We calculated the hazard ratios and C-statistics (area under the curve [AUC]) of 4 risk scores (logistic European System for Cardiac Operative Risk Evaluation [EuroSCORE], EuroSCORE II, Society of Thoracic Surgeons predicted risk of mortality, and German aortic valve score) and 5 biomarkers of inflammation and/or myocardial dysfunction (high-sensitivity C-reactive protein, growth differentiation factor (GDF)–15, interleukin-6, interleukin-8, and N-terminal pro–B-type natriuretic peptide) for the risk of death (n = 80) and the combination of death or rehospitalization (n = 132) during the first year after TAVR in 310 consecutive TAVR patients.
The EuroSCORE II and GDF-15 had the strongest predictive value for 1-year mortality (EuroSCORE II, AUC 0.711; GDF-15, AUC 0.686) and for the composite end point (EuroSCORE II, AUC 0.690; GDF-15, AUC 0.682). When added to the logistic EuroSCORE and EuroSCORE II, GDF-15 enhanced the prognostic performance of the score and enabled substantial reclassification of patients. Combinations of increasing tertiles of the logistic EuroSCORE or EuroSCORE II and GDF-15 allowed the stratification of the patients into subgroups with mortality rates ranging from 4.0% to 49.1% and death/rehospitalization rates ranging from 15.3% to 68.4%.
Our study identified GDF-15 in addition to the logistic EuroSCORE and the EuroSCORE II as the most promising predictors of a poor outcome after TAVR.