120 Rates of Re-hemorrhage, Risk Factors, and Outcomes of Previously Ruptured Arteriovenous Malformations (AVMs)

Abstract

Abstract INTRODUCTION The annual hemorrhage rate of unruptured AVMs is well established. However, hemorrhage risk of a previously ruptured AVM is not well defined. In this study, we describe the rate of AVM re-hemorrhage as well as the risk factors and outcomes associated with AVM re-hemorrhage from an institutional cohort. METHODS We retrospectively reviewed AVM patients from our institution seen from 1990–2015. Patients with hemorrhagic presentation were stratified into 1) patients with subsequent re-hemorrhage prior to treatment and 2) patients without re-hemorrhage. Patient demographic, clinical and angiographic data was compared across study groups. Cumulative and annual re-rupture rates were determined using Kaplan-Meier survival analysis. RESULTS >Of the 247 patients with hemorrhagic presentation, 29(11.7%) re-bled prior to treatment. There were no statistically significant differences between groups for age, gender, race, Spetzler-Martin grade, eloquence involvement, deep venous drainage, venous stenosis, AVM location, feeding artery aneurysms and intranidal aneurysms. AVM size was larger in the re-hemorrhage group (3.5cm+/−2.1 vs. 2.5cm+/−1.4, P = 0.032). Multivariate analysis identified age(HR1.03[CI1.00-1.05], P = 0.043), Hispanic race(HR5.94[CI1.34-26.45], P = 0.019), AVM size (HR1.31[1.02-1.68], P = 0.035) and venous stenosis (HR3.9[1.03-14.73], P = 0.045) as risk factors for re-hemorrhage. Presence of a venous varix (HR0.04[CI0.00-0.37], P = 0.005 and significant venous dilatation (HR0.27[0.09-0.80], P = 0.018) decreased the risk of re-hemorrhage. The cumulative risk of rupture at 2, 4, 12, 26 and 52 weeks was 0.9%, 1.3%, 2%, 3.2% and 5.5% respectively. The annual rate of re-rupture decreased from 7.09%(year 1) to 2.89%(year 10). Outcomes (dichotomized as good((modified Rankin score (mRS) 0–2) or poor (mRS 3–6)) were significantly improved in the patients without re-hemorrhage(79.8%) compared to re-hemorrhage patients(58.6%)(P = 0.010). CONCLUSION The rate of re-rupture is low immediately following initial hemorrhage of an AVM, with increasing cumulative risk over time. Outcomes in patients are significantly improved if there is no re-hemorrhage. This study helps identify the rates and risk factors associated with AVM re-hemorrhage, such that timing of treatment of a ruptured AVM may be optimized.