Treatment of Cerebral Arteriovenous Malformations with Radiosurgery or Hypofractionated Stereotactic Radiotherapy in a Consecutive Pooled Linear Accelerator Series.

OBJECTIVE: To review outcomes after linear accelerator stereotactic radiosurgery (SRS) and hypofractionated stereotactic radiotherapy (hfSRT) of arteriovenous malformations (AVMs) from a consecutive and pooled series of 2 Novalis centers and to analyze the influence of AVM size, Spetzler-Martin (SM) grade, pretreatment, and hemorrhagic versus nonhemorrhagic presentation. A subgroup analysis of A Randomized Trial of Unruptured Brain Arteriovenous Malformations (ARUBA)-eligible patients also was performed.

METHODS: Prospectively collected treatment and outcome data were supplemented by retrospectively collected follow-up data for 93.8% of all patients. A total of 129 patients with AVM had SRS or hfSRT between 2000 and 2014 with the same linear accelerator system in 2 centers. Data analysis included initial presentation, SM grade, occlusion rates assessed by magnetic resonance and/or digital subtraction angiography, neurologic and therapeutic complications, and pretreatments. Statistical analysis was performed for patient demographic data and for factors potentially influencing outcome.

RESULTS: Initial presentation was hemorrhage in 43.8% or seizures/neurologic deficits in 46.2%. The series included 6 SM grade I (5%), 26 SM II (21.5%), 55 SM III (45.5%), 28 SM IV (23%), and 6 SM V cases (5%). Pre-embolization was used in 36 patients (29.8%), 8 patients had previous surgery (6.6%), and 6 patients were irradiated before elsewhere (5%); 5 patients (4.2%) received multimodal pretreatment. Mean follow-up was 43 months. The occlusion rate for the total series was 71.1%, for SM I/II cases 80.6%, and 67.4% for the SM ≥ subgroup. The occlusion rate was 75.0% for the small volume (<4 cc) and 55.6% for the large volume (>10 cc) subgroup. There was no statistical difference between the occlusion rate of patients with or without pretreatment if taken all modalities together (72.7% and 69.7%, respectively). There was only a trend of a belated occlusion of pre-embolized AVMs. The occlusion rate for hemorrhagic AVM was with 77.4% better than for nonhemorrhagic (66.2%) or ARUBA-eligible AVMs (64.8%) but without reaching statistical significance. Neurologic deterioration was seen in 13.2% of the patients. There were 2 re-bleedings within 17-18 months (1.7%), 1 of them without a new neurologic deficit and total occlusion after re-SRS. One patient with pre-existing epilepsy died a sudden unexpected death (mortality rate: 0.8%).

CONCLUSIONS: Overall SRS and hfSRT are valuable therapy options, especially in symptomatic patients with AVM, with a low rate of morbidity and mortality and an acceptable overall complete occlusion rate of >70% and >80% for SM I/II AVMs.