Quantitative assessment of changes in cerebral arteriovenous malformation hemodynamics after embolization.

Abstract

Embolization reduces flow in arteriovenous malformations (AVMs) before surgical resection, but achievement of this goal is determined subjectively from angiograms. Here, we quantify effects of embolization on AVM flow. Records of patients who underwent AVM embolization at our institution between 2007 and 2013 and had flow rates obtained pre- and postembolization using quantitative magnetic resonance angiography were retrospectively reviewed. Total flow was estimated as aggregate flow within primary arterial feeders or flow in single draining veins. Twenty-one patients were included (mean age 35 years, 24% hemorrhagic presentation) with Spetzler-Martin grades 1 to 4. Fifty-four total embolization sessions were performed. The mean AVM flow was 403.4±262.4 mL/min at baseline, 285.3±246.4 mL/min after single session (29% drop, P<0.001), and 102.0±103.3 mL/min after all sessions of embolization (75% drop, P<0.001). Total number of pedicles embolized (P<0.001) and embolization of an intranidal fistula during any session (P=0.002) were significantly associated with total decreased flow postembolization. On multivariate analysis, total pedicles embolized was predictive of total flow drop (P<0.001). However, pedicles embolized per session did not correlate with flow drop related to that session (P=0.44). AVM flow changes after embolization can be measured using quantitative magnetic resonance angiography. The total number of pedicles embolized after multiple embolization sessions was predictive of final flow, indicating this parameter is the best angiographic marker of a hemodynamically successful intervention. The number of pedicles embolized per session, however, did not correlate with flow drop in that session, likely because of flow redistribution after partial embolization.