O-024 Multimodal brain arteriovenous malformation treatment – 12-year experience and comparison of key outcomes to the aruba trial

Introduction and hypothesis Controversy remains about generalizability of the results of the ARUBA trial, which demonstrated superior outcomes of patients with brain arteriovenous malformations (AVMs) treated with medical management compared to intervention. We sought to investigate clinical outcomes in a 12 year cohort of patients treated at our tertiary academic referral center in the US with a multimodal approach, and to compare results to key outcomes of the ARUBA trial. Methods We retrospectively analyzed 256 consecutive patients with AVMs who presented to our center and were treated with embolization, surgery, or proton-beam radiosurgery between 2003 and 2015. Clinical, demographic, and imaging data (angiograms, CT, and MR) were reviewed and Spetzler-Martin grade determined. Good outcome was defined as modified Rankin Scale of 0–2. Sub-analysis was performed on ARUBA eligible patients (baseline mRS 0–1, no history of hemorrhage). Stroke and death rates were compared to the as-treated ARUBA arms. Results A total of 256 patients (101 or 54.9% were males) were identified. Average age at presentation was 40.7±17.7 years, 115 patients (44.9%) presented with hemorrhage, 44 (17.2%) with seizures, 52 (21.1%) with headaches, and 54 (16.8%) had other symptoms at presentation. 140 (54.7%) AVMs were located on the left side, median Spetzler-Martin grade was 2 (interquartile range 2–3). Twenty-two patients (8.6%) were treated with embolization, 34 (13.3%) with surgery, 143 (55.9%) with radiosurgery, 34 (13.3%) with embolization and surgery, and 23 (9.0%) with embolization and radiosurgery. Complete obliteration was achieved in 82 patients (49.3%) treated with radiosurgery. Average follow-up time was 4 years±2 years 6 months. 216 patients (84.4%) had a baseline mRS score of 0–2. At the end of follow-up, 217 patients (84.8%) had a mRS of 0–2. 28 patients (10.9%) had strokes (ischemic or hemorrhagic) after treatment, 6 patients (2.3%) deceased. One hundred thirty-one patients (51.2%) were ARUBA eligible (62 or 47.3% were males) with average age of 40.0±15.9 years. 8 patients (6.1%) were treated with embolization, 10 patients (7.6%) with surgery, 88 (67.2%) with radiosurgery, 17 (13.0%) with embolization and surgery, and 8 (6.1%) with embolization and radiosurgery. Complete obliteration was achieved in 39 patients (40.6%) treated with radiosurgery. One hundred twenty-four patients (94.7%) had a follow-up mRS of 0–2. Thirteen patients (9.9%) suffered from a stroke (ischemic or hemorrhagic) post-treatment, three patients (2.3%) deceased. Compared to the as-treated interventional management group from the ARUBA trial, incidence of stroke was significantly lower (35.7 vs. 9.9%, p Conclusion Our multimodal approach to brain AVM patient selection yields good clinical outcome with key safety endpoints (stroke, death) similar to the medical management arm in ARUBA, and significantly lower stroke incidence than the intervention arm in ARUBA. This suggests that expertise in patient selection to different treatment approaches may allow for better clinical outcome than reported in ARUBA, and may warrant further randomized controlled trials of brain AVM patients. Disclosures B. Pulli: 1; C; RSNA Research Resident Grant. P. Chapman: None. C. Ogilvy: None. A. Patel: None. C. Stapleton: None. T. Leslie-Mazwi: None. J. Hirsch: None. J. Rabinov: None.