Abstract
Higher B-type natriuretic peptide (BNP) levels at discharge predict higher cardiovascular events in transcatheter aortic valve implantation (TAVI) patients. However, it is not known whether the reduction in BNP levels effectively predicts prognosis. The purpose was to examine the predictive power of percentage changes in BNP levels for all-cause death and hospitalization of heart failure (HF) after TAVI in severe aortic stenosis (AS) patients. We analyzed 70 severe AS patients treated with TAVI who had a record of BNP > 200 pg/mL. Receiver operating characteristics (ROC) curves analysis for all-cause death and hospitalization for HF after TAVI revealed the cut-off percentage change in BNP, and we divided the study population into the "responder group" and the "non-responder group". The cut-off level for the percentage change in BNP evaluated by ROC analysis was a 40% decrease in BNP (AUC = 0.733, p < 0.001). There were 48 patients (68.6%) in the responder group and 22 patients (31.4%) in the non-responder group. Kaplan-Meier estimates showed that the responder group had lower all-cause death and hospitalization for HF than the non-responder group by a log rank test (all-cause mortality; p = 0.006, hospitalization rate for HF; p < 0.001). The predictor of the non-responder group using multivariate logistic regression analysis was AF (OR 4.2, 95% CI 1.15-16.2, p = 0.03). A reduction of BNP was associated with improved prognosis after TAVI.