Influence of Glycosylation on Diagnostic and Prognostic Accuracy of N-Terminal Pro–B-Type Natriuretic Peptide in Acute Dyspnea: Data from the Akershus Cardiac Examination 2 Study

Abstract

Abstract BACKGROUND The N-terminal part of pro–B-type natriuretic peptide (NT-proBNP) is glycosylated, but whether glycosylation influences the diagnostic and prognostic accuracy of NT-proBNP measurements is not known. METHODS We measured NT-proBNP concentrations of 309 patients with acute dyspnea by use of standard EDTA tubes and EDTA tubes pretreated with deglycosylation enzymes. The primary cause of dyspnea was classified as heart failure (HF) or non-HF, and the diagnosis was adjudicated by 2 independent physicians. We collected information on all-cause mortality during follow-up. RESULTS In all, 142 patients (46%) were diagnosed with HF. NT-proBNP concentrations in nondeglycosylated samples distinguished HF patients from patients with non-HF related dyspnea [median 3588 (quartiles 1–3 1578–8404) vs 360 (126–1139) ng/L, P < 0.001], but concentrations were markedly higher in samples pretreated with deglycosylation enzymes (total NT-proBNP) [7497 (3374–14 915) vs 798 (332–2296) ng/L, P < 0.001]. The AUC to separate HF patients from patients with non-HF related dyspnea was 0.871 (95% CI 0.829–0.907) for total NT-proBNP compared with 0.852 (0.807–0.890) for NT-proBNP measurements in standard EDTA plasma. During a median follow-up of 816 days, 112 patients (36%) died. Both NT-proBNP and total NT-proBNP concentrations were associated with mortality in separate multivariate models, but only total NT-proBNP concentrations provided added value to the basic risk model of our dataset as assessed by the net reclassification index: 0.24 (95% CI 0.003–0.384). There was a graded increase in risk across total NT-proBNP quartiles, in contrast with the results for NT-proBNP measurements. CONCLUSIONS NT-proBNP concentrations were higher, and diagnostic and prognostic accuracy was improved, by pretreating tubes with deglycosylation enzymes.