Budget impact analysis of extended half-life recombinant factor IX (rFIXFc) in the treatment of haemophilia B

Introduction

Haemophilia B is a rare, potentially disabling congenital haemorrhagic disease caused by deficiency of coagulation factor IX (FIX). The condition mainly affects male subjects and its natural history in the absence of adequate treatment is marked by muscle and joint haemorrhages that lead to disabling arthropathy characterized by functional limitations and chronic pain (1). Severity of bleeding symptoms is related to residual plasma levels of FIX, and development of arthropathy is more frequent in severe forms of the disease (2).

Haemophilia is classified as mild when the biological activity of FIX is between 5% and 40%, with symptoms like bleeding secondary to minor trauma, surgery, or invasive procedures (e.g. dental extraction), although spontaneous bleeding is also possible. When biological activity of FIX is between 1% and 5%, this is considered to be moderate haemophilia with pathological bleeding secondary to minor trauma, surgery, or invasive procedures, while spontaneous bleeding is rare (2).

With biological activity of FIX <1%, haemophilia is considered severe, being characterized by frequent spontaneous bleeding as well as joint bleeding secondary to minor trauma, surgery, or invasive procedures (2). Haemophilia B has an incidence of 3 cases per 100,000 live male births. According to the Associazione Italiana Centri Emofilia (AICE) national registry data, of all patients with a known diagnosis of haemophilia B in Italy in 2017, 43.9% were affected by mild form, 21.1% by moderate form, and 35.0% by severe form (3).

The treatment of haemophilia is based on replacement therapy with administration of FIX concentrates according to prophylactic or on-demand regimens. According to the latest estimates, in Italy, prophylaxis is the most adopted therapeutic regimen in patients with severe haemophilia B (85.1%) and is prescribed in about 40% of patients with moderate haemophilia (3).

The primary goal of prophylaxis is prevention of joint damage. Prophylaxis entails regular intravenous infusions of FIX concentrates in order to prevent the onset of spontaneous bleeding (especially in the joints) by maintaining a minimum level of FIX in plasma (trough level) in the range of moderate to mild haemophilia (1). With prophylaxis, it is possible to prevent the onset of joint damage or at least to slow its progression and to reduce the number of bleeding episodes, which gives patients the opportunity to lead a relatively normal life (1,4,5,6,7,8,9,10).

Given the greater ease of use (e.g. reduced infusion volumes, faster preparation times) and better safety profile (11,12), whenever possible, recombinant products (rFIX) are preferred over plasma-derived factors (pFIX) (3). Indeed, in Italy, rFIX is more commonly used than pFIX (3).

Given the need for repeated intravenous administrations, adherence to prophylaxis can be problematic, which can lead to missed administrations and/or under-dosing (13,14,15,16,17,18). Poor adherence can thus be associated with an increase in bleeding rates due to the reduced protection conferred by intermittent treatment (19,20,21,22). The main barriers to treatment adherence include the time needed for infusion, planning of the infusion, and problems in venous access (19,22).

In recent years, new rFIX molecules were registered and reimbursed, with improved pharmacokinetic profile in order to ensure good haemostatic coverage with a lower number of infusions (3,23,24,25,26). On average, half-life of these new molecules is 4-5 times the normal physiological value, allowing prophylactic infusions every 7, 10, or ≥14 days. This is in contrast to prophylaxis with standard products, requiring 2-3 infusions per week (27,28,29,30).

Thanks to these characteristics, the extended half-life of rFIXs offers greater therapeutic flexibility, with better adherence, greater access to prophylaxis, better long-term outcomes, and potential cost savings (10,17).

Considering the economic impact of these drugs on the National Health Service (NHS), the most recent data published by the National Observatory on the use of Medicines (OsMed) reported that in Italy, in 2018, coagulation factors were among the drug categories with the greatest impact on healthcare expenditures (31).

Among these new molecules, rFIX covalently fused to the Fc domain of human immunoglobulin G1 (rFIXFc) has been available in Italy for the treatment of patients with haemophilia B since 2017 (3,32). Prophylaxis with rFIXFc has been shown to be safe and effective for the prevention and treatment of bleeding events in both paediatric and adult haemophilia B populations (23,25). Fusion with the Fc fragment of human IgG1 leaves the functional capacity of FIX intact, with no observed increase in immunogenicity. Pharmacokinetic and efficacy data support the potential for treatment at prolonged intervals, while providing excellent bleeding control vs. shorter half-life rFIX molecules in patients of all ages (23,25).

The aim of this study was to estimate the economic impact of the use of prophylaxis with rFIXFc compared to other rFIX that are available in Italy for the treatment of moderate to severe haemophilia B.

Methods

The analysis was carried out from the perspective of the NHS considering a 3-year period, including only direct healthcare costs related to use of rFIX for prophylaxis and resolution of bleeding episodes.

The number of eligible patients was estimated from Italian population data and prevalence of the disease (3,33). The haemophilia B population was then stratified by age (<6, 6-11, 12-17, and ≥18 years).

In each age group, the mean per patient body weight was considered: (a) 20 kg in subjects <6 years; (b) 30 kg for ages 6-11 years; (c) 55 kg for ages 12-17 years; (d) 70 kg in adults.

Percentage of haemophilia B adults treated with rFIX prophylaxis was based on consumption data from the AICE national registry (3).

Calculation of percentage of paediatric patients treated with rFIX prophylaxis was based on the following assumptions:

a)In patients <6 years and patients aged 12-18 years, prophylaxis is used in 50% of patients with moderate haemophilia and in 80% of patients with severe haemophilia.

b)In the 6-11 year age group, prophylaxis is used in 70% of patients with moderate haemophilia and in 85% of patients with severe haemophilia.

Two extended half-life rFIX concentrates were included in the model: (1) rFIXFc (eftrenonacog alfa; Alprolix®) and (2) rIX-FP, a recombinant fusion protein containing rFIX fused with recombinant albumin (rIX-FP – albutrepenonacog alfa; Idelvion®); two standard half-life rFIX concentrates were included: (1) nonacog gamma (Rixubis®) and (2) nonacog alfa (Benefix®). Market shares for each rFIX were estimated based on 2018 market data and projected for the next 3 years.

Doses and frequencies of administration by age (<12 years vs. ≥12 years) in prophylaxis were obtained from the respective Summary of Product Characteristics (SPC) of each rFIX concentrate included in the model (27,28,29,30). For extended half-life rFIX, weekly prophylactic administrations were considered in the base case scenario, that is, one administration per week (27,28).

Bleeding rates and the amount of rFIX used for each product were based on literature data (12,34,35). According to clinical practice and guidelines, bleeding episodes are treated with an additional dose of FIX products (1). rFIXFc and rIX-FP bleeding rates and the amount of rFIX used to manage bleedings were retrieved from the pivotal clinical trials (23,24,25,26).

Additional data were estimated as follows:

a)rIX-FP, albutrepenonacog alfa: in order to estimate the amount of drug used to treat a bleeding episode in adults and children, the mean number of administrations derived from the clinical study was multiplied by the dose indicated for prophylaxis (24,26).

b)Nonacog gamma: to estimate the amount of drug used in children <12 years of age, bleeding rates were taken from the EPAR (36), and rFIX dose was considered to be the same as administered in adults (12).

c)Nonacog alfa: mean dose used for prophylaxis was applied to management of bleeding episodes, multiplied by the mean number of infusions needed to treat a bleeding episode (34).

Unit cost for each rFIX was calculated as average price from published regional tenders in seven Italian regions (Piedmont, Veneto, Tuscany, Lazio, Campania, Puglia and Sicily) (37).

Two scenarios are compared in the model: a reference scenario (Reference) and an alternative scenario (Alternative) in which increased rFIXFc market shares replace the other rFIX considered.

The base case analysis considered rFIXFc and rIX-FP prophylaxis to be administered once every 7 days. Sensitivity analysis was also carried out in adult patients, considering administrations once every 10 or 14 days as reported in the respective SPC (27,28). Specifically, administration every 14 days was considered for rIX-FP, and every 10 or 14 days for rFIXFc. This also allowed to identify, given a sample of 100 patients for each arm, in what percentage of patients receiving rFIXFc every 10 days the two treatments reached parity price.

Results

Discussion

Phase III trials for rFIXFc showed that prophylaxis with rFIXFc is safe and effective and associated with significant reductions in bleeding episodes, which represents the desired target in the management of haemophilia B (23,25).

Treatment with extended half-life rFIX has the advantage of fewer administrations, an aspect that is very important for both patients and caregivers (39). To date, only a limited number of studies have evaluated the pharmacoeconomic impact of recombinant therapies in patients with haemophilia B in Italy; only one has assessed the impact of new extended half-life rFIX, with different conclusion compared to our results (34). Pradelli et al. estimated that rFIXFc had higher costs than rIX-FP, since the doses and treatment regimens used were different from those considered in the present analysis (34). In that study, they estimated rIX-FP prophylaxis at a dose of 50 IU/kg, which represents the minimum dose of rIX-FP for once every 14 days administration. This dose was derived from a pharmacokinetic model published by Zhang et al. in 2016 (40); indeed, data were not taken from a clinical trial with efficacy based on patient outcomes (40). In the main Phase III trial in adults (26), the dose investigated was 75 IU/kg administered every 10 or 14 days; if this latter dose had been used, it would likely have given different results, that is, an increase in costs due to an increase in the costs of prophylaxis with rIX-FP 30% higher or more (considering a dose of 50 IU/kg vs. 75 IU/kg) (34).

In the present analysis we used a conservative approach, considering the weekly dosages from the SPC in the base case scenario and 14-day administrations from the SPC in the sensitivity analysis, in which the number of patients receiving rFIXFc and undergoing infusions every 10 or 14 days was assessed, evaluating the impact on mean costs (27,28). As demonstrated in a previous study, over 90% of the costs of haemophilia are attributable to drug therapy (98% of direct healthcare costs and 92% of total costs); accordingly, substantial changes in doses and drug prices would impact the results (41).

In light of the difficulty in carrying out direct comparative studies in a rare disease, to reach the higher numbers needed for statistically and clinically significant evidence, and considering the potential consumption of the two extended half-life drugs, we believe that our approach provides conservative estimates and solid results; the sensitivity analysis, based on an unfavourable hypothesis for rFIXFc, is particularly robust and fully supports the economic benefits of rFIXFc.

The major limitations of our analysis are the lack of direct comparison between therapies, although we used the doses from the SPC and bleeding rates from Phase III trials, which even if dissimilar due to the different populations studied still contribute to the robustness of the economic model (27,28,29,30). A detailed review of the Phase III studies of rFIXFc and rIX-FP clearly demonstrates several differences (23,24,25,26):

a)Patient numbers differ significantly in the rIX-FP and rFIXFc Phase III trials.

b)Different pre-study treatment regimens and baseline study populations in both trials may have influenced medical status.

c)The PROLONG-9FP study design included selection of patients.

d)Pharmacokinetic evaluations have been performed with different methodologies, different pre-study products, sampling time points, and at different timelines during the trials.

Real-world evidence data presented at the American Society of Hematology (ASH) in December 2019 by Malec et al. found important differences in bleeding control for rIX-FP vs. rFIXFc in patients with severe haemophilia B (90 patients including 67 treated with extended half-life products). In 26 patients receiving rIX-FP, 16 (62%) had unexpected or poorly controlled bleeding events, while in 37 patients receiving rFIXFc there were no such events (42). In light of this evidence, our assessment appears conservative and possibly underestimates the potential savings that could be obtained with rFIXFc.

Another possible limitation of this study is that the costs of hospitalization, instrumental/laboratory tests, and management of intracranial bleeding were not included; moreover, the costs directly sustained by patients and caregivers (especially relevant in paediatric patients) and work days lost were not taken into account. Since the study by Kodra et al. reported that 92% of total costs and 98% of all healthcare costs are attributable to drugs, we believe our analysis reflects the true economic impact of greater utilization of rFIXFc over other rFIX (41). In addition, it should be emphasized that the model does not take into account advantages that could be possibly obtained with extended half-life drugs compared to standard half-life drugs in terms of adherence (and therefore efficacy) and quality of life. In fact, considering fewer administrations, extended half-life factors, especially in some patient groups, could provide greater adherence and thus greater efficacy in real-life management; therapies with poor adherence may be less effective, leading to unnecessary and unjustified costs for the NHS and the society as a whole.

Lastly, the sensitivity analysis validated the results by using scenarios that are particularly unfavourable to rFIXFc by assuming infusion every 10 days, while for rIX-FP administration every 14 days was deliberately considered. Administration of rIX-FP every 10 days would produce a cost of prophylaxis in adults of €384,615 per patient per year, well above the cost of the every 14 days option considered in the model. The 10 days administration interval considered for rFIXFc is certainly the most unfavourable for the product: in fact, in the pivotal study, 53.8% of patients, in the group with an interval-adjusted prophylaxis, had reached an interval of 14 days or more in the last 3 months of therapy (25). This further confirms the validity of the present economic analysis: despite the unfavourable hypotheses for rFIXFc, there is still economic benefit for the NHS compared to the other factors with extended half-life.

Conclusions

In conclusion, an increase in the use of rFIXFc, replacing other rFIX concentrates for treatment of haemophilia B, can represent an economically advantageous choice for the NHS with overall savings, at the national level, over the 3-year period of €1,988,068, while ensuring effective control of bleedings. Sensitivity analyses confirmed the robustness of the results.

Disclosures

Conflict of Interest: AA and BP are employees of Certara Italy S.r.l.; MME reports personal fees from Swedish Orphan Biovitrum (Sobi), Bayer Healthcare, CSL Behring, Grifols, Kedrion, Novo Nordisk, Octapharma, Pfizer, Roche; CA and TC are employees of Sobi S.r.l., Milan, Italy.

Financial support: Sobi S.r.l., Milan, Italy, has provided financial support to cover the cost of this project and the editorial assistance, but has not influenced the manuscript content.