RT-38 * MAJOR CRITERIA TO SELECT PATIENTS FOR STEREOTACTIC RADIOSURGERY FOLLOWING RESECTION OF BRAIN METASTASES

Abstract

PURPOSE: Stereotactic radiosurgery (SRS) following resection of brain metastases is well-tolerated, provides excellent local control and reserves whole-brain radiation therapy for high volume CNS recurrence. We identified criteria to facilitate appropriate candidate selection for SRS based on CNS recurrence and overall survival (OS). METHODS: From 2002-2010, individuals with a KPS >70 who received SRS to the resection cavity for brain metastases were evaluated. The target encompassed the entire cavity defined on MRI, and was treated using Gamma Knife radiosurgery. Follow-up included clinical exam and MRI at 2-3 month intervals. CNS failure was defined by imaging or pathology. Factors predicted to impact OS and CNS control were assessed using log-rank and Cox-regression analyses. RESULTS: 151 consecutive patients with a median follow-up of 5.3 months were identified. Local control was >90% at 1 year and >80% at 2 years. Median OS was 14.2 months with 19% 5-year survival. Patients without systemic disease and those with single metastasis survived significantly longer. Median time to regional CNS failure was 14.5 months. Systemic disease overwhelmingly predicted regional CNS recurrence. Multivariate analysis showed that absence of systemic disease, a single brain metastasis and age were associated with improved OS and regional CNS control. We identified 3 groups: single brain lesion without systemic disease, >2 brain lesions without systemic disease, and those with systemic disease, that predicted OS and CNS control. Median OS was 16.5, 18.3, and 8.1 months, respectively (p < 0.001) and median time to CNS recurrence was 32.8, 15.7 and 8.7 months, respectively (p < 0.001). CONCLUSION: This represents the largest study of SRS following resection of brain metastases. Excellent long-term local control, survival and regional CNS control can be achieved in select patients. Major criteria to identify treatment candidates should include systemic disease status as well as number of brain lesions.