Delayed intracerebral hemorrhage after uneventful embolization of brain arteriovenous malformations is related to volume of embolic agent administered: multivariate analysis of 13 predictive factors.

Abstract

The mechanisms and management of delayed intracerebral hemorrhage (dICH) after treatment of brain arteriovenous malformations (AVMs) are poorly understood and widely debated. Many clinical predictive factors have been theorized for dICH after an otherwise uneventful AVM embolization, but there is an absence of data to discern their significance. To analyze 13 proposed predictive factors and to assess their potential in guiding prevention strategies. One hundred sixty-eight embolization procedures were performed on 67 patients with brain AVMs by a single surgeon. Patients were divided into 2 groups: those with symptomatic dICH and control subjects. Thirteen factors were analyzed: age, sex, race, previous ICH, Spetzler-Martin grade, AVM size, eloquence, embolic volume, embolic agent, percent obliteration, and timing, number, and stage of embolizations. Univariate and multivariate analyses were performed on these factors to determine significance. Six procedures were complicated by dICH; 5 (83%) occurred after the final planned procedure. The volume of embolic agent was significantly higher in the dICH group (4.5 ± 1.0 mL) compared with control subjects (1.7 ± 0.2 mL) in both univariate and multivariate analyses (P < .01), even after controlling for AVM size. AVM size was significant in univariate analysis but not multivariate analysis. There were no statistically significant differences between the groups for any of the other possible predictive factors. High volume of embolic agent administered per procedure is an independent predictive factor for dICH. Limiting the injected volume for each procedure may reduce this poorly understood complication.