Ferric carboxymaltose and exercise capacity in heart failure with preserved ejection fraction and iron deficiency: the FAIR-HFpEF trial

Trial procedures

With FCM being a brown substance and normal saline serving a comparator, the trial was conducted by unblinded and blinded investigators. Infusion sites and treatment vials were covered, and if considered needed, also eye masks were used during infusions to ensure patient blinding. Following screening (visit 1), study-related assessments, including the 6MWTD, patient global assessment (PGA), QoL questionnaires, and NYHA class, were performed by blinded investigators at baseline, after 8 ± 3 days, and after 4, 16 ± 1, 24 ± 1, 32 ± 2, and 52 ± 2 weeks (visits 2–7).¹³ Health-related quality of life (HRQoL) was evaluated using the HRQoL questionnaire, which integrates the European QoL-5 Dimensions (EQ-5D) questionnaire as a generic instrument with a disease-specific cardiology tool, the Kansas City Cardiomyopathy Questionnaire (KCCQ). Unblinded investigators performed infusions of intravenous placebo (i.e. normal saline) or FCM according to the dosing regimen in Supplementary data online, Table S1 as well as certain trial-related procedures (e.g. drug accountability, monitoring patients for elevated iron parameters or haemoglobin levels).

Primary and secondary endpoints

The primary endpoint of the trial was the change in 6MWTD from baseline to week 24, assessed in metres. Secondary endpoints included the change in 6MWTD from baseline to visits 3–7 and the change in PGA, NYHA class, blood parameters of kidney function and inflammation, and HRQoL from baseline to visits 3–7, and as safety-related outcome the rate of recurrent cardiovascular hospitalizations.¹³

Statistical analysis

A sample size of 86 patients per group gave a power of 90% for a two­sample t­test at the usual one­-sided level of 2.5% if the standardized mean difference (Cohen's d) is 0.50. Accounting for 10% dropout the original aim was to recruit 100 patients per group. Enrolment, however, proved to be more difficult than expected for reasons outlined above.

The primary analysis was based on the modified intention-to-treat (ITT) population including all randomized patients who received at least one dose of study medication and had at least one post-baseline assessment of the primary endpoint; they were analysed in the group they were randomized to. The primary endpoint was analysed using a mixed-model repeated measures (MMRM) approach adjusted for the presence of AF, anaemia, age at baseline, visit, and baseline 6MWTD; visit-by-treatment interactions were also included. Least square group differences were reported with standard error, 95% confidence intervals (CIs), and P-values testing the null hypothesis of no intervention effect using all available data without any imputation method. The analyses of continuous secondary endpoints followed the same lines as the analysis of the primary endpoint. Patient global assessment at all visits was compared between both groups using Fisher's exact tests. To adjust for covariates, a mixed-effects proportional odds model including a random intercept for subject and the main effects for treatment, visit, treatment–visit interaction, presence of AF at baseline, haemoglobin status at baseline, and age at baseline, as fixed effects was calculated. The change in NYHA class from baseline to the respective visit in the FCM group was tested using the Bowker symmetry test. Between groups, the NYHA class was compared using the Wilcoxon–Mann–Whitney test. As with PGA, mixed-effects proportional odds models were used for adjusted analyses. Recurrent events were modelled by negative binomial regression. Adverse events were summarized as frequencies and percentages by treatment group. Event rates were compared using negative binomial regression accounting for over-dispersion. All analyses were carried out using the statistical software R (version 4.3.1).

Follow-up

Of the 18 patients assigned to receive FCM, one did not complete the 24 weeks and the 52 weeks of follow-up. Of the 21 patients assigned to receive placebo/saline, 3 did not complete the 24 weeks of follow-up and 4 did not complete the 52 weeks of follow-up.

Primary and secondary endpoints

The 6MWTD improved in the FCM group from a median of 308 m (IQR: 198–378) at baseline to 403 m (IQR: 306–416) at week 24, with a change of 45 ± 62 m compared to baseline (P = .014) The respective values in the placebo/saline group were 325 m (IQR: 250–342) at baseline and 308 m (IQR: 285–354) at week 24, with a change of −8 ± 61 m vs. baseline (P = .62). The difference in least square means between the two groups at 24 weeks was 49 ± 22 m (mean ± SEM; 95% CI 5–93, P = .029). Of note, in total, 31 patients performed 6MWTD at week 24 (FCM: 15, placebo 16; see Supplementary data online, Table S2), but all 39 randomized patients contributed data to the result in the MMRM analysis for this endpoint. Detailed data are provided in Figure 2A and Table 2, which also show that the efficacy was somewhat further enhanced at week 32 (treatment effect between groups: 65 ± 22 m, P = .005) and then mostly lost at 52 weeks (treatment effect 13 ± 23 m, P = .57). Improvements in 6MWTD were similar in patients with and without anaemia (i.e. haemoglobin values of < 12.0 or ≥ 12.0 g/dL) at baseline (Figure 2B). When key subgroups were considered (including sex, ischaemic aetiology of HF, NYHA class, glomerular filtration rate, and haematinics at baseline), no significant interaction was found for subgroup of patients with HFpEF to respond differently to FCM vs. placebo with regard to changes in 6MWTD between baseline and 24 weeks (all P > .3; Supplementary data online, Figure S1); however, all subgroups were small. Due to these small subgroup sizes, the absence of interaction should not be considered as evidence of no interaction.

By week 24, 5 of 17 patients in the group receiving FCM reported moderate or much improvement in the PGA (in the placebo/saline group 3 of 18) with no significant difference compared to the group that received placebo/saline where 8 of 18 patients reported little or moderate improvement (in the FCM group 5 of 17) (P = .20) (see Supplementary data online, Figure S2).

At baseline, 11 (61%) of 18 patients treated with FCM were in NYHA class II and 7 (39%) in NYHA class III. At week 24, 11 (65%) of 17 patients were in NYHA class II and 6 (35%) in NYHA class III. For patients receiving placebo/saline, the respective numbers at baseline were 10 (48%) of 21 patients in NYHA class II and 11 (52%) in NYHA class III. Like in the FCM group, values were essentially unchanged in week 24 with 9 (50%) of 18 patients in NYHA class II and 9 (50%) patients in NYHA class III.

Between baseline and week 24 or other time points, no significant difference was noted between FCM and placebo/saline with regard to changes in the values of the EQ-5D or KCCQ (Figure 3). We observed a non-significant difference in the KCCQ overall summary score of 6.5 ± 5.1 points between FCM- and placebo-treated patients (P = .21).

Significant increases were noted from baseline to week 24 in patients treated with FCM vs. placebo/saline for haemoglobin (P = .028), ferritin (P ≤ .001), and TSAT levels (P ≤ .001) (Table 2). Serum levels of creatinine, estimated glomerular filtration rate, bilirubin, blood urea nitrogen, aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transferase, or C-reactive protein were not different between treatment groups at week 24 (all P > .05).

Safety

Survival status was available for all patients through week 52 with no patient having died during the conduct of the trial. A total of 5 of 21 patients receiving placebo/saline and 1 of 18 patients receiving FCM were hospitalized for cardiovascular reasons during the conduct of the trial—in total 8 vs. 2 cardiovascular hospitalization events were observed in the two treatment groups, respectively (P = .045). Nine patients in the placebo group and 3 patients in the FCM group reported at least one serious adverse event (P = .085). The total number of adverse events (76 vs. 114) and serious adverse events (5 vs. 19) was lower in patients treated with FCM compared to placebo. To account for the number of events and observation time per patient, we used a negative binomial regression showing a rate ratio of 0.38 (95% CI: 0.17, 0.88; P = .023) for the number of adverse events and of 0.27 (95% CI: 0.07, 0.96; P = .043) for the number of serious adverse events. Results for serious adverse events were primarily driven by adverse events that required hospitalization or prolongation of an existing hospitalization (16 vs. 4 events) and by events that were deemed life-threatening (3 vs. none). Detailed data on serious and non-serious adverse events are shown in Supplementary data online, Table S3. The study treatment was stopped prematurely in 3 of the 18 patients assigned to receive FCM and in 1 of the 21 patients assigned to receive placebo. No severe allergic reactions related to the study treatment were reported.

Limitations

The study was smaller than originally planned, and the large changes in 6MWTD are based on an overall small sample size. Interestingly, similar effect sizes for 6MWTD have been reported with testosterone in men and women with HFrEF.^37,38 Furthermore, the study suffered from some premature study discontinuations and not all patients retained in the study performed the 6MWT at all visits. Clearly, these results require confirmation overall and in key subgroups including those with and without anaemia, renal dysfunction, coronary artery disease, and TSAT < 20%.³⁹

Disclosure of Interest

S.v.H. has been a paid consultant for and/or received honoraria payments from Amomed, AstraZeneca, Bayer, Boehringer Ingelheim, BRAHMS, Edwards Lifesciences, MSD, Novo Nordisk, Novartis, Pfizer, Pharmacosmos, Respicardia, Roche, Servier, Sorin, and Vifor. S.v.H. reports research support from Amgen, AstraZeneca, Boehringer Ingelheim, Pharmacosmos, IMI, and the German Center for Cardiovascular Research (DZHK). W.D. reports consulting fees from Aimediq, Bayer, Boehringer Ingelheim, Boston Scientific, Lilly, Medtronic, Pfizer, Sanofi-Aventis, Sphingotec, Vifor Pharma, travel support from Pharmacosmos, and research support to the Institute from EU (Horizon2020), German Ministry of Education and Research, German Center for Cardiovascular Research, Boehringer Ingelheim, from Vifor Pharma related to this trial, and ZS Pharma. M.K. reports consulting fees and honoraria payments within the last 3 years from Adrenomed, 4TEEN4 Pharmaceuticals, Pharmacosmos, Sphingotec, and CSLVifor, all not related to the submitted work, and research support within the last 3 years from the German Research Foundation (DFG), the German Ministry of Education and Research (BMBF), the German Center for Cardiovascular Research (DZHK), the European Union (Horizon 2020), the Else Kroener-Fresenius Foundation (EKFS Clinician Scientist Professorship), Adrenomed AG, and CSL Vifor, all not related to the submitted work. G.F. reports lecture fees and /or advisory and/or trial committee membership by Bayer, Boehringer Ingelheim, Servier, Novartis, Impulse Dynamics, Vifor, Medtronic, Cardior, Novo Nordisc and Research Grants from the European Union. M.L. reports personal fees from AstraZeneca, Boehringer Ingelheim, Pfizer, and Bayer. J.B. serves as a consultant to Abbott, Adrenomed, Amgen, Array, AstraZeneca, Bayer, Berlin Cures, Boehringer Ingelheim, Bristol-Myers Squib, CVRx, G3 Pharmaceutical, Innolife, Janssen, LivaNova, Luitpold, Medtronic, Merck, Novartis, Novo Nordisk, Occlutech, Relypsa, Roche, Sanofi, SC Pharma, V-Wave Limited, and Vifor. P.P. reports consulting fees from Boehringer Ingelheim, AstraZeneca, Vifor Pharma, Amgen, Servier, Novartis, Bayer, MSD, Pfizer, Cibiem, Impulse Dynamics, Renal Guard Solutions, and BMS and has also received honoraria from Boehringer Ingelheim, AstraZeneca, Vifor Pharma, Amgen, Servier, Novartis, Berlin Chemie, Bayer, Pfizer, Impulse Dynamics, Renal Guard Solutions, BMS, and Abbott Vascular for lectures, presentations, speakers' bureaus, manuscript writing, or educational events. M.B. was supported by the Deutsche Forschungsgemeinschaft (German Research Foundation; TTR 219, project no. 322900939) and reports personal fees from Abbott, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Cytokinetics, Medtronic, Novartis, Servier, and Vifor. T.F. reports personal fees for statistical consultancies (including data monitoring committees) from Actimed, Bayer, BiosenseWebster, BMS, Coherex Medical, CSL Behring, Enanta, Fresenius Kabi, Galapagos, IQVIA, Janssen, KyowaKirin, LivaNova, Minoryx, Novartis, r-connect, Recardio, Relaxera, Roche, Servier, Viatris, and Vifor, all outside the submitted work. S.D.A. reports grants and personal fees from Vifor and Abbott Vascular and personal fees for consultancies, trial committee work and/or lectures from Actimed, Amgen, Astra Zeneca, Bayer, Boehringer Ingelheim, Bioventrix, Brahms, Cardiac Dimensions, Cardior, Cordio, CVRx, Edwards, Farraday, Impulse Dynamics, Janssen, Novartis, Occlutech, Pfizer, Respicardia, Servier, Vectorious, and V-Wave and declares that he is named co-inventor of two patent applications regarding MR-proANP (DE 102007010834 and DE 102007022367), but he does not benefit personally from the related issued patents. All other authors report no conflict of interest.

Data Availability

Requests for data analyses or sharing can be made towards S.v.H. and S.D.A. Such requests should be accompanied by an outline of the proposed study and a detailed statistical analysis plan. Proposals will be considered on a case by case basis, based on their scientific merit and availability of data and resources, but they also need to keep in mind German data protection laws and related requirements.

Funding

The study was funded by Vifor Pharma (Glattbrugg, Switzerland) by providing an unrestricted grant to Charité—Universitätsmedizin Berlin, Germany. The latter is the legal sponsor of the trial.

Ethical Approval

An independent ethics committee approved the protocol at all participating centres, and all subjects provided written informed consent.

Pre-registered Clinical Trial Number

The trial was registered at clinicaltrials.gov (NCT03074591).