Results of Endovascular-First Approach for Treatment of Scalp Arteriovenous Malformations and the Effect of Simplified Angiographic Classification in Therapeutic Decision-Making and Outcomes

Abstract

Scalp arteriovenous malformation (SAVM) is primarily treated by surgery and reports on endovascular treatment are scarce. We report the results of an endovascular-first approach in the treatment of SAVM. We also have proposed a simple dichotomized classification using the angiographically determined morphology findings and discussed its effect on therapeutic decision-making and outcomes. A consecutive series of 25 patients with SAVM treated by initial or endovascular-only methods were included in the final analysis. The SAVM was categorized as a plexiform or fistulous type according to the angiographic morphology. The demographic data, treatment variables, and final clinical outcomes were analyzed and correlated with the proposed classification. Of the 25 patients, 21 were treated with n-butyl cyanoacrylate (n-BCA) and 3 with liquid ethylene vinyl alcohol. One patient had been treated initially with polyvinyl alcohol particles and later with n-BCA. Overall, complete or near complete obliteration (>90%) was achieved in 72%. Surgical excision was performed in 76%, most often after n-BCA embolization. Fistulous-type SAVMs required greater n-BCA concentrations (median, 33% vs. 20%; P= 0.024) and achieved greater rates of complete or near-complete obliteration (90% vs. 63%). Two patients with fistulous-type SAVMs treated with liquid ethylene vinyl alcohol showed complete resolution, and further treatment was not necessary. Overall, regardless of the embolic material used, the fistulous type demonstrated a significantly greater rate of complete obliteration (75% vs. 33%; P= 0.041) compared with plexiform type. High rates of complete and durable obliteration of SAVM are achievable with endovascular embolization. The proposed simplified classification is easy to implement and can aid in choosing the appropriate embolic agent and predicting the therapeutic outcome.