Comparative treatment outcomes of management modalities for unruptured brain arteriovenous malformations: a systematic review and meta-analysis of 6,124 patients

BACKGROUND: The optimal management of unruptured brain arteriovenous malformations (AVMs) remains controversial, with conflicting evidence regarding conservative versus interventional approaches, particularly following the ARUBA trial. This meta-analysis aims to provide comprehensive, grade-individualized outcomes across treatment modalities to guide clinical decision-making. METHODS: We conducted a PRISMA-compliant meta-analysis of 16 studies (n = 6,124 patients) comparing conservative management, microsurgery, stereotactic radiosurgery (SRS), embolization, and hybrid therapy. Primary outcomes were hemorrhage risk and obliteration rates; secondary outcomes included functional status (mRS ≥2), new-onset seizures, and mortality. . RESULTS: Microsurgery achieved the highest obliteration rates (99% for Spetzler-Martin [SM] I-II, 86% for SM-III) and lowest hemorrhage risk (0.6%/yr for SM I-II, 1.3%/yr for SM-III), with superior functional outcomes (8% disability [95% CI 5-11%] vs. 11% for conservative management; p = 0.04). SRS showed moderate efficacy (76% obliteration for SM I-II, 47% for SM III-IV) with higher latency-period hemorrhage risk (1.1-2.1%/yr) and disability rates comparable to conservative management (13% [95% CI 10-16%]; p = 0.18). New-onset seizures occurred in 9.1% of microsurgery and 4.9% of SRS cases. Mortality was low across treatments (1-2%), with conservative management at 1.1% and SRS at 1.6%. Conservative management had a 1.9-2.6%/yr hemorrhage risk without procedural morbidity. . CONCLUSIONS: Among the available published evidence, microsurgery was associated with the most favourable pooled outcomes for operable SM I-II AVMs, including high obliteration rates, lower haemorrhage rates, and favourable functional outcomes, despite a higher seizure risk. SRS remains an important treatment option for inoperable lesions, while conservative management appears reasonable for high-grade AVMs in the absence of high-risk features.