Introduction/purpose The optimal management of unruptured AVMs (Medical vs. Interventional) remains unclear. The recent ARUBA trial suggested superiority of medical management, but has been criticized for poor generalizability and low numbers. In this analysis, we evaluate real world population-level rates of outcomes in unruptured AVMs by management strategy.

Materials and methods Using the Healthcare Cost and Utilization Project database on all adult (age ≥18) discharges from acute care hospitals in California (2005–2011) and Florida (2005–2014), we identified patients with unruptured AVMs. Patients were excluded for any preceding diagnosis of intracerebral hemorrhage (ICH), subarachnoid hemorrhage or trauma. The primary endpoint was any ischemic stroke, hemorrhage, or death during the 3 year follow up period.

• Download figure
• Open in new tab
• Download powerpoint

Abstract O-015 Figure 1

Results Among 7241 patients with unruptured cerebral AVMs, median age was 53 [IQR 40–67], 3764 (54%) were female, 4189 (62%) were white, 714 (11%) were black and 1369 (20%) were Hispanic. Overall, 1454 (20%) underwent surgical treatments, 1343 (19%) embolization, and 4849 (67%) were treated medically. Three-year rates of stroke, cerebral hemorrhage or death were 10.9% in the medically treated group and 16.5% in the procedurally-managed group. In multivariate analysis, likelihood of stroke, hemorrhage or death was associated with age (OR 2.7 [95% CI 2.2–3.2]≥70 years vs. <50 years), female sex (OR 0.86 [95% CI 0.74–0.99]), black race (OR 1.44 [95% CI 1.14–1.81] vs. white race), and interventional management (OR 2.2 [95% CI 1.92–2.60] vs. medical). In three-year survival analysis adjusted for age, sex, race, the difference in outcome between the two management strategies is driven primarily by peri-procedural events (figure 1a). Excluding patients with peri-procedural events, the risk of stroke, hemorrhage or death grew at a slower rate for patients in the procedurally-managed group (figure 1b, 1.5%/year vs. 2.4%/year, intervention vs. medical).

Conclusions In this large, real world cohort of patients with unruptured cerebral AVMs, rates of stroke, death or hemorrhage were greater in patients managed with intervention relative to medical approaches. These differences were driven largely by peri-procedural events, after which the rates of these outcomes decreased relative to the medical arm, suggesting patients with longer life expectancy and lower peri-procedural risk may benefit from treatment.

Disclosures H. Saber: None. A. Yoo: None. P. Chen: None. F. Vahidy: None. S. Savitz: None. S. Sheth: None.