Multimarker approach in risk stratification of patients with decompensated heart failure

Abstract

Purpose: to study prognostic value of various biomarkers and their combinations in patients who survived decompensation of chronic heart failure. Materials and methods. Patients (n=159) who were hospitalized with diagnosis of heart failure (HF) decompensation were included in a prospective single-center study. Examination on admission and the day of hospital discharge, included measurement of concentrations of N-terminal pro-brain natriuretic peptide (NT-proBNP), high-sensitivity troponin T (hsTnT), copeptin, soluble suppression of tumorigenicity 2 (sST2), kopetin, neutrophil gelatinase-associated lipocalin (NGAL), and galectin-3. The combined primary endpoint comprised cardiovascular (CV) death, first hospitalization because of HF heart failure decompensation, episodes of HF deterioration which required additional i/v diuretics, and CV death with successful resuscitation. Results. During one-year follow-up 56 pts (35.2%) reached the combined primary endpoint. There were 78 (49.1%) cardiovascular events. During hospitalization, patients with the decompensation of heart failure experienced a decrease of sST2, NT-proBNP, galectin-3, kopetin, hsTnT and an insignificant increase of NGAL. ROC analysis identified significant relation between concentrations of NT-proBNP, sST2, copeptin and, to a lesser degree, hsTnT, determined at hospital discharge, and risk of combined primary endpoint during 1-year follow-up: area under the curve (AUC) was 0.733 [95% CI 0.645–0.820], p<0.0001, 0.772 [95% CI 0.688–0.856], p<0.0001, 0.735 [95% CI 0.640–0.830], p<0.0001, and 0.659 [95% CI 0.553–0.764], p=0.005, respectively. Patients who during hospitalization did not achieve cut-off values of NT-proBNP ≤1696 rg/ml, sST2≤37.8 hg/ml, copeptin≤28.31 rmol/L and hsTnT≤28.37 rg/ml, had higher risk of reaching adverse events during 1 year; OR and 95% CI were 2.96 [1.61, 5.42] p<0.0001, 4.31 [2.34, 7.93] p<0.0001, 3.06 [1.59, 5.89] and 2.19 [2.12, 4.27]), respectively. According to Cox regression analysis, risk of the combined primary end point was the highest in patients with 3 or more elevated markers (OR = 6.6 [3.584, 12.158], p<0.0001), average in patients with 2 elevated markers (OR = 1.123 [0.51, 2.48]), p=0.7), and the lowest in patients with no markers increase or increase of only one marker (OR = 0.11 [0.049, 0.241], p<0.0001). In the Kaplan-Mayer survival analysis all three groups were statistically different. In order to identify the most prognostically strong model, a reclassification analysis was performed. According to this analysis, the combination of sST2 and NT-proBNP concentrations determined at hospital discharge, exceeded one NT-proBNP (reclassification = –8.1%). At the same time, predictive value of only sST2 just insignificantly less than value of sST2 and NT-proBNP combination (reclassification = –1.9%). Conclusion. Patients with three and more elevated markers at hospital discharge have high risk of adverse events. The biggest prognostic value has combination of sST2 and NT-proBNP concentrations. In order to determine the long-term prognosis of a patient with HF decompensation, it is sufficient to measure concentrations of sST2 and NT-proBNP at hospital discharge. Alternatively, it is possible to limit to sST2 only, which is just insignificantly inferior to the sST2 and NT-proBNP combination. Patients with concentrations of sST2 ≥37.8 hg/ml and NT-proBNP ≥1696 rg/ml at hospital discharge have maximal 1year risk of death due to recurrent HF decompensation.