Diagnostic utility of total NT-proBNP testing by immunoassay based on antibodies targeting glycosylation-free regions of NT-proBNP.

ABSTRACT

OBJECTIVES: The N-terminal fragment of pro-B-type natriuretic peptide (NT-proBNP) is a widely used heart failure (HF) biomarker. Commercial NT-proBNP immunoassays detect only a subfraction of endogenous NT-proBNP, as the antibodies target a region of NT-proBNP that could be glycosylated at Ser44. The diagnostic utility of immunoassays measuring total NT-proBNP remains unclear. METHODS: NT-proBNP was measured in 183 HF and 200 non-HF patients diagnosed by two independent cardiologists blinded to NT-proBNP results. Plasma samples either non-treated or treated with a mixture of glycosidases were analyzed by the Elecsys proBNP II assay (Roche Diagnostics, based on antibodies targeting a glycosylated region of NT-proBNP) and the SuperFlex NT-proBNP assay (PerkinElmer, based on antibodies targeting regions of NT-proBNP that are free of O-glycans). The diagnostic accuracy of the two assays was analyzed by comparison of ROC curves. RESULTS: The ROC-AUC for the proBNP II assay was 0.943 (95% CI 0.922-0.964) for NT-proBNP measured in untreated samples and 0.935 (0.913-0.958) for NT-proBNP measured in glycosidase-treated samples. The SuperFlex NT-proBNP assay in untreated samples gave a ROC-AUC of 0.930 (95% CI 0.907-0.954). The median percentage of non-glycosylated NT-proBNP to total NT-proBNP was 1.5-1.6-fold lower in the non-HF group compared to that in the HF group. CONCLUSIONS: The clinical value of total NT-proBNP for HF diagnosis was similar to the subfraction of NT-proBNP that was non-glycosylated at Ser44. The lower percentage of non-glycosylated NT-proBNP to total NT-proBNP in non-HF patients suggests that total NT-proBNP might be more sensitive in individuals without current or prior symptoms of HF.