Impact of paravalvular leak following transcatheter aortic valve replacement on one-year mortality: analysis of the combined PARTNER cohorts

Purpose: To examine severity of paravalvular leak (PVL) and association with 1yr mortality and other outcomes after transcatheter aortic valve replacement (TAVR) with the Edwards Sapien valve in the combined PARTNER cohorts. Methods: 2553 patients underwent TAVR in the randomized or non-randomized cohorts of the PARTNER trial. Discharge or 7-day echos were available for core lab analysis in 2270 patients. PVL was graded using semi-quantitative criteria as none/trace, mild, moderate or severe. Patients were followed for ≥ 1yr; clinical events were CEC adjudicated. The impact of PVL on mortality was evaluated using a cox proportional hazards model. Results: PVL was graded as none/trace in 53.1%, mild in 38.1% and mod/sev in 8.8%. There were no differences in mean STS risk score between groups, but in mod/sev and mild (vs none/trace) there were more males, larger BSA, and more pre-treatment renal disease. 30-day/in-hospital mortality (3.4% vs 3.9% vs 5.5%, p=0.33) and stroke (3.2% vs 3.7% vs 2.5%, p=0.64) were similar in all groups (none/trace, mild, mod/sev). At one year, there was increased all-cause mortality (14.1% vs 22.1% vs 34.1%, p<0.0001), cardiac mortality (5.1% vs 7.4% vs 18.2%, p<0.0001) and rehospitalization (12.5% vs 22.2% vs 30.7%, p<0.0001) with worsening PVL (figure). Multivariate analysis indicated that the presence of mod/sev (HR 3.58 2.42, 5.29, p<0.0001) or mild PVL (HR 1.75 1.28, 2.39, p<0.001) was associated with higher late mortality. All clinical outcomes, echocardiographic results, and results stratified by approach (transfemoral or transapical) will be presented. Figure 1 Figure 1 Conclusions: Following TAVR, even mild PVL is associated with increased one-year all-cause and cardiac mortality. Refinements to reduce PVL following TAVR may improve clinical outcomes in the future.