RESUMO
OBJECTIVES: Determine whether moderate or greater paravalvular leak (PVL) after transcatheter aortic valve replacement quantified using intraoperative transesophageal echocardiography (TEE) is associated with mortality and investigate the correlation between PVL grading using intraoperative TEE and postoperative transthoracic echocardiography (TTE). DESIGN: Retrospective, observational study. SETTING: Single academic institution. PARTICIPANTS: The study comprised adult patients undergoing elective transcatheter aortic valve replacement between April 2011 and February 2014. INTERVENTIONS: Patients were grouped by amount of PVL on intraoperative TEE into "significant" (moderate or greater) and "nonsignificant" (no, trivial, or mild) PVL groups. Demographics and patient characteristics were compared. Continuous variables were assessed with t-tests or Wilcoxon rank sum tests and categorical variables with the chi-square or Fisher exact test. A Cox proportional hazards model adjusted for EuroSCORE was used to test the independent association of PVL with late mortality, and covariate-adjusted survival curves were constructed. A Fleiss-Cohen-weighted kappa value was used to assess agreement between PVL grading using intraoperative TEE and postoperative TTE. MEASUREMENTS AND MAIN RESULTS: One hundred ninety-six patients were grouped into the "significant" (n = 22) or "nonsignificant" (n = 174) PVL group. Twenty patients (10%) died during the follow-up period. Significant PVL on either TTE (p = 0.62, hazard ratio 1.68, 95% confidence interval [CI] 0.22-12.85) or TEE (p = 0.49, hazard ratio 0.49; 95% CI 0.06-3.68) was not associated with a survival difference. Modest agreement was found between PVL on intraoperative TEE and postoperative TTE (kappa = 0.47, CI 0.37-0.57, p < 0.0001). CONCLUSIONS: Larger studies are needed to evaluate the association of PVL graded on intraoperative TEE with survival. There is modest agreement between the degree of PVL found on TEE and TTE.