Frameless stereotactic radiosurgery for brain metastases: a review of outcomes and prognostic scores evaluation.

Abstract

Stereotactic brain radiosurgery provides good local control in patients with limited brain metastases. A newly developed frameless system allows pain-free treatment. We reviewed the effectiveness of this frameless stereotactic brain radiosurgery and identified prognostic factors that may aid better patient selection. Medical records of patients with brain metastases treated with linear accelerator-based frameless stereotactic brain radiosurgery between January 2010 and July 2015 in a university affiliated hospital in Hong Kong were reviewed. Outcomes including local and distant brain control rate, progression-free survival, and overall survival were analysed. Prognostic factors were identified by univariable and multivariable analyses. Association of outcomes with four common prognostic scores was performed. In this study, 64 patients with 94 lesions were treated with a median dose of 18 Gy (range, 12-22 Gy) in a single fraction. The median follow-up was 11.5 months. One-year actuarial local and distant brain control rates were 72% and 71%, respectively. The median overall survival was 13.0 months. On multivariable analysis, Karnofsky performance status score (>50 vs ≤50) and number of lesions (1-2 vs ≥3) were found to associate significantly with distinct brain progression-free survival (P=0.022, hazard ratio=0.20, 95% confidence interval 0.05-0.80 and P=0.003, hazard ratio=0.31, 95% confidence interval 0.14-0.68, respectively). Overall survival was associated significantly with Basic Score for Brain Metastases (P=0.031), Score Index for Radiosurgery in Brain Metastases (P=0.007), and Graded Prognostic Assessment (P=0.003). Improvement in overall survival was observed in all groups of different prognostic scores. Frameless stereotactic brain radiosurgery is effective in patients with oligo-metastases of brain and should be increasingly considered in patients with favourable prognostic scoring.