Impact of ischaemic aetiology on the efficacy of intravenous ferric carboxymaltose in patients with iron deficiency and acute heart failure: insights from the AFFIRM-AHF trial

Abstract

Aims: In AFFIRM-AHF, intravenous ferric carboxymaltose (FCM) reduced heart failure (HF) hospitalisations and improved quality of life versus placebo in iron-deficient patients stabilised after an acute HF episode. This analysis explored the effects of FCM versus placebo in patients with ischaemic and non-ischaemic HF aetiology.

Methods and results: We included 1082 patients from AFFIRM-AHF: 590 with ischaemic HF (defined as investigator-reported ischaemic HF aetiology and/or prior acute myocardial infarction and/or prior coronary revascularisation) and 492 with non-ischaemic HF. The prevalences of male sex, comorbidities, and history of HF were higher in the ischaemic versus non-ischaemic HF subgroup. Annualised event rates for the primary composite outcome of total HF hospitalisations and cardiovascular death with FCM versus placebo were 65.3 versus 100.6 per 100 patient-years in the ischaemic HF subgroup (rate ratio [RR] 0.65, 95% confidence interval [CI] 0.47-0.89, p = 0.007) and 58.3 versus 52.5 in the non-ischaemic HF subgroup (RR 1.11, 95% CI 0.75-1.66, p = 0.60) (p_interaction = 0.039). An interaction between HF aetiology and treatment effect was also observed for the secondary outcome of total HF hospitalisations (p_interaction = 0.038). A nominal increase in quality of life, assessed using the 12-item Kansas City Cardiomyopathy Questionnaire, was observed with FCM versus placebo, within each subgroup.

Conclusions: Heart failure hospitalisations and cardiovascular deaths occurred at a higher rate in patients with ishaemic versus those with non-ischaemic HF and were reduced by FCM versus placebo only in ischaemic patients. Further studies are needed to assess the role of aetiology in FCM efficacy.

Keywords: AFFIRM-AHF; Acute heart failure; Ferric carboxymaltose; Iron deficiency; Ischaemic heart failure.

Figure 1

(A) Annualised event rates per 100 patient‐years for recurrent event‐based primary and secondary outcomes; (B) annualised number of days lost per 100 patient‐years; (C) Kaplan–Meier (KM) estimates for time to cardiovascular (CV) death; (D) KM estimates for time to first hospitalisation or death in the ischaemic versus non‐ischaemic heart failure (HF) aetiology subgroups (modified intention‐to‐treat population). Annualised event rate ratio (RR) for ischaemic HF versus non‐ischaemic HF subgroups analysed using a negative binomial model. Hazard ratio (HR) for ischaemic versus non‐ischaemic HF subgroups analysed using Cox regression model. Both models were adjusted for the following baseline covariates: sex, age, HF history, country and subgroup of ischaemic aetiology of HF. Respective number of patients with ischaemic versus non‐ischaemic HF at baseline were 590 and 492. CI, confidence interval.

Figure 2

Subgroup analyses showing (A) cumulative primary outcome events (total heart failure [HF] hospitalisations and cardiovascular [CV] death) over time and (B) relative rate/risk for primary and secondary outcomes at week 52 with ferric carboxymaltose (FCM) versus placebo by HF aetiology (modified intention‐to‐treat population). *Rate ratio (RR) or hazard ratio (HR) for FCM versus placebo in each subgroup. ^†Annualised event rate per 100 patient‐years and RR analysed using a negative binomial model. ^‡HR for treatment difference analysed using Cox regression model. Negative binomial model and Cox regression model were adjusted for baseline covariates: sex, age, HF aetiology, HF duration, country, and included interaction between treatment group and HF aetiology. Respective number of patients with ischaemic and non‐ischaemic HF at baseline were 265 and 282 for FCM, and 257 and 275 for placebo. CI, confidence interval; NA, not applicable.

Figure 3

Adjusted mean change from baseline to week 52 in the 12‐item Kansas City Cardiomyopathy Questionnaire (KCCQ‐12) (A) overall summary score (OSS) and (B) clinical summary score (CSS) by heart failure (HF) aetiology and treatment arm, and (C) interaction of HF aetiology with ferric carboxymaltose (FCM) treatment effect at week 24 (modified intention‐to‐treat population). *p < 0.05 for difference in change versus baseline with FCM versus placebo within the non‐ischaemic HF subgroup only (no significance seen in the ischaemic HF subgroup). Estimates are from analysis based on mixed‐effect model of repeated measures (MMRM) using unstructured covariance matrix: change score = baseline score + subgroup of ischaemic aetiology of HF (sensitivity analysis) + visit + treatment + visit × treatment + subgroup of ischaemic aetiology of HF (sensitivity analysis) × visit + subgroup of ischaemic aetiology of HF (sensitivity analysis)*treatment + subgroup of ischaemic aetiology of HF (sensitivity analysis) × visit × treatment + baseline covariates. CI, confidence interval.

Figure 4

Mean change from baseline to week 52 in (A) serum ferritin and (B) transferrin saturation (TSAT) by heart failure (HF) aetiology (safety analysis set). **p < 0.01, ***p < 0.0001 for ferric carboxymaltose (FCM) versus placebo within each subgroup.