Background: Heart failure with preserved ejection fraction (HFpEF) has been assessed extensively, but few studies analysed the predictive value of the NT-proBNP in patients with de novo and acute HFpEF. We sought to identify NT-proBNP at admission as a predictor for all-cause mortality and rehospitalisation at 12 months in patients with new-onset HFpEF. Methods: We analysed 91 patients (73 ± 11 years, 68% females) admitted for de novo and acute HFpEF, using the Cox proportional hazard risk model. Results: An admission NT-proBNP level above the threshold of 2910 pg/mL identified increased all-cause mortality at 12 months (AUC = 0.72, sensitivity = 92%, specificity = 53%, p < 0.001). All-cause mortality adjusted for age, gender, medical history, and medication in the augmented NT-proBNP group was 16-fold higher (p = 0.018), but with no difference in rehospitalisation rates (p = 0.391). The predictors of increased NT-proBNP ≥ 2910 pg/mL were: age (p = 0.016), estimated glomerular filtration rate (p = 0.006), left atrial volume index (p = 0.001), history of atrial fibrillation (p = 0.006), and TAPSE (p = 0.009). Conclusions: NT-proBNP above 2910 pg/mL at admission for de novo and acute HFpEF predicted a 16-fold increased mortality at 12 months, whereas values less than 2910 pg/mL forecast a high likelihood of survival (99.3%) in the next 12 months, and should be considered as a useful prognostic tool, in addition to its utility in diagnosing heart failure.
Keywords: NT-proBNP at admission; TAPSE; acute heart failure; all-cause mortality; atrial fibrillation; estimated glomerular filtration rate; left atrial volume index; preserved ejection fraction; risk stratification.