Prognostic value of sST2 added to BNP in acute heart failure with preserved or reduced ejection fraction

Natriuretic peptides and suppression of tumorigenicity 2 (ST2) represent two different physiopathological pathways. We evaluated the prognostic accuracy and complementarity of B-type natriuretic peptide (BNP) and soluble ST2 (sST2) plasma levels at discharge from a hospital admission for acute heart failure, both in patients with preserved (HFpEF) and depressed (HFrEF) systolic function. We enrolled 195 consecutive patients discharged alive and followed them prospectively for 6 months. The endpoint was all-cause death or hospital readmission for heart failure. Seventy-six patients had HFpEF and 119 had HFrEF, of whom 23 (30.3 %) and 43 (36.1 %) reached the combined endpoint, respectively. In both HFpEF and HFrEF, having the two biomarkers into account added prognostic information, with the highest risk in patients with both biomarkers above the median in their group (approximately 40 % hospitalization-free survival in both groups at 6 months). These associations translated into a significant fourfold increase in risk of the endpoint for one elevated biomarker and sevenfold for both biomarkers elevated in HFrEF, and no association for one elevated biomarker and fivefold increase in risk for both biomarkers elevated in HFpEF. Considering the reclassification of risk added to BNP by measurement of sST2, net reclassification index was 0.31 (p = 0.21) among patients with HFpEF and 0.70 (p < 0.001) among patients with HFrEF. sST2 provides robust prognostic information in acute heart failure with HFrEF, while this pattern was less clear in HFpEF. When sST2 was measured together with BNP, it improved prognostic accuracy in both groups, more clearly in HFrEF.