Abstract

Objective Recent studies have shown that plasma levels of brain natriuretic peptide (BNP)-32 and proBNP-108 are increased in heart failure (HF) and that the BNP-32 assay kit in current clinical use cross-reacts with proBNP-108. We investigated why proBNP is increased without processing in HF was investigated.

Design, setting and patients Plasma BNP-32 and proBNP-108 in normal individuals (n=10) and in patients with atrial fibrillation (AF) (n=18) and HF (n=132) was measured. BNP-32 and proBNP-108 in ventricular and atrial tissue and in pericardial fluid using a specific fluorescent enzyme immunoassay following Sep-Pak C18 (Waters, Milford, Massachusetts, USA) cartridge extraction and gel filtration was also measured.

Main outcome measures Levels of both BNP-32 and proBNP-108 were higher in HF than in control or AF (both p<0.01), and the levels of these peptides significantly correlated (r=0.94, p<0.001). The proBNP-108/total BNP (BNP-32+proBNP-108) ratio was widely distributed and lower in HF (0.33 (0.17)) than in control (0.41 (0.06), p<0.05) and AF (0.45 (0.04), p<0.002). The proBNP-108/total BNP ratio was higher in HF with ventricular than in HF with atrial overload (0.45 (0.10) vs 0.20 (0.11), p<0.001). Consistent with this finding, the major molecular form were proBNP-108 and BNP-32 in ventricular (n=6, 0.67 (0.04)) and atrial (n=7, 0.76 (0.05), p<0.0001) tissues, respectively. ProBNP-108 was also the major molecular form of BNP in pericardial fluid (n=8, 0.82 (0.05)). The proBNP-108/total BNP ratio increased and decreased with HF deterioration and improvement, respectively.

Conclusion These results suggest that BNP-32 and proBNP-108 is increased in HF and that the proBNP/total BNP ratio increases in association with pathophysiological conditions such as ventricular overload.