Comparison of four novel biomarkers alongside NT-proBNP for risk assessment of heart failure or death within 1 year from admission with acute dyspnoea

Abstract Background We aimed for a comparative analysis of four candidate biomarkers, soluble urokinase plasminogen activator receptor (suPAR), mid-regional pro adrenomedullin (MR-proADM), bioactive-adrenomedullin (bio-ADM) and pro-enkephalin (p-ENK) for prognosticating the composite endpoint of heart failure admission or all-cause mortality (HF/death) within 1 year after admission with undifferentiated acute breathlessness. Methods A total of 982 patients with the primary complaint of acute dyspnoea were recruited at two study centres according to identical protocols. Admission plasma biomarker measurements for suPAR (Virogates), MR-proADM (ThermoFisher), bio-ADM (Sphingotec) and p-ENK (Sphingotec) were compared against or used together with NT-proBNP (Roche) for predicting HF/death within 1 year. Statistical analyses were undertaken using receiver operating characteristics (ROC) area under the curve (AUC) and Cox regression adjusting for clinical risk factors. Biomarkers were assessed as continuous (per doubling concentrations) or stratified by cutoff values (Youden’s index). Results In the whole population (median age 63 years (IQR: 52-75), 36.8% females, 30.9% with acute decompensated heart failure (ADHF)), a total of 243 (24.7%) patients incurred the composite endpoint within 1 year. Admission levels of all biomarkers were higher in those with the endpoint, i.e. suPAR (median: 5.3 vs 3.1 ng/mL), MR-proADM (1.43 vs 0.73 nmol/L), bio-ADM (41.0 vs 26.0 pg/mL), p-ENK (96.0 vs 64.8 pmol/L), and NT-proBNP (1038 vs 284 pg/mL), P0.0001 for all. suPAR (ROC-AUC: 0.78, 95%CI: 0.74-0.82), MR-proADM (0.80 (0.77-0.84)) and NT-proBNP (0.80 (0.77-0.83)) had superior discriminative power over bio-ADM (0.68 (0.64-0.72)) and p-ENK (0.71 (0.67-0.75)) for the clinical endpoint, P0.0001. Each single biomarker, as continuous variables, or at their respective cut-offs (Table 1) were associated with increased hazards for the composite outcome P0.0001. However, NT-proBNP and suPAR were the top two markers when all markers were assessed within fully adjusted models. Both remained independently associated with the clinical endpoint in the presence of all biomarkers within the subgroup with ADHF (Figure 1). Combined suPAR 3.8 ng/mL and NT-proBNP 1000 pg/mL obtained the highest adjusted hazards ratio (aHR) of 4.44 for the clinical endpoint, a marked increment over using NT-proBNP (aHR:2.63) or suPAR (aHR: 2.16) alone (Table 1). Conclusion Circulating suPAR offers supplementary prognostic performance to the well-recognised cardiac-specific marker of NT-proBNP amongst populations with undifferentiated acute breathlessness. Its use with NT-proBNP could augment risk stratification strategies in acute dyspnoea. Figure 1