Results:

The majority of patients were male (75%), median age 68 years, and had HF of ischemic origin (65%); 87% had a left ventricular ejection fraction (LVEF) <40%. Median concentrations of hs-TnT, NT-proBNP, and sST2 levels were 18, 1,360, and 27 ng/ml, respectively. Patients with sST2 ≥27 ng/ml were older, more often men, and less likely to have ischemic HF or a preserved EF (p < 0.001). They also had a higher body mass index, higher frequency of diabetes, worse renal function, and higher NT-proBNP and hs-TnT concentrations (p < 0.001). All-cause death occurred in 31%, cardiovascular (CV) death in 22%, and HF hospitalizations in 24%. Patients with sST2 ≥ the median had an increased risk of all-cause death, CV death, and HF hospitalization by 100%, 50%, and 10%, respectively, compared to those with no biomarker ≥ the median. Having all three biomarkers ≥ the median was also associated with an increased risk of the endpoint, 850%, 640%, and 590%, respectively. The combination of sST2 and another biomarker ≥ the median had an intermediate risk profile, and the combination of sST2 and hs-TnT ≥ the median was associated with increased risk of all-cause death. The best sST2 cutoff for the prediction of all-cause and CV death, and HF hospitalization was 28 ng/ml. sST2 was independently associated with outcomes in models that included prognostically significant baseline characteristics as well as those with all three biomarkers. When all three biomarkers were included, the risk of all-cause death, CV death, and HF hospitalization increased by 26%, 25%, and 30%, respectively, per each doubling of sST2. sST2 was prognostically significant across multiple subgroups, except patients with preserved or mid-range EF.