Diagnostic accuracy of plasma brain natriuretic peptide and aminoterminal‐proBNP in mild heart failure depends on assay and introduction of therapy

OBJECTIVE: A reliable biochemical marker of left ventricular dysfunction (LVD) could improve diagnostic accuracy. The aim of this study was to compare the correlation of measurements of brain natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) with different assays. The diagnostic accuracy of the tests in mild heart failure (HF) was estimated before and after the start of therapy. MATERIAL AND METHODS: Doppler echocardiography and measurements of plasma BNP and NT-proBNP were performed in 150 patients. RESULTS: Systolic dysfunction (LV ejection fraction 0.45) was present in 22 patients, and 58 had only abnormalities in LV filling. P-NT-proBNP based on two different assays demonstrated a moderate correlation (r = 0.57, p<0.0001) and a concentration-dependent systematic difference. Excellent correlation (r = 0.95, p<0.0001) was found between BNP and NT-proBNP based on two-site antibody assays, but was moderate between BNP and a one-site antibody NT-proBNP assay (r = 0.58, p<0.0001). Areas under the receiver operating characteristic (ROC) curves (AUCs) were 0.93 (95 % CI, 0.90-0.98) for BNP, 0.95 (0.91-0.99) for NT-proBNP (two-site antibody assay) and 0.77 (0.70-0.85) for the one-site antibody NT-proBNP assay (p = 0.0001). At re-evaluation of LVD at 6 and 12 months, AUCs of BNP were 0.81 (0.74-0.99) and 0.83 (0.76-0.89), respectively, and AUCs of NT-proBNP (two-site) were 0.84 (0.77-0.91) and 0.87 (0.81-0.93), respectively. Using the baseline threshold reduced the sensitivity and specificity of BNP and NT-proBNP measurements. CONCLUSIONS: BNP and NT-proBNP measurements demonstrated assay-dependent correlations. Measurement of p-BNP or p-NT-proBNP by a two-site antibody assay demonstrated potential as an indicator of mild, incident HF, but the applicability of the index tests was limited over time and was likely influenced by therapeutic interventions.