Hemangioblastoma of the posterior fossa. The role of multimodality treatment.

The authors made a review of a series of patients with hemangioblastomas of the posterior fossa treated between 1973 and 1993. A total of 32 patients were analyzed with 24 patients receiving resection, 8 patients receiving radiosurgery and 2 patients receiving conventional radiotherapy. The mortality in the patients with a resection was considered acceptable with 2 deaths (8%) and with a morbidity of 3 patients (12.5%). A review of the literature suggests that conventional radiotherapy with high doses (45-60 Gy) may have a role in the post-operative control of hemangioblastomas and in some cases could be employed even before the resection in order to facilitate the surgery. The radiosurgical treatment is regarded like adjuvant. Poor results were obtained with radiosurgery in large tumors where low doses (less than 20 Gy) were used. Because of the rarity and complexity of these tumors, mainly when associated with von Hippel-Lindau disease, a multicenter study could be useful with the assessment of the optimal utilization and combination of these treatment modalities.

A multi-institutional experience with stereotactic radiosurgery for solitary brain metastasis.

PURPOSE: A multi-institutional experience in radiosurgery for solitary brain metastases was combined to identify factors associated with safety, efficacy, tumor control, and survival. MATERIALS AND METHODS: A review of 116 patients with solitary brain metastases who underwent gamma knife stereotactic radiosurgery at five institutions was performed. The median follow-up was 7 months following radiosurgery and 12 months following diagnosis. Minimum tumor doses varied from 8-30 Gy (mean, 17.5 Gy). Forty-five patients failed prior radiotherapy and 71 had no prior brain irradiation. Fifty-one patients had radiosurgery alone and 65 underwent combined radiosurgery with fractionated large-field radiotherapy (mean dose, 33.8 Gy). RESULTS: Median survival was 11 months after radiosurgery and 20 months after diagnosis. Follow-up documented local tumor control in 99 patients (85%), tumor recurrence in 17 (15%), and documented radiation necrosis in one (1%). The 2-year actuarial tumor control rate was 67 +/- 8%. Tumor histology affected survival (better for breast cancer, p = .004) and local control (better for melanoma and renal cell, p = .0003) in multivariate analyses. Combined fractionated radiotherapy and radiosurgery improved local control (p = 0.111), but not survival in multivariate testing. CONCLUSION: Radiosurgery is effective in controlling solitary brain metastases with low morbidity. Further study is needed to better define optimum treatment parameters for radiosurgery.

Stereotactic options in the management of craniopharyngioma.

Multimodality stereotactic techniques were used in the management of 39 patients with craniopharyngiomas in a 12-year interval. Monocystic craniopharyngiomas were treated successfully by intracavitary beta-irradiation using 32P (96% cyst control rate). Solid tumor progression or secondary cyst formation required repeat irradiation, radiosurgery or microsurgery in selected patients. In the future, wider and earlier application of stereotactic techniques may further reduce the still unacceptable morbidity currently associated with initial radical microsurgical resection of craniopharyngiomas.

No access surgery: the gamma knife.

Swedish neurosurgeon Lars Leksell, frustrated by the invasiveness of existing surgical tools and the morbidity some neurosurgical patients endured, created the field of stereotactic radiosurgery in 1951. He subsequently pioneered the development of the dedicated multi-source Cobalt 60 Gamma Knife. During the 27 year interval from its first clinical use in 1967 to its latest application in 1994, single fraction, closed skull irradiation of deep intracranial targets has been performed in more than 20,000 patients worldwide. The goals of radiosurgery are obliteration or prevention of further growth of the target coupled with reduced patient risk in comparison to more invasive procedures.

Potential human error in setting stereotactic coordinates for radiosurgery: implications for quality assurance.

PURPOSE: The error frequency in setting stereotactic coordinates for gamma knife radiosurgery was investigated to determine what quality assurance safeguards are necessary. METHODS AND MATERIALS: A prospective study of 200 consecutive isocenter settings for gamma knife radiosurgery was analyzed to identify the frequency of spontaneous errors in setting and checking stereotactic coordinates (corrected prior to treatment). An additional 25 coordinate errors were introduced at random among the next 200 consecutive isocenter settings to provide additional data on identification of errors. RESULTS: Stereotactic coordinates required resetting in 12% (24/200) of the isocenters treated due to errors of 0.25-0.50 mm (8%) and 1-20 mm (4%). This comprised 2.2% (26/1200) of the individual coordinate settings. The frequency of these errors was significantly related to the specific directional coordinate set (p = 0.0004) and experience (p = 0.016). Errors were identified by 83.5% (91/109) of the observers checking the settings (60.0% of 0.25 mm errors, 94.6% of errors > or = 0.5 mm, p = 0.0000). Verification of stereotactic coordinates by two observers reduces the probability of an undetected error > or = 0.25 mm to 1/1,392 and to 1/154,712 for errors > or = 1 mm. CONCLUSION: Errors in setting stereotactic coordinates are common (12% prior to checking) but are corrected with a high degree of confidence by a quality assurance policy requiring coordinate verification by a minimum of two observers.

Cranial nerve length predicts the risk of delayed facial and trigeminal neuropathies after acoustic tumor stereotactic radiosurgery.

PURPOSE: To test the hypothesis that length of cranial nerve irradiated is a major factor predicting the risk of cranial nerve injury following radiosurgery and to identify any other significant related treatment factors. METHODS AND MATERIALS: Ninety-two patients (93 acoustic tumors) were treated with a 201 source Cobalt-60 gamma unit from 1987 to 1990 and prospectively followed. The range of minimum tumor dose was 12-20 Gy and maximum dose 24-50 Gy. Univariate and multivariate analyses were used to evaluate any correlations between tumor measurements and treatment factors, with the development of trigeminal and facial neuropathies following radiosurgery. RESULTS: The risks of trigeminal and facial neuropathy following radiosurgery were associated with the pon-petrous distance and mid porous transverse tumor diameters respectively (anatomically related to the irradiated length of cranial nerves V and VII respectively) in both univariate (p = .002 for V and p = .026 for VII) and multivariate (p = .004 for V and p = .055 for VII) analyses. Tumor volume, other tumor measurements, maximum dose, minimum tumor dose, and tumor dose inhomogeneity were not significantly related to either trigeminal or facial neuropathy in univariate and multivariate analyses. CONCLUSION: Within a minimum tumor dose range of 12-20 Gy, the incidence of delayed trigeminal or facial neuropathy depended more on the estimated length of nerve irradiated than the tumor dose or tumor volume. In the future, the risk of delayed facial or trigeminal cranial neuropathy may be reduced significantly by performing radiosurgery when the tumor still has both a small mid-porous transverse diameter and a small pons-petrous distance.

Radiosurgery for vascular malformations of the brain stem.

The challenges associated with microsurgery of vascular malformations located in the midbrain, pons and medulla have promoted the development of alternative therapeutic techniques. To assess the efficacy and safety of radiosurgery in the management of brain stem vascular malformations we reviewed our 5-year experience in 50 patients evaluated between 4 and 51 months (mean, 25 months) after radiosurgery. Twenty-eight patients (56%) underwent gamma unit radiosurgery for symptomatic arteriovenous malformations (AVMs), and 22 patients (44%) for angiographically occult vascular malformations (AOVMs). Patients varied in age from 7 to 76 years (mean, 39 years). Forty-one patients (82%) had from 1 to 5 hemorrhages prior to gamma knife radiosurgery. Ten (20%) had one or two prior unsuccessful operations, and 37 (74%) presented with a neurological deficit. Of the patients with AVMs, 6 were considered Spetzler Grade III, and 22 (79%) Grade VI (inoperable: major component within the brain stem parenchyma). Forty-four malformations (88%) were adjacent to or within the midbrain and pons; the remainder involved the medulla. Average malformation diameters varied from 6 to 30.4 m (mean, 20.6; mean volume 4614 mm3). The minimal radiation dose to the margin of the malformations ranged from 12 to 25.6 Gy (mean, 18.9 Gy). Of the 28 patients with AVMs, 8 had follow-up angiograms at a minimum of 2-years after radiosurgery (or sooner if their MRIs suggested obliteration). Of these patients, 7 (88%) showed complete obliteration of their malformations. No patients with AOVMs rehemorrhaged if more than 15 months elapsed after radiosurgery.(ABSTRACT TRUNCATED AT 250 WORDS)

Radiosurgery and brain tolerance: an analysis of neurodiagnostic imaging changes after gamma knife radiosurgery for arteriovenous malformations.

In order to analyze complications and the factors responsible for the development of serial imaging changes after stereotactic radiosurgery for intracranial arteriovenous malformations, we reviewed serial post-treatment magnetic resonance imaging scans in 72 patients. Median follow-up was 23 months (range 12 to 35 months). Twenty patients developed post-radiosurgical imaging changes consisting of new regions of increased T2 signal on magnetic resonance imaging in brain surrounding the arteriovenous malformation (two year actuarial incidence of 31%). Imaging changes were associated with headache or new neurological deficits in nine of these 20 (45%) and remained asymptomatic in 11 (55%). Symptoms developed in three of 13 patients with imaging changes in the cerebral cortex or cerebellum, in contrast to six of seven patients who had symptoms with imaging changes in the brainstem (p = .028). The onset of imaging changes varied from five to 18 months after radiosurgery (median, 12 months). Serial follow-up scans four to 25 months after the onset of imaging changes were available for review in 16 patients. Post-radiosurgical imaging changes completely resolved within 4 to 19 months in ten patients and have not yet completely resolved after 6 to 25 months in six patients. The projected actuarial rate for resolution of imaging changes was 88%, 19 months after onset; the median time for resolution was 14 months. Univariate analysis revealed that the development of imaging changes was significantly associated with treatment volume (p = .025), the risk predicted from the integrated logistic formula (p = .042), and the number of isocenters treated (p = .042). In multivariate analysis, volume was the only factor significantly associated with the development of imaging changes.

Neuroimaging of acoustic nerve sheath tumors after stereotaxic radiosurgery.

Using a strict method for measuring tumor size, we evaluated tumor response to radiosurgery in 88 patients with 89 acoustic tumors treated over 3 years with a 201-source cobalt-60 gamma unit. Overall, tumor size was unchanged in 73% of patients and increased in 4%. In 22% of patients, tumor diameter decreased an average of 4.9 mm 3-33 months after treatment. Tumor shrinkage occurred in 36% of 50 patients who were followed for at least 1 year after treatment. Loss of tumor contrast enhancement was seen in 79% of patients 1-18 months after treatment. Delayed communicating hydrocephalus developed in four patients. In eight patients, increased signal on T2-weighted MR images developed in the adjacent cerebellar peduncle (n = 5) or the peduncle and dorsolateral pons (n = 3) 5-15 months after treatment. T1-weighted MR imaging and CT were insensitive to these adjacent brain changes. Stereotaxic radiosurgery is an important alternative treatment for selected patients with acoustic tumors. There is no mortality or major perioperative morbidity, hospitalization time and costs are smaller than for microsurgery, patient employment or functional level is maintained, and hearing preservation and facial neuropathy rates are comparable to those in published microsurgical series. Although the rate of occurrence of trigeminal neuropathy is greater than those reported in published microsurgical series, the majority of cases are mild, transient, and nondebilitating. MR imaging before and after radiosurgery is the most sensitive imaging tool to evaluate tumor response, the presence of adjacent parenchymal signal changes, and ventricular size. With a mean follow-up time of 14.6 months, the rate of complications detected by neuroimaging is low and the tumor control rate is 96%.

Gamma knife radiosurgery of meningiomas.

Fifty patients with meningiomas were treated during the initial 30 months experience using the 201 source cobalt-60 gamma knife at the University of Pittsburgh. The most frequent site of origin was the skull base. Stereotactic radiosurgery was the primary treatment modality in 16 (32%) patients with symptomatic tumors demonstrated by neuroimaging. Thirty-six patients (72%) had undergone at least one craniotomy, and 4 patients (8%) previously had fractionated external beam radiation therapy. The proximity of cranial nerves, vascular, pituitary and brainstem structures to the often convoluted tumor mass was crucial to dose selection. Follow-up imaging studies and clinical analysis of patients were performed at 6-month intervals. The actuarial 2-year tumor control rate was 96%. Only 2 patients have shown delayed tumor growth outside the radiosurgical treatment volume. To date, stereotactic radiosurgery proved to be a relatively safe and effective therapy for selected patients with symptomatic meningiomas, either as an adjuvant treatment to prior resection, or as a primary treatment alternative for patients whose advanced age, medical condition or high-risk tumor location mitigated against surgical resection.

Dose-volume considerations in radiosurgery.

Dose-volume effects are very important in radiosurgery. Functional radiosurgery illustrates brain tolerance at the extremes of small volumes and high radiation doses. The risks of radiation necrosis from radiosurgery of arteriovenous malformations (AVMs) and tumors appear to be reasonably well predicted by the integrated logistic formula and the 1% dose-volume isoeffect line for proton beam irradiation. The two main exceptions to this rule are cranial nerves, which appear to be more sensitive than the rest of the brain, and angiographically occult vascular malformations, where complications appear to occur at lower doses/volumes than in the treatment of tumors or AVMs. Further investigation is needed to better predict complications throughout the entire range of volumes, radiation doses, treatment locations and techniques presently used in stereotactic radiosurgery.

Treatment volume shaping with selective beam blocking using the Leksell gamma unit.

The Leksell gamma unit at the University of Pittsburgh uses 201 highly focused 60Co beams arranged in a hemispherical array. Selective beam blocking can be used to modify the treatment volume into ellipsoid shapes oriented in different directions to match better the shape of the target volume. Dose distributions for different blocking patterns were calculated using specially developed computerized 3-D treatment planning software. The changes in dose distribution with different blocking patterns predicted by computer were verified by film densitometry. Techniques for using selective beam blocking to match more closely the treatment volume to the intended target volume have the potential of reducing the likelihood of complications for radiosurgery with the Leksell gamma unit and need to be further developed.

Physics of gamma knife approach on convergent beams in stereotactic radiosurgery.

The Presbyterian-University Hospital of Pittsburgh installed the first clinically designated Leksell gamma knife in the U.S. in August 1987. Gamma knife radiosurgery involves stereotactic target localization with the Leksell frame and subsequent closed-skull single-treatment session irradiation of a lesion with multiple highly focused gamma ray beams produced from 60Co sources. The hemispherical array of sources, the large number of small-diameter beams, and the steep dose gradients surrounding a targeted lesion make physical characterization of the radiation field complex. This paper describes the physical features and the operation of the gamma knife as well as the calibration procedures of the very small, well-collimated beams. The results of studies using in-phantom ion chamber, diode, film, and lithium fluoride thermoluminescent dosimetry were all in close agreement. Both single-beam and multiple-beam dose profiles were measured and reported for the interchangeable helmets, which have 4-, 8-, 14-, and 18-mm-diameter collimators. We also describe the dose calculation and treatment planning algorithm in the treatment planning system. Measurements of the accuracy of mechanical and radiation alignment are also performed and discussed.

Shielding requirements on-site loading and acceptance testing on the Leksell gamma knife.

On August 14, 1987, the first stereotactic radiosurgical procedure using the gamma knife was performed in North America. Located in a self-contained radiosurgical suite in the basement of Presbyterian-University Hospital in Pittsburgh, Pennsylvania. This device uses 201 highly focused beams 60Co for the single-treatment closed-skull irradiation of brain lesions localized by stereotactic techniques (radiosurgery). One hundred and fifty-two patients with intracranial arteriovenous malformations or brain tumors were treated in the first year of operation. The Presbyterian University Hospital of Pittsburgh gamma knife is the first such unit in which the 60Co sources were loaded on-site. This effort required us to solve some difficult and unusual problems encountered during site preparation, delivery, and loading of the unit in a busy hospital setting. The solutions developed enabled installation and use of the gamma knife with minimal disruption of hospital activities while maintaining acceptable levels of exposure to radiation. Environmental surveys performed during the loading of the 201 radioactive sources (total, 219 TBq) confirmed that on-site loading is possible and practical. Our experience in the design, construction, and implementation of the first North American gamma knife supports the practicality and safety of on-site loading and may be of value in the planning and development of future gamma knife installations.

Stereotactic radiosurgery in children and adolescents.

Stereotactic radiosurgery has had an increasing role in the treatment of selected intracranial lesions in pediatric patients. In our first 44 months experience, 60 of the patients (9%) treated were less than or equal to 18 years of age. Current indications for radiosurgery include primary treatment of high-risk arteriovenous malformations or acoustic neurinomas (usually in patients with neurofibromatosis), adjuvant treatment for recurrent benign tumors after surgery, or as adjuvant treatment to fractionated irradiation for malignant tumors.

An integrated logistic formula for prediction of complications from radiosurgery.

An integrated logistic model for predicting the probability of complications when small volumes of tissue receive an inhomogeneous radiation dose is described. This model can be used with either an exponential or linear quadratic correction for dose per fraction and time. Both the exponential and linear quadratic versions of this integrated logistic formula provide reasonable estimates of the tolerance of brain to radiosurgical dose distributions where there are small volumes of brain receiving high radiation doses and larger volumes receiving lower doses. This makes it possible to predict the probability of complications from stereotactic radiosurgery, as well as combinations of fractionated large volume irradiation with a radiosurgical boost. Complication probabilities predicted for single fraction radiosurgery with the Leksell Gamma Unit using 4, 8, 14, and 18 mm diameter collimators as well as for whole brain irradiation combined with a radiosurgical boost are presented. The exponential and linear quadratic versions of the integrated logistic formula provide useful methods of calculating the probability of complications from radiosurgical treatment.

Manual reconstruction of tumor volumes from CT scans for radiotherapy planning.

Using a computed tomographic scan to construct a tumor volume on a simulation film has become an important part of radiation therapy treatment planning. Techniques for reconstruction of brain tumor volumes from computed tomographic scans onto simulation films are discussed. A technique for reconstructing by hand the tumor volume projection at a different angle from the plane of the computed tomographic scan is outlined. Ignoring beam divergence simplifies reconstruction, allowing it to be done by hand but introduces an additional error of at most 0.06, 0.15, 0.26 and 0.42 cm for tumors with maximum diameters of 5, 7.5, 10 and 12.5 cm, respectively. This technique provides the capability for limited 3-dimensional volume reconstruction of small tumors without the use of the computer.

Gamma knife radiosurgery for intracranial arteriovenous malformations in childhood and adolescence.

Eighteen children or adolescents with intracranial arteriovenous malformations (AVM) underwent stereotactic radiosurgery using the first North American gamma knife. This closed-skull, single-treatment therapy, utilizing 201 ionizing beams of gamma-irradiation, was used as an alternative to microsurgical removal in these selected patients (aged 34 months to 18 years, mean 12.3 years) beginning in August 1987. No significant perioperative morbidity occurred, and no patient rebled or died in the follow-up interval ranging between 7 and 19 months. Computed tomography (CT) and magnetic resonance imaging (MRI) were used to monitor the response to treatment and to determine when postoperative angiography was indicated. Of seven AVMs examined with cerebral angiography 1 year after treatment, three were completely obliterated; three others were significantly smaller, and their complete obliteration is anticipated by 2 years after treatment. Follow-up CT or MRI confirmed attenuation or signal changes suggestive of edema surrounding the treatment volume in 3 patients; 1 had transient worsening of a preexisting neurological deficit. Although a more long-term perspective is still required for this new technology now available in the United States, we believe that gamma knife stereotactic radiosurgery is a safe and effective method to obliterate AVM deemed too risky for microsurgical removal.