Cutoff values of NT-proBNP for the prediction of low functional capacity, decreased ejection fraction and cardiovascular events in patients with heart failure.

BACKGROUND: It has been demonstrated in numerous studies that N-terminal pro-B-type natriuretic peptide (NT-proBNP) is strongly associated with left ventricular ejection fraction (LVEF), functional capacity (FC), and cardiovascular (CV) mortality in heart failure (HF) patients. The aim of the present study was to determine the predictive cutoff values of NT-proBNP for predicting these parameters.

METHODS: One hundred HF patients (88 male, 12 female, mean age 53.6 +/- 8.9 years) with left ventricular (LV) systolic dysfunction and impaired exercise capacity were enrolled into the study. Echocardiographic examination was performed. The NT-proBNP concentration was measured after resting for 20 min in the supine position. The modified Bruce protocol was utilized for exercise testing. The patients were followed for between 690 and 840 days (mean 750 +/- 30 days) for the occurrence of CV events.

RESULTS: There was a strong negative correlation between NT-proBNP concentration and LVEF (p < 0.004). It was found that NT-proBNP is a strong predictor of LVEF < 30% (p < 0.001). When 940 pg/mL was accepted as a cutoff value for NT-proBNP for the prediction of an LVEF < 30%, the sensitivity and the specificity were 89.8% and 71.4%, respectively. NT-proBNP and left atrial diastolic dimension were the most significant parameters for predicting FC (p < 0.001, each one). An NT-proBNP cutoff value of 940 pg/mL responded to 78.8% sensitivity and 81% specificity for the prediction of FC < 5 METs. The observed independent predictors for the CV events were NT-proBNP, LV mass index, and resting heart rate (p < 0.001, p = 0.02 and p = 0.006, respectively). Every 1000 pg/mL elevation in NT-proBNP level resulted in a 27% increase in the occurrence of CV events (p < 0.006). Moreover, 940 pg/mL NT-proBNP cutoff value revealed a sensitivity and specificity of 86.7% and 64.7% respectively for the prediction of incident CV events.

CONCLUSIONS: Use of NT-proBNP cutoff values is easy and reliable method for the prediction of low FC and decreased LVEF, and may aid identification of patients at the highest risk for future CV events. We suggest to use NT-proBNP cutoff value of 940 pg/mL for predicting these parameters.