Do more specific assays for B-type natriuretic peptides better predict heart failure in breathless patients than the currently used assays?

Abstract

Abstract Funding Acknowledgements Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Heart Foundation of New Zealand Health Research Council of New Zealand Introduction The cardioprotective B-type natriuretic peptide (BNP1-32) and its inactive congener N-terminal proBNP1-76 (NTBNP1-76) are produced from their precursor peptide proBNP, proportionate to cardiac dysfunction, underpinning their now-universal endorsement as markers for heart failure (HF) diagnosis. ProBNP concentrations are also increased in patients with HF. BNP1-32 is difficult to measure in plasma due to its low concentration and short half-life. Thus, less specific BNP and NT-proBNP assays are routinely used in the diagnosis and prognosis of HF. However, these assays also variably detect proBNP, BNP metabolites, and glycosylated proBNP or glycosylated NT-proBNP. How well these assays compare to highly specific assays that only detect BNP1-32 or NTBNP1-76 has not been assessed. Purpose We will compare the performance of assays specific for BNP1-32 and NTBNP1-76 only with that of a less specific commercially available NT-proBNP assay that is widely used for the diagnosis of heart failure. Methods Plasma samples obtained from patients presenting to the Emergency Department with breathlessness (n = 195) were assayed using a commercially available NT-proBNP assay and our specific in-house BNP1-32 (1) and NTBNP1-76 (2) assays. The diagnostic performance for clinically adjudicated acute HF (AHF, 60/195 patients, 31%) was assessed for each assay using receiver operator curve (ROC) analysis. Results Median (IQR) concentrations (n = 195) of NT-proBNP (101.7 [29.7, 353] pmol/L) and NTBNP1-76 (108.6 [25, 442] pmol/L) were markedly higher than BNP1-32 concentrations (1.9 [0.3-6.6] pmol/L) (p < 0.001 for both comparisons). Peptide concentrations were higher in patients with heart failure (n = 60) than those without heart failure for NT-proBNP (559 [247,1097] vs 46 [16,123] pmol/L), NTBNP1-76 (441.9 [155,1205] vs 37.9 [15,131] pmol/L) and BNP1-32 (7.8 [4,17] vs 0.5 [0.2,2.4] pmol/L) (all P < 0.001). NT-proBNP assay results were highly correlated with our specific NTBNP1-76 and BNP1-32 assays (Spearman’s rho = 0.91 and 0.92, respectively, both P < 0.001, n = 195). NT-proBNP, NTBNP1-76 and BNP1-32 assays performed similarly in the diagnosis of AHF (AUC [95%CI] = 0.88 [0.84-0.93], 0.85 [0.79-0.91] and 0.89 [0.84-0.94] respectively (all p < 0.001). Discussion Despite the differences in assay specificity all 3 assays performed similarly in the diagnosis of AHF in patients presenting with breathlessness. Conclusions All three assays diagnosed heart failure similarly, providing evidence that the widely used NT-proBNP assay can reliably be used as a proxy for active BNP1-32 in the diagnosis of acute heart failure. Whether this less specific NT-proBNP assay and our specific NTBNP1-76 and BNP1-32 assays (i) offer similar prognostic information and whether (ii) their marker performance is differentially altered by common confounders including age, renal dysfunction, obesity and atrial fibrillation, remain to be ascertained.