Natriuretic pepdides (NPs): automated electrochemiluminescent immunoassay for N-terminal pro-BNP compared with IRMAs for ANP and BNP in human failure patients and healthy individuals

The analytical performance of a fully-automated electrochemiluminescence “sandwich” immunoassay (ECLIA) method for NT-proBNP assay has been evaluated. NT-proBNP results obtained assaying plasma samples from normal subjects and patients with different degrees of heart failure have been compared with those of BNP assay measured using a previously described IRMA method. The between-assay imprecision (20 runs) of NT-proBNP, measured on Elecsys® 2010 analyzer, was 7.6 CV% at 49 pg/ml, 3.1% at 229 pg/ml and 2.3 % at 4195 pg/ml. Functional sensitivity, computed from the imprecision profile, was 22 pg/ml. The dilution test (measurement of 2 plasma samples with elevated NT-proBNP progressively diluted) showed a good linear response to different sample volumes. The cut-off for NT-proBNP, calculated at 97.5 percentile, was 157 pg/ml and was measured with an imprecision of about 3%.
We studied 148 consecutive patients (47 women and 101 men; mean age 64±13 years, range 20-80 years) with chronic cardiomyopathy and a group of 58 healthy subjects (19 women and 39 men; mean age 58±8 years). The different capability of BNP and NT-proBNP assays in discriminating between normal subjects and patients with heart failure (subdivided according to NYHA class) was tested by ROC analysis. The area under the curve (AUC) for NT-proBNP was 0.957 (95% CI 0.925-0.981) versus AUC 0.922 (95% CI 0.884-0.954) for BNP (p<<0.0001); when patients with NYHA class I-II (mild disease) were compared with normal subjects, the AUCs for NT-proBNP and BNP were 0.933 (95% CI 0.880-0.971) and 0.881 (95% CI 0.819-0.935) (p<<0.0001); when patients with NYHA class III-IV (severe disease) were compared with normal subjects, the AUCs for NT-proBNP and BNP were 0.957 (95% CI 0.993-1.000) and 0.922 (95% CI 0.914-1.000) (p=0.04).
On the whole, our data indicate that NT-proBNP ECLIA is significantly better than BNP IRMA in discriminating patients from healthy subjects, especially when only patients with mild disease (NYHA class I and II) are considered. Our study indicates that NT-proBNP can be used as screening test in the first step of an algorithm for the diagnosis of chronic heart failure.