LAUSR

Aims This study aimed to determine the frequency of detectable relaxin-2 concentrations and associated factors in patients with acute heart failure (AHF) and to establish relationships with short-term outcome and clinical and analytical parameters. Methods and results We performed a prospective, analytical, observational, multicentre study with cohort follow-up. We collected sociodemographic variables, previous history, basal status, previous treatment, vital signs, electrocardiographic and analytical data in the emergency department. Relaxin-2 was quantified by enzyme-linked immunosorbent assay (ELISA) (detection limit: 1.5 pg/mL). The primary event was 30-day all-cause mortality. We compared clinical characteristics and outcomes between patients with undetectable and detectable concentrations, and across concentration quartile groups in the latter. We analysed data for 522 patients (mean ± standard deviation age: 80.9 ± 9.1 years; 53.6% women). In 92 (17.6%) patients, relaxin-2 concentrations were undetectable. This finding was associated with male sex, absence of heart valve disease and treatment with aldosterone receptor antagonists. The incidence of 30-day mortality was significantly lower (2.2% vs. 9.1%; P = 0.03) in these patients even after adjustment for differences between groups (hazard ratio: 0.22, 95% confidence interval 0.05–0.92; P = 0.038). There were no associations with age or sex. Among the 430 (82.4%) patients with detectable relaxin-2 concentrations, values did not differ between subjects who remained alive [median: 30.9 pg/mL; interquartile range (IQR) 15.8–74.7 pg/mL] and those who had died (median: 30.9 pg/mL; IQR 15.6–85.2 pg/mL) (P = 0.89). Analysis of patients according to relaxin-2 concentration quartiles showed no differences in mortality. Relaxin-2 was correlated, albeit weakly, with haemoglobin values (R2 = 0.018, P = 0.005), but not with NT-proBNP, age or left ventricular ejection fraction (LVEF). Conclusions Patients with AHF and undetectable relaxin-2 concentrations have lower 30-day mortality. In contrast, there were no linear relationships between blood concentrations within the measurable range and mortality, NT-proBNP or LVEF.