Cost-effectiveness analysis of disease modifiying drugs (interferons and glatiramer acetate) as first line treatments in remitting-relapsing multiple sclerosis patients

Abstract

Objective:

The aim of this study was to assess cost-effectiveness of the different Disease Modifying Drugs (DMD) used as first-line treatments (interferons IM IFNβ-1a, SC IFNβ-1a, SC IFNβ-1b, and glatiramer acetate, GA) in Remitting-Relapsing Multiple Sclerosis (RRMS) in Spain.

Methods:

A Markov model was developed to simulate the progression of a cohort of patients with RRMS, during a period of 10 years. Seven health states, defined by the Expanded Disability Status Scale (EDSS), were considered in the model. Patients with an EDSS score less than 6.0 were assumed to be treated with one of the DMD. In addition, all patients were assumed to receive symptomatic treatment. The monthly transition probabilities of the model were obtained from the literature. The analysis was performed from the societal perspective, in which both direct and indirect (losses in productivity) healthcare costs (€, 2010) were included. A discount rate of 3% was applied to both costs and efficacy results.

Results:

GA was the less costly strategy (€322,510), followed by IM IFNβ-1a (€329,595), SC IFNβ-1b (€ 333,925), and SC IFNβ-1a (€348,208). IM IFNβ-1a has shown the best efficacy results, with 4.176 quality-adjusted life years (QALY), followed by SC IFNβ-1a (4.158 QALY), SC IFNβ-1b (4.157 QALY), and GA (4.117 QALY). Incremental costs per QALY gained with IM IFNβ-1a were €−1,005,194/QALY, €−223,397/QALY, and €117,914/QALY in comparison to SC IFNβ-1a, SC IFNβ-1b, and GA, respectively.

Conclusions:

First-line treatment with GA is the less costly strategy for the treatment of patients with RRMS. Treatment with IM IFNβ-1a is a dominant strategy (lower cost and higher QALY) compared with SC IFNβ-1a and SC IFNβ-1b. However, IM IFNβ-1a is not a cost-effective strategy vs GA, because incremental cost per QALY gained with IM IFNβ-1a exceeds the €30,000 per QALY threshold commonly used in Spain.

Limitations:

The highly-restrictive inclusion criteria of clinical trials limits generalization of the results on efficacy to all patients with multiple sclerosis. Availability of data for head-to-head comparisons is associated with the use of information from clinical trials.

Keywords::

• Relapsing remitting multiple sclerosis (RRMS)
• Markov chain
• Disease modifying drugs (DMD)
• Glatiramer acetate
• Interferons

Introduction

Multiple sclerosis (MS) is an autoimmune neurological disease. It is fundamentally characterized by degenerative inflammation and demyelination of the central nervous system (CNS). The most common form of MS is relapsing-remitting multiple sclerosis, which represents ∼65% of patients with MS1. MS is the chronic neurological disease with the greatest prevalence among young adults; indeed, 70% of patients develop MS between the ages of 15 and 452,3. MS is highly prevalent in Spain, affecting 50–70 of every 100,000 inhabitants4–6.

There is currently no effective cure for MS. However, treatments with Disease Modifying Drugs (DMD) slows disease progression and prevents new relapses in MS patients. Actually, in Spain, the treatments used as a first-line of RRME therapy are glatiramer acetate and the interferons beta-1a and beta-1b; the second-line treatment is natalizumab monoclonal antibody7. Fingolimod is a novel MS-immunosuppressant-treatment recently approved by the Food and Drug Administration and the European Medicines Agency; but at the moment of this analysis not commercialized in Spain.

The introduction of DMDs in the therapeutic arsenal for MS has meant an increase in the management costs of this disease8,9. Various studies have shown that the economic impact of the disease is fundamentally related to the progression of the disability, the appearance of relapses10, and the cost of the treatment8,9.

In order to determine the efficiency of the different DMDs for MS (i.e., the relationship between the incremental cost associated with these medications for disease management and the benefits provided by DMDs to MS patients), several modeling studies have been developed11–17.

Many publications focused their analyses on the comparison between the best supportive care strategy vs the administration of any of the DMDs available. Most of them conclude that the DMDs are not cost-effective compared with the best supportive care12,17.

The management of RRME in Spain implies the use of any of the DMD, as established in the consensus document prepared by the Spanish Neurological Society7. In the actual economical context, it is important to assess the cost-effectiveness relationship among the DMD, and provide reliable information that support clinicians and healthcare stakeholders decisions.

The objective of this study was to assess cost-effectiveness of the different therapies used as first-line treatments (interferons beta-1a, beta-1b, and glatiramer acetate) for patients with relapsing-remitting multiple sclerosis (RRMS) in Spain.

Methods

Model description

In order to calculate the incremental cost-effectiveness ratio of first-line treatments among patients diagnosed with RRMS in Spain, a Markov model was developed using Microsoft Excel® based on previously developed models11,13. Markov models are characterized by representation of the clinical progression of patients by transitions through different health states. The present model considers seven health states (Figure 1) that are defined by the Expanded Disability Status Scale (EDSS)18:

1. EDSS 0.0–2.5: no limitations or small mobility limitations;

2. EDSS 3.0–5.5: moderate mobility limitations;

3. EDSS 6.0–7.5: requiring some help to walk or a wheelchair;

4. EDSS 8.0–9.5: incapable of getting out of bed;

5. Death (natural causes or EDSS 10);

6. Relapse EDSS 0.0–2.5: a relapse with a change in disability within EDSS 0.0–2.5; and

7. Relapse EDSS 3.0–5.5: a relapse with a change in disability within EDSS 3.0–5.5.

Figure 1.  Schematic description of Markov model.

Similar to other models11,13,16,19, the duration of each Markov cycle is set at 1 month. In this way, the patients may, in each month, either remain in the same health state, progress to the next state, or die. Natural death is considered in all heath states excepts for EDSS 8.0–9.5. Transition probability from EDSS 8.0–9.5 to death is considered death due to MS, given that data confirm that 90% of patients with MS are disabled before dying20.

All parameters (transition probabilities and cost) were fitted to 1-month cycles.

Treatments

The model assumes that patients with an EDSS score between 0.0 and 5.5 are treated with one of the DMDs recommended as a first-line treatment for RRMS by the Spanish Neurology Society (Sociedad Española de Neurología)7:

• Intramuscular interferon β-1a (IM IFNβ-1a; Avonex®, Biogen Idec Ltd, Switzerland);

• Subcutaneous interferon β-1a 44 mcg (SC IFNβ-1a 44 mcg; Rebif 44®, Merck Serono Europe Ltd, Switzerland);

• Subcutaneous interferon β-1b (SC IFNβ-1b; Betaferon®, Bayer Schering Pharma AG, Germany and Extavia®, Novartis Europharm Ltd, Switzerland); and

• Subcutaneous Glatiramer Acetate (SC GA; Copaxone®, Teva Pharmaceutical Ltd).

In addition to treatment with a DMD, the model assumes that all patients receive symptomatic treatment for MS. In our model, changes in DMD treatment were not considered, so when a patient discontinues only symptomatic treatment is allowed.

Population

The model simulates a hypothetical cohort of 22,255 patients, equivalent to an estimated number of RRMS patients in Spain in 201021.

With the intention of representing current clinical practices for RRMS patients, patients with varying degrees of disability in terms of EDSS were included22. Table 1 specifies the distribution of the patients with respect to their EDSS scores.

Table 1.  Summary of model inputs.

Parameters	Reference scenario
Initial distribution of EDSS22
EDSS 0.0–2.5	30.0%
EDSS 3.0–5.5	31.7%
EDSS 6.0–7.5	20.6%
EDSS 8.0–9.5	17.7%
Probability of disease progression11
EDSS 0.0–2.5 to 3.0–5.5	0.004438
EDSS 3.0–5.5 to 6.0–7.5	0.009189
EDSS 6.0–7.5 to 8.0–9.5	0.003583
EDSS 8.0–9.5 to 10 (death)	0.000952
One month relapse rate11	0.075500
Incidence of Nabs36
IM IFNβ-1a	4.5%
SC IFNβ-1a	23.5%
SC IFNβ-1b	19.3%
SC GA	0.0%
Utility weights8
EDSS 0.0–2.5	0.777
EDSS 3.0–5.5	0.577
EDSS 6.0–7.5	0.446
EDSS 8.0–9.5	0.085
Relapse EDSS 0.0–2.5	0.747
Relapse EDSS 3.0–5.5	0.547

Perspective, time horizon, and discount rate

The analysis was performed from the societal perspective, whereby both direct and indirect health costs were included (pharmacological, MS management, and Loss of productivity costs).

The model assumes a time horizon equals the maximum follow-up period in the clinical trial, in order to minimize the bias of using short-term data to estimate long-term efficacy. The time horizon was set to 10 years, follow the period of Ford et al.23 study.

Therefore, an annual discount rate of 3% was applied to adjust clinical and economical results24.

Transition probabilities

Transition probabilities for symptomatic treatment were obtained from the literature11. These transitions represent the progression over time of patients with MS (Table 1)3,25–28, and apply to those patients who are not treated with DMDs. In addition, death probability due to natural causes was estimated from the National Statistics Institute (2008 data)29. This death probability was applied for all health states, except for EDSS 8.0–9.5. The death transition probability for patients in EDSS 8.0–9.5 was calculated to be the probability of death by natural causes plus the probability of dying of MS based on Prosser et al.19.

It was assumed that DMDs reduce the transition probabilities to health states with higher EDSS score (greater disability) and the probability of a relapse, i.e., a transition to a health state with a relapse (Relapse EDSS 0.0–2.5 or Relapse EDSS 3.0–5.5). Based on this assumption, to calculate transition probabilities for patients treated with a DMD, symptomatic treatment transition probabilities were used as baseline and adjusted by treatment’s effects on relapses relative risks and disease progression.

Data used for estimating the effectiveness of each of the treatments was obtained from the following clinical trials: for SC GA23,30, for IM IFNβ-1a31, for SC IFNβ-1a32,33, and for SC IFNβ-1b34,35. To avoid the bias derived of the different follow-up period in the clinical trial data, DMDs effectiveness was fitted with log-linear regression curves (Figures 2–4).

Figure 2.  Relapse rate estimation.

Figure 3.  Disease progression estimation for EDSS 0.0–2.5 to EDSS 3.0–5.5.

Figure 4.  Disease progression estimation for EDSS 3.0–5.5 to EDSS 6.0–7.5.

Neutralizing antibodies

The administration of interferons may cause the development of neutralizing antibodies (NAbs). The development of NAbs to β-interferons abrogates biological activity in the short-term, and leads to a loss of clinical benefit of patients with ME36.

The current model assumes that the presence of NAbs modifies the probability of developing a relapse among patients who receive β-interferon as a DMD. Therefore, the probabilities of suffering a relapse were adjusted using NAbs development incidence data36. This fitting was only applied after the second year of treatment with β-interferon. NAbs incidence data are specified in Table 1.

Utilities

Quality-adjusted life years (QALYs) were calculated by multiplying the life years gained (LYG) by a utility value corresponding to each health state (QALY = LYG * Utility).

The utilities (Table 1) were obtained from an observational study performed in Spain8 using a sample of 1626 patients with MS who responded to the EQ-5D questionnaire37–39.

Resource use and cost estimation

The model included DMD treatment costs, symptomatic treatment costs, management costs of each health state, and loss of productivity-related costs. All costs were obtained from the literature, were confirmed by an expert panel, and were updated to €2010 through the annual data of the Consumer Price Index (CPI)40 when necessary.

A five expert panel lead by Dr Arroyo validated the inputs.

Pharmacological cost

DMDs pharmacological costs were expressed as ex-factory prices, and dosages used were those recommended in the summary product characteristics41–45.

Table 2 specifies DMDs pharmacological cost per month.

Table 2.  Costs (monthly cost, €, 2010).

Parameters	Reference scenario
DMD Drug costs19
IM IFNβ-1a	€907.96
SC IFNβ-1a	€1267.95
SC IFNβ-1b	€942.59
SC GA	€848.68
MS management cost
Health state-specific symptomatic treatment MS costs19
EDSS 0.0–2.5	€135.04
EDSS 3.0–5.5	€159.08
EDSS 6.0–7.5	€168.06
EDSS 8.0–9.5	€202.43
Relapse EDSS 0.0–2.5	€135.04
Relapse EDSS 3.0–5.5	€159.08
Health state-specific MS-related costs8
EDSS 0.0–2.5	€936.72
EDSS 3.0–5.5	€1683.79
EDSS 6.0–7.5	€2898.13
EDSS 8.0–9.5	€4350.83
Relapse EDSS 0.0–2.5	€1856.56
Relapse EDSS 3.0–5.5	€2603.13
Cost of lost worker productivity^8,48
No treatment	€211.44
IM IFNβ-1a	€174.87
SC IFNβ-1a	€174.87
SC IFNβ-1b	€197.02
SC GA	€117.02

Costs associated with the management of MS

Medical costs related to MS were obtained from the literature8. To estimate the symptomatic treatment cost of MS, an expert panel was consulted; the panel provided information about symptom frequency (spasticity, pain, seizures, depression, and anxiety) and their clinical management.

The cost of the treatments for these symptoms was calculated from the ex-factory prices as a function of dosage recommended in each of the respective summary product characteristics46. The MS management costs were disaggregated into medical costs and symptomatic treatment costs, per month.

Loss of productivity costs

Costs derived from loss of productivity were calculated according to the cost of lost working days due to MS. To estimate the loss of productivity costs, the percentage of sick leave among MS patients8 and the mean cost of one working day (€21.10/day based on 2010 Spanish cross-professional minimum wage47) were used. It was assumed that treatment with DMDs reduce the percentage of sick leave (53.7% for SC GA, 20.8% for IM IFNβ-1a, and SC IFNβ-1a and 8.2% for SC IFNβ-1b)48. The monthly costs due to loss of productivity are specified in Table 2.

Cost-effectiveness and cost-utility analysis

The life years gained (LYG), the quality-adjusted life years (QALY), the average number of years without a relapse, and the mean percentage of patients with an EDSS score greater than 6.0 during the time horizon considered were used as benefit measures.

The efficiency of comparing the treatments in RRMS was established through the Incremental Cost-Effectiveness Ratio (ICER) and the Incremental Cost-Utility Ratio (ICUR). For each comparison, the reference treatment was the most effective of the treatments compared.

For an appropriate interpretation of these ratios, this study used the commonly accepted efficiency threshold in Spain (€30,000 per QALY)49.

Sensitivity analysis

To evaluate the consistency of the model results, univariate sensitivity analyses were performed. The values modified in the analyses were:

1. Discount rate was set to 0% and 5%;

2. The incidence of NAbs appearance, using data from Rossman50;

3. The monthly cost due to loss of productivity. Assuming a common reduction of 40% in sick leave for all of the DMDs, a monthly cost for productivity loss was set at €141.11 for all of the patients who received treatment with a DMD; and

4. Time horizon considered was set to 2, 4, 6, and 8 years.

Results

Reference scenario

With a time horizon of 10 years, the total costs of multiple sclerosis per patient were €329,595.43, €348,208.20, €333,925.43, and €322,509.96 for IM IFNβ-1a, SC IFNβ-1a, SC IFNβ-1b, and SC GA, respectively (Table 3). The cost per patient for DMDs was, €47,531.94 (14.4%) for IM IFNβ-1a, €65,474.67 (18.8%) for SC IFNβ-1a, €48,751.47 (14.6%) for SC IFNβ-1b, and €42,453,89 (13.2%) for SC GA.

Table 3.  Results.

	Cost/outcome
Reference scenario	IM IFNβ-1a	SC IFNβ-1a	SC IFNβ-1b	SC GA
MS drug costs per patient (€, 2010)	€47,531.94	€65,474.67	€48,751.47	€42,453.89
Total costs (€, 2010)	€329,595.43	€348,208.20	€333,925.31	€322,509.96€
Life years gained (LYG) per patient	8.580766998	8.580462928	8.580450005	8.579813781
Incremental cost–effectiveness ratio (cost/LYG) IM IFNβ-1a vs (SC IFNβ-1a or SC IFNβ-1b or SC GA)	NA	Dominant	Dominant	7,433,218
Incremental cost–effectiveness ratio (cost/LYG) SC IFNβ-1a vs (SC IFNβ-1b or SC GA)	NA	NA	1,105,230,210	39,587,705
Incremental cost–effectiveness ratio (cost/LYG) SC IFNβ-1b vs SC GA	NA	NA	NA	17,942,344
Quality adjusted life years (QALY) per patient	4.17699627	4.15847968	4.15761431	4.116906617
Incremental cost–utility ratio (cost/QALY) IM IFNβ-1a vs (SC IFNβ-1a or SC IFNβ-1b or SC GA)	NA	Dominant	Dominant	117,914
Incremental cost–utility ratio (cost/QALY) SC IFNβ-1a vs (or SC IFNβ-1b or SC GA)	NA	NA	16,504,952	618,146
Incremental cost–utility ratio (cost/QALY) SC IFNβ-1b vs SC GA	NA	NA	NA	280,422
Percentage of surviving patients in EDSS 0.0–5.5	42.52%	40.77%	40.92%	38.28%
Average time spent relapse free (years) for patients on therapy	9.614615	9.879432	9.795767	9.792811
Incremental cost–relapse free years ratio (cost/relapse free years) SC IFNβ-1a vs (IM IFNβ-1a or SC IFNβ-1b or SC GA)	70,285	NA	170,715	296,674
Incremental cost–relapse free years ratio (cost/relapse free years) SC IFNβ-1b vs (IM IFNβ-1a or SC GA)	Dominant	NA	NA	3,861,756
Incremental cost–relapse free years ratio (cost/relapse free years) SC GA vs IM IFNβ-1a	28,497	NA	NA	NA

NA, Not applicable.

The modeling of the disease shows that there are differences in the results (as measured by LYG and QALYs) obtained with the four DMDs. The health benefit that supports these strategies ranges between a minimum of 8.579,81 LYG/patient (SC GA) and a maximum of 8.580,77 LYG/patient (IM IFNβ-1a) or, respectively, 4.116,90 to 4.176,99 QALYs/patient over a time horizon of 10 years. These differences are also reflected in the survival percentages of patients with a score on the EDSS scale between 0.0–5.5. The percentages were 42.52%, 40.77%, 40.92%, and 38.28% for IM IFNβ-1a, SC IFNβ-1a, SC IFNβ-1b, and SC GA, respectively.

The average time free from relapses (in years) per patient was 9.792,81, 9.614,62, 9.879,43, and 9.795,77 for SC GA, IM IFNβ-1a, SC IFNβ-1a, and SC IFNβ-1b, respectively.

The incremental costs per quality adjusted life year (QALY) gained for IM IFNβ-1a in comparison to the other therapies were €−1,005,194/QALY compared with SC IFNβ-1a, €−223,397/QALY compared with SC IFNβ-1b, and €117,914/QALY compared with SC GA. ICURs comparing SC IFN-1a with SC IFNβ-1b or SC GA were €16,504,952/QALY and €618,146/QALY, respectively. ICUR of SC IFNβ-1b compared with SC GA was €280,422/QALY.

Deterministic sensitivity analysis

The change in the incidence of NAbs appearance according to Rossman50 in the baseline case yielded average years without a relapse of 9.6159, 9.8955, 9.7848, and 9.7928 for IM IFNβ-1a, SC IFNβ-1a, SC IFNβ-1b, and SC GA, respectively.

When a monthly cost due to productivity loss of €141.11 was applied for all the DMDs, the total costs of the disease per patient in the baseline case were €326,294.04, €344,863.09, €328,422.06, and €324,805.73 for IM IFNβ-1a, SC IFNβ-1a, SC IFNβ-1b, and SC GA, respectively.

The ICURs obtained by varying the time horizon and discount rates of the analysis are shown in Table 4.

Table 4.  Sensitivity analysis results.

Parameter	Value		Incremental Cost Utility Ratios (ICUR €/QALY)
	BC	SA		IM IFNβ-1a	SC IFNβ-1a	SC IFNβ-1b	SC GA
Discount rate 0%
IM IFNβ-1a	3%	0%	IM IFNβ-1a vs	NA	Dominant	Dominant	109,517
SC IFNβ-1a			SC IFNβ-1a vs	NA	NA	18,299,530	585,518
SC IFNβ-1b			SC IFNβ-1b vs	NA	NA	NA	264,039
SC GA			SC GA vs	NA	NA	NA	NA
Discount rate 5%
IM IFNβ-1a	3%	5%	IM IFNβ-1a vs	NA	Dominant	Dominant	123,798
SC IFNβ-1a			SC IFNβ-1a vs	NA	NA	15,591,890	640,896
SC IFNβ-1b			SC IFNβ-1b vs	NA	NA	NA	291,873
SC GA			SC GA vs	NA	NA	NA	NA
Incidence of NAbs
IM IFNβ-1a	4.50%	2.00%	IM IFNβ-1a vs	NA	Dominant	Dominant	116,419
SC IFNβ-1a	23.50%	12.50%	SC IFNβ-1a vs	NA	NA	12,147,625	612,283
SC IFNβ-1b	19.30%	27.80%	SC IFNβ-1b vs	NA	NA	NA	279,063
SC GA	0.00%	0.00%	SC GA vs	NA	NA	NA	NA
Work cost
IM IFNβ-1a	€174.87	€141.11	IM IFNβ-1a vs	NA	Dominant	Dominant	24,768
SC IFNβ-1a	€174.87		SC IFNβ-1a vs	NA	NA	18,998,848	482,460
SC IFNβ-1b	€197.02		SC IFNβ-1b vs	NA	NA	NA	88,836
SC GA	€117.02		SC GA vs	NA	NA	NA	NA
Time horizon 2 years
IM IFNβ-1a	10 years	2 years	IM IFNβ-1a vs	NA	Dominant	Dominant	692,158
SC IFNβ-1a			SC IFNβ-1a vs	NA	NA	12,781,019	2,549,178
SC IFNβ-1b			SC IFNβ-1b vs	NA	NA	NA	1,313,301
SC GA			SC GA vs	NA	NA	NA	NA
Time horizon 4 years
IM IFNβ-1a	10 years	4 years	IM IFNβ-1a vs	NA	Dominant	Dominant	326,740
SC IFNβ-1a			SC IFNβ-1a vs	NA	NA	9,015,921	1,335,585
SC IFNβ-1b			SC IFNβ-1b vs	NA	NA	NA	662,927
SC GA			SC GA vs	NA	NA	NA	NA
Time horizon 6 years
IM IFNβ-1a	10 years	6 years	IM IFNβ-1a vs	NA	Dominant	Dominant	210,792
SC IFNβ-1a			SC IFNβ-1a vs	NA	NA	8,968,754	940,653
SC IFNβ-1b			SC IFNβ-1b vs	NA	NA	NA	451,683
SC GA			SC GA vs	NA	NA	NA
Time horizon 8 years
IM IFNβ-1a	10 years	8 years	IM IFNβ-1a vs	NA	Dominant	Dominant	152,883
SC IFNβ-1a			SC IFNβ-1a vs	NA	NA	10,808,574	740,466
SC IFNβ-1b			SC IFNβ-1b vs	NA	NA	NA	345,141
SC GA			SC GA vs	NA	NA	NA	NA

NA, Not applicable.

Discussion

In this study, a cost-effectiveness analysis was performed to assess the efficiency of first-line treatments for RRMS available in Spain. The results of the analysis show that IM IFNβ-1a is a dominant strategy compared with SC IFNβ-1a or SC IFNβ-1b. However, the treatment with IM IFNβ-1a is not a cost-effective strategy vs SC GA, with an acceptability threshold of €30,000/QALY49. The differences in effectiveness between both therapies do not set off the incremental drug acquisition costs of IM IFNβ-1a compared with SC GA. However, differences in effectiveness among all DMDs are small compared with the higher differences in costs. This fact is also supported by the sensitivity analyses performed in the model.

The model estimates that treatment with SC GA compared with IM IFNβ-1a renders a ratio of €28,497 per relapse-free year. This result shows that treatment with SC GA can be considered in terms of relapse-free years as a cost effective strategy compared with IM IFNβ-1a, considering a threshold of €30,000 per relapse-free year.

Previous studies^{11,12,13,15,16,51} have suggested that the use of DMDs involves an additional economic investment in disease management that produces the benefits of a reduced incidence of relapses and a delay in the progression of the disease, thus achieving an improvement for patients.

Comparisons with other economic evaluations are difficult due to differences in terms of the methodology and design. The models of Goldberg et al.51 in the US and Nuitjen et al.15 in Germany employ as a main result the number of relapses avoided instead of the cost per QALY gained; however, SC GA was the lowest global cost for managing MS strategy in both studies.

The present Markov model has been previously used to evaluate the efficiency of SC GA in comparison to symptomatic treatment. In the US, Bell et al.11 estimated that the incremental cost-effectiveness ratio of SC GA in comparison to symptomatic treatment was $258,465/QALY (US dollars, valued in 2005), which was considered cost-effective using a threshold of $50,000/QALY gained52. Another study performed in the US13 concluded that therapy with SC GA was also the less costly comparing with the other DMDs, but not a dominant strategy (in terms of 2007 costs saved) in comparison to symptomatic treatment for MS. In both studies, SC GA was the least costly option compared to the other treatments assessed in the study, natalizumab13 and interferons (IM IFNβ-1a SC IFNβ-1a, SC IFNβ-1b)11.

In a cost-utility study performed in Spain considering a time horizon of 54 years53, treatment with SC GA turned out to be the dominant strategy when compared individually with each of the three interferons (IM IFNβ-1a SC IFNβ-1a, SC IFNβ-1b) evaluated or jointly with a combination of the three (in 2001 values).

The current model has some limitations that should be considered when interpreting the results. The inclusion of long-term prospective treatment information in a model is the option preferred by decision-makers for the processes of adopting new therapeutic strategies. Naturalistic design studies with prospective determination of the consumption of medical resources would be ideal54 for obtaining data with an economic focus; however, when this type of data is not available, as in this case, it is necessary to base models on data from clinical trials, with the limitations that implies55 for the external validity of the results. The highly-restrictive inclusion criteria of clinical trials limits generalization of the results on efficacy to all patients with multiple sclerosis.

There is also a limitation in the availability of data for head-to-head comparisons associated with the use of information from clinical trials. Since the comparative studies published to date56,57 do not contain enough information to complete the model, it was necessary to use several sources.

In the model it was assumed that the individuals were always working. This assumption was made because data available about working activity in RRME patients were not split by EDSS levels. However, this issue was tested in a sensitivity analysis.

The use of data from different sources requires the adoption of certain assumptions, and the quality of the results is conditional on the consistency of the studies of origin. In the present model, the influence of the assumptions and of the heterogeneity of these sources was limited by the performance of a sensitivity analysis. Our simulation was also not able to include the individual characteristics of the patients, a common characteristic of all cohort models, and, for this reason, the model results represent an average patient.

Despite the above-mentioned limitations, the model assumptions may be considered reasonable, even conservative, and the results of the sensitivity analysis attest to the robustness of the results.

Conclusions

First-line treatment with GA is the less costly strategy for the treatment of patients with RRMS. Treatment with IM IFNβ-1a is a dominant strategy (lower cost and higher QALY) compared with SC IFNβ-1a and SC IFNβ-1b. IM IFNβ-1a is not a cost-effective strategy vs GA, because incremental cost per QALY gained with IM IFNβ-1a is much above the €30,000 per QALY threshold commonly used in Spain.

Transparency

Declaration of funding

This study was funded by Teva Pharmaceutical.

Declaration of financial/other interest

ES and MAC have disclosed that they are employees of Pharmacoeconomics & Outcomes Research Iberia, Madrid, Spain, an independent contract health economic organization that has received research funding from Teva Pharmaceutical. RS has disclosed that he is employed by Teva Pharmaceutical. RA has disclosed that he serves as a consultant to and has received honoraria from Teva Pharmaceutical, as well as other pharmaceutical companies that market drugs for the treatment of patients with multiple sclerosis and other medical conditions.