Prognostic Value of Emerging Neurohormones in Chronic Heart Failure during Optimization of Heart Failure–Specific Therapy

Abstract

Abstract Background: Serial measurements of neurohormones have been shown to improve prognostication in the setting of acute heart failure (HF) or chronic HF without therapeutic intervention. We investigated the prognostic role of serial measurements of emerging neurohormones and BNP in a cohort of chronic HF patients undergoing increases in HF-specific therapy. Methods: In this prospective study we included 181 patients with chronic systolic HF after an episode of hospitalization for worsening HF. Subsequently, HF therapy was gradually increased in the outpatient setting until optimized. We measured copeptin, midregional proadrenomedullin, C-terminal endothelin-1 precursor fragment, midregional proatrial natriuretic peptide, and B-type natriuretic peptide before and after optimization of HF therapy. The primary endpoint was all-cause mortality at 24 months. Results: Angiotensin-converting enzyme/angiotensin receptor blocker and β-blockers were increased significantly during the 3-month titration period (P < 0.0001 for both). In a stepwise Cox regression analysis adjusted for age, sex, glomerular filtration rate, diabetes mellitus, and ischemic HF, baseline and follow-up neurohormone concentrations were predictors of the primary endpoint as follows (baseline hazard ratios): copeptin 1.92, 95% CI 1.233–3.007, P = 0.004; midregional proadrenomedullin 2.79, 95% CI 1.297–5.995, P = 0.009; midregional proatrial natriuretic peptide 2.05, 95% CI 1.136–3.686, P = 0.017; C-terminal endothelin-1 precursor fragment 2.24, 95% CI 1.133–4.425, P = 0.025; B-type natriuretic peptide 1.46, 95% CI 1.039–2.050, P = 0.029. Conclusions: In pharmacologically unstable chronic HF patients, baseline values and follow-up measures of copeptin, midregional proadrenomedullin, C-terminal endothelin-1 precursor fragment, midregional proatrial natriuretic peptide, and B-type natriuretic peptide were equally predictive of all-cause mortality. Relative change of neurohormone values was noncontributory.