Paravalvular leak (PVL) is an important complication of transcatheter aortic valve implantation (TAVI) and is associated with poor prognosis. We aimed to identify the risk factors for PVL after TAVI including patient (calcium amount or location), device (leakage-proof or not), and procedural (oversizing index (OI)) factors. The primary outcome was mild or greater PVL at 1-month follow-up echocardiography. Overall, 238 patients who underwent TAVI using eight types of valves (Edwards Sapien, Sapien XT, Sapien 3, CoreValve, Evolut R, Evolut PRO, Lotus, and Lotus Edge) were included. The incidence of significant PVL (≥mild PVL) was 24.4%. Although patient factors (calcification of valve) were not predictors of PVL, valve without leakage-proof function (Edwards Sapien, Sapien XT, and CoreValve) was a significant predictor of PVL (adjusted odds ratio, 3.194, 95% CI, 1.620-6.299). Furthermore, OI has a significant protective role against PVL (PVL increased by 45% when OI decreased by 5%). The best cutoff value of OI to predict the absence of PVL was ≥17.6% for the Evolut system and ≥10.2% for the Sapien system. The predictors of PVL after TAVI included factors from the device (valve without leakage-proof function) and procedure (under-sizing). In patients with a high risk of PVL, the procedure should be optimized using valves with leakage-proof function and adequate OI.
Keywords: leakage-proof function; oversizing index; paravalvular leakage; transcatheter aortic valve implantation.