Gamma knife surgery with and without embolization for cerebral arteriovenous malformations: A systematic review and meta-analysis.

Highlights • Embolization prior to GKS significantly may decrease the AVM obliteration rate. • Embolization prior to GKS didn't reduce the risk of hemorrhage compared to CKS alone in treating AVM. • Embolization prior to GKS didn't reduce the risk of permanent neurological deficits compared to CKS alone in treating AVM. Abstract The benefit and risk of gamma knife surgery (GKS) in the treatment of residual cerebral arteriovenous malformations (AVMs) after endovascular embolization remain controversial. The aim of this meta-analysis was to assess current evidence regarding the efficiency and safety of GKS for AVMs with and without prior embolization. To compare GKS in patients with and without embolization, the authors conducted a meta-analysis of studies by searching the literature via PubMed and EMBASE databases for the period between January 2006 and December 2017. Six retrospective studies were finally identified. Outcomes were the rate of AVM obliteration on a 3-year follow-up angiogram, hemorrhage at 3 years after GKS and permanent neurological deficits. Six studies eligible for analysis included 2069 patients: 637 had undergone embolization followed by GKS, and 1432 had undergone GKS alone. The obliteration rate was significantly lower in patients who had undergone embolization followed by GKS than in those who had undergone GKS alone (49.5% vs 70.4%, OR 2.29, 95% CI 1.55–3.38, p < 0.00001). Subgroup analysis also indicated high obliteration rates in 'similar mean nidus volume', 'high quality' and 'sample size over 100 patients' subgroups. However, the rates of rehemorrhage (8.9% vs 4.2%, OR 0.59, 95% CI 0.23–1.57, p = 0.29) and permanent neurological deficits rate (3.6% vs 4.6%, OR 0.51, 95% CI 0.57–3.12, p = 0.51) were not significantly different between the two groups and subgroups. Embolization prior to GKS significantly decreases the AVM obliteration rate and didn't reduce the risk of hemorrhage and permanent neurological deficits. Further evaluation by well-designed prospective or randomized cohort studies is highly needed. [ABSTRACT FROM AUTHOR]