Reimbursement Recommendation

Ferric Carboxymaltose (Ferinject)

Indication: For the treatment of iron deficiency in adult patients with heart failure and New York Heart Association class II or III to improve exercise capacity

Sponsor: CSL Vifor

Recommendation: Reimburse with conditions

What Is the Reimbursement Recommendation for Ferinject?

Canada’s Drug Agency (CDA-AMC) recommends that Ferinject be reimbursed by public drug plans for the treatment of iron deficiency (ID) in adult patients with heart failure (HF) and New York Heart Association (NYHA) class II or III to improve exercise capacity if certain conditions are met.

Which Patients Are Eligible for Coverage?

Ferinject should only be covered to treat adult patients with HF and NYHA class II or III and who have a left ventricular ejection fraction (LVEF) of 40% or less and a serum ferritin level of 300 mcg/L or lower with a transferrin saturation (TSAT) of less than 15%.

What Are the Conditions for Reimbursement?

Ferinject should only be reimbursed if prescribed by a cardiologist or clinician experienced in the management of chronic HF and if the cost of Ferinject is reduced.

Why Did CDA-AMC Make This Recommendation?

• Evidence from 3 studies showed that Ferinject likely improves HF symptoms, health-related quality of life (HRQoL), fatigue, and iron levels at 24 and 52 weeks. However, it is unclear how significant these improvements are.

• Evidence from a meta-analysis showed that patients treated with Ferinject had fewer total hospitalizations, HF hospitalizations, and combined cardiovascular hospitalizations and deaths compared to those given a placebo. These reductions are considered clinically meaningful.

• The Canadian Drug Expert Committee (CDEC) noted that although Ferinject might increase serum ferritin levels, it is uncertain if it will improve functional and exercise capacity, HRQoL, or cardiovascular mortality.

• Based on the CDA-AMC assessment of the health economic evidence, Ferinject does not represent good value to the health care system at the public list price. The committee determined that uncertainty surrounding savings in administration costs and the assumption of comparative safety remain. Therefore, a price reduction is required.

• Based on public list prices, Ferinject is estimated to cost the public drug plans approximately $400,000 over the next 3 years. However, the actual budget impact is highly uncertain.

Additional Information

What Is ID in Adult Patients With HF?

HF causes significant morbidity and mortality, reduces the ability to function, and worsens HRQoL. Low iron levels can lead to ID anemia (IDA), which makes it hard for the blood to carry oxygen due to fewer red blood cells or low hemoglobin levels. Iron deficiency is the most common cause of anemia. Iron deficiency greatly affects patient well-being because iron is essential for transporting oxygen and producing energy, especially in the heart muscle. ID is prevalent in 35% to 55% of HF outpatients and 72% to 83% of hospitalized patients with HF.

What Are the Unmet Needs in Adult Patients With ID and HF?

The clinical experts consulted for this review noted that ID in patients with HF worsens symptoms, speeds up disease progression, and leads to poorer outcomes. It also reduces functional capacity, increases hospitalizations, and raises death rates. The main goal of treatment is to improve hemoglobin levels, replenish iron, and keep iron levels stable to relieve symptoms, improve HRQoL, and boost functional and exercise capacity.

How Much Does Ferinject Cost?

Treatment with Ferinject is expected to cost approximately $225 to $900 per adult (aged 18 years and older) per course of treatment.

Recommendation

CDEC recommends that ferric carboxymaltose be reimbursed for the treatment of ID in adult patients with HF and NYHA class II or III to improve exercise capacity only if the conditions listed in Table 1 are met.

Rationale for the Recommendation

Evidence from 3 studies (FAIR-HF [N = 459], CONFIRM-HF [N = 304], and HEART-FID [N = 3,065]) in patients with chronic heart failure (CHF) demonstrated that treatment with ferric carboxymaltose likely results in an improvement in NYHA class at 24 and 52 weeks, Kansas City Cardiomyopathy Questionnaire (KCCQ) score at 52 weeks, fatigue score at 24 and 52 weeks, and serum ferritin at 24 and 52 weeks. However, CDEC noted that there is uncertainty regarding whether the magnitude of benefit is clinically meaningful for all the aforementioned outcomes. In addition, treatment with ferric carboxymaltose in patients with CHF may also result in an improvement in KCCQ scores at 24 weeks, but the certainty was lower at this time point due to imprecision. Regarding the effect of ferric carboxymaltose on 6-minute walk test (6MWT) distance in patients with CHF at 24 or 52 weeks, the evidence was very uncertain due to inconsistency, imprecision, and missing data. Notably, the FAIR-HF and CONFIRM-HF studies showed likely clinically meaningful improvements when compared with placebo, whereas HEART-FID did not show an improvement in 6MWT distance when compared with placebo. The reason for this discrepancy is not fully clear. Finally, ferric carboxymaltose may result in little to no difference in cardiovascular (CV) mortality when compared to placebo at 26 or 52 weeks. However, the duration of follow-up may be inadequate and studies may be inadequately powered to fully assess this outcome. Finally, 1 study (AFFIRM-AHF [N = 1,132]) that assessed the efficacy and safety of ferric carboxymaltose in patients with acute heart failure (AHF) was also reviewed. However, this study did not assess the impact of ferric carboxymaltose on exercise capacity. CDEC noted that the results from the Ponikowski et al. meta-analysis, which utilized individual patient data from CONFIRM-HF, HEART-FID, and AFFIRM-AHF, demonstrated that compared to placebo, patients treated with ferric carboxymaltose experienced a reduction in total hospitalizations, total HF hospitalizations, and the composite outcome of total CV hospitalization and CV death. CDEC emphasized that these reductions are clinically meaningful.

The clinician group input received and the clinical expert consulted for this review noted that, for patients with HF, ID exacerbates symptoms, accelerates disease progression, and worsens prognosis, and that ID is also associated with reduced functional capacity, more frequent hospitalizations, and higher mortality rates. It was also noted that the main treatment goals are to address low hemoglobin levels, replenish iron stores, and maintain iron stores over time to alleviate symptoms, improve HRQoL, and enhance functional and exercise capacity. CDEC noted that although ferric carboxymaltose might increase serum ferritin levels, it is uncertain whether it will have an impact on functional and exercise capacity, HRQoL, and cardiovascular mortality. Therefore, it is uncertain whether ferric carboxymaltose addresses some of the unmet needs identified for this patient population, such as improving exercise capacity and HRQoL and reducing cardiovascular mortality.

At the sponsor-submitted price for ferric carboxymaltose and the publicly listed prices for all other drugs, ferric carboxymaltose may incur lower total costs compared to iron sucrose. This is primarily due to differences in infusion time and frequency, which may result in reduced administration costs for ferric carboxymaltose (e.g., supplies, chair time, nursing time). However, the total costs of ferric carboxymaltose are relatively similar to those of ferric derisomaltose. Ferric carboxymaltose is not expected to consistently yield savings with respect to administration cost because ferric derisomaltose allows for a higher maximum dose per single infusion. CDEC noted that if patients develop more severe adverse events (AEs) (e.g., severe hypophosphatemia, hypophosphatemia osteomalacia, or fractures), savings associated with the reimbursement of ferric carboxymaltose may be further reduced or eliminated, which was not captured in the economic analysis. Due to the uncertainty associated with savings in administration costs and the comparative safety assumptions, a price reduction for ferric carboxymaltose would be required.

Table 1: Reimbursement Conditions and Reasons

6MWT = 6-minute walk test; CDA-AMC = Canada’s Drug Agency; CV = cardiovascular; HF = heart failure; LVEF = left ventricular ejection fraction; NYHA = New York Heart Association; TSAT = transferrin saturation.

Discussion Points

• Reconsideration request: The sponsor requested a reconsideration of the initial CDEC draft recommendation to not reimburse ferric carboxymaltose for the treatment of ID in adult patients with HF and NYHA class II or III to improve exercise capacity. CDEC discussed 3 issues outlined by the sponsor in the request for reconsideration. The first issue involved the sponsor’s request that CDEC consider additional context when interpreting the HEART-FID results. The second issue was the sponsor’s view that there is evidence indicating that ferric carboxymaltose would reduce hospitalization in patients with ID and HF. The third issue was the sponsor’s indication that there is an unmet need in patients with HF.

• Treatment goals: During the initial meeting and the reconsideration meeting, CDEC discussed that when iron levels are too low to support adequate hemoglobin synthesis, the blood’s ability to carry oxygen efficiently can be impaired, either due to a reduced number of red blood cells or low hemoglobin levels. In patients with HF and ID, correcting iron levels is theorized to improve HRQoL, functional capacity, and exercise capacity and to alleviate symptoms associated with ID. The clinical expert noted to CDEC that the treatment goal of IV iron is not necessarily to improve hospitalization and mortality directly because these are likely more strongly driven by the patient’s underlying HF, which is not modified by iron supplementation.

• GRADE assessment: CDEC discussed the Grading of Recommendations Assessment, Development and Evaluation (GRADE) assessment of selected outcomes from the included studies for patients with CHF and concluded with moderate certainty that treatment with ferric carboxymaltose likely results in an improvement in NYHA class, fatigue score, serum ferritin, and hospitalization when compared with placebo. However, it was noted that there is uncertainty regarding whether the magnitude of benefit is clinically meaningful for these outcomes. Assessment of exercise capacity using 6MWT was not conclusive and did not show a clearly meaningful benefit of ferric carboxymaltose over placebo, and the GRADE assessment of evidence was of very low certainty.

• Uncertainty in the evidence regarding exercise capacity: CDEC discussed the uncertainty in the results for 6MWT, which was mainly driven by the inconsistency between the studies, and that confidence intervals (CIs) and some effect estimates included the potential for both clinically meaningful benefit and little to no benefit. CDEC noted that the largest study, HEART-FID, is also the most recent and enrolled patients with similar characteristics to patients with HF who live in Canada in terms of the geographical location of the sites. This study also found the least apparent benefit in the 6MWT distance, and its 95% CIs did not overlap or minimally overlapped with those of the other studies contributing data to this outcome. There were also some concerns about missing data.

• Evidence in AHF: CDEC discussed the results of the AFFIRM-AHF study, which enrolled patients with AHF. The study showed that ferric carboxymaltose likely results in improvements in NYHA class, KCCQ score, serum ferritin levels, and cardiovascular hospitalization rates. However, CDEC noted that there was uncertainty about whether the observed effects were clinically significant for all the mentioned outcomes. Additionally, there was a lack of data regarding the impact of ferric carboxymaltose on exercise capacity or fatigue in patients with AHF.

• Comparators: During the initial meeting and reconsideration meeting, CDEC discussed that other IV formulations of iron supplementation exist and are used off-label in clinical practice to treat patients with HF; however, no direct or indirect evidence comparing ferric carboxymaltose to other IV irons is available. Therefore, no conclusions can be drawn regarding the relative efficacy and safety of ferric carboxymaltose compared to other commonly used IV iron formulations in patients with HF. During the reconsideration meeting, CDEC noted that although access to IV iron may vary across jurisdictions, it would appear patients with HF and ID do receive IV iron. Evidence to address an unmet need related to differences to infusion time was not provided. Neither was any comparative evidence provided.

• Cardiovascular outcomes: During the reconsideration meeting, CDEC discussed the results from the Ponikowski et al. meta-analysis, which used individual patient data from CONFIRM-HF, HEART-FID, and AFFIRM-AHF. CDEC noted that the results appear clinically important, demonstrating that, compared to placebo, patients treated with ferric carboxymaltose experienced a reduction in the composite outcome of total CV hospitalization and CV death, total hospitalizations, and total HF hospitalizations. However, it was noted that the meta-analysis did not show a reduction in time to CV death or time to all-cause death and did not report on HRQoL outcomes or exercise capacity (e.g., 6MWT). CDEC also discussed that the Ponikowski et al. meta-analysis explored the outcomes related to baseline TSAT, and it was reported that those with a TSAT of less than 15% were the only group that experienced in reduction in the composite outcome of total CV hospitalizations and CV death, whereas those with a baseline TSAT between 15% and 24% and a TSAT at least 24% did not appear to derive a statistically significant benefit from ferric carboxymaltose therapy. The same trends were observed for the composite outcome of total HF hospitalization and CV death.

• Hypophosphatemia risk: During the reconsideration meeting, CDEC noted that ferric carboxymaltose is associated with a risk of hypophosphatemia, as detailed in the product monograph, which is not the case to the same magnitude among the other IV iron formulations. Events of hypophosphatemia were rare in the included studies, when reported, but there are concerns about the reporting given that events (including several severe events) do appear to have occurred without being recorded as TEAEs in CONFIRM-HF. It is unknown whether similar events occurred in the other trials and were unreported. Discussion with clinical experts consulted in the ferric carboxymaltose review pertaining to patients with IDA (outside of the HF context) elucidated that hypophosphatemia is relatively easy to manage with inexpensive oral supplements, although it may incur additional monitoring costs, and untreated hypophosphatemia can result in serious health risks. CDEC also noted that the product monograph also recommended that serum phosphate levels be checked in patients at risk of low serum phosphate who require a repeat course of treatment within 3 months.

• Administration costs: During the reconsideration meeting, CDEC discussed that the cost savings estimated by the sponsor were entirely driven by reduced administration costs; however, other administrative costs exist (e.g., patient treatment registration and check-in). Clinical input indicates variability in clinical practice on how the cumulative iron dosage is determined and administered across ferric carboxymaltose indications. Consequently, there are unlikely to be any significant differences in administration costs compared with ferric derisomaltose. If there are any potential cost savings, these would entirely be driven by reduced administration costs. In such instances, such cost savings would not be observed by public drug plan because such costs may pertain to a different budget holder.

• AE costs: During the reconsideration meeting, CDEC noted that the sponsor did not incorporate AEs into the pharmacoeconomic model despite clinical evidence indicating that there may be an increased risk of hypophosphatemia with ferric carboxymaltose that may require additional monitoring and treatment. CDEC felt that these differences in AEs may be associated with different resource use costs. If rates of hypophosphatemia requiring treatment and monitoring are higher than assumed in the CDA-AMC reanalysis or if patients develop more severe AEs (e.g., severe hypophosphatemia, hypophosphatemia osteomalacia, and/or fractures), savings associated with the reimbursement of ferric carboxymaltose may be further reduced or eliminated.

• Budget impact is highly uncertain: During the reconsideration meeting, CDEC noted that the sponsor’s estimates for the budget impact were based on the historical volume of claims. It is unclear whether these claims included the treatment of ID for patients with HF, or if they only represented patients with HF who also already met reimbursement criteria for the treatment of IDA. Because ferric carboxymaltose is explicitly indicated for the treatment of ID (not necessarily accompanied by IDA), the number of patients with HF that will meet criteria is potentially much larger than the number of historical claims. Hence, the true budget impact is unknown.

Background

HF is a complex and life-threatening syndrome in which abnormal heart function leads to subsequent risk of clinical symptoms and signs of reduced cardiac output and/or pulmonary or systemic congestion at rest or with stress. This condition is marked by significant morbidity and mortality, reduced functional capacity, and poor quality of life. Patients with CHF require continuous medical care, frequent monitoring, hospitalizations, and extensive treatment. Symptoms of HF are classified using NYHA functional classes I to IV, which categorize the severity of symptoms ranging from minimal limitations during physical activity (class I) to severe symptoms even at rest (class IV). These symptoms reflect the progressive impact of HF on daily activities and quality of life. Approximately 60% of patients with HF have anemia, and 40% of those without anemia have ID. Iron levels that are too low to support adequate hemoglobin synthesis can lead to IDA, which impairs the blood’s ability to carry oxygen efficiently, either due to a reduced number of red blood cells or low hemoglobin levels. While anemia can have various causes, ID is the most prevalent. This condition significantly impacts patient well-being and outcomes because iron plays a critical role in oxygen transport and cellular energy metabolism, particularly in high-energy-demanding tissues like cardiac muscle. The prevalence of ID is 35% to 55% in HF outpatients and 72% to 83% in patients admitted to the hospital due to HF. A recent study in Alberta found that among 17,463 patients with acute HF (AHF), 38.5% had their iron status evaluated within 30 days of the index episode compared to 34.2% of 11,320 patients with CHF. Of those tested, 72.6% and 73.9% of patients with acute and chronic HF, respectively, were found to have ID.

Ferric carboxymaltose is a colloidal dispersion that contains iron in a stable ferric state. This complex consists of a polynuclear iron-hydroxide core bound to a carbohydrate ligand. It is specifically formulated to provide easily utilizable iron for the body's iron transport and storage proteins, namely transferrin and ferritin. Ferric carboxymaltose was approved by Health Canada for the treatment of ID in adult patients with HF and NYHA class II or III to improve exercise capacity. The diagnosis of ID must be based on laboratory tests. The dosage of ferric carboxymaltose is expressed as milligrams of elemental iron, with each millilitre containing 50 mg of elemental iron. The recommended dosing of ferric carboxymaltose for adult patients follows a stepwise approach, first determining the individual iron need for repletion based on the patient’s body weight and hemoglobin level. The maximum recommended cumulative dosage of ferric carboxymaltose is 1,000 mg of iron (20 mL ferric carboxymaltose) per week.

Sources of Information Used by the Committee

To make its recommendation, the committee considered the following information:

• A review of 4 placebo-controlled, double-blind, randomized trials, 3 of which enrolled patients with CHF (FAIR-HF, CONFIRM-HF, and HEART-FID) and 1 that enrolled patients with AHF (AFFIRM-AHF)

• Input from public drug programs that participate in the Reimbursement Review process

• Input from 1 clinical specialist with expertise in diagnosing and treating patients with HF

• Input from 1 clinician group comprising 13 independent clinical experts

• A review of the pharmacoeconomic model and report submitted by the sponsor

• Information submitted as part of the sponsor’s request for reconsideration (described subsequently)

• Feedback on the draft recommendation.

Perspectives of Patients, Clinicians, and Drug Programs

Patient Input

No patient group input was submitted.

Clinician Input

Input From Clinical Experts Consulted for This Review

Treatment of ID with oral iron, although widely available and inexpensive, was described by the expert to be of limited utility due to poor absorption in general, but especially among patients with HF due to a variety of physiological factors specific to HF, such as epithelial dysfunction in the gut because of mucosal edema and reduced intestinal blood flow. The clinical expert indicated that IV iron is currently the preferred and guideline-recommended route for treatment of ID in patients with HF, and the intention of iron supplementation for ID in patients with HF is to improve HRQoL, functional capacity, and exercise capacity. The clinical expert said that IV irons typically used in clinical practice include iron sucrose (maximum dosage of 200 mg per sitting), ferric derisomaltose (maximum dose of 1,000 mg per injection), or ferric carboxymaltose (maximum dosage of 1,000 mg per week). Of the IV iron formulations, only ferric carboxymaltose has a Health Canada–approved indication specific to the HF subpopulation. However, the other second- or third-generation IV irons may also be used in clinical practice in this population.

The expert noted that guidelines for the treatment of HF recommend that all patients with HF be tested for ID using serum ferritin and TSAT, and that Canadian treatment guidelines recommend consideration of IV iron therapy for patients with HF with all of the following: a LVEF of 40% or less, a serum ferritin level of less than 100 mcg/L or between 100 mcg/L and 299 mcg/L, and a TSAT of less than 20%. The expert noted that, based on Canadian and multiple international treatment guidelines for patients with HF and ID, treatment with IV iron formulations, including ferric carboxymaltose, may be suitable for patients with HF of any NYHA class.

The clinical expert noted that most patients with HF receiving IV iron supplementation would be expected to continue this therapy for the duration of their lives. As long as the guideline criteria for iron replacement therapy are unchanged, and aside from intolerable AE or patient or clinician decision or preference, there are no specific reasons to require discontinuation of ferric carboxymaltose in a patient with HF and ID. According to the expert, there is no threshold of any laboratory parameter under which the drug should be discontinued due to lack of efficacy, and treatment should be required for as long as dictated by guideline criteria for ID-related IV iron replacement therapy.

IV iron formulations such as ferric carboxymaltose are prescribed in hospitals and can be prescribed by any prescribing clinician managing a patient’s HF and ID in that setting.

Clinician Group Input

A group of 13 independent clinical experts responding to the CDA-AMC call gathered data from product monographs, literature, and personal experience. According to the group, ID is a progressive condition that can lead to IDA if untreated, affecting and impacting patients with HF by worsening disease symptoms and prognosis.

The clinician group emphasized that treatment goals include correcting hemoglobin deficits, replenishing iron stores, and maintaining iron stores over time to alleviate symptoms and enhance HRQoL. While initial therapy often involves oral iron supplements, IV iron is recommended as the first-line treatment for patients with HF due to its rapid efficacy, especially given that up to 50% of HF patients can experience ID, leading to poorer functional capacity and increased hospitalizations and mortality. The clinician group noted that guidelines advocate for initiating IV iron therapy as soon as ID is identified.

The group noted challenges with previous IV iron formulations in Canada in that they required prolonged administration and were not indicated for use in pediatric populations or for the treatment of patients with ID and HF, underscoring the need for more efficient options. Newer products like ferric carboxymaltose can deliver high doses (up to 1,000 mg) in a single session, potentially reducing treatment burden and improving adherence.

The clinician group indicated that treatment response is assessed using hematologic and iron parameters, aiming to normalize hemoglobin and ferritin levels. Clinically meaningful outcomes also include reduced need for blood transfusions, symptom alleviation, enhanced exercise capacity, improved quality of life, and fewer hospitalizations. Monitoring typically occurs 4 to 8 weeks after completing the initial treatment course to track progress and adjust therapy as needed.

According to the input, factors to consider when deciding to discontinue treatment with ferric carboxymaltose include hemoglobin, ferritin, and TSAT levels after iron repletion. Treatment should be immediately discontinued in cases of hypersensitivity reactions or intolerance during administration, and it is contraindicated in patients with iron overload or persistent hypophosphatemia for whom re-evaluation of treatment is warranted. Ferric carboxymaltose is an appropriate treatment in settings equipped to manage anaphylaxis and hypersensitivity reactions. It can also be administered in emergency departments or surgical inpatient units when indicated. Although specialists such as hematologists and other physicians commonly prescribe ferric carboxymaltose, a specialist is not always required for diagnosis, treatment, and monitoring. Family medicine practitioners and specialists in cardiology, gastroenterology, internal medicine, nephrology, and obstetrics/gynecology, among others, may also manage patients requiring IV iron therapy.

Drug Program Input

Input was obtained from the drug programs that participate in the Reimbursement Review process. The following were identified as key factors that could potentially affect the implementation of a recommendation for ferric carboxymaltose:

• Relevant comparators

• Considerations for initiation of therapy

• Considerations for continuation or renewal of therapy

• Considerations for discontinuation of therapy

• Considerations for the prescribing of therapy

• Care provision issues

• System and economic issues.

The clinical expert consulted for the review provided advice on the potential implementation issues raised by the drug programs.

Table 2: Responses to Questions From the Drug Programs

AE = adverse event; AHF = acute heart failure; CDEC = Canadian Drug Expert Committee; CHF = chronic heart failure; GI = gastrointestinal; HF = heart failure; ID = iron deficiency; IDA = iron deficiency anemia; TSAT = transferrin saturation.

Clinical Evidence

Systematic Review

Description of Studies

The studies included were FAIR-HF (N = 459), CONFIRM-HF (N = 304), and HEART-FID (N = 3,065) in patients with CHF, and AFFIRM-AHF (N = 1,132) in patients with AHF, all 4 of which were placebo-controlled, double-blind, randomized phase III (FAIR-HF and HEART-FID) or phase IV (CONFIRM-HF and AFFIRM-AHF) trials in adults with HF and ID. Of these, 2 studies (FAIR-HF and CONFIRM-HF) were focused primarily on clinical efficacy outcomes such as exercise capacity and NYHA class, whereas the remaining 2 (HEART-FID and AFFIRM-AHF) were focused primarily on composite outcomes related to hospitalizations and deaths. The studies ranged in duration from approximately 6 months (FAIR-HF) to 12 months (CONFIRM-HF, HEART-FID, and AFFIRM-AHF).

The 3 CHF studies each had a maximum allowed LVEF at screening or index visit, although the precise threshold varied: 40% or less (NYHA II) or 45% or less (NYHA III) in FAIR-HF, 45% or less in CONFIRM-HF, and 40% or less in HEART-FID (although historically reduced LVEF was also allowed in specific circumstances). In the AFFIRM-AHF study, the inclusion criteria required that patients had a LVEF of less than 50% within 12 months before randomization. All 4 included studies required a serum ferritin of less than 100 ng/mL, or 100 ng/mL to 299 ng/mL or 300 ng/mL with a TSAT less than 20%. At baseline, mean patient ages were approximately 68 to 71 years across the treatment groups, and the proportion of female patients ranged from 33% to 55%. All 4 studies included adult patients with NYHA class II or III HF. Although the HEART-FID study also included NYHA class IV and the AFFIRM-AHF study included NYHA classes I and IV, the overwhelming majority of patients belonged to NYHA class II or III, with very few patients belonging to class IV (< 4% in the AFFIRM-AHF study and < 1% in the HEART-FID study) or class I (< 3% in the AFFIRM-AHF study). Patients who self-identified as white were disproportionately over-represented in all the studies; 86% in the HEART-FID study, 95% in the AFFIRM-AHF study, and 99% to 100% in the FAIR-HF study and the CONFIRM-HF study. Comorbidities were common, including hypertension, dyslipidemia, diabetes, atrial fibrillation, angina pectoris, and others.

Efficacy Results

NYHA Class

The NYHA functional class system is a subjective but widely used classification used to determine CHF severity based on symptoms. NYHA class I suggests little to no symptoms of HF, and class IV is defined by the inability to carry on any physical activity without discomfort and by the presence of symptoms even at rest.

For the outcome of change in NYHA class from baseline, there was a benefit associated with ferric carboxymaltose compared to placebo in the FAIR-HF study at week 24 (odds ratio [OR] = 2.40; 95% CI, 1.55 to 3.72; P < 0.001), for which an OR of greater than 1 indicates a benefit of ferric carboxymaltose, and in CONFIRM-HF at week 24 (OR [████ ████ ███ ████ ██ ████], P = 0.004) and week 52 (OR: ████ ████ ███ ████ ██ ██████ P < 0.001), for which an OR of less than 1 indicates a benefit of ferric carboxymaltose compared to placebo. In AFFIRM-AHF, although the point estimate suggests a benefit associated with ferric carboxymaltose at week 52 based on an OR of greater than 1 (OR: ████ ████ ███ ████ ██ ███████████), there was insufficient evidence to confirm a difference between the treatment arms because the 95% CI crossed the null. The HEART-FID study did not report this outcome.

6-Minute Walk Test

The 6MWT is a common, validated test that measures the distance a patient can walk on a hard, flat surface over a 6-minute period under clinician supervision. Longer distances represent better exercise capacity. This outcome was assessed in the FAIR-HF, CONFIRM-HF, and HEART-FID studies but not in the AFFIRM-AHF study.

In the FAIR-HF study, the mean change from baseline in the 6MWT distance at 24 weeks was ████ ██████ (standard deviation [SD] █████) in the ferric carboxymaltose group compared to ████ ██████ (SD █████) in the placebo group, and the between-group difference was ████ (SD ███) metres (█████). Because no 95% CI was presented for the between-group results, additional data were requested from the sponsor, who provided a post hoc analysis of absolute differences on request to inform the GRADE analysis; according to this additional data, which may not follow the same analysis as described in the study’s Statistical Analysis Plan, the between-group difference was ████ ████ ███ ████ ██ ██████████ metres.

In the CONFIRM-HF primary analysis at week 24, the change from baseline in the 6MWT distance was ████ ███ ██████ ███ ███ ████ ██ ███████ █████ ███ ██████ ███ ███ ██████ ██ ████) metres in the ferric carboxymaltose and placebo groups, respectively. The least squares mean between-group difference was ████ ██████ (standard error [SE] ██████ ███ ███ ████ ██ ██████.

In the HEART-FID study, as a component of the composite primary outcome, the mean change from baseline to 6 months (i.e., 24 weeks) in the 6MWT distance was 8 metres (SD = 60) and 4 metres (SD = 59), respectively. Because no between-group differences were presented numerically, additional information was requested from the sponsor to support the GRADE analysis, which found the following: the mean change from baseline to week 24 in the 6MWT distance was █████ ███ ██████ metres in the ferric carboxymaltose group and █████ ███ ██████ metres in the placebo group, with a between-group difference in change from baseline of █████ ████ ███ ██████ ██ ██████ metres.

In the CONFIRM-HF secondary analyses at week 52, the least squares mean between-group difference was ████ metres (95% CI, ████ ██ █████████). Sensitivity analyses using the per-protocol (PP) set did not include reporting of between-group differences, but the within-group changes from baseline were similar to those of the full analysis set (FAS) analyses; the treatment benefit was also consistent across preplanned and post hoc subgroup analyses on demographic and disease-related features, and in a supportive analysis without primary imputation for deaths and hospitalizations.

In the HEART-FID secondary analyses at week 52, the mean change was 5 metres (SD = 71) in the ferric carboxymaltose group and 4 metres (SD = 72) in the placebo group. Between-group values were not reported numerically, so additional information was requested from the sponsor, in which it was reported that the mean change from baseline to week 52 was █████ ███ ██████ and █████ ███ ██████ metres for the ferric carboxymaltose and placebo groups, respectively, with a between-arm difference in change from baseline of █████ ████ ███ ██████ ██ ██████ metres.

Kansas City Cardiomyopathy Questionnaire

The KCCQ was reported in the FAIR-HF study (24 weeks), the CONFIRM-HF study (24 and 52 weeks), and the AFFIRM-AHF study (24 and 52 weeks) as a secondary outcome in each case. It is a 23-item, self-administered questionnaire that quantifies physical limitation, symptoms (stability, frequency, and burden), self-efficacy, social function, and HRQoL. Scores are transformed to a range of 0 to 100, with higher scores reflecting better health status. Although studies have been performed to assess its measurement properties and validity, the KCCQ is primarily a clinical trial tool and is not typically used in real-world clinical practice.

In the FAIR-HF study at week 24, the study treatment effect of ferric carboxymaltose was █████ ███ ██ greater in change from baseline of the KCCQ overall summary score, compared to placebo (P < 0.001). Because there was no 95% CI provided with the point estimate in the FAIR-HF study, additional data were requested from the sponsor, which showed a between-group difference of ███ ██████ ████ ███ ███ ██ ██████████ favouring ferric carboxymaltose.

In the CONFIRM-HF study at week 24, the least squared mean between-group difference was ███ ████ ███ █████ ██ █████ █████ points. In the AFFIRM-AHF study, the difference was ███ ████ ███ ███ ██ █████████.

In the CONFIRM-HF study at week 52, the between-group difference was ███ ████ ███ ████ ██ ██████████ points, and in the AFFIRM-AHF study it was 1.44 (95% CI, –1.45 to 4.33, P not reported)

Fatigue Score

Only the CONFIRM-HF study assessed the change from baseline in fatigue score, ranked on a visual analogue scale from 0 to 10, in which 0 implies no fatigue and 10 represents very severe fatigue. Some assessments of validity exist for this method of measuring fatigue, but there is no established minimal important difference (MID). At week 24, the between-group difference (as least squares mean) in change in fatigue score was ████ ███ █████ ███ ███ █████ ██ ███████████, and at week 52, it was ████ ███ █████ ███ ███ █████ ██ ███████████.

Serum Ferritin

At baseline, mean serum ferritin values were below the threshold of 100 ng/mL that defines ID in the context of HF in all included studies.

At week 24, across all studies, the ferric carboxymaltose groups had mean serum ferritin levels of greater than 100 (although note that patients may still have “functional ID” if their TSAT is < 20% and their serum ferritin is 100 ng/mL to 300 ng/mL, as previously discussed), while the mean serum ferritin levels in the placebo groups were near or below 100 ng/mL.

In the FAIR-HF study at week 24, the between-group difference in absolute mean serum ferritin was █████ ███ ██████ █████ in the ferric carboxymaltose group and ████ ███ █████ ng/mL in the placebo group (P < 0.0001).

In the CONFIRM-HF study, between-group values were not reported, but the mean serum ferritin at week 24 was ██████ █████ ███ ████████ in the ferric carboxymaltose group, representing a mean change from baseline of ██████ █████ ███ ████████, and in the placebo group, the mean serum ferritin was █████ █████ ███ ███████ representing a change from baseline of ████ █████ ████████. Additional data were provided by the sponsor upon request, in which it was reported that the between-group difference in change from baseline was ██████ █████ ███████ ██ ███████.

Numerical values for serum ferritin were not reported in the HEART-FID study, although a graphical representation can be found in the supplementary documents. Additional data were provided by the sponsor upon request, in which it was reported that the between-group difference in change from baseline was ██████ █████ ███████ ██ ███████ at week 24.

In the AFFIRM-AHF study, at week 24, the mean (SD) change from baseline was ██████ █████ █████████ in the ferric carboxymaltose group compared to █████ █████ █████████ in the placebo group; no between-group differences were reported. Additional data were provided by the sponsor upon request, in which it was reported that the between-group difference in change from baseline was ██████ █████ ███████ ██ ███████ at week 24.

At week 52, the mean serum ferritin levels in the ferric carboxymaltose groups of the CONFIRM-HF, HEART-FID, and AFFIRM-AHF studies were all in excess of 100 ng/mL but were lower than 300 ng/mL. FAIR-HF was a 24-week study, so there are no 52-week values. In the placebo groups, the levels were close to or below 100 ng/mL.

In the CONFIRM-HF study at 52 weeks, between-group values were not reported, but the mean serum ferritin ng/mL (SD) at week 52 in the ferric carboxymaltose group was ██████ █████ █████████ representing a change from baseline of ██████ █████ █████████, and in the placebo group was █████ █████ █████████ representing a change from baseline of █████ █████ █████████. Additional data were provided by the sponsor upon request, in which it was reported that the between-group difference in change from baseline was ██████ █████ ███████ ██ ███████ at week 52.

Values for serum ferritin were not reported in HEART-FID, although graphical representation can be found in the supplementary documents. Additional data were provided by the sponsor upon request, in which it was reported that the between-group difference in change from baseline was ██████ █████ ███████ ██ ███████ at week 52.

In the AFFIRM-AHF study at week 52, the mean (SD) change from baseline was ██████ █████ █████████ in the ferric carboxymaltose group compared to █████ █████ █████████ in the placebo group; no between-group differences were reported. Additional data were provided by the sponsor upon request, in which it was reported that the between-group difference in change from baseline was ██████ ███████ ██ ███████.

CV Hospitalizations

De novo analyses of the CV hospitalization rate were provided by the sponsor upon request to assist in the review. Through both 26 and 52 weeks, the between-group difference in event rate per 100 patient-years was lower in the ferric carboxymaltose groups than the placebo groups. Through 26 weeks, the between-group difference in event rate per 100 patient-years (95% CI) for ferric carboxymaltose versus placebo was ██████ ███████ ██ ██████ in the FAIR-HF study, ██████ ███████ ██ █████ in the CONFIRM-HF study, █████ ███████ ██ ██████ in the HEART-FID study, and ██████ ███████ ██ ██████ in the AFFIRM-AHF study. Through week 52, the between-group differences were ██████ ███████ ██ ██████ in the CONFIRM-HF study, █████ ███████ ██ ██████ in the HEART-FID study, and ██████ ███████ ██ ██████ in the AFFIRM-AHF study.

The HEART-FID and AFFIRM-AHF studies both had composite primary efficacy end points that included hospitalization-related outcomes; these will be discussed briefly here.

In the hierarchical composite primary end point of the HEART-FID study, at 12 months, 131 patients (8.6%) in the ferric carboxymaltose group and 158 patients (10.3%) in the placebo group had died, 297 patients in the ferric carboxymaltose group and 332 patients in the placebo group had been hospitalized for HF, and the mean change in the 6MWD from baseline to 6 months was 8 metres (SD = 60) in the ferric carboxymaltose group and 4 metres (SD = 59) in the placebo group (overall P = 0.02). The unmatched win ratio for the hierarchical composite outcome in the ferric carboxymaltose group compared with the placebo group was 1.10 (99% CI, 0.99 to 1.23). Results of prespecified sensitivity analyses that included different imputation methods was reported to be consistent with those of the primary analysis. Because more than half the patients underwent randomization after March 2020, the censoring of data after the onset of the COVID-19 pandemic would have excluded the majority of follow-up data from the analyses.

In the primary efficacy end point of the AFFIRM-AHF study, which was a composite of recurrent HF hospitalizations and CV deaths up to 52 weeks after randomization, the annualized event rate ratio for ferric carboxymaltose versus placebo was 0.79 (95% CI, 0.62 to 1.01; P = 0.059). In the prespecified COVID-19 sensitivity analysis, the annualized event rate ratio for ferric carboxymaltose versus placebo was 0.75 (95% CI, 0.59 to 0.96; P = 0.024).

CV Mortality

De novo analyses of CV mortality were provided by the sponsor upon request to assist in the review. Through both 26 and 52 weeks, minor and inconsistent differences were observed, with 95% CIs that always crossed the null. Through 26 weeks, the between-group risk difference (95% CI) comparing ferric carboxymaltose to placebo was █████ ██████ ██ ██████ in the FAIR-HF study, ████ ██████ ██ ██████ in the CONFIRM-HF study, █████ ██████ ██ ██████ in the HEART-FID study, and █████ ███████ ██████ in the AFFIRM-AHF study. Through 52 weeks, the values were █████ ███████ ██████ in the CONFIRM-HF study, █████ ███████ ██████ in the HEART-FID study, and █████ ███████ ██████.

Harms Results

Adverse Events

In the FAIR-HF study, the proportion of patients who experienced at least 1 AE was █████ in the ferric carboxymaltose group and █████ in the placebo group, while in the CONFIRM-HF study, the values were 79.6% and 75.7%, respectively. Overall. AEs of any severity were not reported in HEART-FID. In AFFIRM-AHF, at least 1 AE was experienced by ███ ███████ patients in the ferric carboxymaltose group and ███ ███████ patients in the placebo group.

Common AEs that occurred in at least 5% of any 1 treatment group across the FAIR-HF and CONFIRM-HF studies included cardiac failure, atrial fibrillation, angina pectoris, bronchitis, respiratory tract infection (viral), nasopharyngitis, influenza, increased blood pressure, hypertension, hypotension, headache, dizziness, and skin or subcutaneous tissue disorders. For the most part, the proportion of patients experiencing these events was relatively similar between treatment groups. Cardiac failure (chronic) appeared to be slightly less common in the ferric carboxymaltose group than the placebo group (FAIR-HF: ████ ███ █████ CONFIRM-HF: ████ ███ ████). Common AEs that occurred in at least 5% of either treatment group in the AFFIRM-AHF study included cardiac failure, infections, gastrointestinal (GI) disorders, nervous system disorders, metabolism and nutrition disorders, renal and urinary disorders, vascular disorders, injury, and musculoskeletal disorders.

Serious Adverse Events

In the FAIR-HF study, a total ██ ██ ████████ ███████ in the ferric carboxymaltose group and ██ ████████ █████ in the placebo group reported at least 1 serious adverse event (SAE). The most common SAE was cardiac disorders (████ ███ ████ in the ferric carboxymaltose and placebo groups, respectively).

In the CONFIRM-HF study, 43 patients (28.3%) in the ferric carboxymaltose group and 53 patients (34.9%) in the placebo group reported at least 1 SAE. The most common SAE was cardiac disorders (█████ ███ █████ in the ferric carboxymaltose and placebo groups, respectively).

In the HEART-FID study, SAEs were reported in 581 patients (37.9%) in the ferric carboxymaltose group and 537 patients (35.0%) in the placebo group. The most common SAEs during the treatment period were pneumonia, reported in 57 patients (3.7%) in the ferric carboxymaltose group and 35 patients (2.3%) in the placebo group; acute kidney injury, reported in 46 patients (3.0%) in the ferric carboxymaltose group and 40 patients (2.6%) in the placebo group; and COVID-19, reported in 39 patients (2.5%) in the ferric carboxymaltose group and 37 patients (2.4%) in the placebo group. One event in the ferric carboxymaltose group was classified as hypophosphatemia. This event resolved and ferric carboxymaltose treatment was continued.

In the AFFIRM-AHF study, SAEs occurred in 250 (44.7%) of 559 patients in the ferric carboxymaltose group and 282 (51.2%) of 551 patients in the placebo group. The most common were cardiac disorders (█████ ███ █████ in the ferric carboxymaltose and placebo groups, respectively), followed by infections and infestations (████ ███ ████ in the ferric carboxymaltose and placebo groups, respectively), followed by general disorders and administration site conditions (████ ███ ████ in the ferric carboxymaltose and placebo groups, respectively). Other SAEs were less common.

Withdrawals Due to AEs

In the FAIR-HF study, ██ ████████ ██████ in the ferric carboxymaltose group and ██ ████████ ██████ in the placebo group withdrew from the study treatment due to AEs; in the CONFIRM-HF study, this occurred among ██ ████████ ██████ in the ferric carboxymaltose group and ██ ████████ ███████ in the placebo group; in the HEART-FID study, ██ ███████ and ██ ███████; and in the AFFIRM-AHF study, ██ ███████ ███ ██ ███████, respectively.

Deaths

In the FAIR-HF study, 5 patients in the ferric carboxymaltose group died (████) and 4 patients in the placebo group died (████). During the study period, ██ patients from the ferric carboxymaltose group and ██ from the placebo group died. In the ferric carboxymaltose group, 3 patients died suddenly, ███ died due to ischemic stroke, and ███ died due to severe anemia after terminating the study early. In the placebo group, █████ patients died due to myocardial infarction, pulmonary edema, and sudden death.

In the CONFIRM-HF study, a total of ██ ████████ ██████ in the ferric carboxymaltose group and ██ ████████ ██████ in the placebo group died during the study period. The majority of deaths (██) were related to cardiac disorders and cardiac-related TEAEs in other system organ classes (e.g., sudden cardiac death and cardiac death in the category of general disorders and administration site conditions). Two patients in the placebo group died of noncardiac disorders (staphylococcal sepsis and acute renal failure).

In the HEART-FID study, 361 patients (23.6%) in the ferric carboxymaltose group and 376 patients (24.5%) in the placebo group died from any cause (hazard ratio, 0.90; 95% CI, 0.78 to 1.05). The hazard ratio for death from any cause through month 12 was 0.82 (95% CI, 0.65 to 1.05). It is unknown how many deaths were due to AE.

In the AFFIRM-AHF study, ██ ███████ patients in the ferric carboxymaltose group and ██ ███████ in the placebo group had TEAEs resulting in death. The majority of deaths were related to cardiac disorders and cardiac-related TEAEs in other organ classifications (e.g., sudden cardiac death and cardiac death in the category of general disorders and administration site conditions).

Notable Harms

Hypophosphatemia is a notable harm associated with ferric carboxymaltose treatment that is not as highly associated with other IV iron formulations.

There were no reported cases of hypophosphatemia as TEAEs in both the FAIR-HF and CONFIRM-HF studies.

In the FAIR-HF study, transient decreases in phosphate levels were observed in the ferric carboxymaltose group, and this was reported to be most pronounced at week 4, but there were no clinical consequences, sequelae, or interventions associated with this change. Differences between treatment arms in the percentage of patients with values outside the normal range were observed for phosphate during follow-up (██ ███████ in the ferric carboxymaltose group versus ██ ███████ in the placebo group (P = 0.008).

In the CONFIRM-HF study, the minimum recorded serum phosphorus value was ████ ██████ █████ ██████ in 2 patients (1 in the ferric carboxymaltose group and 1 in the placebo group), for which hypophosphatemia is typically defined as a serum phosphorus value of less than 2.5 mg/dL. Among all patients, ██ experienced severe hypophosphatemia based on the clinical events classification threshold of 0.3 to less than 0.6 mmol/L, although the investigators did not report these as TEAEs.

The overall incidence of hypophosphatemia was not reported in the HEART-FID study, but 1 SAE of hypophosphatemia reportedly occurred in 1 patient in the ferric carboxymaltose group and none in the placebo group. This hypophosphatemia event was considered by the investigator to be unrelated to ferric carboxymaltose; the event resolved and ferric carboxymaltose was continued.

In the AFFIRM-AHF study, there were ██ cases of hypophosphatemia reported, ██ of which was in the ferric carboxymaltose group and ██ was in the placebo group.

Critical Appraisal

The overall risk of bias regarding internal validity was low for the randomization process, allocation concealment, and maintenance of blinding.

Concerns for potentially important missing data were present for most outcomes assessed in at least 1 contributing study.

Regarding the outcomes assessed in this review, the primary efficacy analyses of the FAIR-HF study (i.e., NYHA class) and the CONFIRM-HF study (i.e., 6MWT) were adjusted for multiplicity, but other outcomes from these studies were not.

In supportive analyses of the CONFIRM-HF study using the PP set instead of the FAS, the sample size was substantially reduced due to a high number of protocol violations, especially in the active treatment arm, which may result in a risk of bias due to deviations from the intended interventions.

The external validity and applicability of the results of this review are limited by the absence of any direct or indirect evidence comparing ferric carboxymaltose with other IV irons, which are its direct comparators in patients with HF and ID despite having no specific indication among the population of people with HF in Canada at this time. No conclusions can be drawn from any of the submitted evidence on the relative efficacy or safety of ferric carboxymaltose with any other available IV iron formulation in patients with HF and ID. Additionally, there may be generalizability concerns regarding the demographics of the studies. The proportion of patients identified as white was disproportionately high across the trials, particularly in the FAIR-HF and CONFIRM-HF studies (98% to 100% across the treatment arms), but this was also generally true in the other 2 studies (85% and 95%). The trials were primarily conducted in countries other than Canada, and so demographic features as well as clinical practice related to both ID and HF individually and together may differ. The HEART-FID study included US sites, but the other 3 included studies were conducted primarily in Eastern Europe with some sites in Asia, Oceania, South America, and Western Europe, and so the clinical practices and patient characteristics may differ from those in Canada. Additionally, the HEART-FID study was the most recently conducted study and had the most inconsistent results, especially regarding 6MWT, when compared to the other included studies. The reason for this inconsistency is not certain, but it may be related to changes in clinical practice and standard of care over time.

The proposed reimbursement request for HF is specific to patients of NYHA class II or III. However, this does not reflect the treatment guidelines, which do not specify any particular NYHA class in its recommendations for IV iron repletion therapies. The AFFIRM-AHF study enrolled patients with a broader range of NYHA classes (I to IV inclusive), and the HEART-FID study in patients with CHF also enrolled patients of NYHA class IV. However, the number of patients of NYHA class I and/or IV was proportionally very low in both studies.

GRADE Summary of Findings and Certainty of the Evidence

The selection of outcomes for GRADE assessment was based on the sponsor’s Summary of Clinical Evidence, consultation with the clinical expert, and input received from the clinician group and public drug programs. The following list of outcomes was finalized in consultation with expert committee members:

• HF disease severity as measured by NYHA score

• Exercise capacity as measured by change in the 6MWT distance from baseline

• HRQoL as measured using change from baseline in KCCQ score

• Fatigue score

• Change from baseline in serum ferritin

• CV hospitalization rate

• CV mortality risk.

Table 3: Summary of Findings for Ferric Carboxymaltose Versus Placebo for Patients With CHF and ID

6MWT = 6-minute walk test; CHF = chronic heart failure; CI = confidence interval; CV = cardiovascular; FCM = ferric carboxymaltose; HF = heart failure; HRQoL = health-related quality of life; ID = iron deficiency; KCCQ = Kansas City Cardiomyopathy Questionnaire; LSM = least squared mean; MID = minimal important difference; NR = not reported; NYHA = New York Heart Association; OR = odds ratio; PBO = placebo; RCT = randomized controlled trial.

Note: Study limitations (which refers to internal validity or risk of bias), inconsistency across studies, indirectness, imprecision of effects, and publication bias were considered when assessing the certainty of the evidence. All serious concerns in these domains that led to the rating down of the level of certainty are documented in the table footnotes.

^aThis outcome is subjective, but there was no suspected risk of bias due to adequate maintenance of blinding.

^bNo known MID, so the target of certainty appraisal was any effect.

^cOnly FAIR-HF adjusted for multiplicity for this outcome (NYHA). With that exception noted, no studies adjusted for multiplicity for this or any other outcome in this table.

^dRated down 1 level for risk of bias due to missing data.

^eThis value was provided by the sponsor upon request to assist in the interpretation of the evidence. Note that the analysis is post hoc and not necessarily represented in the Statistical Analysis Plan of the relevant study.

^fRated down 2 levels for very serious inconsistency due to differences in magnitude of effect between the studies, wherein the 95% CIs of some studies are minimally overlapping or not overlapping.

^gRated down 1 level for serious imprecision. CI crosses thresholds between 1 set of: harm, no difference, or benefit.

^hThe total number of events for all patients in the treatment group was divided by the total patient-years of follow-up in that treatment group multiplied by 100. Follow-up duration is equal to time on study. Time on study (weeks) = (last known date - randomization date + 1)/7.

ⁱRisk difference is FCM – placebo with 95% Miettinen-Nurminen CI.

^jRated down 1 level for indirectness. Based on clinical expert opinion, the duration of assessment is likely insufficient to identify a difference between treatment groups for this outcome.

Sources: Additional information request for absolute differences results data, clinical study reports for FAIR-HF and CONFIRM-HF, and publication and supplementary appendix of HEART-FID.

Table 4: Summary of Findings for Ferric Carboxymaltose Versus Placebo for Patients With AHF and ID

6MWT = 6-minute walk test; AHF = acute heart failure; CI = confidence interval; CV = cardiovascular; FCM = ferric carboxymaltose; HF = heart failure; HRQoL = health-related quality of life; ID = iron deficiency; KCCQ = Kansas City Cardiomyopathy Questionnaire; NA = not available; NYHA = New York Heart Association; OR = odds ratio; PBO = placebo; RCT = randomized controlled trial.

Note: Study limitations (which refers to internal validity or risk of bias), inconsistency across studies, indirectness, imprecision of effects, and publication bias were considered when assessing the certainty of the evidence. All serious concerns in these domains that led to the rating down of the level of certainty are documented in the table footnotes.

^aRated down 1 level for serious risk of bias caused by a high proportion of potentially important missing data.

^bThis outcome was not adjusted for multiplicity.

^cNo known minimal important difference, so the target of certainty appraisal was any effect.

^dRated down 1 level for risk of bias due to missing data.

^eThis value was provided by the sponsor upon request to assist in the interpretation of the evidence. If this value represents an analysis outcome, note that the analysis is post hoc and not necessarily represented in the Statistical Analysis Plan of the relevant study.

^fRated down 1 level for serious imprecision. There is no known minimal important difference, so the target of certainty appraisal was any effect.

^gThe total number of events for all patients in the treatment group was divided by the total patient-years of follow-up in that treatment group multiplied by 100. Follow-up duration is equal to time on study. Time on study (weeks) = (last known date - randomization date + 1)/7.

^hRisk difference is FCM – placebo with 95% Miettinen-Nurminen CI.

ⁱRated down 1 level for indirectness. Based on clinical expert opinion, the duration of assessment is likely insufficient to identify a difference between treatment groups for this outcome.

^jAlthough the CI includes the possibility of both benefit and harm, the CI is relatively narrow, actually around the null, so it was subjectively judged that this was not imprecise enough to warrant rating down a second time for imprecision.

Source: Additional information request for absolute differences results data and Clinical Study Report of AFFIRM-AHF.

Long-Term Extension Studies

No long-term extension studies were submitted for this indication.

Indirect Comparisons

No indirect comparisons were submitted for this indication.

Studies Addressing Gaps in the Evidence From the Systematic Review

Summary and Critical Appraisal of the Ponikowski et al. (2024) Meta-Analysis

Objective

The objective of the meta-analysis was to evaluate the effects of ferric carboxymaltose treatment on clinical events, such as hospitalizations and mortality, in patients with HF and ID using patient-level data from randomized, placebo-controlled trials of ferric carboxymaltose that enrolled adults with HF and ID.

Methods

Inclusion Criteria

Ponikowski et al. performed a pooled analysis of patient-level data from trials that met the following criteria: adult patients with HF and ID (defined as ferritin < 100 ng/mL or ferritin 100 ng/mL to 300 ng/mL and TSAT < 20%); patients used ferric carboxymaltose as an active treatment for ID; studies were double-blind, randomized, placebo-controlled trials; patients had at least 52 weeks of follow-up; prospectively recorded clinical outcomes were first and recurrent HF and CV hospitalizations, CV death, and all-cause death.

Included Studies

The meta-analysis ultimately included CONFIRM-HF, AFFIRM-AHF, and HEART-FID. Searches were conducted to identify any additional studies, but none were added per the eligibility criteria. The authors were able to access individual patient data for the included studies.

End Points

The prespecified coprimary efficacy end points were:

• a composite of recurrent (total) CV hospitalizations and death for any CV reason (CV death)

• a composite of recurrent (total) HF hospitalizations and CV death.

All the outcomes were based on events adjudicated independently by blinded event committees. All 3 ferric carboxymaltose trials used consistent criteria for adjudication, which were prespecified in each trial.

The key secondary efficacy end points were time to first CV hospitalization or CV death, time to first HF hospitalization or CV death, rate of total HF hospitalizations, time to first HF hospitalization, time to CV death, time to all-cause death, total CV hospitalizations, time to CV death, time to all-cause death, total CV hospitalizations, time to first CV hospitalization, and total all-cause hospitalizations.

All primary and key secondary end points were examined through 52 weeks of follow-up (primary end points: set with a time window of up to 408 days).

Data Analysis

Efficacy analyses were conducted on the full analysis population defined as all randomly assigned patients who received at least 1 dose of study medication and had at least 1 post-baseline efficacy assessment. The safety population comprised all patients who were randomly assigned and received at least 1 dose of study medication, and it was used to assess baseline characteristics and analyze the frequency of AEs.

A negative binomial regression model was used to analyze event rates, including recurrent hospitalizations. The models were adjusted for baseline hemoglobin and region as fixed effects. Study was included as a random effect. The between-trial heterogeneity was explored by including a treatment-by-study interaction and a Cochrane Q test. Time-to-event outcomes used Cox proportional hazard analyses, and the models were adjusted for hemoglobin at baseline and region. To explore between-trial heterogeneity, the study effect was included as a fixed effect.

Several preplanned subgroup analyses and sensitivity analyses were conducted. Because the purpose of the sponsor presenting this meta-analysis was in part to discuss the impact of TSAT on efficacy results, the TSAT subgroup will be discussed briefly in this subgroup, but other subgroups will not be.

A sensitivity analysis incorporating the IRONMAN trial (which did not include ferric carboxymaltose) was also conducted to evaluate IV irons versus placebo, which will not be discussed in depth here. An additional sensitivity analysis was also conducted using all available follow-up data (i.e., beyond 52 weeks).

Results

Efficacy

The results were in favour of ferric carboxymaltose without the 95% CI overlapping null for the coprimary composite end point of CV death and total CV hospitalizations (rate ratio = 0.86, 95% CI, 0.75 to 0.98; P = 0.029). Similarly ferric carboxymaltose was associated with a 17% relative rate reduction in total CV hospitalizations (rate ratio = 0.83; 95% CI, 0.73 to 0.96; P = 0.009) and a 16% relative rate reduction in total HF hospitalizations (rate ratio = 0.84, 95% CI, 0.71 to 0.98; P = 0.025). For the outcome of total HF hospitalizations and CV death, the result was in favour of ferric carboxymaltose. However, the 95% CI overlapped the null (0.75 to 1.01). For the outcome of time to CV death, the 95% CI also crossed the null, indicating that there was no statistically significant difference between the treatment arms. Similarly, there was no statistically significant difference for time to all-cause death. Rate reductions in the primary composite end points were mainly driven by the treatment effect on HF hospitalizations and CV hospitalizations, with no apparent effect on CV or all-cause mortality.

In terms of subgroup results, there was a significant interaction effect identified between TSAT tertile and the composite of CV hospitalization and CV death (interaction P = 0.019) and between TSAT tertile and CV death (interaction P = 0.035), with patients in a lower TSAT tertile being more likely to experience benefits from treatment than those in a higher TSAT tertile. Although not statistically significant, a similar pattern was observed for the effect of TSAT on total HF hospitalizations and CV death (interaction P = 0.095). There were some other subgroups identified regarding “numerically” different treatment effects by subgroup (e.g., across hemoglobin tertiles and HF etiology); other than these, the effects of ferric carboxymaltose therapy on both of the primary efficacy end points, CV death, and all-cause death were similar across other subgroups examined.

Safety

The incidences of investigator-reported serious treatment-emergent adverse events (TEAEs), serious TEAEs leading to death, and serious TEAEs leading to study discontinuation were similar across treatment groups through week 52. No deaths were judged to be the cause of serious treatment-related TEAEs. The rate of serious treatment-emergent infections was 9.9 per 100 patient-years and 9.6 per 100 patient-years in the ferric carboxymaltose and placebo groups, respectively. Treatment appeared to be safe and well tolerated.

Critical Appraisal

The meta-analysis appears to be conducted appropriately. All the included trials were placebo-controlled, double-blind, randomized phase III or IV trials in adults with HF and ID. CONFIRM-HF was focused on clinical efficacy outcomes such as exercise capacity and NYHA class, although the study did report survival and HF- or CV-related outcomes as well. AFFIRM-AHF and HEART-FID were focused primarily on composite outcomes related to hospitalizations and death. All 3 trials were at least 12 months in duration. The studies differed in minor ways in terms of inclusion criteria, such as whether NYHA class I or IV was included; in CONFIRM-HF, only class II and III were included, while HEART-FID additionally included class IV, and AFFIRM-AHF did not specify any exclusions based on NYHA class. This was not expected to represent an important difference in patient populations, in part because prescription of IV iron is not dependent on NYHA class and in part because very few patients outside of class II and III were included in any study overall. There were also minor differences regarding the upper limit of LVEF and the range of included hemoglobin levels, but altogether these were considered very similar. The definitions of ID were the same across the studies.

A key difference in study design is that AFFIRM-AHF required patients to be hospitalized for acute HF during enrolment, whereas CONFIRM-HF and HEART-FID required a hospitalization within the prior year (or, in the case of HEART-FID, hospitalization within the prior 12 months or elevated N-terminal pro-brain natriuretic peptide within 90 days of randomization). Although these differences are important to consider when interpreting the results, they were not expected to pose a concern in conducting a meta-analysis such as the one presented in Ponikowski et al. (2024), and all the studies do represent the patient population in question. The inclusion of AFFIRM-AHF, which involves hospitalized patients with acute HF, introduces a significant between-trial difference in design and patient population. However, HEART-FID is a substantially larger trial, so any skewing of results due to the more at-risk population in AFFIRM-AHF may not be substantial. A sensitivity analysis excluding AFFIRM-AHF could have been useful to address this difference in patient population. Despite this minor concern, the treatment effect in AFFIRM-AHF was not consistently the highest in magnitude. In fact, CONFIRM-HF generally showed the largest treatment effect but also had the widest 95% CIs across all coprimary composite outcomes and their components. Further exploration into this observation may be warranted, but it does not discredit the meta-analysis results. In the publication, between-trial heterogeneity in treatment effect was explored and models were adjusted for baseline hemoglobin and region, which was considered appropriate. The study authors state that there was no identified heterogeneity between the trials for any primary or key secondary outcomes and that the treatment arms appeared to be balanced by demographics.

Economic Evidence

Cost and Cost-Effectiveness

Table 5: Summary of the Economic Evaluation

CDA-AMC = Canada’s Drug Agency; CKD = chronic kidney disease; HF = heart failure; IBD = inflammatory bowel disease; ID = iron deficiency; IDA = iron deficiency anemia; NYHA = New York Heart Association.

Budget Impact

CDA-AMC identified several key limitations with the sponsor’s analysis. The sponsor’s use of a claims-based approach to estimate market size and market shares introduces uncertainty in the anticipated budget impact of ferric carboxymaltose. The market capture of ferric carboxymaltose was also uncertain. The sponsor’s approach to including dispensing and mark-up fees was inappropriate, and the submitted model was not user-friendly. Due to the limitations with the sponsor’s claims-based analysis that could not be adequately validated or addressed, CDA-AMC did not conduct base-case reanalyses. It should be noted that the sponsor’s estimated incremental budget impact of $404,491 over 3 years is highly uncertain.

Request for Reconsideration

The sponsor filed a request for reconsideration of the draft recommendation for ferric carboxymaltose for the treatment of ID in adult patients with HF and NYHA class II or III to improve exercise capacity. In their request, the sponsor identified the following issues:

• The sponsor requested that CDEC consider additional context when interpreting the HEART-FID results.

• The sponsor was of the view that there is evidence to indicate that ferric carboxymaltose would reduce hospitalization in patients with ID and HF.

• The sponsor indicated that there is an unmet need in patients with HF.

In the meeting to discuss the sponsor’s request for reconsideration, CDEC considered the following information:

• Information from the initial submission related to the issues identified by the sponsor

• New information provided by the sponsor (meta-analysis by Ponikowski et al.) to address an important, clear gap in the evidence identified by CDEC

• Feedback from 1 clinical specialist with expertise in the diagnosis and treatment of patients with HF

• Feedback on the draft recommendation from 1 clinician group formed of 16 clinical specialists with diverse expertise and other input from 1 cardiologist

• Feedback on the draft recommendation from the public drug plans that participate in the Reimbursement Review process

• Feedback on the draft recommendation from the sponsor.

All feedback received in response to the draft recommendation is available on the CDA-AMC website.

CDEC Information

Members of the Committee

Dr. Peter Jamieson (Chair), Dr. Sally Bean, Daryl Bell, Dan Dunsky, Dr. Trudy Huyghebaert, Morris Joseph, Dr. Dennis Ko, Dr. Christine Leong, Dr. Kerry Mansell, Dr. Alicia McCallum, Dr. Srinivas Murthy, Dr. Nicholas Myers, Dr. Krishnan Ramanathan, Dr. Marco Solmi, Dr. Edward Xie, and Dr. Peter Zed.

Initial meeting date: October 23, 2024

Regrets: Four expert committee members did not attend.

Conflicts of interest: None

Reconsideration meeting date: February 27, 2025

Regrets: One expert committee member did not attend.

Conflicts of interest: None