Natriuretic Peptides (NPs): Automated Electrochemiluminescent Immunoassay for N-Terminal pro-BNP Compared with IRMAs for ANP and BNP in Heart Failure Patients and Healthy Individuals

We evaluated the performance and diagnostic accuracy of an electrochemiluminescence immunoassay (ECLIA) method for N-terminal pro-B-type natriuretic peptide (NT-proBNP) in healthy persons and in patients with cardiac disease, and then compared the results obtained with the ECLIA method with results from IRMA methods for BNP and atrial natriuretic peptide (ANP). We studied 58 healthy individuals [mean (SD) age, 58 (8) years; 19 women and 39 men] and 148 consecutive patients [mean age, 64 (13) years; range, 20–80 years; 47 women and 101 men] with cardiomyopathy, admitted to the Department of Cardiovascular Medicine of our Institute. The study was done from November 2001 to October 2002. All healthy participants were nonobese, normotensive, and free from acute diseases, and all denied the use of any drug during the 4 weeks before the study. All had normal values for the main plasma indices and nonpathologic erythrocyte and leukocyte counts and urine analysis. In all of the participants, a complete cardiologic examination, including electrocardiogram and echocardiographic investigation (left ventricular ejection fraction >55%), was performed; in patients >50 years of age, an effort stress test was performed to exclude asymptomatic heart disease. Cardiac morphology and function were assessed in all patients by Doppler echocardiography, radionuclide ventriculography, or cardiac catheterization, when needed. Primary dilated cardiomyopathy was found in 95 patients, whereas the other 53 patients suffered from secondary cardiomyopathy; of these, 38 had ischemic cardiomyopathy. A total of 22 patients were in functional New York Heart Association (NYHA) class I, 72 in class II, and 54 in class III-IV; the mean (SD) left ventricular ejection fraction was 31.4 (9.6)%. NT-proBNP was measured by a fully automated “sandwich” ECLIA method using an Elecsys® 2010 analyzer (Roche Diagnostics). This ECLIA is based on two polyclonal antibodies: a biotinylated antibody and a ruthenium derivative-labeled antibody. Total duration of assay was 18 min. Plasma ANP and BNP were measured with two-site IRMA methods, previously set up in our laboratory, as described elsewhere in detail (1)(2). We assayed blood samples (10 mL), collected into ice-chilled disposable polypropylene tubes containing aprotinin (500 kIU/mL of plasma) and EDTA (1 g/L of plasma), in such a way as to use the same samples for all three assays, even if the addition of plasma protease inhibitors (such as aprotinin) was not necessary for NT-proBNP assay. Plasma samples were rapidly separated by centrifugation for 15 min at 4 °C, and then were frozen and stored at −20 °C in 1-mL aliquots in polypropylene tubes until assay. Within-run and total imprecisions were tested following the guideline approved by NCCLS (3) by repeatedly measuring two plasma samples on 20 different working days with NT-proBNP concentrations in the upper part of the reference interval (e.g., 103.8 ng/L) and the lower part of abnormal values (e.g., 601.7 ng/L; typical of patients with only moderate heart failure), in such a way as to better evaluate the imprecision in the discrimination zone between healthy individuals and patients with cardiac disease. Within-run imprecisions (CV) were 1.7% and 1.8%, and total imprecisions were 4.0% and 3.8%, for the two samples, respectively. These data indicate that ECLIA for NT-proBNP measures peptide concentrations in the discrimination zone between healthy participants and patients with cardiac disease with an imprecision better than that of ANP and BNP assays (1)(2). Peptide concentrations of patients with cardiomyopathy were much higher than those of healthy participants and progressively increased with disease severity (as measured: in healthy participants, 51.6 (34.6) ng/L; in NYHA class I, 367.0 (349.9) ng/L; NYHA class II, 1376.4 (1590.4) ng/L; NYHA class III, 5297.9 (6373.6) ng/L; NYHA class IV, 8421.1 (9231.1) ng/L). Moreover, we tested and compared the different degrees of diagnostic accuracy for ANP, BNP, and NT-proBNP assays in discriminating between the group of healthy participants and the group of patients with heart failure by ROC analysis using the “bootstrap percentile method” with 1000 bootstrap replications (Table 1 ). Our data indicate that NT-proBNP assay performance is better than that of both ANP and BNP assays in discriminating diseased from healthy individuals. In particular, the better performance of NT-proBNP ECLIA is more evident when patients with only mild heart failure (NYHA class I and II) are considered in the ROC analysis (Table 1 ). Of course, this finding is of great clinical relevance when considering that NT-proBNP could be used as a screening test for the diagnosis of chronic heart failure. However, it is important to note that a recent report (4) indicated pitfalls in community screening for left ventricular dysfunction when IRMA methods were used for the assay of BNP and NT-proANP in asymptomatic individuals, particularly in women. ROC analysis.1 Ref. (10). AUC, area under the curve; CI, confidence interval. AUC comparisons: ANP vs BNP, P = 0.037; ANP vs proBNP, P ≪0.0001; BNP vs proBNP, P ≪0.0001. AUC comparisons: ANP vs BNP, P = 0.028; ANP vs proBNP, P ≪0.0001; BNP vs proBNP, P ≪0.0001. AUC comparisons: ANP vs BNP, P = 0.72 (not significant); ANP vs proBNP, P = 0.078 (not significant); BNP vs proBNP, P = 0.04. ROC analysis.1 Ref. (10). AUC, area under the curve; CI, confidence interval. AUC comparisons: ANP vs BNP, P = 0.037; ANP vs proBNP, P ≪0.0001; BNP vs proBNP, P ≪0.0001. AUC comparisons: ANP vs BNP, P = 0.028; ANP vs proBNP, P ≪0.0001; BNP vs proBNP, P ≪0.0001. AUC comparisons: ANP vs BNP, P = 0.72 (not significant); ANP vs proBNP, P = 0.078 (not significant); BNP vs proBNP, P = 0.04. Our findings cannot indicate whether the improved clinical performance of NT-proBNP compared with the two other cardiac natriuretic hormone (CNH) assays is a result of its better assay precision or a true larger separation of NT-proBNP distribution values in healthy and patient populations than those of ANP and BNP. To address this important issue, we need to compare the performance of the NT-proBNP ECLIA method with that of a new generation of BNP assays (5)(6). In this regard, a previous study (7), which used a different immunoassay for NT-proBNP, showed no significant differences from the clinical results respectively obtained with NT-proBNP and BNP assays, thus suggesting that the performance of the immunoassays used may a very crucial point in determining the results of a clinical study comparing different CNH assays.