Abstract
Background: Brain metastases are a major driver of cancer-associated morbidity and mortality, occurring in 20-40% of all malignancies. Objective: Systematic review and meta-analysis of randomized controlled trials (RCTs) comparing whole brain radiotherapy (WBRT), stereotactic radiosurgery (SRS), or observation after resection of 1-4 brain metastases. Methods: PRISMA-compliant search of Embase and MEDLINE from inception to August 2024. Overall survival (OS), cognitive decline (CD), surgical bed control (SBC), and intracranial control (IC) were assessed by pooling hazard ratios (HR) or 12-month post-SRS odds ratios (OR) via meta-analysis of proportions with random-effects modeling. Results: Of 1319 unique, English-language abstracts, 37 underwent full-text review, and 7 were included, representing 812 patients in three paired comparisons: WBRT vs. observation (4 RCTs, n = 545), WBRT vs. SRS (2 RCTs, n = 253), and SRS vs. observation (1 RCT, n = 132). Pooled HR were not significant for OS (WBRT vs. observation, HR = 1.01, 95% CI = 0.87-1.18; WBRT vs. SRS, 0.77, 95% CI = 0.47-1.26) or CD (WBRT vs. observation, HR = 0.98, 95% CI = 0.81-1.17; WBRT vs. SRS, 1.31, 95% CI = 0.48-3.60). Pooled OR favored WBRT over both observation and SRS with respect to SBC (SBC: WBRT vs. observation, OR = 1.35, 95% CI = 1.05-1.74; WBRT vs. SRS, OR = 1.46,95% CI = 1.00-2.13) and IC (IC: WBRT vs. observation, OR = 1.22, 95% CI = 1.06-1.40; WBRT vs. SRS, OR = 1.22, 95% CI = 1.03-1.59). GRADE assessment demonstrated very low to moderate certainty for all reported outcomes. Conclusions: Even among RCTs, the best available evidence regarding adjuvant treatment for oligometastatic brain disease is heterogeneous, imprecise, and inconclusive. WBRT improves SBC and IC, but not OS. SRS may decrease CD risk, which requires validation in double-blind RCTs incorporating precise neurocognitive metrics.