High soluble transferrin receptor in patients with heart failure: a measure of iron deficiency and a strong predictor of mortality.

Abstract

Iron deficiency (ID) is frequent in heart failure (HF), linked with exercise intolerance and poor prognosis. Intravenous iron repletion improves clinical status in HF patients with left ventricular ejection fraction (LVEF) ≤45%. However, uncertainty exists about the accuracy of serum biomarkers in diagnosing ID. The aims of this study were (i) to identify the iron biomarker with the greatest accuracy for the diagnosis of ID in bone marrow in patients with ischaemic HF, and (ii) to establish the prevalence of ID using this biomarker and its prognostic value in HF patients. Bone marrow was stained for iron in 30 patients with ischaemic HF with LVEF ≤45% and 10 healthy controls, and ID was diagnosed for 0-1 grades (Gale scale). A total of 791 patients with HF with LVEF ≤45% were prospectively followed up for 3 years. Serum ferritin, transferrin saturation, soluble transferrin receptor (sTfR) were assessed as iron biomarkers. Most patients with HF (n=25, 83%) had ID in bone marrow, but none of the controls (P < 0.001). Serum sTfR had the best accuracy in predicting ID in bone marrow (area under the curve 0.920, 95% confidence interval 0.761-0.987, for cut-off 1.25 mg/L sensitivity 84%, specificity 100%). Serum sTfR was ≥1.25 mg/L in 47% of HF patients, in 56% and 46% of anaemics and non-anaemics, respectively (P < 0.05). The reclassification methods revealed that serum sTfR significantly added the prognostic value to the baseline prognostic model, and to the greater extent than plasma N-terminal pro B-type natriuretic peptide. Based on internal derivation and validation procedures, serum sTfR ≥1.41 mg/L was the optimal threshold for predicting 3-year mortality, independent of other established variables. High serum sTfR accurately reflects depleted iron stores in bone marrow in patients with HF, and identifies those with a high 3-year mortality.