Radiosurgery for ruptured intracranial arteriovenous malformations

Abstract

Object Ruptured intracranial arteriovenous malformations (AVMs) are at a significantly greater risk for future hemorrhage than unruptured lesions, thereby necessitating treatment in the majority of cases. In a retrospective, single-center study, the authors describe the outcomes after radiosurgery in a large cohort of patients with ruptured AVMs. Methods From an institutional review board–approved, prospectively collected AVM radiosurgery database, the authors identified all patients with a history of AVM rupture. They analyzed obliteration rates in all patients in whom radiological follow-up data were available (n = 639). However, to account for the latency period associated with radiosurgery, only those patients with more than 2 years of radiological follow-up and those with earlier AMV obliteration were included in the analysis of prognostic factors related to obliteration and complications. This resulted in a cohort of 565 patients with ruptured AVMs for whom data were analyzed; these patients had a median radiological follow-up of 57 months and a median age of 29 years. Twenty-one percent of the patients underwent preradiosurgery embolization. The median volume and prescription dose were 2.1 cm3 and 22 Gy, respectively. The Spetzler-Martin grade was III or higher in 56% of patients, the median radiosurgery-based AVM score was 1.08, and the Virginia Radiosurgery AVM Scale (RAS) score was 3 to 4 points in 44%. Survival and regression analyses were performed to determine obliteration rates over time and predictors of obliteration and complications. Results In the overall population of 639 patients with ruptured AVMs, the obliteration rate was 11.1% based on MRI only (71 of 639 patients), 56.0% based on angiography (358 of 639), and 67.1% based on combined modalities (429 of 639 patients). In the cohort of patients with 2 years of follow-up or an earlier AVM obliteration, the cumulative obliteration rate was 76% and the actuarial obliteration rates were 41% and 64% at 3 and 5 years, respectively. Multivariate analysis identified the absence of preradiosurgery embolization (p < 0.001), increased prescription dose (p = 0.001), the presence of a single draining vein (p = 0.046), no postradiosurgery-related hemorrhage (p = 0.007), and lower Virginia RAS score (p = 0.020) as independent predictors of obliteration. The annual risk of a hemorrhage occurring during the latency period was 2.0% and the rate of hemorrhage-related morbidity and mortality was 1.6%. Multivariate analysis showed that decreased prescription dose (p < 0.001) and multiple draining veins (p = 0.003) were independent predictors of postradiosurgery hemorrhage. The rates of symptomatic and permanent radiation-induced changes were 8% and 2.7%, respectively. In the multivariate analysis, a single draining vein (p < 0.001) and higher Virginia RAS score (p = 0.005) were independent predictors of radiation-induced changes following radiosurgery. Conclusions Radiosurgery effectively treats ruptured AVMs with an acceptably low risk-to-benefit ratio. For patients with ruptured AVMs, favorable outcomes are more likely when preradiosurgical embolization is avoided and a higher prescription dose can be delivered.