Objective In heart failure populations without aortic stenosis (AS), the prognostic utility of multiple biomarkers in addition to clinical factors has been demonstrated. We aimed to determine whether multiple biomarkers of cardiovascular stress are associated with mortality in patients with AS undergoing aortic valve replacement (AVR) independent of clinical factors.

Methods From a prospective registry of patients with AS, 345 participants who were referred for and treated with AVR (transcatheter (n=183) or surgical (n=162)) were included. Eight biomarkers were measured on blood samples obtained prior to AVR: growth differentiation factor 15 (GDF15), soluble ST2 (sST2), amino-terminal pro-B-type natriuretic peptide (NTproBNP), galectin-3, high-sensitivity cardiac troponin T, myeloperoxidase, high-sensitivity C reactive protein and monocyte chemotactic protein-1. Biomarkers were evaluated based on median value (high vs low) in a Cox proportional hazards model for all-cause mortality and a parsimonious group of biomarkers selected. Mean follow-up was 1.9±1.2 years; 91 patients died.

Results Three biomarkers (GDF15, sST2 and NTproBNP) were retained in the model. One-year mortality was 5%, 12%, 18% and 33% for patients with 0 (n=79), 1 (n=96), 2 (n=87) and 3 (n=83) biomarkers elevated, respectively (p<0.001). After adjustment for the Society of Thoracic Surgeons (STS) risk score, a greater number of elevated biomarkers was associated with increased mortality (referent: 0 elevated): 1 elevated (HR 1.47, 95% CI 0.60 to 3.63, p=0.40), 2 elevated (HR 2.89, 95% CI 1.24 to 6.74, p=0.014) and 3 elevated (HR 4.59, 95% CI 1.97 to 10.71, p<0.001). Among patients at intermediate or high surgical risk (STS score ≥4), 1-year and 2-year mortality rates were 34% and 43% for patients with three biomarkers elevated versus 4% and 4% for patients with 0 biomarkers elevated. When added to the STS score, the number of biomarkers elevated provided a category-free net reclassification improvement of 64% at 1 year (p<0.001). The association between a greater number of elevated biomarkers and increased mortality after valve replacement was similar in the transcatheter and surgical AVR populations.

Conclusions These findings demonstrate the potential utility of multiple biomarkers to aid in risk stratification of patients with AS. Further studies are needed to evaluate their utility in clinical decision-making in specific AS populations.