Despite major technological advances, 1-year morbi-mortality after transcatheter aortic valve implantation (TAVI) is still high. Further outcome improvements may be obtained provided proper identification of prognostic factors. A change of prognostic value over time should be the hallmark of an outcome-implicated factor. In 1,425 patients treated by TAVI, the examined prognostic factors were: demographic factors and co-morbidities (age, male gender, glomerular filtration rate, and chronic obstructive pulmonary disease), cardiac function (left ventricular ejection fraction, pulmonary pressure, aortic gradient, dyspnea, and mitral regurgitation), and vascular factors (coronary artery disease, peripheral vascular disease (PVD), previous stroke, and thoracic aortic calcium-TAC-as assessed by CT scan). Cox models were used to analyze cardiovascular and all-cause mortalities over 3 years of follow-up. The time-dependent effects of the factors were analyzed using the distribution of Schoenfeld residuals. During the study period, 375 (26.3%) deaths occurred of whom 248 (17.4%) from cardiovascular causes. Only 2 factors associated with cardiovascular or all-cause mortality showed significant changes over time: dyspnea and PVD. The effect of dyspnea on cardiovascular mortality decreased over time (first- and third-year hazard ratios [95% confidence intervals]: 1.47 [1.10; 1.96] and 0.94 [0.55; 1.63], respectively), whereas the effect of PVD increased (first- and third-year hazard ratios: 0.87 [0.56; 1.35] and 2.58 [1.25; 5.33], respectively). TAC had a stable effect. In conclusion, the detrimental effects of vascular factors remained stable (TAC) or increased (PVD) over time. These factors should be targeted by specific measures to improve post-TAVI outcomes.
Copyright © 2020 Elsevier Inc. All rights reserved.