Value of the “TAVI2-SCORe” Versus Surgical Risk Scores for Prediction of One Year Mortality in 511 Patients Who Underwent Transcatheter Aortic Valve Implantation

Abstract

A bedside-available transcatheter aortic valve implantation (TAVI)-dedicated prognostic risk score is an unmet clinical need. We aimed to develop such a risk score predicting 1-year mortality post-TAVI and to compare it with the performance of the logistic EuroSCORE (LES) I and LES-II and the Society of Thoracic Surgeons' (STS) score. Baseline variables of 511 consecutive patients who underwent TAVI that were independently associated with 1-year mortality post-TAVI were included in the "TAVI2-SCORe." Discrimination and calibration abilities of the novel score were assessed and compared with surgical risk scores. One-year mortality was 17.0% (n = 80 of 471). Porcelain thoracic aorta (hazard ratio [HR] 2.56), anemia (HR 2.03), left ventricular dysfunction (HR 1.98), recent myocardial infarction (HR 3.78), male sex (HR 1.81), critical aortic valve stenosis (HR 2.46), old age (HR 1.68), and renal dysfunction (HR 1.76) formed the TAVI2-SCORe (all p <0.05). According to the number of points assigned (1 for each variable and 2 for infarction), patients were stratified into 5 risk categories: 0, 1 (HR 2.6), 2 (HR 3.6), 3 (HR 10.5), and ≥4 (HR 17.6). TAVI2-SCORe showed better discrimination ability (Harrells' C statistic 0.715) compared with LES-I, LES-II, and STS score (0.609, 0.633, and 0.50, respectively). Cumulative 1-year survival rate was 54% versus 88% for patients with TAVI2-SCORE ≥3 versus <3 points, respectively (p <0.001). Contrary to surgical risk scores, there was no significant difference between observed and expected 1-year mortality for all TAVI2-SCORe risk strata (all p >0.05, Hosmer-Lemeshow statistic 0.304), suggesting superior calibration performance. In conclusion, the TAVI2-SCORe is an accurate, simple, and bedside-available score predicting 1-year mortality post-TAVI, outperforming conventional surgical risk scores for this end point.