Cost-effectiveness and budget impact analysis of infliximab and its biosimilar in patients with refractory moderate-to-severe Crohn’s disease using real world evidence in Thailand

Abstract

Aims

This study aimed to conduct a cost-effectiveness analysis of infliximab and its biosimilar compared to conventional therapy in refractory moderate-to-severe Crohn’s disease (CD) in Thailand.

Materials and methods

A Markov model was used to estimate lifetime costs and health benefits of infliximab from a societal perspective. Our analyses consisted of three choices of treatment (conventional therapy, infliximab originator, and biosimilar) and three treatment scenarios (infliximab 2 years and 3 years if relapse, infliximab 2 years and lifelong if relapse, and infliximab lifelong). The input parameters were obtained from the CD registry and systematic literature reviews. The results were reported as incremental cost-effectiveness ratios (ICERs) in 2017 USD per quality-adjusted life year (QALY) gained. The sensitivity analyses were performed to assess the influence of parameter uncertainty. Threshold sensitivity analyses were carried out to determine the optimal drug prices. Finally, budget impact analyses were conducted.

Results

None of the scenarios was cost-effective at Thai willingness-to-pay threshold (4,706 USD/QALY gained). The lowest ICER of 30,121 USD/QALY gained was reported in the scenario that included only standard dose of infliximab biosimilar with the maximum of 5-year treatment. The drug prices need to be reduced by at least 72% to allow infliximab biosimilar to be cost-effective. The 5-year budget impact was only 695,958 USD for the current biosimilar price.

Conclusions

Infliximab for the treatment of refractory moderate-to-severe CD in Thailand would be cost-effective if the drug prices were significantly decreased. The best value for money strategy was infliximab biosimilar with a restricted duration of treatment.

Key points

• The use of infliximab and its biosimilar in a restricted duration of maximum 5-year is not cost-effective for patients with moderate-to-severe Crohn’s disease refractory to conventional therapy, unless their price was lowered around 72–90% in Thailand.

• The estimated budget impact for adopting infliximab or its biosimilar for such indication has potential financial feasibility. Policy makers may consider cost-effectiveness and budget impact findings as well as other aspects such as rarity of disease as a part of the decision making process.

Keywords:

• Cost-effectiveness analysis
• Crohn's disease
• infliximab
• biosimilar
• Thailand

JEL CLASSIFICATION CODES::

• I15
• H51

Introduction

Crohn’s disease (CD), especially refractory CD, is a rare chronic inflammatory bowel disease that has a significant long-term effect on health-related quality of life (HRQoL)¹ The prevalence and incidence of CD were quite low among Asian countries, but recently, it has been continuously increasing. The incidence of CD in Thailand was 0.3 per 100,000 population and recently increased to 0.34 per 100,000 population, which is still considered very low when compared with what is reported in western countries^2–5. CD is also clearly known to be associated with increased risk of mortality compared to general population⁶. There is a standard conventional therapy (CT) for CD patients but for those who suffer with conventional-drug refractory moderate-to-severe CD require treatment such biological treatment in order to control the disease^7–10. Anti-tumor necrosis factor (anti-TNF) agents, such as infliximab, have ability to increase to the rate of disease remission and decrease the rate of surgery^7,8,11. In 2016, the European Crohn’s and Colitis Organization (ECCO) concluded that anti-TNF based treatment was appropriate for treating patients whose disease still relapse after the optimal dose of CT was reached¹¹. Currently, in Thailand, infliximab is only accessible by those who can pay by themselves. Therefore, those who cannot afford biological products have to maintain the disease using CT even in the refractory state.

Infliximab biosimilar has been another viable option for biological products given availability of a proof of equivalence or non-inferiority evidence. In 2016, the first biosimilar product of infliximab (CPT-13, Remsimaⁱ) was launched in Thailand. Recent evidence of effectiveness and safety of this biosimilar has demonstrated the equivalence to the originator product¹². The National List of Essential Medicine (NLEM) Selection Committee requires cost-effectiveness and budget impact analyses evidence as part of the Health Technology Assessment (HTA) process before determining whether these biological products should be adopted as treatment of refractory moderate-to-severe CD patients. Despite a number of previously published cost-effectiveness studies of infliximab, no study was conducted in low- and middle-income countries (LMICs) or even considered biosimilar as their therapeutic option^13–16. In the era of biosimilar availability equipped with strong clinical effectiveness evidence, this study aims to assess cost-effectiveness and budget impact analyses of infliximab and its biosimilar compared to CT in refractory moderate-to-severe CD in Thailand.

Methods

Overall description

A cost-effectiveness analysis was performed to determine lifetime health benefits and cost from using infliximab in adult patients with conventional-drug refractory moderate-to-severe CD. The scope of study including framework and options was defined after considering comments from all parties involved in a stakeholder meeting (Supplementary Appendix 1) which consists of representatives of patients, payers, clinicians, medical associations, and national list of essential medicine subcommittee. A decision tree and Markov model were adopted to capture total cost and health outcomes through patient’s lifespan using societal perspective as recommended in Thai HTA guidelines¹⁷. Our results were presented as incremental cost-effectiveness ratios (ICERs) in 2017 US dollars (USD) per quality-adjusted life year (QALY) gained. The willingness-to-pay (WTP) threshold of 4,706 USD/QALY (160,000 Thai Bath (THB)/QALY) was used to determine whether the option was cost-effective¹⁸. The discount rate of 3% was applied to all costs and outcomes. This article was reported in accordance with the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) Statement. We also estimated the financial burden with a budget impact analysis using a payer perspective. This study was approved by ethical committee of Faculty of Medicine Siriraj Hospital, Mahidol University.

Interventions

We compared 5 therapeutic options including (1) Infliximab originator allowing high dose if non-responsive to standard dose, (2) Infliximab biosimilar allowing high dose if non-responsive to standard dose, (3) Infliximab originator standard dose, not allowing high dose, (4) Infliximab biosimilar standard dose, not allowing high dose, and (5) Usual care or CT. The CT in our model included immunomodulators and corticosteroids titrated to the maximum doses that are safe and tolerable. Infliximab was the additional treatment to the CT in all options except usual care. We did not incorporate other biological products in the analyses as they were not in the scope of HTA consideration defined by NLEM given the lack of local data. Standard dose and high dose of infliximab referred to infliximab 5 and 10 mg/kg/dose, respectively. Patients were not allowed to switch between originator and biosimilar products. Our base-case analysis was based on the use of infliximab for 2 years then stop and then reinitiate for another 3 years only among those with relapse (IFX 2 years and 3 years if relapse). The other 2 scenarios include (1) the use of infliximab for 2 years then stop and reinitiate for a lifelong in those who relapse (IFX 2 years and lifelong if relapse) and (2) lifelong treatment with infliximab (IFX lifelong). IFX biosimilar (CPT-13, Remsima) was chosen in our analysis because it was the only available biosimilar product in Thailand at the time of the conduct of this study. These options were chosen taking into account perspective of clinicians and policy makers concerning clinical practicality and feasibility for policy implementation in a stakeholder meeting. The evidence of the potential benefits of IFX which may last longer than 2 years after the use of IFX for 2 years has been brought up in discussion. It was clear that after treatment for 2 years, more than half of the patients did not relapse within 2 years after discontinuation^19,20. The relapse rates were only 30% and 40% at 1- and 2-year after the first-time withdrawal of infliximab, respectively¹⁹. Thus, base-case strategy (IFX 2 years and 3 years if relapse) is the most appropriate in clinical practice for preventing overuse of infliximab^19,20. However, the efficacy of infliximab after reinitiation was reduced to 92% of the first-time using¹⁹ and there was no evidence about the long-term effect of infliximab. After infliximab second discontinuation, its effect would be gradually reduced until no effect left within 28 months²¹.

Economic model

A decision tree coupled with Markov model was used as a tool to estimate lifetime cost and health outcomes. In the decision tree, patients were cascaded to different pathways based on the treatment options during the induction phase. The patients who responded to standard dose infliximab without any serious adverse event (SAE) would be treated with infliximab, besides, for those who did not respond or encounter with SAE would discontinue infliximab. In addition, the patients who did not respond to standard dose infliximab would receive the treatment by CT, surgery (or high dose infliximab; only allowing the high dose of infliximab arm) as shown in Supplementary Figure S3. A Markov model was simulated to capture the process after we could determine the eligible treatment for patients from the decision tree until the end of their life. It consisted of 5 health-states that represented the natural course of CD grouped by the severity of disease that classified by Crohn’s disease activity index (CDAI) i.e. remission CD (CDAI <150), mild CD (CDAI 150-220), moderate-to-severe CD (CDAI 220-600), surgery, and death. This Markov model was adapted from the previous study by Bodger et al.²². The arrows between health states indicated the ability of transition of health status in each cycle of the model (Figure 1). The 8-weekly cycle length was used according to the frequency of medical assessment and the dosage regimen of infliximab.

Figure 1. Markov model. Abbreviation. CDAI, Crohn’s disease activity index.

Input parameters

Overall, the efficacy and safety data were elicited from RCT, meta-analysis, and recent equivalence study^7,12,19, while other input parameters including transitional probabilities, costs, and, utilities were obtained from real world evidence such as hospital electronic medical records. We assumed patient weight of 60 kg and the age of 41 years based on the data from the study of Thai CD patients by Rerknimitr et al.²³ and the Thai CD registry data. All input parameters were listed in Table 1.

Table 1. Input parameters.

Input parameters	Base-case	Range	Source
Induction phase
Response to CT	0.200	0.180–0.220	[7,24]
Response to standard dose infliximab	0.778	0.700–0.856	[7,24]
Infliximab originator discontinuation due to AEs	0.031	0.018–0.056	[12]
Infliximab biosimilar discontinuation due to AEs	0.036	0.024–0.056	[12]
Surgery of infliximab and 2^nd CT non-responder	0.567	0.518–0.633	[25]
Efficacy of infliximab vs. CT
Relative risk of induction to remission	3.70	0.87–15.80	[7]
Relative risk of induction to response	4.01	1.29–12.44	[7]
Efficacy of high dose infliximab
Moderate-to-severe → remission	0.279	0.251–0.307	[26]
Moderate-to-severe → mild	0.344	0.310–0.379	[26]
Moderate-to-severe → moderate-to-severe	0.354
Moderate-to-severe → surgery	0.023	0.021–0.026	[26]
Efficacy of CT
Moderate-to-severe → remission	0.080	0.072–0.088	[7,24]
Moderate-to-severe → mild	0.120	0.108–0.132	[7,24]
Moderate-to-severe → moderate-to-severe	0.800
Moderate-to-severe → surgery	0.000		Assumption
Efficacy of 2^nd CT after failure of infliximab or 1^st CT
Moderate-to-severe → remission	0.000		[25]
Moderate-to-severe → mild	0.061	0.055–0.067	[25]
Moderate-to-severe → moderate-to-severe	0.364	0.327–0.400	[25]
Moderate-to-severe → surgery	0.576	0.518–0.633	[25]
Efficacy of surgery
Surgery → remission	0.007	0.006–0.007	[27]
Surgery → mild	0.915
Surgery → moderate-to-severe	0.009	0.008–0.010	[8]
Surgery → surgery	0.070	0.063–0.076	Database
Maintenance phase
Efficacy of infliximab vs. CT
Relative risk of maintenance of remission	1.86	1.21–2.86	[7]
Relative risk of maintenance of response	1.86	1.21–2.86	Assumed from[7]
Odds ratio of surgery reduction	0.26	0.10–0.61	[28]
Transitional probability of disease progression
Remission → mild	0.121	0.108–0.133	[27]
Mild → moderate-to-severe	0.400	0.360–0.440	[27]
Moderate-to-severe → surgery	0.008	0.007–0.009	Database
Surgery → surgery	0.064	0.063–0.076	Database
Transitional probability of disease responds to CT
Surgery → moderate-to-severe	0.008	0.002–0.014	[8]
Surgery → remission	0.007	0.063–0.076	[27]
Moderate-to-severe → mild	0.020	0.018–0.022	[27]
Moderate-to-severe → remission	0.054	0.044–0.065	[7,29]
Mild → remission	0.000		Assumption
Relapse phase
Relapse rate after 1 year of infliximab 1^st discontinuation	0.30	0.27–0.33	[19,21]
Relapse rate after 2 years of infliximab 1^st discontinuation	0.40	0.33–0.48	[19]
Relapse rate after infliximab 2^nd discontinuation	0.79	0.73–0.85	[21]
Efficacy of infliximab retreatment vs. 1^st treatment
Relative risk of efficacy of infliximab retreatment	0.92	0.70–0.95	[19]
Mortality data
Age-specific mortality in Thai population	Supplementary Appendix 3		[30]
Hazard ratio of mortality rate of the patients with CD vs. general population
Remission CD	1.055	0.653–1.673	[6,31]
Mild CD	1.131	1.084–1.159	[6,31]
Moderate-to-severe CD	1.792	1.705–1.853	[6,31]
Utility
Remission CD	0.908	0.737–1.000	Survey
Mild CD	0.658	0.586–0.730	Survey
Moderate-to-severe CD	0.425	0.197–0.653	[1]
Surgery	0.425	0.197–0.653	[1]
Cost data
Direct medical costs
Drug costs (USD/1 cycle)
Infliximab originator	1,650	1,439–1,860	NLEM
Infliximab biosimilar	990	865–1,117	NLEM
Cost of infliximab administration (USD/visit)	29	23–35	Database
OPD and IPD service (USD/cycle)
Remission CD	36	5–365	Database
Mild CD	195	10–787	Database
Moderate-to-severe CD	383	3–1,671	Database
Cost of surgery (USD/episode)	2,617	1,882–3,352	Database
Cost of colonoscopy (USD/cycle)	16	14–19	[32]
Direct non-medical costs
Food and transportation (USD/visit)	7	6–7	[32]

Abbreviations. AE, adverse event; CD, Crohn’s disease; CT, conventional treatment; NLEM, National List of Essential Medicine; USD, United State dollars.

Effectiveness and safety of infliximab, biosimilar, conventional therapy, and surgery

The infliximab treatment regimen consisted of 2 phases; induction (at week 0, 2, and 6) and maintenance phases (every 8 weeks). The response of standard and high dose infliximab in both phases shown in Table 1 was based on the meta-analyses^7,8, RCTs^24,29, retrospective US and Denmark studies^25,26, and previous economic evaluation study²⁷. The effectiveness and safety of biosimilar were based on the equivalence study by Meyer et al.¹². However, this study reported different infliximab discontinuation rates due to SAE, thus, we used these data to define the proportion of patients switching from biological therapy to CT during the induction phase. High dose infliximab would provide the benefit over standard dose infliximab in terms of higher response in the induction phase. In allowing the high dose of infliximab arm, the patients who failed from standard dose infliximab, they could achieve remission and mild CD in 28% and 34%, respectively²⁶. We assumed no loss of response in case of lifelong infliximab therapy. The relapse rate after the first-time withdrawal of infliximab were 30% and 40% at 1- and 2-year¹⁹. The efficacy of infliximab after reinitiation was reduced to 92% of the first-time using¹⁹. After the second discontinuation of infliximab, its effect would be gradually reduced until no effect left within 28 months²¹. Regarding the effects of CT and surgery, we adopted the response rate of CT in refractory CD patients from meta-analyses, RCTs, and previous economic evaluation studies^7,8,24,27,29. The effectiveness of surgery was calculated based on a previous economic evaluation study²⁷, meta-analyses^8,28, and Thai CD registry.

Probability data

Age-specific mortality rate (ASMR) of Thai population³⁰ was used for calculating the mortality rate of the patients with CD (Supplementary Appendix 3). Hazard ratios of mortality of the patients with CD in each severity were calculated from previous studies^6,31. The calculation details were shown in Supplementary Appendix 2.

Cost data

Direct medical costs such as costs of standard care, adverse event care, drug administration, laboratory analyses, surgery, and hospital service were obtained from the university hospital electronic database, which included 110 patients with CD who visited the hospital between January 2012 and December 2018 (Supplementary Appendix 5). Colonoscopy for follow-up cost and direct non-medical cost (food and transportation) per visit were obtained from the standard unit cost list from published literature in Thailand³². We used infliximab costs from the prices submitted to the NLEM 2018 committee. The cost of infliximab originator and biosimilar were 550 and 330 USD per vial, respectively. Based on patient weight of 60 kg and infliximab dose of 5 mg/kg (100 mg/vial), the cost of infliximab originator and biosimilar per administration would be 1,650 USD and 990 USD, respectively. Since the societal perspective was applied in this analysis, indirect costs were not included because we assumed that lost or impaired ability to work or engage in leisure activities due to morbidity would be captured in the disutility of QALY as per recommendation of HTA guideline³³. All cost data were converted and reported in 2017 USD (1 USD = 34 THB) by using the consumer price index^34,35.

Utility data

We collected HRQoL using the Thai-validated EQ-5D-5L questionnaire from 75 patients enrolled in Thai CD registry and calculated utility based on Thai algorithm³⁶. Classified by CDAI, most of them (70/75) were in remission state, while the remaining 5 patients were in mild CD state. We estimated the utility of patients in moderate-to-severe CD state based on the equation established by Buxton et al.¹ and assumed that patients in surgery state had the same utility as patients in moderate-to-severe CD state. The details were shown in Supplementary Appendix 4.

Cost-effectiveness analysis

Base-case and scenario analyses

We assigned a scenario of IFX 2 years and 3 years if relapse which patient would be treated with infliximab for 2 years then stop and retreated for another 3 years if relapse as the base-case analysis. Scenario analyses include (1) treatment with infliximab for 2 years then stop and retreated for a lifelong if relapse (IFX 2 years and lifelong if relapse) and (2) treatment with infliximab for a lifelong (IFX lifelong). The costs and outcomes of the base-case and two scenario analyses were compared with CT and reported as incremental costs, incremental QALY, and ICERs (USD/QALY).

Sensitivity analyses

One-way sensitivity analyses were performed by varying the values of input parameters within ranges indicated in Table 1. We also varied the discount rate within the range of 0–6% per annum. The results of one-way sensitivity analyses were presented in Tornado diagrams. A threshold analysis was used to determine the cost-effective price of options. In addition, multivariate probabilistic sensitivity analyses (PSA) were conducted with 1,000 Monte Carlo simulations by using Microsoft Excel 2019 (Microsoft Corp, Redmond, WA)³⁷ to explore the uncertainties of parameters. The transitional probabilities and utilities were assigned beta distribution. All cost data were assigned gamma distribution. Relative risks and hazard ratios were assigned log-normal distribution³⁸. The results of PSA were presented as cost-effectiveness acceptability curves (CEAC) to show the probability of treatment with infliximab and its biosimilar being cost-effective.

Budget impact analyses

The 5-year and 10-year budget impact analyses were conducted to estimate the budget that government needed to bear if infliximab is included in NLEM. The prevalence of CD in Thailand from Thai CD registry was used to estimate the total number of patients that have indication for infliximab. We used both infliximab and biosimilar costs based on prices submitted to NLEM and the reduced prices from threshold analyses to calculate budget impact. We assumed the access rate up to 95% within 10 years with 3% disease growth rate.

Results

Base-case analysis

The total costs of each option in base-case analysis (IFX 2 years and 3 years if relapse) were 67,377, 54,344, 63,726, 52,226, and 37,606 USD for regimen (1) Infliximab originator allowing high dose if non-responsive to standard dose, (2) Infliximab biosimilar allowing high dose if non-responsive to standard dose, (3) Infliximab originator standard dose, not allowing high dose, (4) Infliximab biosimilar standard dose, not allowing high dose, and (5) CT, respectively. The incremental cost of those 4 treatment strategies compared with CT was 29,771, 16,738, 26,120, and 14,620 USD. Addition of IFX treatment gained 0.50 QALY in infliximab originator and its biosimilar allowing high dose strategies and 0.49 QALY in infliximab originator and its biosimilar standard dose strategies.

None of the infliximab options was cost-effective when compared to CT as the ICERs were above 4,706 USD/QALY. Regimen of infliximab biosimilar standard dose, not allowing high dose provided the ICER of 30,121 USD/QALY gained. The ICER of 33,651 USD/QALY gained, was shown in the regimen of infliximab biosimilar allowing high dose if non-responsive to standard dose. The ICERs of infliximab originator treatment were much higher than biosimilar at 59,786 and 53,751 USD/QALY gained for the regimens of allowing and not allowing high dose, respectively. The results of base-case analysis are shown in Table 2.

Table 2. Result of base-case analysis.

Regimen	Cost (USD)	Effectiveness (QALYs)	Incremental cost (ΔUSD)	Incremental effectiveness (ΔQALYs)	ICER ΔUSD/ ΔQALYs)
CT	37,605.88	11.73
IFX originator allowing high dose	67,376.54	12.23	29,770.66	0.50	59,786.06
IFX biosimilar allowing high dose	54,343.52	12.23	16,737.64	0.50	33,650.64
IFX originator	63,725.86	12.21	26,119.98	0.49	53,751.46
IFX biosimilar	52,226.06	12.21	14,620.18	0.49	30,121.03

Abbreviations. CT, conventional treatment; IFX, infliximab; QALY, quality-adjusted life year; USD, United State dollars.

Scenario analyses

The similar results were found in both scenario analyses (IFX 2 years and lifelong if relapse and IFX lifelong). The ICERs of all regimens compared to CT for both scenarios were shown in Supplementary Table S10.

Sensitivity analyses

One-way sensitivity analyses

The results of one-way sensitivity analyses of regimen (3) Infliximab originator standard dose and (4) Infliximab biosimilar standard dose are displayed in Tornado diagrams, Supplementary Figures S4 and S5. Three most influential parameters were utility of moderate-to-severe CD, utility of remission CD, and transitional probability of moderate-to-severe CD to remission. Nevertheless, the sensitivity analyses showed that within the ranges of any parameters, the results were robust and remained higher than Thai WTP threshold.

Threshold analyses

Threshold analysis was performed for base-case analysis (IFX 2 years and 3 years if relapse), the cost of infliximab originator needed to be reduced by 83% or 72% for biosimilar to be cost-effective. For the scenario of IFX 2 years and lifelong if relapse; the cost of infliximab needed to be reduced by 87% for originator and 78% for biosimilar. Lastly, for IFX lifelong scenario; the cost of infliximab needed to be reduced by 94% for originator product price and 91% for biosimilar product price as shown in Supplementary Table S11.

Probabilistic sensitivity analyses (PSA)

In base-case analysis, the probability of infliximab being cost-effective was zero when compared to CT. However, when the WTP threshold was increased, the probability of infliximab being cost-effective was gained (Supplementary Figure S6). When comparison among regimens containing infliximab (CT was not included), regimen (4) Infliximab biosimilar standard dose had the highest probability of being cost-effective of 71.1% at Thai WTP threshold. For regimen (2) Infliximab biosimilar allowing high dose, (3) Infliximab originator standard dose, and (1) Infliximab originator allowing high dose had the probability of being cost-effective of 28.5%, 0.4%, and 0%, respectively (Figure 2).

Figure 2. Cost-effectiveness acceptability curve comparing between four regimens. Abbreviations. IFX, infliximab; WTP, willingness-to-pay; USD, United State dollars.

Budget impact analyses

Based on Thai CD registry data, about 14% of them had an indication for infliximab, with the assumption of yearly increasing access rate 48% to 95% and 3% disease growth rate, the number of treated adult patients with CD in the first year would be 25 cases. In base-case analysis, the 5-year budget impacts were about 0.70 and 0.21 million USD if the full and reduced prices from threshold analyses were used, respectively. The 10-year budget impacts were only 1.22 and 0.37 million USD for the full and reduced prices, respectively (Table 3). In the scenario of IFX 2 years and lifelong if relapse, the 5-year and 10-year budget impact were 0.70, 0.17 million and 1.51, 0.37 million USD if the full and reduced prices were used, respectively (Supplementary Table S12).

Table 3. Result of budget impact of base-case analysis in adult patients with moderate-to-severe Crohn’s disease.

CD, adult	Y1	Y2	Y3	Y4	Y5	Y6	Y7	Y8	Y9	Y10
Number of treatment cases	25	29	33	37	41	45	49	53	58	63
BIA of IFX biosimilar 2 years and 3 years if relapse (USD/year)
a. IFX biosimilar full price (330 USD/vial)	201,610	190,044	58,626	113,379	132,299	139,307	105,172	93,186	95,892	99,394
	Total (USD/5 years)				695,958	Total (USD/10 years)				1,228,909
b. IFX biosimilar reduced price (92 USD/vial)	60,569	57,094	17,613	34,062	39,746	41,852	31,596	27,996	28,809	29,861
	Total (USD/5 years)				209,085	Total (USD/10 years)				369,198

Abbreviations. CD, Crohn’s disease; IFX, infliximab; USD, United State dollars.

Discussion

This is the first cost-effectiveness study that investigates the value for money of infliximab and its biosimilar product in adult CD patients compared to CT in LMICs. Several cost-effectiveness analyses^14,22,39 have been conducted, but none of them evaluated biosimilar product of infliximab based on real world evidence of effectiveness and safety. Our base-case analysis suggests that even though biosimilar provided equivalent effectiveness with the originator with less cost, they are not cost-effective at their current price based on WTP threshold in Thailand. These findings are consistent with those from sensitivity and scenario analyses, reflecting the robustness of results. Importantly, the threshold analyses reveal that the substantial reduction in the costs of infliximab, both originator and biosimilar products, is required to be considered cost-effective in Thai context. In order to make infliximab treatment cost-effective, the prices needed to be reduced at least 83% for infliximab originator and 72% for biosimilar products.

The economic value of infliximab, especially the originator product, has been assessed by several cost-effectiveness studies performed in high-income countries. For example, in the UK¹⁴, where the economic evaluations of pharmacological and surgical interventions were systematically reviewed, the study included a total of 12 individual studies associated with moderate-to-severe CD management by infliximab. It concluded that infliximab originator was not considered cost-effective when compared to CT in developed western countries when the threshold of 50,000 USD purchasing power parity (PPP)/QALY gained was applied. Moreover, it claimed that using infliximab for a shorter period of time or evaluated the model with shorter time horizon might potentially make CD treatment with infliximab become more cost-effective when compared to CT^14,22,39. This was consistent with our finding in the base-case analysis which showed lowers ICERs as it evaluated the shorter period of infliximab treatment, as compared to other scenario analyses. Even if our study showed unfavorable cost-effectiveness results of infliximab, the maximum 10-years budget impact is very low, not more than 1.5 million USD. As CD is a rare disease which is associated with poor HRQoL, the policy makers should consider every aspect, such as the value for money together with ethical issue, whenever they make a decision.

To our knowledge, this is the first cost-effectiveness analysis of infliximab and its biosimilar in patients with refractory moderate-to-severe Crohn’s disease that using efficacy and safety data of biosimilar agent based on the equivalent study published in 2019¹². Unlike previous economic evaluation studies of infliximab biosimilar which used efficacy data from RCTs of originator product and assumed the equivalency^40–42. Our findings are highly valid and contextually relevant due to three main reasons. First, various gastroenterologists were involved throughout our cost-effectiveness analysis. They also suggest the feasible choices such stopping rule in order to balance between clinical benefits and resource consumption. Second, the local data were used as much as possible. Thai CD registry data which included majority of adult CD patients in the country were used in our analysis. The quality of life among Thai patients with CD was directly collected by the EQ-5D-5L questionnaire along with CDAI. Moreover, as the evidence suggested the CD patients have higher mortality rate than general population. We adjusted the mortality rates of these patients by incorporating Thai ASMR with mortality rate from the meta-analyses to reflect Thai context^6,31. We also used the costs of CT of each disease severity from local registry data which included most CD patients in Thailand. The unit cost data from a previous study in Thailand was used as well³². Third, other input parameters used in the model were collected from the most recent systematic reviews, meta-analyses, and randomized controlled trials. In addition, we comprehensively searched literature to identify relevant probabilities and utilities, such that all estimates in the model incorporated the majority of data that were currently available for conventional-drug refractory moderate-to-severe CD. Our results should be robust and had good quality for policy decisions.

There are several limitations in this study. First, this study was based on an assumption of no loss of response when using infliximab in a lifelong period. Nonetheless, the results of lifelong infliximab therapy without consideration of loss of response showed that infliximab was not cost-effective. If the loss of response was applied, the result would be even less cost-effective. However, we did consider the loss of effectiveness of infliximab after reinitiation as well as the loss of effect after the second infliximab withdrawal in base-case analysis²¹. Second, there was no evidence about the difference between the natural course of disease of infliximab naïve patients and patients who had exposed to infliximab. So, we assumed that after 28 months of infliximab second withdrawal, the disease progression rate was equal to the patient who received only CT that had never exposed to infliximab. Third, there are several other biological products, such as adalimumab, vedolizumab, ustekinumab, and others, which had an indication for treating refractory moderate-to-severe CD that we did not include in our analysis, due to NLEM policy on selection of interventions for conducting HTA and the limited local data of these products. We recommend to consider them when the data are well established. Last, it is important to caution the generalizability of our findings to other countries given the differences in healthcare systems, decision-making criteria, and cost data⁴³. However, our study could be considered a model for other LMICs to adopt our methodological approach in their analyses. The finding of being not cost-effective and the magnitude of price reduction may somehow play roles in guiding policy makers in considering infliximab and its biosimilar in their settings.

The actual implication of the use of our research for policy formulation deserved to be mentioned. Our research work was presented to the National List of Essential Medicine Selection Committee in 2019. Despite being not cost-effective at the current price of all options, the threshold analysis suggests potential price that would guide the price negotiation process which might be required in the next step. The magnitude of budget implication seems to be manageable from the financial point of view. The issues of ethical consideration and the rarity of the disease have been taken into account during policy discussion. The process of using HTA work as part of policy decision making demonstrates the importance of linking research to users, a crucial aspect of health technology assessment. Consideration of various aspects other than health economics and budget implications is part of the policy formulation practice in Thailand and can be part of practice in other countries especially LMICs that are using HTA to support policy decision making to achieve universal health coverage⁴⁴.

Conclusions

In conclusion, based on Thai willingness-to-pay threshold of 4,706 USD/QALY, infliximab and its biosimilar were not cost-effective for the treatment of refractory moderate-to-severe CD in Thailand. These treatment options would be cost-effective if the prices were significantly decreased. The best value for money treatment strategy was standard dose infliximab biosimilar with a restricted duration of treatment. Policy makers and clinicians may consider this study as part of key evidence to support the role of infliximab in managing refractory moderate-to-severe CD patients in Thailand.

Transparency

Declaration of funding

This work was supported by Thai Food and Drug Administration (Thai FDA), Ministry of Public Health to support National List of Essential Medicine Selection Process of Thailand [grant numbers 119/2562 and 249/2562].

Declaration of financial/other relationships

PP, CK, JL, SA, PC, PiP, TK, ST, and NC declare no potential conflicts of interest with respect to the research, organization, and publication of this work.

JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

PP, CK, and NC participated in the study concept and design, data acquisition, data analysis, data interpretation, manuscript drafting, critical revision of the manuscript, and the final review of the manuscript.

JL and SA participated in the study concept and design, data acquisition, critical revision of the manuscript, and the final review of the manuscript.

PC, PiP, TK, and ST participated in the study concept and design, data acquisition, critical revision of the manuscript, and the final review of the manuscript.

Previous presentations

This work was presented at International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Europe conference in Copenhagen, Denmark on November 6, 2019.

Acknowledgements

The authors thank Thai FDA, Ministry of Public Health, Thailand for providing the opportunity and financial support for this work. The authors thank Faculty of Medicine Siriraj Hospital, Mahidol University and Faculty of Medicine, Chulalongkorn University for the data and Natchamol Tophol, research assistant, for collecting the data.

Notes

i Remsima, Celltrion, Inc., Incheon, Republic of Korea.