Can we predict postprocedural paravalvular leak after E dwards SAPIEN transcatheter aortic valve implantation?

BACKGROUND: Postprocedural paravalvular leak (PVL) ≥ 2 has been shown to be associated with worse outcomes after transcatheter aortic valve implantation (TAVI). This study sought to identify predictive factors of postprocedural PVL ≥ 2 after TAVI with the Edwards valve. METHODS AND RESULTS: A total of 176 patients with Edwards TAVI (aged 83.4 ± 7.4 years, Logistic EuroSCORE [the Logistic European System for Cardiac Operative Risk Evaluation] 18.8 ± 12.0, transfemoral 54.5%) who had preprocedural multidetector computed tomography (MDCT) were studied. A PVL ≥ 2 was observed in 12.5% of cases. By multivariate analysis, only the valve calcification index (VCI) defined as aortic root calcification volume/body surface area (odds ratio [OR] = 1.002, 95% confidence interval [CI] = 1.001-1.004, P = 0.006) and the valve diameter/the calculated average annulus diameter (CAAD) by MDCT (OR = 0.683, 95% CI = 0.474-0.984, P = 0.041) were identified as independent predictors of postprocedural PVL ≥ 2. A score predicting postprocedural PVL ≥ 2 (PVL score) was determined by allotting one point when the valve diameter/CAAD ratio was <1.055 and one point when VCI was >418.4 mm(3) /m(2) , and summing all points accrued. Area under receiver-operator characteristic curves of PVL score was 0.71 (95% CI = 0.59-0.83, P < 0.01). The incidence of PVL ≥ 2 was 5.3% in patients with a PVL score of 0, 11.8% for a PVL score of 1 and 37.5% for a PVL score of 2. CONCLUSIONS: The only predictors of PVL ≥ 2 after Edwards valve implantation are the valve diameter/CAAD and VCI. The PVL score could prove to be an excellent tool for predicting the risk of PVL.