Embolization of Brain Arteriovenous Malformations With Versus Without Onyx Before Stereotactic Radiosurgery.

Abstract

Embolization of brain arteriovenous malformations (AVMs) using ethylene-vinyl alcohol copolymer (Onyx) embolization may influence the treatment effects of stereotactic radiosurgery (SRS) differently than other embolysates. To compare the outcomes of pre-SRS AVM embolization with vs without Onyx through a multicenter, retrospective matched cohort study. We retrospectively reviewed International Radiosurgery Research Foundation AVM databases from 1987 to 2018. Embolized AVMs treated with SRS were selected and categorized based on embolysate usage into Onyx embolization (OE+SRS) or non-Onyx embolization (NOE+SRS) cohorts. The 2 cohorts were matched in a 1:1 ratio using de novo AVM features for comparative analysis of outcomes. The matched cohorts each comprised 45 patients. Crude AVM obliteration rates were similar between the matched OE+SRS vs NOE+SRS cohorts (47%vs 51%; odds ratio [OR]=0.837, P=.673). Cumulative probabilities of obliteration were also similar between the OE+SRS vs NOE+SRS cohorts (subhazard ratio=0.992, P=.980). Rates of post-SRS hemorrhage, all-cause mortality, radiation-induced changes, cyst formation, and embolization-associated complications were similar between the matched cohorts. Sensitivity analysis for AVMs in the OE+SRS cohort embolized with Onyx alone revealed a higher rate of asymptomatic embolization-associated complications in this subgroup compared to the NOE+SRS cohort (36%vs 15%; OR=3.297, P=.034), but the symptomatic complication rates were similar. Nidal embolization using Onyx does not appear to differentially impact the outcomes of AVM SRS compared with non-Onyx embolysates. The embolic agent selected for pre-SRS AVM embolization should reflect both the experience of the neurointerventionalist and target of endovascular intervention.