Development and validation of a multibiomarker model for 12-month adverse outcomes in acute heart failure

To develop a multibiomarker model for predicting 12-month major adverse cardiovascular events (MACE) in patients with acute heart failure (AHF) and assess incremental value.

This study included 675 patients with AHF, comprising a development cohort (n = 303) and a validation cohort (n = 372). MACE comprised all-cause death or heart failure rehospitalization. Least absolute shrinkage and selection operator Cox regression and multivariable Cox regression were used for model development. Model performance was assessed using discrimination, calibration, decision curve analysis (DCA), and incremental value.

The final model included monocyte-to-lymphocyte ratio (MLR), N-terminal pro-B-type natriuretic peptide (NT-proBNP), heparin-binding protein (HBP), and soluble suppression of tumorigenicity 2 (sST2). Harrell's concordance index, the 12-month area under the time-dependent receiver operating characteristic curve (AUC), and Brier score were 0.710, 0.804, and 0.180 in the development cohort and 0.710, 0.760, and 0.203 in the validation cohort, respectively. Calibration demonstrated overall agreement between predicted and observed risks, and DCA suggested potential clinical net benefit. Adding HBP, sST2, and MLR to clinical variables and NT-proBNP increased the 12-month AUC from 0.753 to 0.824.

The multibiomarker model showed moderate predictive performance and may support post-discharge risk stratification in patients with AHF.