Biomarker-driven prognostic models in chronic heart failure with preserved ejection fraction: the EMPEROR-Preserved trial.

Abstract

Biomarker-driven prognostic models incorporating N-terminal pro-B-type natriuretic peptide (NT-proBNP) and high-sensitivity cardiac troponin T (hs-cTnT) in heart failure (HF) with preserved ejection fraction (HFpEF) are lacking. We aimed to generate a biomarker-driven prognostic tool for patients with chronic HFpEF enrolled in EMPEROR-Preserved. Multivariable Cox regression models were created for (i) the primary composite outcome of HF hospitalization or cardiovascular death, (ii) all-cause death, (iii) cardiovascular death, and (iv) HF hospitalization. PARAGON-HF was used as a validation cohort. NT-proBNP and hs-cTnT were the dominant predictors of the primary outcome, and in addition, a shorter time since last hospitalization, New York Heart Association (NYHA) class III or IV, history of chronic obstructive pulmonary disease (COPD), insulin-treated diabetes, low haemoglobin, and a longer time since HF diagnosis were key predictors (eight variables, all p< 0.001). The consequent primary outcome risk score discriminated well (c-statistic=0.75) with patients in the top 10th of risk having an event rate >22× higher than those in the bottom 10th. A model for HF hospitalization alone had even better discrimination (c= 0.79). Empagliflozin reduced the risk of cardiovascular death or hospitalization for HF in patients across all risk levels. NT-proBNP and hs-cTnT were also the dominant predictors of all-cause and cardiovascular mortality followed by history of COPD, low albumin, older age, left ventricular ejection fraction ≥50%, NYHA class III or IV and insulin-treated diabetes (eight variables, all p< 0.001). The mortality risk model had similar discrimination for all-cause and cardiovascular mortality (c-statistic=0.72 for both). External validation provided c-statistics of 0.71, 0.71, 0.72, and 0.72 for the primary outcome, HF hospitalization alone, all-cause death, and cardiovascular death, respectively. The combination of NT-proBNP and hs-cTnT along with a few readily available clinical variables provides effective risk discrimination both for morbidity and mortality in patients with HFpEF. A predictive tool-kit facilitates the ready implementation of these risk models in routine clinical practice.