Image-guided robotic radiosurgery for the treatment of arteriovenous malformations

Background Cerebral arteriovenous malformations (AVMs) are challenging lesions, often requiring multi-modal interventions; however, data on the efficacy of stereotactic radiosurgery for cerebral AVMs are limited. This study aimed to evaluate the clinical and radiographic results following robotic radiosurgery, alone or in combination with endovascular treatment, and to investigate factors associated with obliteration and complications in patients with AVM. Methods We retrospectively analyzed the clinical and imaging characteristics of 123 patients with AVMs of all Spetzler-Martin grades treated at two institutions by robotic radiosurgery in single-fraction doses (CyberKnife). Embolization was performed before radiosurgery in a subset of patients to attempt to downgrade the lesions. Factors associated with AVM obliteration and complications (toxicity) were identified via univariate and multivariate analyses. Results The median follow-up time was 48.1 months (range, 3.6-123 months). Five patients were lost to follow-up. The obliteration rate in the 59 patients with a follow-up period exceeding four years was 72.8%. Complete obliteration and partial remission were achieved in 67 (56.8%) and 31 (26.3%) cases, respectively, whereas no change was observed in 20 cases (17.8%). Embolization was performed in 54/123 cases (43.9%). Complete and partial obliteration were achieved in 29 (55.7%) and 14 (26.9%) embolized patients, respectively. In the multivariate analysis, the factors associated with obliteration were age (p = .018) and the Spetzler-Martin grade (p = .041). Treatment-induced toxicity (radiation necrosis and/or edema) was observed in 15 cases (12.7%), rebleeding occurred in three cases (2.5%), and the rate of mortality associated with rebleeding was 1.7%. Conclusions CyberKnife radiosurgery is a valid approach for treating AVMs of all Spetzler-Martin-grades, with satisfactory obliteration rates, low toxicity, and a relatively rare incidence of rebleeding.