Abstract 18537: Aortic Valve Repair Improves Mid-term Outcomes versus Replacement for Aortic Insufficiency: A Propensity Matched Study

Abstract

Objectives: Aortic valve (AV) repair has emerged as a feasible and attractive alternative to AV replacement in patients with aortic insufficiency (AI). However, little data exists comparing outcomes following repair versus replacement. We performed a single-center comparative study in patients undergoing AV repair or replacement for AI. Methods: Patients undergoing AV replacement (n=263) or repair (n=153) for AI between 2000 and 2014 were included. Patients ineligible for both repair and replacement surgeries and patients lacking perioperative data were excluded (n=155). To adjust for baseline differences, non-parsimonious logistic regression models were used to generate propensity scores. A 1:1 greedy matching algorithm was used to create 70 propensity matched pairs. Peri-operative and long-term clinical and echocardiographic outcomes were assessed, focusing on survival and valve-related events, and analyzed using Kaplan-Meier techniques. Primary endpoints included late mortality, myocardial infarction, stroke, and AV reoperation. Total follow-up exceeded 1100 patient-years in the entire cohort and 800 patient-years in the matched cohort. Results: The matched pairs were similar in baseline attributes including age (57±14 vs. 56±15, p=0.75), sex (Female: 17% vs. 25%, p=0.22), incidence of LV dysfunction (27% vs. 24%, p=0.38), severe AI (67% vs. 74%, p=0.42) and presence of cardiac risk factors. Replacements consisted of 53% mechanical, 41% bioprosthetic, and 6% homograft valves. Repairs included 61% valve-sparing root replacements, 70% cusp repairs with 43% bicuspid AVs. In hospital mortality (1% vs. 4%, repair vs. replacement, p=0.62) and perioperative outcomes were similar between groups. Four-year survival was 97±2% and 87±4% in repair vs. replacement groups (p=0.11). A total of 5 vs. 17 valve-related events occurred in repair vs. replacement groups. Freedom from primary endpoints at 4 years was 97±2% and 82±8% in the repair vs. replacement groups (p=0.03). Conclusions: In a propensity matched cohort, AV repair for AI was associated with a mid-term reduction in death and serious valve related complications compared to AV replacement. AV repair should be considered, when technically feasible, as the preferred treatment of AI.