Spontaneous Improvement in Iron Parameters and the Relation With Changes in Functionality in Heart Failure

Abstract

Iron deficiency is associated with reduced functional status , poor exercise performance, and increased cardiovascu-lar risk in patients with heart failure (HF). 1 Treatment with ferric carboxymaltose (FCM) has been shown to improve functional status, cardiac function, and clinical outcomes. 2,3 Although the European and American HF guidelines recommend the use of intravenous iron, periodic screening of patients with HF for the presence of iron deficiency has only been endorsed by European guidelines. 4 It is becoming increasingly recognized that iron parameters can be dynamic and may fluctuate over short periods of time (e.g., from screening to randomization in clinical trials or from discharge to 30 days follow-up after acute HF). 5 Clinicians have been challenged with the question as to what these spontaneous changes in iron parameter mean, why they occur, if they are associated with functional improvement, and if they would alter the treatment effect of FCM. This is a post hoc analysis of 2 published trials of iron supplementation, IRONOUT-HF (The Iron Repletion Effects on Oxygen Uptake in Heart Failure) and IRON-CRT (Effect of Intravenous Ferric Carboxymaltose on Reverse Remodelling Following Cardiac Resynchroniza-tion Therapy), in symptomatic patients with HF and reduced ejection fraction (left ventricular ejection fraction <40% to 45%) with iron deficiency. 6,7 Specifically, we investigated the relation between changes in iron parameters (serum iron and transferrin saturation [TSAT]), their spontaneous improvement (defined as an improvement in TSAT or serum iron above the coefficient of variation [5%]), and exercise capacity (assessed by peak oxygen consumption [VO 2 ] or 6-minute walk test) and quality of life (assessed by the Kansas City Cardiomyopathy Questionnaire [KCCQ]) from baseline to 16 weeks. Given the null finding of oral iron in IRONOUT-HF, changes in spontaneous iron parameters after randomization from baseline to 16 weeks could be assessed and the variables associated with spontaneous improvement could be determined. 6 For IRON-CRT, the change in spontaneous iron parameters was done from screening to baseline visit laboratory analysis the day of randomization (before study drug administration). The change from baseline in peak VO 2 and KCCQ were analyzed using an analysis of covariance model with a fixed effect of treatment allocation and the interaction between treatment allocation and spontaneous TSAT improvement (as defined previously). The analyses were adjusted for the individual baseline value. All analysis were performed using SPSS version 25 (IBM Corp, Armonk, NY) or STATA version 12 (StataCorp LLC, College Station, TX). A 2-sided p value of 0.05 was considered significant for all analyses. Of the 225 patients with HF with reduced ejection fraction randomized in the IRONOUT-HF trial, 44% of the patients had spontaneous improvement in serum iron and 37% in TSAT. Overall, the patients were well balanced in baseline characteristics in those with versus without spontaneous improvement in iron parameters. Patients with (vs without) spontaneous improvement had a significant improvement in peak VO 2 (absolute group difference for serum iron definition +0.73, TSAT definition +0.61 ml/kg/min, both p <0.05) (Figure 1). The between-group difference for serum iron was 52.9 mg/dl (95% confidence interval [CI] 45.4 to 59.4) or 9.5 mmol/L (95% CI 8.1 to 10.6) and 12% (95% CI 10 to 14) for TSAT. Larger improvements in serum iron and TSAT were correlated with larger improvements in functional status (KCCQ) (both p <0.007). In a multivariable model including covariates known to be associated with the development of iron deficiency (including hepcidin), improvement in N-terminal pro−B-type natriuretic peptide was the only variable independently associated with spontaneous iron parameter improvement (Table 1). There were positive, albeit modest, correlations between the proportional change (change from baseline to 16 weeks divided by baseline) in serum iron (Pear-son r = 0.200, p = 0.004) or change in TSAT (Pearson r = 0.203, p = 0.006) and improvement in KCCQ at 16 weeks. In the sensitivity analysis, using a cutoff of >10% improvement , similar findings were documented for serum iron but less pronounced for the TSAT definition. The addition of interaction fixed effect (spontaneous iron improvement and treatment allocation of oral iron supplementation) generated