A Single-Institution Analysis of 126 Patients Treated with Stereotactic Radiosurgery for Brain Metastases

BACKGROUND: The objective of this study was to report our institutional experience with Gamma Knife(®) Radiosurgery (GKRS) in the treatment of patients with brain metastases. METHODS: Retrospectively collected demographic and clinical data on 126 patients with intracranial metastases were reviewed. The patients in our study underwent GKRS at Vidant Medical Center between 2009 and 2014. Kaplan–Meier curves were used to compare survival based on clinical characteristics for univariate analysis, and a Cox proportional hazards model was used for multivariate analysis. RESULTS: The median age of the patient population was 62 years. Medicare patients constituted 51% of our patient cohort and Medicaid patients 15%. The most common tumor histologies were non-small cell lung cancer (50%), breast cancer (12.7%), and melanoma (11.9%). The median overall survival time for all patients was 5.8 months. Patients with breast cancer had the longest median survival time of 9.15 months, while patients with melanoma had the shortest median survival time of 2.86 months. On univariate analysis, the following factors were predictors for improved overall survival, ECOG score 0 or 1 vs. 2 or greater (17.0 vs. 1.8 months, p < 0.001), controlled extracranial disease vs. progressive extracranial disease (17.4 vs. 4.6 months, p = 0.0001), recursive partitioning analysis Stage I vs. II–III (18.2 vs. 6.2 months, p < 0.007), multiple GKRS treatments (p = 0.002), prior brain metastasectomy (p = 0.012), and prior chemotherapy (p = 0.021). Age, ethnicity, gender, previous external beam radiation therapy, number of brain metastases, and hemorrhagic vs. non-hemorrhagic tumors were not predictors of longer median survival time. Number of metastatic brain lesions of 1–3 vs. ≥4 (p = 0.051) and insurance status of Medicare/Medicaid vs. commercial insurance approached significance (13.7 vs. 6.8 months, p = 0.08). On multivariate analysis, ECOG performance status 0–1 (p < 0.001), multiple GKRS treatments (p = 0.003), and control of extracranial disease (p = 0.001) remained significant predictors of survival. CONCLUSION: ECOG score, control of extracranial disease, and multiple GKRS treatments are predictors of longer median survival following GKRS in our patient population. GKRS is an effective treatment for brain metastases, but these factors may be considered in patient selection for GKRS.