Abstract
Background Tools predicting intracranial dural arteriovenous fistulas (dAVFs) treatment outcomes remain scarce. This study aimed to use a multicenter database comprising more than 1000 dAVFs to develop a practical scoring system that predicts treatment outcomes. Methods Patients with angiographically confirmed dAVFs who underwent treatment within the Consortium for Dural Arteriovenous Fistula Outcomes Research-participating institutions were retrospectively reviewed. A subset comprising 80% of patients was randomly selected as training dataset, and the remaining 20% was used for validation. Univariable predictors of complete dAVF obliteration were entered into a stepwise multivariable regression model. The components of the proposed score (VEBAS) were weighted based on their ORs. Model performance was assessed using receiver operating curves (ROC) and areas under the ROC. Results A total of 880 dAVF patients were included. Venous stenosis (presence vs absence), elderly age (<75 vs ≥75 years), Borden classification (I vs II-III), arterial feeders (single vs multiple), and past cranial surgery (presence vs absence) were independent predictors of obliteration and used to derive the VEBAS score. A significant increase in the likelihood of complete obliteration (OR=1.37 (1.27-1.48)) with each additional point in the overall patient score (range 0-12) was demonstrated. Within the validation dataset, the predicted probability of complete dAVF obliteration increased from 0% with a 0-3 score to 72-89% for patients scoring ≥8. Conclusion The VEBAS score is a practical grading system that can guide patient counseling when considering dAVF intervention by predicting the likelihood of treatment success, with higher scores portending a greater likelihood of complete obliteration. © Author(s) (or their employer(s)) 2024. No commercial re-use. See rights and permissions. Published by BMJ.