Abstract
Background and Aims
In the AFFIRM-AHF trial, treatment with intravenous (IV) ferric carboxymaltose (FCM) reduced the risk of heart failure (HF) hospitalisations vs placebo in patients with iron deficiency after an episode of acute HF. Of these patients, 41% had a medical history of chronic kidney disease (CKD). This prespecified subanalysis of AFFIRM-AHF data was performed to investigate the effect of renal function on FCM efficacy.
Methods
In AFFIRM-AHF, patients stabilised following hospitalisation for acute HF with concomitant iron deficiency (defined as ferritin <100 μg/L, or 100–299 μg/L with transferrin saturation <20%) were randomised to receive either IV FCM or placebo before discharge for the index hospitalisation. In this analysis, patients who had received at least one dose of the study drug, and who had at least one post-randomisation data point and a baseline value for estimated glomerular filtration rate (eGFR; calculated using the CKD-EPI formula and baseline creatinine value), were stratified into tertiles according to baseline eGFR. The primary outcome was a composite of total HF hospitalisations and CV death. Secondary outcomes included total HF hospitalisations, CV death, time to first HF hospitalisation or CV death, composite of total CV hospitalisations and CV death, and days lost due to HF hospitalisations or CV death. All outcomes were evaluated up to 52 weeks post-randomisation.
Results
Of the 1,108 patients included in primary AFFIRM-AHF analyses, 967 (FCM: 487; placebo: 480) had a baseline eGFR value and were included in this analysis. In both groups, 60% of patients had an eGFR <60 mL/min/1.73 m2 following the index acute HF episode. Patients were divided into eGFR tertiles 1, 2 and 3, with corresponding respective baseline eGFR values of <42.96, 42.96 to <64.32, and ≥64.32 mL/min/1.73 m2. At baseline, the mean age, proportion of females, and proportions of patients with ischaemic HF aetiology, a documented history of HF, and a medical history of percutaneous coronary intervention, coronary artery bypass graft and/or cardiac resynchronisation therapy, were highest in eGFR tertile 1 and lowest in eGFR tertile 3. In eGFR tertiles 1, 2 and 3, the number of total HF hospitalisations and CV deaths in the FCM group vs placebo group were, respectively, 115 vs 152, 76 vs 83, and 56 vs 79, with respective annualised rate ratios (95% confidence interval [CI]) of 0.76 (0.50, 1.16), 0.76 (0.48, 1.22) and 0.69 (0.42, 1.12) (Figure). In eGFR tertile 3, the total number of CV hospitalisations and CV deaths was significantly lower in the FCM group vs the placebo group (69 vs 107; rate ratio [95% CI] 0.60 [0.39, 0.93]), with a nominally lower number of total HF hospitalisations with FCM vs placebo (44 vs 66; rate ratio [95% CI] 0.62 [0.38, 1.01]). In the time to first event analysis, FCM significantly reduced HF hospitalisation or CV death vs placebo in eGFR tertile 3 (hazard ratio [95% CI] 0.64 [0.42, 0.98]). In eGFR tertiles 1 and 2, differences between FCM and placebo arms for secondary endpoints did not reach statistical significance. The p-trend for treatment by baseline eGFR subgroup was non-significant for the primary outcome (0.941) and also for the secondary outcomes specified here.
Conclusion
In patients with iron deficiency who were stabilised after an episode of acute HF, numerically fewer primary and secondary events, endpoints or outcomes were consistently observed with FCM vs placebo across the eGFR tertiles. In addition, no significant interaction between kidney function and FCM efficacy was noted. Given that this analysis was limited by small patient numbers following subgroup stratification, further studies in larger cohorts with CKD may help to clarify the effect of IV FCM in this patient population.
Intravenous ferric carboxymaltose for heart failure with iron deficiency
