Objective: The effective treatment of giant cerebral arteriovenous malformations (gAVMs) is challenging. The aim of this study was to determine the risk factors for 1-stage resection of gAVM and develop a reliable indicator for patient selection.
Methods: A prospectively maintained database of patients with AVM in our hospital was reviewed. The neuroradiological findings and clinical characteristics of the patients and lesions were analyzed with respect to postoperative functional deficits (FD). A novel blood oxygen level-dependent functional magnetic resonance imaging score and a diffusion tensor imaging (DTI) score were created to predict surgical outcomes. Furthermore, the long-term outcomes of gAVMs treated by other methods in the literature were reviewed.
Results: A total of 35 patients with 35 gAVMs were included. The mean diameter of the gAVMs was 64.8 ± 4.9 mm. In the univariate analysis, the functional magnetic resonance imaging score (P = 0.022) and DTI score (P = 0.003) were both significantly associated with long-term FD. The Spetzler-Martin score (P = 0.092) trended toward significance. Multivariate analysis revealed that a high DTI score (odds ratio, 2.19; 95% confidence interval, 1.08-4.46; P = 0.030) was the only independent risk factor that was correlated with long-term FD. The predictive effect of the DTI score (area under the curve = 0.822) is superior to that of the Spetzler-Martin score (area under the curve = 0.640) according to the receiver operating characteristic analysis, and the cutoff point was 2.5 (sensitivity = 0.860 and specificity = 0.867).
Conclusions: One-stage surgical resection of gAVMs in patients with a low DTI score (0-2) seems to be feasible. The DTI score could be a reliable indicator for patient selection.
Keywords: Diffusion tensor imaging; Functional magnetic resonance imaging; Giant cerebral arteriovenous malformations; Surgical outcomes.
Copyright © 2019. Published by Elsevier Inc.