Serum transferrin concentration: a new prognostic biomarker in heart failure with reduced ejection fraction

Abstract Introduction Iron deficiency (ID) is a common comorbidity in patients with reduced heart failure (HFrEF), and is associated with poor clinical status and outcomes, but there is not enough data investigating serum transferrin concentration (STC) as a biomarker HFrEF to predict prognosis and mortality. Purpose We investigated relationship between STC at admission and all-cause mortality during follow-up in HFrEF. Methods We reviewed the medical data of 420 patients hospitalised in our institution for HFrEF over a period of 18 months. Laboratory tests were performed on the first 24 hours of admission. Patients were divided into two groups according to STC, using the value of 199 μg/dl as the cut-off. This value was determined from the ROC analysis to be the value with the highest sensitivity and specificity. Then we analysed relationship between STC at admission and all-cause mortality during follow-up between the two groups, after adjustment of the various variables described in the literature. Results Of 420 patients with HFrEF, 18.6% had STC ≤ 199 μg/dl (group 1) and 81.4% had STC > 199 μg/dl (group 2) with no difference in age and sex distribution (p = 0.873 ; p = 0.304). Patients with low STC were more likely to be diabetic (70.5% vs 46.5% ; p < 0.001), have dyslipidemia (52.6% vs 33%; p = 0.002), and history of chronic renal failure (6.4% vs 2.0% ; p = 0.045). There was no relation between STC and NYHA class (p = 0.776), hospital stay (p = 0.491) nor EF (p = 0.146). Biologically, lower STC was significantly associated with higher serum ferritin (292.77 vs 198.17 ; p < 0.001), TSAT (68.30 vs 22.21 ; p = 0.020) and C-reactive protein (CRP) (124.91 vs 39.98 ; p = < 0.001), lower plasma B-type natriuretic peptide (BNP) (5798.32 vs 6449.97 ; p = 0.002), with no difference in haemoglobin level (p = 0.272) and estimated glomerular filtration rate (eGFR) (p = 0.260). The prevalence of iron deficiency by the ESC criteria was lower in those with low STC; ferritin < 100 μg/l (15.4% vs 29.2% ; p = 0.015) or TSAT < 20% (33.3% vs 51,8% ; p = 0.004). There was no difference between the two groups in term of NYHA class and EF at three months (p = 0.87, p = 0.108). However, all-cause mortality during follow-up was significantly associated with low STC (37.2% vs 17.0% ; p <0.001). (Figure 1) In multivariable analysis, all-cause mortality was independently associated with STC ≤ 199 μg/dl (HR 1.71; 95% CI : 1.08 ; 2.7 ; p = 0.02). This is the same finding for the ischaemic origin of HFrEF (HR 2.1 ; p = 0.001), therapeutic inertia during follow-up (HR 5.53 ; p <0.001) and EF < 36% (HR 1.03 ; p = 0.047) (Figure 2). Conclusion More data are required before the routine use of STC as a biomarker of prognosis in HFrEF, but it may represent a reliable and cost-effective option to predict mortality.FIGURE 1FIGURE 2