LAUSR

BACKGROUND We reviewed frequency, microbiological pattern, predictors, and outcomes of early infections following transcatheter aortic valve replacement (TAVR). METHODS Five hundred thirty-nine patients who underwent successful TAVR at a single, high-volume center between January 2014 and December 2019 were enrolled. We defined early infections as occurring within 30-day from TAVR. RESULTS Mean age was 83.5 ± 5.4 years; 230 (42.7%) patients were men. Median follow-up was 12.0 (5.7-18.3) months; 30-day and 1-year death rates were 8/539 (1.5%) and 30/539 (5.6%), respectively. Early infections occurred in 61/539 (11.3%) patients, of whom 2 had infections in two sites. Of the 63 infections, 10 were bloodstream infections (BSI), 5 urinary tract, 27 pulmonary (2 with sepsis), 6 access site infections, 1 enterocolitis, and 14 were clinically diagnosed (no specific site). We observed 31/63 (49.2%) microbiologically-documented infections: Gram+ bacteria were isolated in 12/31 (38.7%), Gram- in 17/31 (54.3%), both Gram+ and Gram- in 2/31 (6.5%); in thirty-two infections no specific pathogen could be isolated (clinically-documented infections). Early infections were more prevalent in patients who died within 30-day (8.2% vs. 0.6%, p < 0.001) or 1-year (14.8% vs. 4.4%, p < 0.001) from TAVR. At multivariable analysis, early infections were independently associated with 30-day (HR: 8.82, 95%CI: 1.11-19.83, p = 0.035) and 1-year mortality (HR: 2.10, 95%CI: 1.28-6.21, p = 0.041). The predictive value for 1-year mortality was maintained even restricting the analysis to documented infections only, or to more severe infections (BSI and pneumonia only) (all p < 0.05). CONCLUSIONS Early infections occur in 1/10th of TAVR and are associated with increased short- and long-term mortality. Whereas a causal relationship between early infections and the risk of death cannot be unequivocally proven, careful surveillance of infected patients may improve TAVR results.