Brain Metastases Treated With Surgical Resection Followed by Frame-Based Stereotactic Radiosurgery

Abstract

To evaluate outcomes and quality of life (QOL) for patients with brain metastases treated with surgical resection followed by frame based radiosurgery. 94 patients with brain metastases were underwent surgical resection followed by frame based radiosurgery between 12/2006 and 6/2013. 56 patients were treated to the resection cavity only; 38 had synchronous lesions (median 3, 2-9) treated definitively with frame based radiosurgery along with the surgical bed. The most common histologies were non-small cell lung cancer (59%) and breast (18%). The median marginal dose was 18 Gy (12.5-18 Gy). Median resection cavity volume was 8.9 cc (0.5-41 cc); median treatment volume and conformity index (CI) were 17.6 cc and 1.9, respectively. Clinical endpoints were assessed using the Kaplan-Meier method; univariate (UVA) and multivariate (MVA) analyses using Cox proportional hazard regression models. QOL data was collected prospectively using the EORTC QLQ-C30 and BN20 questionnaires. Median follow-up was 13 mos (1.5-79 mos). 6- and 12-mo actuarial local control rates (LC) for resected metastases were 86% and 78%, with a median time to local recurrence (LR) of 8.5 mos (2 - 55 mos). 6- and 12-month LC rates for unresected metastases were 97% and 93% with median time to LR of 6.1 mo (1.2-22 mo). 6- and 12-month freedom from elsewhere brain failure (EBF) were 65 and 50%, with a median time to EBF of 5.4 mos (1.5-49 mos). Median overall survival (OS) was 15 mos (1.5 - 79 mos). 28 patients (30%) received salvage whole brain RT; 35 (37%) underwent multiple frame based radiosurgery treatments (mean: 2.5, 2-5) for metachronous lesions. On UVA, marginal dose, cavity volume, max cavity dimension, pre-surgical lesion volume, CI, and tumor location did not predict for LR for resected cavities. While tumor volume, max linear dimension, and dose predicted LR for unresected lesions on UVA; max linear dimension remained significant on MVA (HR 8.05, p = 0.02). Compared to smaller resection cavities, larger cavities (> 9.5 cm3) had similar rates of LR but significantly higher rates of radionecrosis (RN) (20% versus 4%, p = 0.048). Freedom from leptomeningeal failure (LMF) at 6- and 12- mos were 89% and 84%; LMF predicted for OS on UVA and MVA (HR 2.3;p<.05), as did controlled primary, extracranial disease and KPS. Compared to baseline, there was no significant difference was seen in global health status, fatigue, seizures or physical/motor/cognitive functioning 6 mos post-frame based radiosurgery; 6-mo nausea was worse (p = 0.02), related to chemotherapy. frame based radiosurgery achieves acceptable LC of resected brain metastases and synchronous lesions with no impact on QOL. Treatment of larger resection cavities increases RN risk. Identifying factors predictive of LMF may aid in patient selection. Future comparisons of post-resection radiosurgery versus whole brain RT with respect to QOL, patterns of failure and toxicity are warranted given these findings.