Prognostic factors for survival after stereotactic radiosurgery vary with the number of cerebral metastases

Abstract

Background: Little is understood about the factors that influence survival in patients who undergo gamma-knife stereotactic radiosurgery (SRS) for brain metastases.

Methods: Demographic, disease, treatment, and survival data on 334 patients with intracranial metastases who underwent initial SRS from 1998 to 2004 were abstracted from treatment records and from the Connecticut Tumor Registry. Multivariate survival analysis was used to identify factors that independently affected survival.

Results: The median age of the patient population was 57.3 years. The median number of lesions treated in a single session was 2 (range, 1-36 lesions treated). The most common tumor histologies were nonsmall cell lung carcinoma (36%), breast cancer (16%), and melanoma (16%). Three hundred patients (90%) had confirmed deaths; the median survival after SRS was 8.1 months. Survival was significantly better in patients who had from 1 to 3 metastases (median, 8.5 months) compared with patients who had > or =4 metastases (median, 6.3 months; hazard ratio [HR], 0.65; P = .003). In the subgroup of patients who had from 1 to 3 metastases, systemic control (HR, 0.49; P < .001), breast cancer (HR, 0.57; P = .003), and total tumor volume < 5 cc (HR, 0.65; P = .002) were associated independently with increased survival, and esophageal cancer (HR, 2.36; P = .042) was associated with decreased survival. In the subgroup of patients who had > or =4 metastases, only age <45 years was associated independently with increased survival (HR, 0.39; P = .006); and melanoma (HR, 2.32; P = .008) and the receipt chemotherapy (HR, 2.59; P = .077) were associated with decreased survival. Sex, race, metastases location, whole-brain radiation, and cranial surgery had no independent associations with altered survival.

Conclusions: The data from this study suggested that different factors affected survival in patients who had from 1 to 3 metastases and patients who had > or =4 metastases. Further research into this area may clarify causes for this discrepancy and improve prognostication.