T he imaging response of brain metastases after stereotactic radiosurgery is complex and dynamic. Stereotactic radiosurgery leads to tumor control in 70%-90% of metastatic brain tumors. 1,3,4,8,10 The clinical effectiveness of radiosurgery relates in part to the histological tumor source. Tumor control is adequate if serial imaging after radiosurgery reveals stable or reduced tumor volumes. A transient increase in tumor volume may not preclude the possibility of long-term tumor inactivation 5 and is inadequate to predict the eventual radiosurgical response at a single follow-up imaging time point. In an effort to better define the effect of radiosurgery on different tumor pathologies, we evaluated the quantitative volumetric response of lung, breast, and melanoma metastases in the brain following Gamma Knife radiosurgery (GKRS). We then correlated this response to patient survival.
MethodsInstitutional review board approval was acquired at the University of Pittsburgh for data review and presentation. We conducted a retrospective review of patients who had undergone GKRS for brain metastases from nonsmall cell lung cancer (NSCLC), breast cancer, and melanoma between 2002 and 2010 at our institution. To be included in the study, patients had to have undergone follow-up imaging at our institution so that volumetric measurements could be made on the same imaging software. Object. The aim of this study was to evaluate the imaging response of brain metastases after radiosurgery and to correlate the response with tumor type and patient survival.Methods. The authors conducted a retrospective review of patients who had undergone Gamma Knife radiosurgery for brain metastases from non-small cell lung cancer (NSCLC), breast cancer, or melanoma. The imaging volumetric response by tumor type was plotted at 3-month intervals and classified as a sustained decrease in tumor volume (Type A), a transient decrease followed by a delayed increase in tumor volume (Type B), or a sustained increase in tumor volume (Type C). These imaging responses were then compared with patient survival and tumor type.Results. Two hundred thirty-three patients with metastases from NSCLC (96 patients), breast cancer (98 patients), and melanoma (39 patients) were eligible for inclusion in this study. The patients with NSCLC were most likely to exhibit a Type A response; those with breast cancer, a Type B response; and those with melanoma, a Type C response. Among patients with NSCLC, the median overall survival was 11.2 months for those with a Type A response (76 patients), 8.6 months for those with a Type B response (6 patients), and 10.5 months for those with a Type C response (14 patients). Among patients with breast cancer, the median overall survival was 16.6 months in those with a Type A response (65 patients), 18.1 months in those with a Type B response (20 patients), and 7.5 months in those with a Type C response (13 patients). For patients with melanoma, the median overall survival was 5.2 months in those with a Type A response (26 patients) an...