A comparative study of biomarkers for risk prediction in acute coronary syndrome—Results of the SIESTA (Systemic Inflammation Evaluation in non-ST-elevation Acute coronary syndrome) study

Abstract Objective We compared the 1-year predictive value of several inflammatory and non-inflammatory biomarkers in ACS patients. Methods In 610 patients (73.0% male) – 36.0% unstable angina (UA) and 64.0% NSTEMI – we assessed high-sensitivity C-reactive protein (hs-CRP), interleukins 6, 10 and 18, soluble CD40 ligand, P- and E-selectin, NT-proBNP, fibrinogen and cystatin C at hospital admission. Two outcomes at 1-year follow up were selected for analysis: (1) all-cause death, MI, UA, or coronary revascularization, and (2) all-cause death, and non-fatal MI. The effect of biomarker levels on endpoints was examined by the Cox proportional hazards model, and their discrimination ability with the C statistic (AUC). Results Of 549 patients (90.0%) who completed the 1-year follow up, 206 (37.5%) and 54 (8.9%) reached the first and second composite endpoints, respectively. None of the biomarkers studied improved prediction of the first endpoint. However, considered as continuous variables, and in combination, NT-proBNP and fibrinogen, increased the AUC from 0.64 (95% CI 0.55–0.72) to 0.73 (95% CI 0.64–0.81; p = 0.02) for prediction of the second endpoint. Cut-off values for NT-proBNP and fibrinogen, regarding best sensitivity and specificity for prediction of the secondary endpoint were 1043.9 ng/L and 4.47 mg/dL, respectively. For these cut-off points, sensitivity, specificity, positive predictive value and negative predictive value were 40.5% vs 59.5%, 83.3% vs 67.1%, 18.8% vs 14.9% and 93.5% vs 94.4% for NT-proBNP and fibrinogen, respectively. Conclusion In ACS patients, inflammatory biomarkers offer modest incremental information to that provided by clinical risk markers. Fibrinogen and NT-proBNP measurements, however, improve cardiovascular risk prediction.