Outcome of early versus late primary embolization in ruptured brain arteriovenous malformations

To determine whether patients with a ruptured brain arteriovenous malformation (rBAVM) would benefit from an early embolization.

rBAVM treated first by embolization between March 2002 and May 2022 were included. Embolization was defined early (Group 1) when performed within 10 days postbleeding. If later, embolization was considered late (Group 2). Demographic and rBAVM data were compared between the groups. High-risk bleeding components and reasons for deferring embolization were retrieved. Primary endpoint was rebleeding. Secondary endpoints were good functional outcome (FO, modified Rankin Scale mRS ≤ 2) and angiographic occlusion. Predictors of rebleeding and FO were determined by multivariate analysis.

105 patients were recruited (N = 34 in Group 1; N = 71 in Group 2). No rebleeding was noted before, during or after the first embolization session in the early embolization group. Late embolization depended on missed diagnosis and referral pattern. Eleven patients (10.5%) suffered a rebleeding, of whom N = 3 before embolization (only in Group 2), N = 5 periembolization (N = 2 at the second embolization session in Group 1) and N = 3 spontaneous more than 30 days postembolization. More high-risk components were embolized in Group 1 (19/34; 55.9 vs 17/71; 23.9%; p = .011). Rebleeding rates, FO at last FU (90.9% vs 74.3%) and occlusion rates (80.8% vs 88.5%) did not differ between the groups. Glasgow coma scale ≤ 8 predicted rebleeding, rebleeding correlated with poor FO.

Early embolization did prevent rebleeding. The overall rebleeding risk was linked to bleeding before late embolization and bleeding at the second embolization. Rebleeding predicted the final FO.