[Heart fatty acid binding protein in patients hospitalized because of worsening heart failure. Relation to prognosis of death].

PURPOSE We tested the hypothesis that serum heart fatty-acid binding protein (FABP), an early marker of myocardial necrosis, is related to prognosis of patients hospitalized because of worsening heart failure (HF). METHODS Sixty nine patients (64% men, age 66.6 +/- 11.0 years) with NYHA class II, III, IV HF (1, 18, and 50 patients, respectively) at hospital admission were followed for 6-12 (mean 11.6 +/- 1.3) months. Forty seven patients (68.1%) had history of myocardial infarction (MI), 56 (81.2%) - hypertension, 15 (21.7%) -- diabetes, and 17 (24.6%) had echocardiographical signs of aortic stenosis. Median left ventricular ejection fraction was 28%. Serum FABP, cardiac troponin I (Tn I) and N-aminoterminal pro brain natriuretic peptide (NT proBNP) were measured within 3 days after admission ( " admission " levels) and 2 weeks later (minimal hospital stay). Manufacturer recommended upper limits of norm (ULN) were 4.0 ng/ml for FABP, 0.35 ng/ml for Tn I, 0.1 ng/ml for NT proBNP. RESULTS Median admission FABP was insignificantly higher than level measured 2 weeks later (4.17 vs 4.03 ng/ml, p=0.069). FABP exceeded ULN in 38 (55.1%) patients and in 35 (50.7%) patients at admission and in 2 weeks, respectively (p=0.65). Median admission NT proBNP was significantly higher than 2 weeks level (13.23 vs 6.02 ng/ml, p < 0.0001). Median admission and 2-weeks levels of Tn I were similar and greatly lower than ULN. There were 27 all cause deaths (39.1%) during follow up. Median admission levels of TnI, FABP and NT proBNP were similar in patients who died and survived. Two weeks NT proBNP was significantly higher in patients who died (8.65 vs 3.62 ng/ml, p=0.012). ROC curve derived cut-off levels of FABP and NT proBNP (3.31 ng/ml and 3.5 ng/ml, respectively) were used in univaritate regression analysis. According to this analysis FABP >or= 3.31 ng/ml was related to occurrence of death (OR 3.54; 95% CI 1.03-12.17, p=0.044). FABP and variables with p > 0.1 (age, history of MI and diabetes, regular treatment with nitrates, signs of aortic stenosis, pulmonary rales at admission, and 2 weeks level of NT proBNP >or = cut-off) were included into multivariate logistic regression model. Independent predictors of death were aortic stenosis (OR 31.67; 95% CI 6.11-164.00) and NT proBNP >or= 3.5 ng/ml (OR 5.75; 95%CI 1.69- 19.52). CONCLUSION In this group of patients hospitalized due to worsening of HF admission values of neither FABP nor other biomarkers studied were predictors of death during about 1 year of follow up. FABP level after 2 weeks of hospital stay was related to occurrence of death but as predictor was inferior to NT-proBNP measured at the same time point.