Serial soluble ST2 for the monitoring of pharmacologically optimised chronic stable heart failure

Background: Soluble ST2 (sST2) is an emerging biomarker of cardiac remodelling and fibrosis. Studies indicate that it is predictive of mortality in acutely decompensated heart failure. The role of sST2 in chronic heart failure (CHF) is less well defined. No studies have examined serial measurements in optimised patients as a potential monitoring tool. This study aimed to prospectively determine the prognostic utility of serial sST2 in patients with pharmacologically optimised stable CHF.

Methods: 41 patients with pharmacologically optimised CHF and left ventricular ejection fraction ≤40% were recruited. Clinical review and blood sampling took place at baseline, and one, three and six months. N-terminal pro-brain natriuretic peptide (NTproBNP), sST2 and renal profile were measured on all samples. 12 lead electrocardiogram (ECG) was performed at baseline. Decompensation was defined as a composite endpoint of cardiovascular admission or worsening renal function (≥25% increase in serum creatinine from baseline).

Results: Receiver operator curve analysis of percentage change in sST2 from baseline to six months was strongly reflective of decompensation with area under the curve (AUC) of 0.778. This was significantly better than NTproBNP (AUC 0.425; p=0.013). Correlation of baseline concentrations to surface ECG showed that both sST2 and NTproBNP were positively correlated with duration of the QRS wave, with higher level of significance demonstrated by sST2 (0.415 (p=0.007) and 0.362 (p=0.020) respectively).

Conclusions: Percentage changes in sST2 are better able to predict cardiovascular admission or worsening renal function in patients with pharmacologically optimised CHF than NTproBNP. Compared with NTproBNP, sST2 appears to be a promising candidate for monitoring these patients.