One-year outcomes of a European transcatheter aortic valve implantation cohort according to surgical risk subanalysis of the SOURCE 3 registry

Abstract

BACKGROUND: Transcatheter aortic valve implantation is increasingly being used in patients at lower risk for surgery. We aimed to assess the distribution of surgical risk score categories in current clinical practice and their relationship with clinical outcomes and the calibration and discrimination power of both the logistic EuroSCORE (logES) and EuroSCORE II. The SOURCE 3 study is a European prospective registry of patients with severe aortic stenosis treated with the commercially available SAPIEN 3 transcatheter heart valve. METHODS AND RESULTS: Out of 1785 patients, 518 patients (low-surgical risk) had a baseline logES <10%, 691 (intermediate-surgical risk) had a logES 10% to 20%, and only 576 patients (high-surgical risk) had a logES ≥20%. Even if low-risk patients were younger compared with the other groups, the mean age was about 80 years old in each risk category. At 1 year, all-cause mortality was 10.3%, 11.4%, and 17.1% in low-, intermediate-, or high-surgical risk patients, respectively, while cardiac mortality was 5.3%, 7.7%, and 11.4%, respectively. Observed mortality rates were substantially lower than that predicted with logES. The observed/predicted mortality ratio was 0.26 in low-surgical risk patients, 0.08 in intermediate-surgical risk patients, and 0.12 in high-surgical risk patients. Similar observations were obtained with EuroSCORE II. CONCLUSIONS: In this real-world setting, two-thirds of SAPIEN 3 transcatheter heart valve treated transcatheter aortic valve implantation patients had a logES <20 but were still considered appropriate transcatheter aortic valve implantation candidates by the heart team, mainly because of older age and less frequently because of conditions not captured by risk scores. logES and EuroSCORE II had poor discrimination and calibration power in this transcatheter aortic valve implantation cohort.