Nicolas Vodovar

Background Elevated BNP and the N-terminal fragment of the proBNP (NT-proBNP) are hallmarks of heart failure (HF). Generally, both biomarkers parallel each other. In patients receiving sacubitril/valsartan, BNP remained stable while NT-proBNP decreased. As BNP and NT-proBNP assays have limited specificity due to cross-reactivity, we quantified by mass spectrometry (MS) the contributing molecular species. Methods We included 356 healthy volunteers, 100 patients with acute dyspnoea (49 acute decompensated HF; 51 dyspnoea of non-cardiac origin), and 73 patients with chronic HF and reduced ejection fraction treated with sacubitril/valsartan. BNP and NT-proBNP immunoreactivities (BNPir and NT-proBNPir) were measured by immunoassays (Abbott ARCHITECT and Roche Diagnostics proBNPII) and proBNP-derived peptides and glycosylation at serine 44 by MS on plasma samples. Results BNPir corresponded to the sum of proBNP1–108, BNP1–32, BNP3–32, and BNP5–32 (R2 = 0.9995), while NT-proBNPir corresponded to proBNP1–108 and NT-proBNP1–76 not glycosylated at serine 44 (R2 = 0.992). NT-proBNPir was better correlated (R2 = 0.9597) than BNPir (R2 = 0.7643) with proBNP signal peptide (a surrogate of proBNP production). In patients receiving sacubitril/valsartan, non-glycosylated NT-proBNP1–76 remained constant (P = 0.84) despite an increase in NT-proBNP1–76 and its glycosylation (P < 0.0001). ProBNP1–108 remained constant (P = 0.12) while its glycosylation increased (P < 0.0001), resulting in a decrease in non-glycosylated proBNP1–108 (P < 0.0001), and in NT-proBNPir. Conclusions Glycosylation interfered with NT-proBNPir measurement, explaining the discrepant evolution of these 2 biomarkers in patients receiving sacubitril/valsartan. Both BNPir and NT-proBNPir are surrogates of proBNP1–108 production, NT-proBNPir being more robust in the clinical contexts studied.