Prediction of One-Year Mortality Based upon A New Staged Mortality Risk Model in Patients with Aortic Stenosis Undergoing Transcatheter Valve Replacement

Background: In-depth knowledge about potential predictors of mortality in transcatheter aortic valve replacement (TAVR) is still warranted. Currently used risk stratification models for TAVR often fail to reach a holistic approach. We, therefore, aimed to create a new staged risk model for 1-year mortality including several new categories including (a) AS-entities (b) cardiopulmonary hemodynamics (c) comorbidities, and (d) different access routes. Methods: 737 transfemoral (TF) TAVR (84.3%) and 137 transapical (TA) TAVR (15.7%) patients were included. Predictors of 1-year mortality were assessed according to the aforementioned categories. Results: Over-all 1-year mortality (n = 100, 11.4%) was significantly higher in the TA TAVR group (TF vs. TA TAVR: 10.0% vs. 18.9 %; p = 0.0050*). By multivariate cox-regression analysis, a three-staged model was created in patients with fulfilled categories (TF TAVR: n = 655, 88,9%; TA TAVR: n = 117, 85.4%). Patients in “stage 2” showed 1.7-fold (HR 1.67; CI 1.07–2.60; p = 0.024*) and patients in “stage 3” 3.5-fold (HR 3.45; CI 1.97–6.05; p < 0.0001*) enhanced risk to die within 1 year. Mortality increased with every stage and reached the highest rates of 42.5% in “stage 3” (plogrank < 0.0001*), even when old- and new-generation devices (plogrank = n.s) were sub-specified. Conclusions: This new staged mortality risk model had incremental value for prediction of 1-year mortality after TAVR independently from the TAVR-era.