Stereotactic Radiosurgery

Updated: Mar 19, 2015
  • Author: Simon S Lo, MD; Chief Editor: David C Spencer, MD  more...
  • Print


Stereotactic radiosurgery (SRS) refers to the precise and focused delivery of a single, high dose of radiation in a single session and has been used to treat various intracranial and skull base lesions. The dosimetric characteristics of SRS, namely a highly conformal isodose distribution and a very steep dose gradient of dose fall-off beyond the prescribed isodose line, lend themselves well to the delivery of an ablative dose of radiation to intracranial and skull base lesions.

In radiobiology, tissues are divided into 2 broad categories, namely, early- and late-responding tissues. Early-responding tissues, such as skin, mucosa, and gastrointestinal epithelium, tend to respond acutely to radiation exposure, whereas radiation-induced effects are not immediately observed in late-responding tissues, such as vascular tissue, nerves, brain parenchyma, and spinal cord. Most malignant tumors behave like early-responding tissue, whereas benign tumors behave like late-responding tissue. [1] Late-responding tissues are more susceptible to a single, high dose of radiation compared with early-responding tissues, and this factor has to be considered in the delivery of SRS.

Table. Target Separation into 4 Different Categories (Open Table in a new window)

Category Target Tissue Target Contains Normal Brain Tissue Examples
1 Late-responding tissue Yes Arteriovenous malformation (AVM)
2 Late-responding tissue No Meningioma, acoustic neuroma
3 Early-responding tissue Yes Low-grade glioma
4 Early-responding tissue No Brain metastases, glioblastoma multiforme
Adapted from Larson et al (1993) [1]

SRS is performed by using 1 of the 3 forms of high-energy radiation. SRS can be delivered with a linear accelerator, a Gamma Knife unit, or with charged particles. In this article, Gamma Knife–based SRS is emphasized.The Gamma Knife device was invented by Leksell, a Swedish neurosurgeon, in Sweden in 1951. It was the first device used to deliver SRS. The first Gamma Knife device was installed in the Karolinska Institute, Sweden in 1967. The device has been modified and upgraded throughout the past few decades. Previous versions contained 201 cobalt-60 sources. The current version is PerfexionTM (Elekta Instrument AB, Stockholm, Sweden), which contains 192 cobalt-60 sources. A stereotactic head frame compatible with MRI is used for immobilization of the head.



Gamma Knife–based SRS is, in general, not suitable for tumors 4 cm or larger in diameter or for those immediately adjacent to eloquent structures such as the optic apparatus and brainstem. A 3- to 5-mm clearance from the optic pathway is usually needed for a lesion to be treated with SRS. The most commonly quoted tolerance for the optic apparatus is 8 Gy, but a study from the Mayo Clinic showed that a point dose of 12 Gy to the optic apparatus might be safe.

In this section, the term SRS will be used instead of Gamma Knife–based SRS, since several studies included patients treated with linear accelerator–based SRS for the same indications. Gamma Knife should be regarded as a special treatment device that delivers SRS, rather than a separate treatment modality of its own. Data in the literature show similar outcomes for each individual condition treated with SRS using different treatment devices.

Arteriovenous malformation

Arteriovenous malformation (AVM) can be managed by microsurgery, endovascular embolization, or SRS. Microsurgery is regarded as the standard first-line treatment, but SRS is offered to patients who have residual AVM after microsurgery or those who are deemed not to be good candidates for microsurgery. AVMs are regarded as category 1 targets for SRS, as described in the above table. Target delineation is done using a combination of angiography and stereotactic MR or CT imaging. SRS to a target volume containing the AVM nidus induces progressive thrombosis of lesions via fibrointimal hyperplasia and subsequent luminal obliteration. These events usually take 1-3 years to occur, and the time between treatment and obliteration is referred to as the latency period. Once the lesion is completely obliterated, the risk of hemorrhage from the AVM is very low, but not eliminated. [2, 3, 4, 5, 6, 7, 8]

Evidence concerning the hemorrhage risk during the latency period between treatment and obliteration is conflicting. Reports show unchanged, decreased, or increased bleeding rates during the latency period. A study of 500 patients showed that the AVM hemorrhage rate following SRS was reduced by 54% during the latency period and by 88% after obliteration. The benefit was greater in patients who presented with hemorrhage compared with those who presented without hemorrhage. [9]

Successful AVM obliteration with SRS depends on lesion size and radiation dose. An overall 80% obliteration rate by 3 years occurs with lesions 3 cm or smaller, while larger lesions have obliteration rates of 30-70% at 3 years. [6, 8, 9, 10, 11, 12, 13] However, some amount of lesion volume reduction (mean, 66%) typically occurs in larger lesions (>3 cm) treated with SRS, and retreatment is effective in about 60% of patients with residual AVMs. [14, 15] A dose response has been demonstrated for radiographic AVM obliteration, with doses of 16, 18, and 20 Gy associated with obliteration rates of about 70%, 80%, and 90%, respectively. [15, 16, 17]

Treatment of AVMs with SRS may improve seizure control in patients with comorbid epilepsy. In a study of 65 patients who had seizures before SRS for AVM, 51% were free of seizures and auras 3 years after treatment. [18] Of the 23 patients with disabling seizures before treatment, 61.1% had an excellent outcome (defined as seizure-free or nondisabling seizures) at 3-year follow-up. A low frequency of seizures before treatment and smaller size AVM predicted better outcomes.

The incidence of complications is related to the AVM location, the volume treated, and the radiation dose, mainly based on Gamma Knife SRS experience. Thalamic, basal ganglionic, and brainstem locations are particularly prone to the development of deficits after SRS. [19, 20] The risk of radiation necrosis with permanent neurologic deficit is approximately 2-3%. [10, 12, 15] In a very large study of patients undergoing SRS for cerebral AVMs, treatment-related complications developed in 8% of the patients and included radiographic parenchymal lesions, cranial nerve deficits, seizures, headaches, and cyst formation. [21]

Trigeminal neuralgia

Trigeminal neuralgia (TN) is defined by the International Association for the Study of Pain (IASP) as sudden, usually unilateral, severe, brief, stabbing or lancinating, recurrent episodes of pain in the distribution of one or more branches of the trigeminal nerve. Treatment options include medications, surgical procedures including microvascular decompression, radiofrequency rhizotomy, glycerol rhizolysis and balloon compression, and SRS (Gamma Knife– or linear accelerator–based radiosurgery). SRS targets the proximal trigeminal root. MRI is used for targeting. The maximum doses used range from 70-90 Gy and the goal is to cause axonal degeneration and necrosis. [22] One study found that treatment response did not correlate with the dose of radiation, but a study from the Mayo Clinic showed that a higher dose of radiation resulted in better facial pain control, although with increased risk of trigeminal nerve dysfunction. [23, 24]

Pain relief with SRS usually occurs after a lag time of about 1 month. [22, 25] Lopez et al performed a systematic review on SRS for primary trigeminal neuralgia and found that approximately 75% of patients reported complete relief within 3 months, but this rate decreases to 50% by 3 years. [26] Less than 50% of patients could permanently stop drug therapy after SRS. Sensory disturbances are the most frequent complications. New or worsened facial sensory impairment occurred in 20% of patients in one series, with a median lag time of 35 months after SRS. [27] Patients who do not respond to radiosurgery or who have recurrence of symptoms may respond to repeat SRS. [28, 29] Linear accelerator–based SRS appeared to yield results similar to Gamma Knife–based SRS. [30, 31]

These treatment outcomes are similar to those achieved with other ablative treatments. However, the efficacy of SRS should be studied in a randomized, controlled trial, and long-term treatment results are needed.


Meningiomas constitute approximately 20% of all primary intracranial tumors, and tumor grade bears clear prognostic importance. Complete surgical resection, if feasible, is the standard of care. However, meningioma may recur after surgery, and safe curative resection is often not possible, especially for skull base tumors. SRS can deliver a single, high dose of radiation to a localized area in the brain for the treatment of meningiomas that are unresectable, recurrent, or residual after surgery. Meningiomas are category 2 targets and extra-axial tumors with a sharp demarcation from normal brain parenchyma. Therefore, they are ideal targets for SRS. Data on SRS for the treatment of meningioma have emerged over the last 2 decades and the overall reported local control ranges from 75-100%. [32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51]

Radiation-induced necrosis or edema may occur after SRS for meningioma. In an early SRS series from Heidelberg, the reported radionecrosis/edema rate was about 30% with the mean dose being 29 Gy (range, 10-50 Gy). [35] More recent series using lower SRS doses (20 Gy or less) yielded much lower radionecrosis/edema rates, which were typically a few percent. [32, 42, 47] Patients with meningiomas of the convexity, parasagittal region, or falx cerebri may have a higher incidence of peritumorous imaging changes after SRS than those of the skull base. [48]

Patients with meningiomas close to eloquent structures such as the optic pathway and the brainstem (especially in skull base locations) or large meningiomas are not suitable for SRS due to the normal tissue constraints. Stereotactic radiotherapy using 2-5 fractions is often offered to these patients as an alternative to conventional fractionated radiation therapy.

Acoustic neuroma

Acoustic neuroma is also known as acoustic schwannoma, acoustic neurinoma, vestibular schwannoma, and vestibular neurilemoma. It is a Schwann cell–derived tumor arising from the vestibular portion of the eighth cranial nerve. Treatment options include observation, microsurgery, SRS, and stereotactic radiotherapy. [52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63] A few series have compared the outcomes of microsurgery and SRS and showed that SRS yielded similar tumor control rates but enabled better preservation of cranial nerve function. [52, 53, 61]

SRS is a viable treatment option for selected patients with smaller tumors (< 3 cm) or for enlarging tumors in patients who are not candidates for surgery. Acoustic neuromas are category 2 targets and therefore they are ideal targets for SRS. One of the largest series of 829 patients on SRS, Gamma Knife–based, for acoustic neuroma from the University of Pittsburgh showed a tumor control rate of 97%. [54] An older series from the same group consisting of patients treated to SRS doses of 18-20 Gy showed greater than 20% trigeminal and facial neuropathy. [55]

Recent series from various centers that used low doses of 12-13 Gy showed similar tumor control rates but better hearing preservation and much lower rates of trigeminal and facial neuropathies. [54, 56, 62, 63] However, longer follow-up of patients treated with this low-dose approach is necessary to ascertain that this is an adequate treatment, given the indolent nature of this disease. SRS also appears to be a reasonable treatment alternative for patients with acoustic neuroma associated with neurofibromatosis type 2 (NF-2).

SRS confers a significant advantage over the natural history of the disease, although the hearing preservation rate in those patients is lower than that of patients without NF-2. [57, 58] Subach et al reported a tumor control rate of 98% in patients with NF-2. [58]

SRS provides excellent tumor control, and, with the use of low doses of 12-13 Gy, the hearing preservation rates are significantly improved and the rates of trigeminal and facial nerve deficits are significantly decreased. The hearing preservation rates are lower for patients with NF-2.

Pituitary adenoma

Pituitary adenomas are histologically benign tumors, but the potential neurological and physiological complications can be devastating. In patients with acromegaly, excessive growth hormone production can lead to life-threatening cardiovascular and respiratory complications, diabetes mellitus, and, possibly, an increased risk of colon cancer. [64] In patients with Cushing disease, prolonged hypersecretion of ACTH can lead to uncontrolled hypertension and osteoporosis. [64]

Galactorrhea and infertility may occur in patients with prolactinomas, as a result of hypersecretion of prolactin. Treatment options include microresection, medical therapy, fractionated radiotherapy, SRS, and stereotactic radiotherapy. SRS has been used in the treatment of endocrine-inactive as well as secretory pituitary adenomas. SRS is suitable for patients with a gap of at least 2-5 mm between the pituitary adenoma (≤3-4 cm in diameter in general) and the optic pathway. [64]

For endocrine-inactive tumors, margin doses of 14-25 Gy were used for SRS, which yielded tumor control rates of 92-100% in various studies with follow-up times ranging from 16-58 months. [64] For patients with acromegaly, margin doses of 15-25 Gy were used for SRS, which yielded tumor control rates of mostly greater than 90%. [64] However, the endocrine improvement rates were much lower and were in the range of 20-82%. While SRS for prolactinoma and Cushing disease yielded very similar results in terms of tumor control (in general >90%), the rates of endocrine normalization ranged from 0-100% in patients with prolactinoma and 10-100% in patients with Cushing disease. [64]

For secretory pituitary adenomas, a decrease in hormone hypersecretion could be seen as early as 3 months but could extend up to 8 years after SRS. If normalization is going to occur, it frequently does so within the first 2 years. [64] Some studies suggested that there might be radiation dose-response in terms of hormone normalization. The reported toxicities associated with SRS for pituitary are, in general, low. In a systematic review of the literature (1255 patients in total), the rate of optic neuropathy was 0.9%. The rate of permanent cranial nerve (III, IV, V, and VI) deficits was 0.4%. New trigeminal neuropathy was reported in only 0.2% of the patients. [64]

Brain metastases

Brain metastases tend to be spherical and have a sharp demarcation from normal brain parenchyma. These characteristics are ideal for SRS because spherical dose distributions can readily be generated by the radiosurgical systems and the use of tight margins is feasible. [65] Brain metastases are regarded as category 4 targets where the radiologically defined targets contain only tumor cells, which are an early-responding tissue. Compared with surgical resection, SRS has the advantage of being able to treat surgically inaccessible lesions and multiple lesions. In general, lesions smaller than or equal to 3-4 cm are regarded as suitable for SRS. Local tumor control rates with SRS are consistently greater than 80%. The SRS doses used by the Radiation Therapy and Oncology Group (RTOG) for lesions smaller than or equal to 2 cm, 2.1-3 cm, and 3.1-4 cm were 24 Gy, 18 Gy, and 15 Gy, respectively. [66, 67]

The American Society for Therapeutic Radiology and Oncology (ASTRO) published an evidence-based review of the role of SRS in the management of brain metastases. There were 3 randomized controlled trials and 7 retrospective series for patients with newly diagnosed brain metastases, treated with whole-brain radiotherapy alone versus whole-brain radiotherapy and SRS boost. One of the conclusions was that for patients with up to 3(< 4 cm) newly diagnosed brain metastases (and in one study up to 4 brain metastases), SRS boost with whole-brain radiotherapy significantly improves local brain control rates as compared with whole-brain radiotherapy alone. [68]

In a large phase III trial, an overall increased ability to taper steroid dose and an improvement in Karnofsky performance status was observed in patients treated with SRS boost compared with patients treated with whole-brain radiotherapy alone. Improved local control was shown with the addition of SRS to whole-brain radiotherapy in all randomized trials. However, there is no overall survival benefit with the use of SRS boost to whole-brain radiotherapy for patients with multiple brain metastases.

There were 2 randomized trials, 2 prospective cohort studies, and 16 retrospective series evaluating patients treated with SRS alone as initial treatment. There is level I to level III evidence to suggest that the use of SRS alone does not alter survival compared with the use of whole-brain radiotherapy. However, there is level I to level III evidence that omission of whole brain radiotherapy results in poorer intracranial disease control. [68]

SRS as salvage for patients with brain metastases was reported in one prospective study and 7 retrospective series. Radiographic response was well documented. The use of SRS is associated with brain tumor response and 1-year survival rates ranging from 26-40%.

SRS is associated with a small risk of early or late toxicity in all series. In a phase III randomized trial comparing SRS and SRS combined with whole-brain radiotherapy for patients with 1-3 brain metastases from the M.D. Anderson Cancer Center, the addition of whole-brain radiotherapy resulted in decreased intracranial failure but did not result in improved survival. Whole-brain radiotherapy also resulted in a much higher incidence of decline in learning and memory function by 4 months. [69] The ASTRO brain metastasis taskforce is in the process of generating a guideline document on radiotherapy and radiosurgery for brain metastasis(es).

Malignant glioma

Malignant glioma is highly lethal. Despite best treatment, nearly all patients eventually succumb to their disease. Two randomized studies showed a radiation dose response for survival. Because the majority of recurrences occur within 2 cm of the enhancing edge of the original tumor, the possibility of dose escalation to improve tumor response and potentially survival has been explored. Multiple strategies, including SRS, have been used to execute dose escalation. Malignant gliomas are category 4 targets, where there is no normal brain parenchyma within the radiologically defined targets, and are suitable for SRS. Patients with larger tumors (in general, diameter >4 cm) or tumors close to eloquent structures may not be suitable for SRS. [70]

Data on SRS as boost therapy after external beam radiation therapy have been emerging in the literature. The role of SRS in recurrent malignant gliomas has also been studied. The ASTRO published a systematic review of the evidence for the use of SRS in adult patients with malignant glioma. [70] There was one randomized trial, 5 prospective cohort studies, and 7 retrospective series examining patients with newly diagnosed malignant glioma treated with SRS as boost therapy with conventional external beam radiation therapy. There is level I evidence to show that the use of SRS boost followed by external beam radiotherapy and carmustine (BCNU) does not confer benefit in regard to overall survival, quality of life, or patterns of failure compared with external beam radiotherapy and BCNU alone. [70]

There is level I-III evidence of toxicity associated with radiosurgery boost compared with external beam radiotherapy alone. Use of SRS as salvage therapy for recurrent or progressive malignant glioma after conventional external beam radiotherapy failure was reported in 3 prospective cohort studies and 5 retrospective series. The reported median survival ranged from 6-12 months. [70] Overall, there is insufficient evidence to support a survival benefit in the use of SRS for progressive or recurrent malignant glioma compared with competing treatment strategies such as debulking surgery, chemotherapy, or best supportive care. [70]

Glomus tumor

Glomus tumors are locally aggressive tumors that are often difficult to resect. Glomus tumors are ideal targets for SRS since they are well demarcated and are readily identified on MRI. Glomus tumors are most likely composed of late-responding tissue, which should be more responsive to a single ablative dose of radiation delivered via SRS.

Data from the literature show promising results with the use of SRS for glomus tumors. [71, 72, 73, 74] Reported tumor control rates ranged from 63-100% and complication rates from 4-40%. The median follow-up intervals of those studies were 20.5-51 months. In a meta-analysis from University of California, San Francisco, patients (n=97) treated with subtotal resection plus SRS for glomus tumors had a tumor control rate of 71% at a median follow-up interval of 96 months, whereas patients (n=337) treated with SRS alone had a tumor control rate of 95% at a median follow-up interval of 71 months.

Rates of IX, X, XI, and XII nerve deficits for patients who underwent SRS alone were 9.7%, 9.7%, 12%, and 8.7%, respectively, and were much lower than the corresponding rates for patients who underwent gross total resection. [71] Another meta-analysis from Johns Hopkins University showed similar findings. [72]

Low-grade astrocytoma

The standard treatment for low-grade astrocytoma is surgery and/or conventional radiotherapy. SRS has been used to treat low-grade astrocytoma in the recurrent, boost, and primary setting. Low-grade astrocytomas are perceived not to be good targets for SRS because the tumor is an early-responding tissue with intermingling normal brain parenchyma, which is a late-responding tissue, rendering a higher risk of radiation injury to the normal brain parenchyma than to the tumor tissue from an ablative dose of radiation. [1] However, data in the literature, albeit limited, do not seem to corroborate the theoretical risk. Various retrospective series of SRS for low-grade astrocytoma show local tumor control rates ranging from 67-90% with low incidence of treatment-related toxicities. [75, 76, 77, 78]

For low-grade astrocytoma, SRS most likely has the strongest indication in the recurrent setting, especially when there are no other local therapy options, as in previously irradiated patients within operable recurrent disease, and more study is needed to define its role in other settings.


The standard treatment for intracranial ependymoma is maximal safe resection followed by postoperative radiotherapy. The extent of surgery is the most important prognostic factor. Ependymoma is a suitable target for SRS because the brain parenchyma–tumor interface is sharp and there is no late-responding tissue in the tumor. SRS has been used to treat intracranial ependymoma mainly in the recurrent setting, and to a lesser extent, in the adjuvant or primary setting. [79]

Patients with recurrent intracranial ependymoma frequently have prior history of radiotherapy, and their options are frequently limited to surgery or chemotherapy. However, a complete resection is not always feasible, and the presence of gross residual disease is associated with a high risk of disease progression. Chemotherapy alone has limited success with recurrent ependymoma. SRS provides an attractive treatment option for patients with recurrent disease.

Data in the literature show mixed results in terms of local control. Most series have a small number of patients, so it is difficult to draw a firm conclusion on the efficacy of SRS for recurrent ependymoma. [79] In one of the relatively larger series from the Mayo Clinic, 12 patients with 17 recurrent intracranial ependymomas were treated with Gamma Knife–based SRS to a median dose of 18 Gy. The reported 3-year local control rate was 68%. [79]

A radiation dose response in terms of local control has been reported for patients with gross residual disease in ependymoma. SRS has been used as a boost after external beam radiotherapy. In a pooled analysis of 13 patients, mostly children, who received SRS as a boost after external beam radiotherapy, 11 were alive with disease at a median follow-up of 40 months. [79]

SRS is seldom used as the sole treatment of intracranial ependymoma. There are scattered reports with mixed results. This is most likely related to differences in patient selection and the widely varying doses used in various series.

Uveal melanoma

Historically, uveal melanomas were primarily managed by enucleation. Eye plaques (using iodine-125) and proton beam therapy have since been established as treatment alternatives for the disease. Use of Gamma Knife–based SRS has been explored in uveal melanoma. [80, 81, 82]

Several retrospective studies reported mixed results. [80, 81, 82] This may be related to differences in patient selection and the widely varying doses used in various series. Some series included patients with large uveal melanomas who would not have been eligible for eye plaque therapy. In a study from Indiana University, where 19 patients were treated with Gamma Knife–based SRS to a uniform dose of 40 Gy, the 5-year local tumor control rate was 94.4% at a median follow-up interval of 40.1 months. [81] Three had improved, 4 had stable, and 12 had worse visual acuity in the treated eye by formal ophthalmologic examination. Two of the 12 patients with decreased visual acuity did not notice any difference in vision subjectively. Two, one, and one patients developed vitreous hemorrhage, vitreitis, and conjunctivitis.

Other studies in Europe using a similar dose range yielded similar local tumor control. Additionally, a study from Tufts University showed similar local tumor in 14 patients with choroidal melanoma treated with Gamma Knife–based SRS with de-escalation of the dose to 22.2 Gy at a median follow-up interval of 32.2 months. [82]


SRS has been used to manage seizure disorders such as mesial temporal epilepsy (MTE) and seizures caused by AVMs, cavernomas, and hypothalamic hamartomas. Remission rates after SRS for mesial temporal epilepsy ranged from 0-86%. Mean remission rates after SRS for seizures caused by AVMs, cavernomas, and hypothalamic hamartomas were 71%, 31%, and 16%, respectively. [83, 84, 85] In a multi-institutional pilot study, 30 patients (13 treated with 20 Gy and 17 with 24 Gy) were treated with Gamma Knife SRS for mesial temporal epilepsy. At the 36-month follow-up, 67% of patients were free of seizures for the prior 12 months, 58.8% and 76.9% for those receiving 20 Gy and 2 Gy, respectively. Verbal memory impairment and improvement was observed in 15% and 12% of patients, respectively. Gamma Knife SRS has also been used to treat heterotopia-based epilepsy, with achievement of complete remission. [86]



In general, SRS is not suitable for tumors or lesions 4 cm or larger in diameter or immediately adjacent to eloquent structures such as the optic apparatus and brainstem if a dose of higher than 12 Gy is needed to control the tumor. According to the QUANTEC (Quantitative Analysis of Normal Tissue Effects in the Clinic), the maximum brainstem dose of 12.5 Gy is associated with low (< 5%) risk of injury to the brainstem. A 2- to 5-mm clearance from the optic pathway is usually needed for a lesion to be treated with SRS. The most commonly quoted tolerance for the optic apparatus is 8 Gy, but a study from the Mayo Clinic showed that a point dose of 12 Gy to the optic apparatus might be safe.

In the pediatric age group, Gamma Knife–based SRS is, in general, not suitable for patients whose skulls are not completely fused, because invasive pin placement for securing the stereotactic head frame is required in Gamma Knife surgery. Witt et al from Indiana University described the use of a modified Aquaplast for immobilization of a 14-month-old child who was successfully treated with Gamma Knife–based SRS, but routine use of this method based on this limited experience is not encouraged. Alternative SRS methods using a frameless system such as CyberKnife may be suitable for these very young pediatric patients.

In circumstances in which the tumor is located inferior to the skull base to a degree that the stereotactic head frame cannot be placed to enable adequate stereotactic localization, a frameless system such as CyberKnife can be used.



In general, local anesthesia of the pin-sites is performed for head frame placement. In pediatric and some teenage patients, as well as patients who cannot tolerate the procedure awake, general anesthesia is used. For patients, especially children, under general anesthesia, it is important to take into account the amount of time needed for treatment delivery when treatment planning is performed. A balance must be struck between a more-conformal plan with a longer delivery time and a less-conformal plan with a shorter delivery time.



The new Leksell Gamma-Knife PerfexionTM (Elekta Instrument AB, Stockholm, Sweden) is an entirely redesigned unit released in 2006. The old-generation machines were designed to have 201 fixed sealed sources. The radiation coming from these sources converge to a point (isocenter) to produce a high-dose region, spherelike shape collimated by 4 heavy helmets (4, 8, 12, and 18 mm). By moving the tumor spatial position around these isocenters, it is possible to cover the entire tumor volume with the prescribed radiation dose.

The first and second models (U and B) were well known for their very labor-intensive manual frame x, y, z, and g head positioning system (Trunnions). The subsequent models (C and 4C), both equipped with Automatic Positioning System (APS), are less labor intensive because the positioning of the head in the x, y, and z dimensions is controlled automatically. However, those models are still limited by the core size.

The new collimator channel system PerfexionTM consists of 8 movable sectors containing 24 sealed sources of cobalt-60 each (total of 192 sources). During treatment, this new build-in collimator allows 3 different-sized isocenters, providing greater flexibility of hybrid shots design. The head frame adapter has 3 g angles (70°, 90°, and 110°) attached to the new positioning system. This positioning system thus is designed to move the whole body in x, y, z direction (instead only the head for the old systems). The direct implication of using this new unit is the potential decrease in overall treatment time because of minimal need to change configurations such as helmets, trunnions, or gamma angles.



Positioning of the head frame relative to the skull

The positioning of the head frame relative to the patient's head depends on the location of the target to be treated. The head frame is secured to the patient's skull bone using titanium pins.

Positioning of the head frame relative to the Gamma Knife unit

The positioning of the head frame relative to the Gamma Knife for each shot is documented by the x-, y- and z-coordinates and the Gamma angle and is determined by treatment planning parameters required to satisfy the therapeutic goals, namely, target coverage and sparing of critical structures.



Improvements in software accompanying PerfexionTM include DICOM (Digital Imaging and Communications in Medicine) implementation in their treatment planning software and automatic radiation dose–balancing algorithm. The implementation improved the compatibility with other treatment software and radiation treatment record and verification software.

The PerfexionTM treatment planning system allows exporting of treatment plans and images in DICOM format to other systems. When a patient is treated at another treatment machine, the DICOM image and dose file can be used to generate a composite dose. The automatic dose–balancing algorithm weights the radiation doses at multiple targets by considering the contribution from scattering and entering/exiting radiation from other targets. Previously, this had to be manually performed, which was very cumbersome.

The "dynamic shaping" was implemented in version 8.0 and higher to protected critical structures by blocking sectors. The co-registration feature (version 8.2 and higher) allows frameless MRI images (taken previously) to be manipulated and analyzed. On the day of treatment, these preplanned images can be registered to the MRI images obtained with the frame and considered for treatment. Re-treatment and follow-up also are possible using this co-register feature.

The prescribed radiation dose and isodose line (in general, 50%) are determined by multiple factors, including target size, target type, and prior radiotherapy or SRS and should be to the discretion of the treating radiation oncologist and neurosurgeon.

Software versions 10.0 and higher (available in 2011), after many years of testing, provide for a robust inverse planning system. Based on automatic shot filling for each target and automatic sector optimization, the degree of conformality of a defined target with treated volume and the rapidity of fall-off (gradient index) can be specified by the user. Optimized solution takes minutes to achieve and provides for plan regeneration through an interactive user platform.

An atlas of common structure is also included, along with a convolution algorithm, which accounts for structure heterogeneity for dose computation. Notably, hardware releases in recent years include insulated frame posts to prevent heating effects from high field strength MR imaging and a frameless, fractioned head positioning system. The latter release is expected to provide for the irradiation of larger target volumes by reducing single-fraction treatment toxicity similar to indications used for CyberKnife intracranial irradiation.

Regulatory considerations for Gamma Knife Perfexion

In 2009, the United States Nuclear Regulatory Commission (NRC) considered the PerfexionTM Gamma Knife to be significantly different from the other gamma SRS units and of such complexity that a special Regulatory Guide to Chapters 10 CFR 35.1000 and Subpart K was issued entitled “Leksell Gamma Knife Perfexion – Licensing Guidance.”

Most current users of previous Gamma Knife treatment models (U to 4C) had to undergo a recertification process. This process consisted of either a 3-day didactic and clinical course at a manufacturer training center or similar structured training sessions held at a 4C licensed center with a recertified AMP (Authorized Medical Physicist) and manufacturer clinical specialist on site.

The licensing guide specifies that all associated personnel (AU – Authorized User, AMP) receive training in hands-on device operation, safety procedures, clinical use, and the operation of a treatment planning system for the PerfexionTM unit. If the individual is already an AU or AMP for a gamma stereotactic unit, in accordance with 10 CFR 35.51(c), this training must also include instruction in the differences in the device operation, safety procedures, clinical use, and the operation of a treatment planning system of the PerfexionTM unit.

After July 1, 2009, all new and existing users need to acquire a Preceptor Attestation through a PerfexionTM AU or AMP. The licensing guide stipulates that only a PerfexionTM AMP can train another physicist, whereas an AU preceptorship could be conducted by a PerfexionTM AMP or AU. The licensing guide goes on to precisely describe daily spot checks and monthly, biannual, and annual tests that must be performed, highlighting the differences the PerfexionTM brings to the user. These include (1) changes in the helmet design to sector delivery, (2) frame adapter mounted to the computer-controlled couch instead of APS (Automatic Positioning System) motion, (3) lack of visual inspection of helmet clearance now replaced with collision detection device, and (4) changes in specification and confirmation of the AU written directive.

At the authors’ institution, all physicist, radiation oncologists, and neurosurgeons who wish to have privileges at the PerfexionTM center are required to complete a minimum of 15 hours of PerfexionTM -specific training and actively participant in the delivery of treatment to at least 10 patients. In addition, Homeland Security concerns have encouraged the user to limit access to the Gamma Knife PerfexionTM treatment room, requiring all unaccompanied personnel have a background clearance conducted at the Federal level.



See the list below:

  1. SRS refers to the precise and focused delivery of a single, high dose of radiation in a single session.
  2. SRS has been used to treat various intracranial and skull base lesions, with results comparable to surgery in some tumors.
  3. The dosimetric characteristics, namely a highly conformal isodose distribution and a very steep dose gradient of dose fall-off beyond the prescribed isodose line, of SRS lend themselves well to the delivery of an ablative dose of radiation to intracranial and skull base lesions.
  4. The serious late toxicities associated with SRS are, in general, low.


Head frame placement

Pain, bleeding, and infection can occur at the pin-sites. To minimize the risk of infection after head frame removal, antibiotic cream can be used in the surgical dressing.

Early effects


Occasionally, some patients may develop seizures within the first 24-72 hours after Gamma Knife–based SRS for supratentorial lesions. For this reason, it is prudent to inform patients of this risk and to recommend against driving during that time interval. Recommendation against driving based on seizure risk other than that associated with SRS is made separately at the discretion of the neurosurgeon or neurologist.

Brain edema

Acute brain edema can occur shortly after SRS and can present as headaches. Some centers, such as the authors’, routinely administer an intravenous steroid in high-risk patients on the day of the procedure to decrease this risk.

Other acute symptoms

Acute symptoms such as headaches, nausea/vomiting, fatigue, and vertigo (in patients treated for vestibular schwannoma) have been observed.

Late effects

Optic neuropathy

Optic neuropathy can occur after SRS. [87, 88] The reported incidence of optic neuropathy is low. This may be related to the fact that extra caution is taken when planning SRS for tumors close to the optic apparatus, based on the early report from Harvard University of optic neuropathy after a single SRS dose of 8 Gy, which is likely a conservative constraint. [87]

In a more recent report from the Mayo Clinic, where 3-dimensional dose computation was used, dose plans and clinical outcomes of 218 Gamma Knife procedures in 215 patients for tumors of the sellar and parasellar region were reviewed. The median follow-up was 40 months. The median maximum radiation dose to the optic nerve was 10 Gy (range, 0.4-16 Gy). Four patients (1.9%), all of whom had prior surgery and 3 of whom also had external beam radiotherapy, developed radiation-induced optic neuropathy at a median of 48 months after SRS. [88] The risk of developing a clinically significant radiation-induced optic neuropathy was 1.1% for patients receiving 12 Gy or less. In spite of the fact that 73% received greater than 8 Gy to a short segment of the optic apparatus, radiation-induced optic neuropathy occurred in less than 2% of patients. [88]

Other cranial nerve toxicities

The cranial nerves in the cavernous sinus appear to be quite resistant to SRS. [64] Based on the SRS literature for skull base meningiomas, the III, IV, V, and VI nerves seem to be able to tolerate a high single dose of radiation, although the exact doses delivered to those nerves individually were not quantified, which is not surprising given the difficulty identifying those nerves on the planning CT or MR images. The risk of permanent III, IV, or VI neuropathy and V neuropathy was 0.4% and 0.2%, respectively, based on a review of 1255 patients who underwent SRS for pituitary adenomas. [64]

The incidence of V and VII nerve deficit after SRS is typically lower than 10% with reduced-dose SRS (12-13 Gy) for vestibular schwannoma. [54, 56, 62, 63] Hearing preservation (VIII nerve) ranged from 33-88%. [52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63]

Based on 2 meta-analyses for glomus tumors, SRS (15-16 Gy) resulted in approximately 10% risk of injury to IX, X, XI, and XII nerves. [71, 72, 73]

Radiation necrosis

Based on data on Gamma Knife–based SRS for AVM, the risk of symptomatic radiation necrosis is related to the volume encompassed by the 12 Gy and the location of the AVM. [19] The volume encompassed by the 12 Gy also predicts risk of symptomatic necrosis for non-AVM intracranial tumors. [89]

Vascular injury

Brain infarction can occur if the internal carotid artery is damaged by radiation. Witt from Indiana University reviewed the information on 1255 patients who underwent SRS for pituitary adenoma and found that the incidence of SRS-induced internal carotid artery stenosis was very low. Based on limited data in the literature, the prescribed dose should cover less than 50% of the diameter of the internal carotid artery or the maximum dose to the internal carotid artery should be limited to 30 Gy or less. [64]



SRS is an effective noninvasive treatment for various intracranial, ocular, and skull base tumors, as well as non-neoplastic and functional disorders such as AVM, trigeminal neuralgia, and seizures. It can be delivered with different devices such as Gamma Knife, CyberKnife, modified linear accelerator, and protons. Compared with other devices, Gamma Knife has the most rigid hardware and fixation system, rendering it very suitable for SRS. Other photon-based systems use multiple static or dynamic beams to aim the target from different angles, by manipulating the gantry, couch, and collimator angles, except for the CyberKnife system, with which the beam angles are controlled by a robotic arm mounted with a linear accelerator. However, the experience and expertise of the treatment team, including neurosurgeon, radiation oncologist, medical physicist, and therapists, are likely more important than the treatment device itself.