NT-proBNP is an established diagnostic and prognostic marker of heart failure (HF), which is a complex clinical syndrome and health problem that is associated with high morbidity and mortality rates. The detection of HF in its early stages and appropriate treatment are key factors in terms of improving quality of life.
• New antibody is not sensitive to glycosylation status of NT-proBNP
• Prototype assay measures the total concentration of NT-proBNP including fully glycosylated molecules
• Prototype assay exhibits the same diagnostic accuracy as the Roche NT-proBNP assay
NT-proBNP is the N-terminal fragment of proBNP that is formed in the processing of proBNP into BNP and NT-proBNP. In blood, NT-proBNP circulates as a pool of molecules heterogeneously modified by O-glycans. Immunoassays that utilize antibodies which are specific to the central, glycosylated region of NT-proBNP have been shown to miss a fraction of the protein in blood and this is most likely due to the fact that glycosylation hides the epitope from the antibody and thus prevents the antibody from recognizing the molecule.
The NT-proBNP immunoassay which is manufactured by Roche employs monoclonal antibodies (MAbs) that are specific to the epitopes 27-31 and 42-46 in the central region of NT‑proBNP. Epitope 42-46 comprises Ser44 which is known to be partially glycosylated. The Roche assay is known to detect only a subfraction of endogenous NT-proBNP as the assay is sensitive to glycosylation.
Since O-glycosylation is known to be heterogeneous, its pattern and extent might vary significantly among individuals. It is not known if underestimating the NT-proBNP concentration due to the influence of glycosylation could affect the clinical significance of the biomarker.
29D12cc-NT34cc is not sensitive to glycosylation
We have developed a new monoclonal antibody NT34cc (Cat.# 4NT1cc) that is specific to the non-glycosylated part of NT-proBNP. By using NT34cc together with the MAb 29D12cc, which also recognizes a non-glycosylated epitope, we have designed a prototype immunoassay that allows for the measurement of the total concentration of NT-proBNP regardless of its glycosylation status.
29D12cc-NT34cc assay exhibits the same diagnostic accuracy as the Roche NT-proBNP assay
The diagnostic accuracy of the Roche NT-proBNP assay (Roche Cobas e411 analyzer) and the HyTest prototype NT-proBNP assay 29D12cc-NT34cc in identifying patients with HF was compared by using ROC curve analysis. NT-proBNP levels were measured by these assays in EDTA-plasma samples that were obtained from 51 patients who had been diagnosed with HF and 53 healthy individuals (age-matched). Recombinant non-glycosylated NT-proBNP (Cat.# 8NT2) was used as a calibrator in the HyTest assay.
ROC-AUC for the prototype assay 29D12cc-NT34cc was 0.951 (sensitivity 0.86 and specificity 0.93) compared to 0.965 (sensitivity 0.86 and specificity 0.98) for the Roche NT-proBNP assay (see Figure 1). Differences were considered to be statistically insignificant (p-value = 0.365). A similar result was previously obtained for another HyTest prototype immunoassay that utilizes MAbs which are specific to non-glycosylated regions of NT-proBNP (15C4cc-13G12cc; see p. 9 in our BNP TechNotes).
Figure 1. ROC curves showing the ability of NT-proBNP measured by the Roche NT-proBNP assay and HyTest’s prototype NT-proBNP assay to identify patients with HF.
The Roche assay underestimates the NT-proBNP concentration
We measured the concentration of NT-proBNP in 15 HF patients using three different assays: the Roche assay, HyTest’s 29D12cc-NT34cc prototype assay and HyTest’s 15C4cc-13G12cc prototype assay. The results show that the Roche assay underestimates the NT-proBNP concentration as it only recognizes a subfraction of the biomarker in blood (see Figure 2).
Figure 2. The difference in NT-proBNP levels obtained by the Roche NT-proBNP assay and the HyTest prototype NT-proBNP assays 29D12-NT34* and 15C4-13G12* in 15 representative samples of HF patients. The Roche NT-proBNP assay underestimates NT-proBNP levels due to the presence of O-glycans in the epitopes recognized by the assay antibodies.
The data obtained with our two prototype immunoassays (ROC curves) shows that NT-proBNP immunoassays which are based on MAbs that are specific to non-glycosylated regions of the NT-proBNP molecule are expected to have at least a similar clinical value for HF diagnosis as the Roche NT-proBNP assay that detects only a subfraction of endogenous NT-proBNP. Taking into account the known high variability in the level and site occupancy of O-glycosylation, one might expect that immunoassays which measure the “total” NT-proBNP levels (e.g. HyTest prototype assays) could be advantageous for HF diagnostics and/or therapy monitoring in certain groups of patients and disease states due to their ability to detect endogenous NT-proBNP independently of its glycosylation status.
Cat# 4NT1cc: NT-proBNP, human, antibody, in vitro