Clinical aspects

Description

There are five primary techniques for stereotactic radiosurgery delivery: Gamma Knife, LINAC, tomotherapy, CyberKnife®, and heavy particle (proton beam) (52).

Gamma Knife. Gamma Knife radiosurgery typically involves placing the patient in a stereotactic frame, which is attached to the patient under local anesthesia by pins that penetrate the scalp into the outer table of the skull. The Gamma Knife focuses 192 cobalt 60 radiation sources, mounted in a 18,000 kg shield, through a collimated helmet on to a target 403 mm from the cobalt 60 radiation sources.

Collimation allows for creation of up to 192 16-, 8-, and 4-mm beams that can each be overlaid and blocked to conform to each isocenter. This independently moving sector design allows for significantly enhanced flexibility, which leads to shorter treatment times and the ability to treat multiple intracranial lesions with a single session and frame placement (150). The new Leksell Gamma Knife system, Icon, approved by the FDA in August 2016, uses a cone beam computed tomography scanner and infrared camera to perform Gamma Knife radiosurgery using the thermal plastic mask head immobilization, which is standard for external beam radiation (185; 176). Use of the thermal plastic mask also allows for multiple sessions to be performed without having to apply and reapply a stereotactic frame, which allows for fractionation of the radiation, increasing the number of patients eligible for Gamma Knife.

LINAC. LINAC utilizes high-energy photons in a series of converging arcs to focus radiation onto a specific target. The newest method of LINAC-based radiosurgery is volumetric arc therapy (VMAT). In VMAT therapy, the LINAC modulates the shape, dose rate, and energy of the exiting beams via a multileaf collimator while rotating continuously around the target. Adjustments can be made to the multileaf collimator, gantry rotation speed, and dose delivery rate, which allows for greater customization of the conformal dose distribution (125). This gives the VMAT the advantage of allowing highly conformal treatments to multiple intracranial lesions with reduced treatment times. New LINAC-based techniques have outputs as high as 1400 monitor units per minute and can complete a high-dose stereotactic treatment in less than 10 minutes. A study of VMAT for the treatment of multiple intracranial metastases has demonstrated effective 1- and 2-year local control rates (86.7% and 71.6% rates, respectively). The intracranial progression-free survival at 6 months and 1 year were 48.5% and 19.4%, respectively (05). In the case of multiple intracranial metastatic lesions, VMAT typically is performed by using multiple isocenters, one for each lesion. Alternative approaches for VMAT using multiple noncoplanar treatment arcs have been suggested as a method for reducing overall treatment time (145).

Tomotherapy. Tomotherapy uses a LINAC mounted within the gantry of a CT scanner, allowing for imaging just prior to treatment delivery (178). In this approach, a continuous, helical-shaped beam with a single isocenter is used to provide radiotherapy to an entire region. This appears to be advantageous as it can treat a large volume without multiple isocenters. Still, it unfortunately requires a longer treatment time (171). In previous studies, tomotherapy resulted in tumor control and a low incidence of toxicities when used as treatment for single or multiple brain metastases (123; 85). It was also found to be a safe treatment for brain metastases over 2 cm that could not be surgically removed.

CyberKnife. The CyberKnife system combines a lightweight LINAC optimized for stereotactic radiosurgery with an articulated highly maneuverable robotic arm.

The system uses image registration techniques from neuroimaging studies prior to treatment and combines them with radiographs of skeletal features associated with the treatment site obtained from two fixed diagnostic fluoroscopes mounted on either side of the patient. This information determines the treatment site, coordinates with respect to the CyberKnife LINAC, and actuates the robotic arm to accurately direct the beam to the treatment site. When the target moves (such as with respiration), the process detects the change and realigns the aim of the beam to the new coordinates in near real-time. This system obviates the need for rigid fixation, allows the treatment of lesions throughout the body, and allows for fractionation of treatment without recurrent patient discomfort (02). The flexibility of the CyberKnife has encouraged treatment of extracranial pathology, including spinal lesions and cancers of the liver, lung (41), prostate, and kidney (62).

Heavy particle (proton beam). The proton beam system employs the ionization portion of the Bragg peak, which describes deposition of radiation from heavy particle beams such as helium. Charged particles produce a triphasic pattern of radiation with an intermediate dose on entrance of the head and a narrow peak of high-dose radiation (Bragg ionization peak) that can be adjusted to target tissue penetration depth and a minimal exit dose. Although the need for a cyclotron limited proton therapy to only a few centers in the United States for many years, there are now many proton therapy centers. The added cost of refinements for improved accuracy, such as pencil beam scanning, as well as expected reductions in reimbursement for proton therapy, will make the financial viability of many proton therapy centers questionable in the future (46). However, as the cost of this technology lowers with time, there is an important role for proton beam, particularly in the treatment of pediatric patients due to a significantly lower rate of secondary malignancy when compared to photon-based treatments (35). In addition to several solid organ-based tissue cancers, promising results have also been obtained in treating challenging intracranial tumors, such as chordomas (155) and glioblastoma (121), and several trials and investigations are ongoing (120).

FLASH radiotherapy. FLASH radiotherapy is a new approach that has significant promise in the treatment of brain tumors. FLASH radiotherapy treatment provides irradiation at a singular high dose over 40 Gy/s, reducing radiation-induced damage in healthy tissues without reducing tumor treatment efficiency (104). Animal studies have found that FLASH RT can delay glioblastoma growth while also sparing learning and memory deficits typically induced by radiation (122). FLASH RT is currently gaining attention in the clinical realm. The first clinical trial for FLASH RT has been completed and concluded that FLASH RT resulted in mild toxic effects and therapeutic benefit in the treatment of symptomatic bone metastases (113). Additional animal studies and clinical trials must be completed before FLASH RT can be used in treating brain metastases, but this method has strong potential to improve the efficacy of stereotactic radiosurgery.

Although capable of treating lesions as small as 2 to 3 mm diameter, all Gamma Knife, LINAC, and proton beam–based techniques require accurate target localization preoperatively, typically with a contrast MRI for primary and metastatic brain tumors, and CT arteriography for arteriovenous malformations. Isodose plots are constructed from the neuroimaging studies to calculate dosimetry. Treatment planning is a team effort including a radiation oncologist, radiation physicist, and neurosurgeon. This effort is aided by advancements in computer programs that help to automate much of the treatment planning and delivery. Stereotactic radiosurgery treatment time varies from under 30 minutes to an hour or more, depending on the number of isocenters treated and the stereotactic radiosurgery technique (system) employed. The patient is usually discharged the same day or within 24 hours. Corticosteroids, antiemetics, or anticonvulsants may be administered.

The goal of stereotactic radiosurgery is the delivery of a lethal radiation dose to the target tissue without damage to the adjacent normal tissue. Most stereotactic radiosurgery devices have total clinical accuracy on the order of 1 to 1.5 mm.

Treatment goals depend on the type and location of the lesion treated. For certain benign tumors, eg, meningiomas and vestibular schwannomas, lesion stabilization may be sufficient; for most malignant tumors, especially symptomatic metastatic tumors, shrinkage or obliteration of the lesion may be necessary. Arteriovenous malformations and other vascular lesions usually require obliteration, which may take 2 years or more after treatment before the risk of bleeding is completely eliminated. For tumors, histological tissue diagnosis and determination of the lesion margins are essential. For functional disorders, notably trigeminal neuralgia and epilepsy, the goal is symptom improvement (143; 170).

Indications

Stereotactic radiosurgery was initially used for functional neurosurgery (eg, lesioning the thalamus for cancer pain) (99); functional neurosurgery indications have expanded to include trigeminal neuralgia, movement disorders such as Parkinson disease and essential tremor, epilepsy, and certain affective disorders (128; 86; 12). Current primary indications for stereotactic radiosurgery include malignant tumors (tumors metastatic to the brain or spine, malignant gliomas), benign tumors (meningiomas, vestibular schwannomas, and pituitary adenomas), and arteriovenous malformations.

The terms stereotactic radiosurgery and stereotactic body radiotherapy are often used interchangeably. However, stereotactic radiosurgery, as sanctioned by the American Association of Neurological Surgeons, the Congress of Neurological Surgeons, and the American Society for Radiation Oncology, is strictly defined as the utilization of externally generated ionizing radiation to inactivate or eradicate defined target(s) in the head or spine (11) in one treatment fraction only. Stereotactic body radiotherapy is defined as the delivery of radiation to an extracranial target in multiple fractions, typically 2 to 5 (27). It operates under the same principles as stereotactic radiosurgery, with both practices using external beam radiotherapy with high precision and accuracy to deliver a biologically high radiation dose with locally curative intent (57). Stereotactic body radiotherapy is currently used to manage a wide array of disease sites, including oligometastases and spine, lung, prostate, liver, renal cell, pelvic, and head and neck tumors (109). For the treatment of oligometastases, stereotactic body radiosurgery can lead to high-dose irradiation of tumor sites, delaying the use of toxic systemic therapies and leading to longer survival without disease symptoms (21).

The Congress of Neurological Surgeons has updated its guidelines on the use of stereotactic radiosurgery for the treatment of adults with metastatic brain tumors (55). Although no level 1 evidence exists, stereotactic radiosurgery is recommended on the basis of level 3 evidence as an alternative to surgical resection in cases of solitary brain metastasis when surgery is likely to induce new neurologic deficits (55). If a metastatic brain lesion is treated with surgery, stereotactic radiosurgery is also indicated to reduce local recurrence rates (124). Results from clinical trials continue to support the use of stereotactic radiosurgery for patients with two to four brain metastases instead of whole-brain radiation therapy as long as their cumulative volume is less than 7 mL (55). Stereotactic radiosurgery alone also has been shown to improve median overall survival for patients with more than four metastases as long as the cumulative tumor volume is less than 7 mL (55). There is level 1 evidence, however, to demonstrate that in patients with two to three brain metastases not amenable to surgery, the addition of stereotactic radiosurgery to whole-brain radiation does not improve survival relative to whole-brain radiation therapy alone (07). Salvage stereotactic radiosurgery is not inferior to whole-brain radiation in patients with four or fewer brain metastases (80). Although not well studied in a controlled study, stereotactic radiosurgery for patients with 5 to 15 brain metastases has been shown to be well tolerated without an increase in toxicity, treatment failure, or need for salvage therapy (70). Patients with 10 or more brain metastases treated with LINAC radiosurgery also were able to maintain pretreatment neurocognitive function in the majority of cases (118).

There is also interest in using stereotactic radiosurgery preoperatively (144). This strategy has the theoretical advantage of preventing dissemination of tumor cells that may occur during surgery to prevent leptomeningeal spread. It also avoids the potential problem of defining residual disease postoperatively when image interpretation may be challenging. Preoperative radiosurgery has been associated with good local tumor control in retrospective studies. When preoperative radiosurgery was utilized for brain metastases, local control at 1 year was achieved in 84% to 91% of cases (08; 133; 172). Furthermore, in a multi-institutional study of preoperative radiosurgery for brain metastases, local recurrence rates at 1 and 2 years were found to be 15% and 17%, respectively. Median overall survival with preoperative radiosurgery was 16.9 to 17.2 months, with a 2-year overall survival rate of 36.7% to 38.4% when treated (140; 139). Rates of radiation necrosis and leptomeningeal disease were 5% and 4%, respectively (172). Histologically, brain metastases that were irradiated preoperatively were more likely to have evidence of tumor necrosis on resection than non-irradiated tumor, though there was no statistical association between the amount of tumor necrosis and the time interval from radiosurgery to surgery (89). When examining the rate of a composite endpoint of local failure, meningeal disease, and symptomatic radiation necrosis resulting from preoperative fractionated stereotactic radiation therapy, the composite endpoint event rate was 8% per patient, which is much lower than the reported composite rates after postoperative stereotactic radiosurgery (49% to 60%) (131).

These results indicate that preoperative radiosurgery may decrease the risk of leptomeningeal disease and radiation necrosis. Preoperative radiosurgery has the potential to become an effective treatment strategy, but it requires further investigation. Two additional studies are currently in progress to study the outcomes of using preoperative stereotactic radiation therapy for brain metastasis (53; 39).

Postoperative radiosurgery is conventionally used in the treatment of both brain and spinal metastases. The International Stereotactic Radiosurgery Society recommends stereotactic radiosurgery as a first-line treatment for postoperative brain metastases resection cavities, with a meta-analysis of 13 studies finding a median local control rate of 80.5% and improved neurocognitive outcomes versus whole-brain radiation therapy (141). Local recurrence rates increase with subtotal resection, larger treatment volume, lower margin dose, and a greater than 3-week delay between surgery and stereotactic radiosurgery (110). Fractionated radiotherapy may be more effective in patients with large tumor cavity volumes or diameters and a preoperative diameter greater than 2.5 cm (141). Postoperative stereotactic body radiotherapy is a safe and efficacious method for increasing local control of resected spinal metastases (04; 17). In a study of 63 patients with spinal metastases, postoperative radiotherapy resulted in a local control rate of 81% at a median follow-up of 12.5 months; local control rate was significantly improved by preoperative embolization and radiotherapy less than 40 days after surgical intervention (17).

Cancer immunotherapy is a rapidly developing and wide-reaching treatment strategy for local and metastatic cancer patients. As this field advances, it will undoubtedly converge with radiotherapy. Timing of immuno- and radiotherapy and indications for each treatment modality will continue to change as further knowledge regarding how these treatments interact emerges. The role of adaptive immunity in immunotherapy raises the possibility that radiotherapy could locally interact with the immune system by triggering the local production of inflammatory cytokines and release of tumor antigens. Radiotherapy may, therefore, have an important role in local treatment as a component of a systemic treatment strategy when combined with immunotherapy. Preclinical and early clinical studies are examining the safety and efficacy of local radiotherapy dosing, fractionation, and timing in combination with systemic immunotherapy (03).

To be considered candidates for radiosurgery, patients should have an established pathological diagnosis with a clearly visible lesion on radiographic imaging to facilitate accurate targeting of the lesion. Lesions with diameters of 4.5 to 5 cm have been treated by CyberKnife in staged procedures. However, most centers consider 2.5 to 3.0 cm to be the upper limit of radiosurgical treatment size. The proximity of the lesion to critical nervous system structures such as the anterior visual pathways, vestibulocochlear nerve, brain stem, and spinal cord must be taken into consideration when evaluating a lesion for treatment with stereotactic radiosurgery, as it may affect the dosage that can be delivered safely.

Contraindications

The principal contraindication to stereotactic radiosurgery is lesion diameter greater than 2.5 to 3.0 cm because radiation-induced complications increase rapidly with larger volumes treated. Major contraindications to stereotactic radiosurgery include the following:

Results

Assuming a competent planning team and stereotactic radiosurgery system, a good outcome depends primarily on the lesion’s sensitivity to radiation, shape, and location (ie, proximity to critical structures).

Brain metastases. Tumor stabilization or reduction occurs in 75% to 90% of cases (160; 64). A meta-analysis of 32 retrospective studies found that treatment of brainstem metastases with a median of 16 Gy resulted in an 86% local control rate with 2.4% treatment-related grade 3 to 5 toxic effects (30). Using mono-isocentric, non-coplanar, LINAC-based stereotactic technique minimizes the amount of radiation to healthy brain tissue, allowing for the safe and efficacious treatment of multiple brain metastases (65; 05). Hypofractionated radiosurgery was found favorable compared to single-fraction radiosurgery for lesions larger than 3 cm (45). Another analysis found both a reduction in radiation necrosis and an improvement in 1-year local control for hypofractionated versus single-fraction radiosurgery (97). A pooled analysis of the literature found lesions with a diameter smaller than 20 mm exhibited excellent 1-year local control rates, over 85%, with a single dose of radiation (18 Gy) (142). However, medium (20-30 mm) and large (> 30 mm) tumors had lower local control rates with such a small dose. As such, they may benefit from fractionated treatments to provide a higher dose without increased rates of radiation necrosis (117; 142). In the treatment of the resection cavity of brain metastases, adding a 2-mm margin improves local control without increasing toxicity and avoids whole-brain radiation (33). In a secondary analysis of a randomized controlled trial, radiosurgery was noted to have improved early local control relative to surgical resection, but the relative benefit lessened over time (36).

Spine metastasis. Efficacy has been demonstrated at all levels of the spine, with cervical spine localization referenced to skull base landmarks and thoracic and lumbosacral spine localization referenced to percutaneously placed fiducials (51). Patients with benign spinal tumors, such as meningiomas, neurofibromas, and schwannomas, were found to have comparable control rates relative to their intracranial counterparts (32). Radiosurgery combined with open surgery (vertebroplasty or kyphoplasty) has been utilized to successfully treat spinal cord compression as a result of metastasis (147). Treatment with a single dose of 12 to 24 Gy resulted in adequate control of local progression in spinal metastases with epidural cord compression (84). In an analysis of 54 patients treated with a median of 24 Gy in 3 fractions, local failure occurred in only 8% with a mean follow-up of 14.36 months, and pain improvement was seen in 76% (91). Metastases to the sacral spine had lower rates of local control and pain improvement than metastases to the thoracolumbar spine (90). A review of 67 studies found that single-fraction stereotactic body radiation therapy, whether performed preoperatively or postoperatively, resulted in 1- and 2-year local control rates of 71% to 95% and 70% to 96%, respectively (58). A majority (85%) of the lesions treated with single-fraction stereotactic body radiation therapy experienced partial or complete pain relief. Stereotactic body radiation therapy was also found to have an acceptable risk profile, as vertebral column fractures and radiation-induced myelopathies occurred in only 6% to 36.1% and in 0 to 10.8% of patients, respectively. In long-term follow-up, increased cumulative dose and point dose to the spinal cord or cauda equina was associated with late toxicity. Vertebral compression fractures were the most common and were observed in 17% of patients at both 5 and 10 years after treatment (105). However, patients who took antiresorptive agents, such as bisphosphonates, before stereotactic body radiation therapy had only a 4% risk of vertebral compression fractures 2 years after radiation, indicating this may be helpful to mitigate the risk of vertebral compression fractures after stereotactic body radiation therapy (134). Schaub and colleagues provide a review of literature that defines toxicity endpoints and actionable dosimetric guidelines to limit toxicity secondary to radiosurgery of the spine (154).

Vestibular schwannoma. Tumor stabilization or reduction occurs in more than 90% of cases (108; 63; 163; 76). In patients who underwent repeat stereotactic radiosurgery for recurrent tumors, tumor control rates at 2, 5, and 10 years were 98.6%, 92.2%, and 92.2%, respectively (74). Radiosurgery is ideal for small- and medium-sized vestibular schwannomas (106; 54). For larger tumors, surgical resection should be performed prior to radiosurgery, with a goal of reducing tumor volume to less than 6 cm³. However, in patients with larger tumors without life-threatening or debilitating symptoms, a stereotactic radiosurgery dose below 13 Gy without surgery was a safe and effective treatment for tumor control (127; 136). Tumor control rates of 92%, 91%, and 91% at 5, 10, and 15 years after radiosurgery were observed in a review of 618 patients (49). In the same study, hearing preservation rates of 53%, 34%, and 34% at 5, 10, and 15 years after radiosurgery were observed (49). In a series of 871 patients, hearing preservation rates were 89.8% at 1 year, 76.9% at 3 years, 68.4% at 5 years, 62.5% at 7 years, and 51.4% at 10 years after radiosurgery (76). Hearing preservation is correlated with the protection of the cochlea from radiation (68); tumor margin doses lower than 12 Gy and cochlear doses lower than 4 Gy maximize hearing preservation (167). Decreased rates of hearing preservation were associated with increased tumor size (23; 153) and the diagnosis of neurofibromatosis 2 (153).

Meningioma. The International Stereotactic Radiosurgery Society recommends stereotactic radiosurgery as an effective treatment for grade 1 meningiomas (111). In a review of studies of large intracranial meningiomas treated with stereotactic radiosurgery, radiographic tumor control ranged from 84% to 100% at follow-up of 2 to 7.5 years (48). However, only 45.1% showed clinical improvement; a two-fold increase in likelihood of improvement was found in non-single-session stereotactic radiosurgery versus single-session stereotactic radiosurgery (OR 2.47) (48). In the treatment of 37 patients with asymptomatic skull base meningiomas with single-session Gamma Knife radiosurgery, tumor size decreased, and there was no mortality caused by either the tumor or the procedure for a mean follow-up of 58.5 months (135). A review of 722 cases found that 94.8% of patients with petroclival meningioma exhibited tumor control following stereotactic radiosurgery, while having 5- and 10-year progression-free survival rates of 91% to 100% and 69.6% to 89.9%, respectively (14). Asymptomatic meningioma patients receiving stereotactic radiosurgery were found to have a tumor control rate of 99.4%, whereas conservatively managed patients had a rate of 62.1%. Importantly, stereotactic radiosurgery was not associated with an increased risk of developing new neurologic deficits (157). Similarly, intraventricular meningiomas were found to respond well to stereotactic radiosurgery; a small study of 11 patients found that 100% of tumors were locally controlled and 55% of tumors decreased in size (173). Radiosurgery may also be an effective treatment option for radiation-induced meningiomas (72).

Arteriovenous malformation. The International Stereotactic Radiosurgery Society cautiously recommends stereotactic radiosurgery as a safe and effective treatment for grade 1 to 2 arteriovenous malformations, particularly in deep or eloquent regions, based on analysis of 1102 arteriovenous malformations in which total obliteration without hemorrhage was achieved in 78% (56). Overall, radiosurgery reduces the risk of arteriovenous malformation hemorrhage, although the benefit is strongly driven by obliteration (44). Success is inversely proportional to the volume and flow of the arteriovenous malformation, with nearly complete success for arteriovenous malformations less than 1 cm in diameter, 75% to 90% success for lesions 1 to 3 cm in diameter, and 35% to 65% success for larger lesions (158). For very large arteriovenous malformations (larger than 10 cm3), a second radiosurgery or surgical intervention treatment within 3 years from the first may be necessary (187; 189). The use of endovascular embolization prior to stereotactic radiosurgery for arteriovenous malformation is also a controversial topic as there are mixed results on the effect of the combined treatments on obliteration (28; 75; 186; 47; 66). For patients with similar arteriovenous malformations, surgical resection and radiosurgery demonstrate similar rates of deficit-free obliteration. Obliteration of the arteriovenous malformation nidus was more frequently achieved with surgery but at the risk of increased new permanent neurologic deficit (29). Long-term obliteration rates with radiosurgery appear to be good as well; the 5- and 10-year nidus obliteration rates were 63% and 82%, respectively, in a large cohort of over 1200 patients (67). The annual hemorrhage rate was 1.5% for the first 5 years and 0.5% after that. Surgical treatment of an arteriovenous malformation frontloads the hemorrhage control and risk of neurologic deficit to the postoperative period, whereas risk of neurologic deficit after radiosurgery can be delayed. Incidence rates of adverse radiation effect at 10 and 15 years after treatment were 4.2% and 10.6%, respectively, so long-term care is necessary (67). In a review of eight studies of spinal arteriovenous malformations treated with 18 to 21 Gy over 2 to 4 fractions, good outcomes were seen in 92.2% with no post-treatment hemorrhage over an average follow-up period of 46.8 months (190).

Dural arteriovenous fistulas. In the treatment of 124 low-risk and 73 high-risk cases of dural arteriovenous fistulas, stereotactic radiosurgery was found to be a safe option with few complications (168). A cohort of 114 patients was treated with a mean margin dose of 21.8 Gy, resulting in obliteration rates of 41.3%, 61.1%, 70.1%, and 82.0% at 3, 5, 7, and 10 years, respectively (162). In a study of 131 patients, cavernous sinus arteriovenous fistulas were more likely to be obliterated than noncavernous sinus fistulas (OR = 4.189) (71). A review of 705 patients with dural arteriovenous fistulas receiving treatment found the complete obliteration rate to be 68.6%, with symptom improvement in 97.2% of patients (159). Fistulas that were not associated with the cavernous sinus and higher grade were less likely to have their symptoms respond to radiosurgery.

Hemangioblastomas. A quantitative meta-analysis of 14 studies of stereotactic radiosurgery use for CNS hemangioblastomas showed an 88.4% 5-year progression-free survival (132). In a large international study of 517 participants with a mean target volume of 0.2 to 0.7 cm3, local tumor control rate after stereotactic radiosurgery was 92% at 3 years, 89% at 5 years, and 79% at 10 years (188). A cohort of 15 patients with 101 lesions was treated with Gamma Knife stereotactic radiosurgery with a median pretreatment volume of 28 mm3 and a median marginal dose of 17.8 Gy. Ninety-seven percent of patients were free from new hemangioblastoma formation at 1 year, 80% at 3 years, and 46% at 5 years. Local failure occurred in only 4% of the treated lesions (103).

Pituitary adenomas. The International Stereotactic Radiosurgery Society recommends stereotactic radiosurgery as an effective and safe treatment for nonfunctional pituitary adenomas based on a meta-analysis of 27 studies that found 5- and 10-year random effects local control estimates after stereotactic radiosurgery (average 15.0 Gy) of 94% and 83%, respectively (88). Post-stereotactic radiosurgery panhypopituitarism occurred in 21% (88). Stereotactic radiosurgery has also been used to treat patients with acromegaly, with low-dose radiosurgery to 15.8 Gy leading to remission in 43.4% of 76 patients treated (130). Another study found similar endocrine remission in acromegaly with 83.3% local tumor control in 21 patients treated (06). In a separate meta-analysis of residual or recurrent tumor control, stereotactic radiosurgery with a mean radiation marginal dose of 19.6 Gy resulted in 95% overall tumor control and 67% hormonal control in secreting tumors, with few side effects (69). A meta-analysis found that stereotactic radiosurgery achieved tumor control in 97%, 92%, and 93% of cases of acromegaly, Cushing disease, and prolactinomas, respectively (114). However, endocrine remission was found in only 44%, 48%, and 28% of patients, respectively.

Chordoma and chondrosarcoma. Stereotactic radiosurgery is currently used in both primary tumor management of small tumors and restricting local progression following surgery (174; 79). A review of the National Cancer Database of 1478 patients with skull base, sacral, or mobile spine chordoma showed a 5-year survival rate of 82.3% with surgical resection and adjuvant radiotherapy versus 70.6% for surgical resection alone in patients with positive tumor margins; stereotactic radiosurgery resulted in higher survival rates than traditional external beam radiation therapy. There was no survival benefit in patients with negative margins (43). Long-term studies utilizing stereotactic radiosurgery with a mean marginal dose of 17 Gy found 5- and 10-year progression-free survival rates of 54.7% and 34.7%, respectively. Age greater than 65 at the time of stereotactic radiosurgery and presence of neurologic deficits before stereotactic radiosurgery were associated with lower local tumor control (137). Tumor volume also likely plays a role; patients with chordoma volumes less than 7 cm³ had significantly longer overall survival rates following gamma knife radiosurgery (22). Preoperative stereotactic body radiation therapy is also promising as all patients receiving preoperative stereotactic body radiation therapy before en bloc resection had negative margins (31).

High-grade glioma. The use of stereotactic radiosurgery in glioma is controversial. The standard treatment of high-grade gliomas includes surgical resection, external beam radiation, and adjuvant temozolomide. Most patients experience tumor progression within the initial tumor margin, and there is no standard treatment for this recurrence. For those who cannot undergo redo resection, stereotactic radiosurgery has been used for the treatment of small volume recurrence or poor surgical candidacy. Results are varied, presumably because of the infiltrative nature of gliomas; however, with refinements in targeted molecular agents, there may be a role for radiosurgery in the treatment of recurrent high-grade gliomas (15; 116). A completed randomized, controlled trial of the addition of stereotactic radiosurgery to conventional radiation and chemotherapy versus conventional radiation and chemotherapy alone did not improve outcome, quality of life, or cognitive function (161). A major cause of disease progression was beyond local failure outside of the stereotactic radiosurgery target in one single-center retrospective study (34). On the other hand, a phase 1 trial of stereotactic radiosurgery with bevacizumab in patients with recurrent glioblastoma using higher doses of stereotactic radiosurgery than the prior study, including only nine patients, demonstrated a median progression-free survival of 7.5 months and overall survival of 13 months, consistent with prior retrospective and prospective studies stating that stereotactic radiosurgery benefits treatment of high-grade gliomas (16; 01; 151). There is increasing interest in the use of stereotactic radiosurgery alongside immunotherapy (169). In a phase I clinical trial, 32 patients with recurrent high-grade gliomas were treated with bevacizumab, an anti PD-1 antibody, and hypofractionated stereotactic radiosurgery; progression-free survival was 7.9 months, and overall survival was 13.4 months (148). Use of stereotactic radiosurgery as a multimodal treatment for these heterogenous and complex high-grade tumors may be an ideal option in the future. Use of preoperative stereotactic radiosurgery is currently under investigation (98). The use of stereotactic radiosurgery and hyperfractionated radiosurgery for the treatment of newly diagnosed and recurrent glioblastoma has generated interest in the field despite the lack of prospective data (156). Overall, the use of stereotactic radiosurgery has been controversial in the treatment of primary or recurrent glioma because data on outcome improvement have been mixed.

Low-grade glioma. In regards to low-grade gliomas, safe surgical resection is the first choice of treatment, but stereotactic radiosurgery has been used for lesions that are too deep to access, particularly pilocytic astrocytomas and grade 2 astrocytomas. For pilocytic astrocytomas, stereotactic radiosurgery was used in both adults and children as postoperative adjuvant and in recurrent or unresectable lesions. Progression-free survival at 1, 3, and 5 years was 84%, 32%, and 32%, respectively, for adult patients with stereotactic radiosurgery, and those who underwent surgical resection had better outcomes (77). In pediatric patients, the progression-free survival was 92%, 83%, and 71% at 1, 3, and 5 years, respectively, with better outcomes in those with small residual solid tumors (78). Long-term progression-free survival in 44 patients receiving stereotactic radiosurgery for juvenile pilocytic astrocytomas at 1, 5, and 10 years was 95.4%, 79%, and 61.4%, respectively (180). In grade 2 astrocytomas, tumor control rates were 67% to 92% with stereotactic radiosurgery in single-center retrospective studies, with up to 41% noted to have radiation-related edema (82; 59; 179). Stereotactic radiosurgery may be an alternative or supplement to surgery in deep or difficult to access locations for these low-grade gliomas.

Other indications. Indications for stereotactic radiosurgery have continued to expand. Studies have demonstrated the utility of stereotactic radiosurgery for a variety of conditions. In a retrospective study consisting of 73 patients with medically refractory essential tremor, Gamma Knife thalamotomy was performed with a median central dose of 140 Gy (range, 130-150) (126). Overall, 93.2% of patients showed improvement in tremor with 60.3% reporting tremor cessation or barely perceptible tremor. Martinez Moreno and colleagues retrospectively analyzed the use of Gamma Knife radiosurgery (86.5 Gy [range: 80-90 Gy]; median: 90 Gy) in 117 patients with medically refractory trigeminal neuralgia (112). The pain-free rate without medications at 3, 5, and 7 years was 85%, 81%, and 76% of patients. Compared to microvascular decompression surgery, stereotactic radiosurgery was found to be less effective than surgery for increasing pain-free intervals (177; 102). Although the use of stereotactic radiosurgery for the treatment of glossopharyngeal neuralgia has not been extensively investigated, some studies suggest that it may be a safe and effective option (138; 18). A meta-analysis found that it had a lower incidence of pain relief but also a lower rate of complications than nerve section (107). Stereotactic radiosurgery resulted in pain reduction when a dose of 75 Gy or higher was used, but long-term rates of pain relief were less promising (13). Radiosurgery has also been compared to open surgical resection of the anterior temporal lobe for the treatment of mesial temporal sclerosis. Although surgery demonstrated an advantage in terms of proportion of seizure remission (78% vs. 52%), radiosurgery also appears to be safe and effective as an alternative for patients who are unable to undergo open surgery (10). In the treatment of epilepsy-causing hypothalamic hamartomas, radiosurgery is a safe and efficacious alternative to open microsurgery (19).

Adverse effects

Adverse reactions to stereotactic radiosurgery can be defined as acute (days), early (weeks to months), or late (> 3 months). Both acute and early reactions are usually self-limiting with no permanent neurologic injury.

Nausea and vomiting are common acute reactions after treatment of posterior fossa lesions, with symptoms occurring in 10% to 16% of patients within 6 hours of treatment. Pretreatment with antiemetics and steroids is useful in minimizing or preventing these symptoms.

Much of the pathology treated with stereotactic radiosurgery has an inherent seizure association. After stereotactic radiosurgery, the risk for seizure is probably elevated for the first 48 hours. Maintenance of therapeutic levels of appropriate antiepileptic drugs is recommended to prevent seizures during the immediate post-treatment period. Further complications that may occur include vertigo, paresthesias, and trigeminal neuralgia (24).

Late reactions to radiosurgery are associated with variable degrees of transient and permanent neurologic sequelae that occur 6 to 9 months after treatment and may have visible changes on CT or MRI (95). In normal tissue near the lesion or in the treatment area, the high dose of radiation may cause blood-brain barrier alterations, cytotoxic edema, myelin lesions, atrophy, calcification, cyst formations, thrombosis of regional vessels (generally in the case of tortuous veins, varices, or venous anomalies), and vasculopathy. Tumoral cytotoxic edema in combination with post-treatment edema may create problems with increased intracranial pressure. These often present several months after stereotactic radiosurgery, with evidence of headaches and perilesional edema on neuroimaging studies. Treatment consists of corticosteroids and diuretics. The condition is usually self-limiting but particularly on occasion when large volumes have been treated with stereotactic radiosurgery, these areas may require surgical decompression. Treatment consists of steroids, often requiring a protracted course of 12 months or longer for resolution. Severe cases may require surgical resection of necrotic tissue to reduce mass effect (164). Data from the treatment of arteriovenous malformations suggest that repeat radiosurgery treatments and deep location of a lesion are risk factors for symptomatic radiation-induced changes (73). Finally, pituitary insufficiency is a potential complication of stereotactic radiosurgery for lesions adjacent to the pituitary gland or hypothalamus.

Though rare, radiation-induced vasculopathy may perforate key branches of cerebral vasculature surrounding targets of radiosurgery, such as the middle cerebral artery (83). Visual deficits can be encountered with targeting areas surrounding the post-geniculate visual pathway, but they are relatively rare, with the risk of sustaining a deficit being 3% at 3 years, 5% at 5 years, and 8% at 10 years (20).

Radiation necrosis is another severe, late complication that occurs in 5% to 26% of cases and can result in symptomatic edema and reduced quality of life (40; 175). In the treatment of brain metastasis, radiation necrosis typically appears between 3 and 9 months following stereotactic radiosurgery, although it can appear up to 5.3 years after treatment (50; 152). Increased risk of radiation necrosis is associated with larger tumor diameter, increased radiation dosage, and advanced age (81). Treatment options for radiation necrosis include observation for asymptomatic lesions and steroids, vitamin E, vascular endothelial growth factor inhibitors (bevacizumab), and surgery for symptomatic lesions (26). Steroids are often the first line of treatment, but they result in numerous side effects when taken at high doses for an extended period of time, which may be required in radiation necrosis. Laser interstitial thermal therapy (LITT) has been used to reduce steroid use in patients with symptomatic radiation necrosis (152). Low doses of bevacizumab have been effective at treating symptomatic radiation necrosis after stereotactic radiosurgery. Bevacizumab reduced the radiation necrosis lesion volume and edema while also improving symptoms and neurologic function (101; 181). Additionally, it is important to note that differentiating between tumor progression and radiation necrosis following stereotactic radiosurgery can be very challenging (93). Occasionally tissue sampling is required to differentiate between the two.

Treatment of lesions of the skull base may result in cranial neuropathies. Sensitivity to radiation injury varies among cranial nerves, with the optic nerve being the most sensitive (single-fraction threshold for injury of approximately 800 cGy) (87). However, specific contouring and targeting strategies that avoid the optic nerve and pathways are successful at avoiding this complication. The cochlear and trigeminal nerves are the next most frequently injured nerves.

Malignant transformation or secondary malignancy following stereotactic radiosurgery is rare (115). Transformation of grade 1 meningiomas to high-grade meningioma and new diagnoses of post-radiation glioblastoma, osteosarcoma, and chondrosarcoma following stereotactic radiosurgery have been reported (129; 96; 92). The time from initial radiosurgery to detection of the secondary malignancy ranged from 58 months (glioblastoma) to 14 years (chondrosarcoma). Of 14,168 patients treated with stereotactic radiosurgery for a variety of intracranial pathologies, the cumulative incidence of radiation-associated malignancy was 0.045% over 10 years, which is similar to the risk of independently developing a malignant CNS tumor (183).

Prognosis varies with the pathological diagnosis and neuroanatomical location of the lesion, as discussed above.

Special considerations

Using stereotactic radiosurgery to treat a woman harboring a brain lesion during pregnancy is feasible because, with abdominal protection, the fetal radiation exposure can be limited to less than 1 rad. Thoracolumbar spinal lesions, however, represent a situation where extremely careful stereotactic radiosurgery treatment planning is essential to minimize fetal radiation exposure. Elderly patients harboring brain lesions or cranial neuralgias may be candidates for stereotactic radiosurgery because it carries lower risks compared to surgical alternatives (37; 119; 138). Special considerations should be given in this patient population, however, due to increased risks of neurologic toxicity (119). Discussions on the goals of care and a neurologic assessment should be performed (119).

Clinical vignette

Case 1. A 59-year-old female presented to the neurosurgery clinic with a known diagnosis of invasive lung adenocarcinoma. She developed a severe headache and difficulty with balance 1 month prior to presentation, and work-up demonstrated two new intracranial metastases. She was found to have a dominant 2.6 x 1.8 x 2.2 cm left lateral cerebellar hemisphere lesion with surrounding perilesional edema as well as a much smaller right cerebellar lesion. She was started on dexamethasone, and her symptoms improved. She underwent stereotactic radiosurgery to both lesions, 19 to 22 Gy in a single fraction. Later that same day she underwent craniotomy for tumor resection of the left lateral cerebellar lesion. She tolerated surgery well and was discharged to home on postoperative day 2. Postoperatively, she remained neurologically intact. At her 3-month follow-up, her right cerebellar lesion was no longer present, and a new area of contrast enhancement was seen within the resection cavity. This was favored to represent radiation effects over tumor progression due to imaging characteristics. She continues to be followed by neurosurgery, radiation oncology, and medical oncology as an outpatient.

Case 2. A 41-year-old female developed progressive and disabling spastic quadriparesis over 15 years--Brown-Sequard syndrome with significant loss of right-hand function in particular. Spinal arteriogram demonstrated a C2-C3 spinal cord arteriovenous malformation, which had hemorrhaged on three occasions. Due to the intramedullary location of the arteriovenous malformation, the patient was not a good candidate for microsurgical resection or embolization. Initial stereotactic radiosurgery was with the CyberKnife (10 mm collimator, 3 fractions, total peripheral dose 21 Gy). Clinically, the patient steadily improved, in particular her gait and right-hand function. Spinal arteriogram at 3 years post-stereotactic radiosurgery showed 70% obliteration of the arteriovenous malformation. She underwent a second CyberKnife stereotactic radiosurgery (7.5 mm collimator, 2 fractions, 85% isodose line total 15 Gy). MRI at 4 years following initial stereotactic radiosurgery showed continued improvement, which paralleled her clinical course.

Case 3. A 76-year-old man presented with a 5-year history of progressive hearing loss in the left ear. An MRI revealed a 12 x 15 x 9 mm contrast-enhancing lesion in the cerebellopontine angle that, on close inspection, followed the course of the facial nerve. The lesion did not compress the brainstem and he demonstrated intact facial function. The patient had no history of neurofibromatosis. Over time, the lesion continued to demonstrate a slow, linear growth pattern. The patient was referred for stereotactic radiosurgery and received a total dose of 12 Gy to the 80% isodose surface with a total treatment tumor volume of 1.77 cm3. The patient tolerated the procedure well and the tumor decreased in size on 1-year interval follow-up imaging. He continues to have intact facial function. Central to this case is the recognition that the cerebellopontine angle lesion was likely a facial nerve schwannoma and not a vestibular schwannoma. Open surgical resection would expose the patient to a high risk of facial nerve dysfunction. Although this is also a risk with stereotactic radiosurgery, the patient has not experienced this complication and continues to do well clinically.

Case 4. A 77-year-old woman presented to the emergency room with confusion, difficulty concentrating, and short-term memory loss. An MRI was ordered, which demonstrated a large left frontal parafalcine contrast-enhancing mass with surrounding vasogenic edema most concerning for a meningioma. She underwent a bifrontal craniotomy for tumor resection, which was complicated by significant cerebral edema, necessitating early closure and subtotal resection. Pathology confirmed a grade 1 meningioma. Initially, she did quite well after this surgery and returned to her neurologic baseline. She was observed on an outpatient basis with slow growth of the tumor. About 1 year after surgery, she developed confusion again and was noted to have rapid growth of her meningioma. She returned to the operating room for another gross total resection of her meningioma, which resulted in memory improvement yet again. She was followed on an outpatient basis, and 5 years after her second surgery was noted to have progression of a new enhancing small lesion within the resection cavity. Stereotactic radiosurgery, 12 Gy in one fraction, was recommended given her advanced age. Three years after stereotactic radiosurgery, the size of her meningioma remained stable and she remained neurologically intact.

Scientific basis

The goal of radiation therapy is to destroy target cells through the delivery of lethal doses of high-power energy. Radiation beams impart energy to all tissues traversed, invariably damaging surrounding nontarget (normal) tissue, resulting in complications such as impaired wound healing, injury to skin, arteries, and the central or peripheral nervous system, and the induction of secondary malignancies.

Stereotactic radiosurgery seeks to deliver a high dose of radiation to the target while minimizing radiation to surrounding normal tissues. A heavy particle beam accurately delivers much of its energy at a depth predictable by Bragg-peak calculations. Gamma Knife stereotactic radiosurgery relies on multiple sources of cobalt 60 high-energy photons, collimated to limit beam divergence, to focus the radiation on a target. In addition to movement of the gantry and the patient, LINAC employs collimated high-energy photons in a series of converging arcs to focus the radiation on a target. Other LINAC techniques include the use of multiple converging shaped beams (“shaped-beam radiosurgery”), intensity-modulated fixed beams (IMRS), and intensity-modulated arc therapy (IMAT).

The frame-based stereotactic radiosurgery techniques generate an isocentric spherical region of high-dose volume (184). Nonspherical or irregularly shaped targets are treated with multiple isocenters. Multiple isocenters create central necrosis of the target, surrounded by a thin intermediate area of less severely damaged tissue (penumbra); the thinner the penumbra, the more accurate the targeting. A sublethal radiation dose causes vascular intima hyperplasia leading to capillary occlusion, allowing for obliteration of arteriovenous malformations and tumor neovascularity (42).

The CyberKnife delivers non-isocentric spherical radiation beams to lesions that are nonspherical in shape. Mounting the LINAC on a robotic arm created an unconstrained delivery system, making the CyberKnife free to generate beams that originate at arbitrary points within the workspace and target arbitrary points within the lesion.

There is no qualitative difference between cobalt 60 gamma radiation and high-energy x-rays, as relative radiobiological efficiency is equivalent in both. Accelerators derive a theoretical advantage from deeper tissue penetration, especially in 10 to 18 MeV. This limits adherence problems of superficial planes, which are characteristic after cobalt 60 therapies. At present, there is no clinical difference between results obtained with either method.