PURPOSE: The implications of margin expansions on local control(LC) and radionecrosis(RN) for treating brain metastases with stereotactic radiosurgery(SRS) remain unclear. We performed a systematic review and meta-analysis to compare LC and RN between patients with brain metastases treated with stereotactic radiosurgery(SRS) planned with no margin vs. a margin. METHODOLOGY: We utilized PICOS/PRISMA/MOOSE selection inclusion criteria for studies of patients with brain metastases treated with SRS with no margin or a margin. Primary outcomes were 1-year LC and radiographic and symptomatic RN incidences. Weighted random effects meta-analyses were performed to compare effect sizes. RESULTS: Across 17 studies, we identified 5,015 lesions treated with SRS (1,360 lesions with no margin and 3,684 with a margin). The median total margin was 1.5 mm (range: 1-3). Single fraction SRS was most common with a median prescription dose of 21 Gy (range: 15-24 Gy). The estimated 1-year LC rate was similar with a margin (88.4% (95% CI: 83.7-92.4%) vs. without (83.0% (95% CI: 69.3-93.2%
p = 0.28)). Estimated incidences of radiographic RN following SRS with no margin vs. a margin were similar at 9.2% (95% CI: 0.2-29.6%) and 7.0% (95% CI: 4.1-10.7%
p=0.56), respectively. Estimated incidences of symptomatic RN following SRS with no margin vs. with a margin were 8.6% (95% CI: 5.2-12.7%) and 4.1% (95% CI: 0.9-9.3%), respectively (p=0.24). CONCLUSIONS: Significant differences were not noted in either LC or RN for patients treated with margin expansions vs. without. Prospective evaluations are warranted to further assess this question while controlling for other relevant treatment planning and metastasis considerations.