Local Control Following Postoperative Stereotactic Radiation Therapy for Brain Metastases

Abstract

Stereotactic radiosurgery (SRS) and fractionated stereotactic radiation therapy (FSRT) have emerged as viable alternatives to whole brain radiation therapy (WBRT) in patients with resected brain metastases. We performed a systematic review and quantitative analysis of the published experiences with adjuvant SRS and FSRT to determine local control (LC) rates and any relationship with dose-response, fractionation, and treatment delivery parameters. We identified published articles that reported local control rates following post-operative SRS or FSRT to the resection cavity in patients with brain metastases. For series in which uniform dosing schedules were used, biologically effective doses (BED) were calculated using the linear quadratic model and assuming an α/β ratio of 10 Gy. Local control data for the individual resection cavities treated in each study were extracted from actuarial survival curves and aggregated to form a single dataset. Kaplan-Meier curves for local control in the entire dataset and after grouping by treatment platform, fractionation, use of margin expansion, and BED were generated, and statistical comparisons were made using the log-rank test. Seventeen studies (1104 patients with 1155 resection cavities) met all inclusion criteria. Median dose for single fraction regimens was 14-20 Gy, while FSRT was used in 8 studies with regimens ranging from 9 Gy x 3 to 4 Gy x 10 (BED range of 41 Gy10 to 56 Gy10). One and two-year actuarial LC rates for the entire cohort were 83% and 75%, respectively. FSRT was associated with slightly higher one-year (87% v. 81%) and two-year (78% v. 73%) actuarial LC rates than SRS (p=0.049). For patients treated with SRS, use of Gamma Knife was associated with inferior LC compared to treatment with linear accelerator-based platforms (p=0.002). There was no association between use of a planning margin and LC. Most SRS series utilized a wide range of doses, so dose-response analysis could not be performed. In FSRT series, there was no significant association between BED and LC. The overall toxicity rate, which predominantly included radiation necrosis, was approximately 7% and not significantly different between treatment platforms or fractionation schemes. Adjuvant SRS and FSRT provide high rates of local control following resection of brain metastases. FSRT should be considered in cases where therapeutic doses cannot be safely delivered in a single fraction. Prospective trials are warranted to further characterize the optimal regimen and treatment parameters.