinus meningiomas are a distinct variety with SRS being a favorable treatment option because of its anatomic location. The 5-year actuarial PFS is reported to be between 93% and 100%. In a series by Pollock and colleagues, 49 patients with dural-based masses of the cavernous sinus presumed to be meningiomas (mean tumor volume of 10.2 mL) were treated with a mean tumor margin dose of 15.9 Gy. No tumor enlargement was reported after SRS. Twelve of 38 patients (26%) with preexisting diplopia or facial symptoms improved in cranial nerve function. Although more data are needed to assert the efficacy of SRS as the primary treatment of cavernous sinus meningiomas, SRS can be an effective option for these tumors after the risk of open surgery is discussed with patients. Open surgery may be more useful in cases with a symptomatic mass effect or atypical imaging features that may suggest alternative or malignant tumor characteristics to confirm the histologic diagnosis.

Although SRS is a reasonable option for the treatment of meningiomas, limitations exist in the management of higher-grade tumors. In a review by Kondziolka and colleagues of 972 patients over an 18-year interval, even though the 10-year actuarial tumor control rate for WHO grade I tumors was 87%, the 5-year actuarial tumor control rates for grades II and III tumors were only approximately 30% and 10%, respectively. The risk of intracranial tumor recurrence increased approximately 4 times with increasing WHO grade. The decision to use SRS as a sole primary treatment of seemingly benign meningiomas may, therefore, be better guided by detailed imaging and available histology of the tumors when available.

At the authors’ institution, small meningiomas that are not adjacent to critical or sensitive neurologic structures are typically treated with single-session radiosurgery. Multi-session radiosurgery for meningiomas is typically reserved for larger meningiomas (greater than 3 cm), meningiomas located in parasagittal locations where the risk of edema has been shown to be substantially higher than other meningiomas, or for meningiomas located next to the optic or cochlear nerves.

Pituitary Adenomas

Pituitary adenomas are categorized as functioning and nonfunctioning subtypes. Functioning adenomas are hormone producing and result in exaggerated endocrine dysfunctions. Otherwise, patients may present with visual defects secondary to mass effect on the optic nerves and chiasm. The goals of management in patients with pituitary adenomas are multifaceted and include biochemical remission in functioning adenomas, preservation of normal pituitary production of other hormones, improvement in visual function, and local tumor control. A stepwise approach is taken in the management of these patients, starting with medical therapy (dopamine agonist for prolactin-secreting adenomas), surgical resection if a tumor is causing significant symptomatic mass effect, and radiotherapy (fractionated or SRS). The decision to undergo medical and surgical therapy often falls onto the endocrinologist and neurosurgeon. The decision to administer SRS or external beam radiation therapy (EBRT) is largely caused by the anatomic characteristics of a tumor. Poorly defined tumors or those in close proximity to the optic apparatus are best treated with EBRT because the margins cannot be well defined in SRS planning. Later, the authors focus on the use of SRS in controlling these tumors and compare its outcomes with those of fractionated radiotherapy.

Dose selection, much like other intracranial pathologic conditions, is based on histology, tumor size, the distance from adjacent normal structures (optic apparatus in the case of pituitary adenomas), and history of prior radiotherapy. For patients with hormone-secreting tumors, the minimum tumor margin dose is generally from 20 to 25 Gy. The dose prescribed for nonfunctioning adenomas generally is slightly lower than those that are functional, with a reported mean dose of 18.4 Gy (range 8–25 Gy) by Gopalan and colleagues. When the maximum dose received by the optic apparatus is less than 12 Gy, postradiosurgical visual deficits are reported to be less than 2%. At Stanford University, the authors recognized the restricted radiation tolerance of the anterior visual pathways as a limitation in single-session radiosurgery. The authors reported their experience with multi-session CyberKnife radiosurgery for perioptic lesions, which included 19 cases of pituitary adenomas. Two to 5 sessions of SRS to an average tumor volume of 7.7 cm ³ and a cumulative average marginal dose of 20.3 Gy were used. After a mean visual field follow-up of 49 months (range 6–96 months), vision was unchanged in 38 patients, improved in 8 (16%), and worse in 3 (6%). In each instance, visual deterioration attributed to tumor progression was confirmed with imaging. Seemingly, a multi-session SRS approach can result in high rates of tumor control and preserve visual function in perioptic pituitary adenomas.

SRS is an attractive approach for nonfunctional adenomas with a wealth of data supporting its efficacy and safety. Pollock and colleagues performed a retrospective review of 62 patients with nonfunctional adenomas undergoing radiosurgery between 1992 and 2004, of whom 59 (95%) underwent prior tumor resection. The median treatment volume was 4.0 cm ³ (range 0.8–12.9). The median treatment dose to the tumor margin was 16 Gy (range 11–20). The median maximum point dose to the optic apparatus was 9.5 Gy (range 5.0–12.6). At a median follow-up interval of 64 months, the tumor size decreased for 37 patients (60%) and remained unchanged for 23 patients (37%). Two patients (3%) had tumor growth outside the prescribed treatment volume and required additional treatment. Tumor growth control was 95% at 3 and 7 years after radiosurgery. The risk of developing new anterior pituitary deficits at 5 years was 32%. The investigators reported no decline in visual function in any patient. Similarly, the University of Virginia reported a high rate of tumor control and a low rate of neurologic deficits in their series. They reported on 140 consecutive patients with nonfunctioning pituitary macroadenomas treated using Gamma Knife surgery. Thirteen patients were treated with SRS as the primary therapy, and 127 patients had undergone at least 1 open resection before Gamma Knife surgery. The mean maximal dose was 38.6 Gy (range 10–70 Gy), with a mean marginal dose of 18 Gy (range 5–25 Gy). They reported PFS at 2, 5, 8, and 10 years to be 98%, 97%, 91%, and 87%, respectively. In multivariate analysis, they found tumor volume greater than 5 cm ³ (hazard ratio = 5.0, 95% confidence interval [CI] 1.5–17.2, P = .023) to be the only factor predictive of tumor growth. In their follow-up, delayed hypopituitarism occurred in 30.3% of patients. As expected, visual decline was the most common neurologic deficit (12.8%) in their series; but all patients with visual decline had evidence of tumor progression. In these large series, SRS seems to be an effective and safe option for tumor control with reasonable rates of posttreatment pituitary and visual deficits.

SRS has been reported to be a safe and effective treatment of patients with hormone-secreting pituitary adenomas. In patients with medically refractory prolactin (PRL)-producing adenomas or those who are intolerant to dopamine agonist therapy, Tanaka and colleagues reported biochemical remission off medications and clinical improvement in 4 patients (18%), normal serum PRL concentrations and clinical improvement on dopamine agonist therapy in 3 patients (14%), improved symptoms off medications but continued elevated serum PRL levels in 7 patients (32%), and 8 patients (36%) continued to be symptomatic with elevated PRL levels either on (n = 3) or off (n = 5) dopamine agonist therapy. A meta-analysis by Yang and colleagues evaluated the rates of remission in acromegaly treated with SRS and found that following SRS, the rate of cure was approximately 48% to 53%. At Stanford University, the authors have reported their initial experience of using SRS to treat acromegaly caused by growth hormone hypersecretion by a pituitary somatotroph adenoma. CyberKnife SRS, delivered as a single session (n = 5), 2 sessions (n = 3), or 3 sessions (n = 1), resulted in complete biochemical remission in 4 (44.4%) patients and in biochemical control with the concomitant use of a somatostatin analogue in an additional patient. Rates of complications following SRS for acromegaly, including new anterior pituitary deficits (0%–50%) and worsening of visual acuity or fields (0%–4%), were comparable with other tumor types receiving SRS. SRS can be a useful adjunct to surgical treatment in Cushing disease resulting from adrenocorticotrophic hormone (ACTH)–producing adenoma because failure to achieve remission or tumor recurrence occurs in up to 30% following successful transsphenoidal resection. Many reports showed ACTH normalization rates of approximately 40% to 65% following SRS. Jagannathan and colleagues reported on 90 patients with Cushing disease undergoing SRS with a mean dose of 23 Gy (median 25 Gy) and a mean endocrine follow-up of 45 months. They reported normalization of 24-hour urinary-free cortisol in 54% of patients, with an average time to remission of 13 months (range 2–67 months). Twenty percent of their patients demonstrated tumor recurrence between 6 and 60 months after SRS, suggesting the need for extended follow-up in patients with Cushing disease to monitor for tumor recurrence. Although there are ample data demonstrating the efficacy of single-session SRS for the treatment of the aforementioned functional adenomas, there remains room for the study of multi-session SRS scheduling in comparing efficacy and complication rates.

Overall, SRS is preferable to EBRT for a variety of reasons. SRS is more convenient for patients because it is performed as a single-day, outpatient-based procedure as opposed to EBRT, which is typically performed over 5 to 6 weeks. Biochemical remission seems to be shorter after SRS for patients with hormone-secreting tumors. Kong and colleagues compared the efficacy of EBRT (64 patients, mean dose 50.4 Gy) and single-fraction SRS (61 patients, mean dose 25.1 Gy) in 125 patients with pituitary adenomas and showed a shorter time to biochemical remission after SRS (median time to remission 26 months vs 63 months). The potential to normalize hormone levels more rapidly makes SRS more preferable in patients with either acromegaly or Cushing disease in order to minimize the metabolic consequences of these conditions. The incidence of new pituitary deficits can be higher after EBRT compared with SRS (60%–80% vs 10%–40%). SRS has a lower incidence of radiation-induced neoplasms compared with EBRT. In fact, Rowe and colleagues, from the National Center for Stereotactic Radiosurgery in Sheffield, found no difference in incidence in their radiosurgical patients compared with the age- and sex-adjusted national cohort of 5000 patients and more than 30,000 patient-years of follow-up.