Relationship Between Soluble Transferrin Receptor and Clinical Outcomes in Patients With Heart Failure According to Ejection Fraction Phenotype: The New Zealand PEOPLE Study

Abstract

Background: Iron deficiency (ID) is highly prevalent in patients with heart failure (HF) but its impact on prognosis in HF with preserved ejection fraction (HFpEF) remains unclear. We assessed whether ID defined by soluble transferrin receptor (sTfR) criteria is independently associated with all-cause mortality in patients with HFpEF, and evaluated its comparative prognostic performance to ID definitions in common clinical use. Methods and Results: Data were analyzed from 788 patients (36% HFpEF) in a prospective multicenter HF cohort study. Baseline plasma samples were analyzed with respect to 4 definitions of ID: sTfR of ≥1.59 mg/L (IDsTfR1), sTfR of ≥1.76 mg/L (IDsTfR2), ferritin of <100 µg/L, or ferritin of 100–300 µg/L + transferrin saturation of <20% (IDFerritin), and transferrin saturation of <20% (IDTsat). In multivariable Cox models IDsTfR2 (hazard ratio [HR] 1.84, 95% confidence interval [CI] 1.23–2.75) and IDTsat (HR, 1.69, 95% CI 1.10–2.59) were both independently associated with all-cause mortality in patients with HFpEF, whereas IDsTfR1 (HR 1.41, 95% CI 0.92–2.16) and IDFerritin (HR 1.19, 95% CI 0.77–1.85) were not. On inclusion of patients with HF with reduced EF, IDsTfR1 (HR 1.45, 95% CI 1.13–1.86) gained significance, but IDFerritin (HR 1.21, 95% CI 0.95–1.54) did not. For each pair of definitions intra-patient concordance was approximately 65%. Conclusion: ID defined by sTfR criteria is independently associated with all-cause mortality in patients with HFpEF. Poor concordance between ID definitions suggests that iron biomarkers do not reflect the same pathological process in the complex relationship between iron and HF. Therefore, which definition should guide iron replacement needs further evaluation.