Cost-minimization analysis of fingolimod compared with natalizumab for the treatment of relapsing–remitting multiple sclerosis in Sweden

Abstract

Background:

Fingolimod and natalizumab have the same European Union licence for the treatment of relapsing multiple sclerosis, and are considered by the Committee for Medicinal Products for Human Use (CHMP) to have broadly similar efficacy.

Objective:

A cost-minimization analysis was performed to compare differences in treatment costs between fingolimod and natalizumab from a societal perspective in Sweden.

Methods:

This analysis included costs associated with initiating and following treatment (physician visits and monitoring), continuing therapy (drugs and administration), and lost patient productivity and leisure time. Unit costs (in Swedish krona [SEK]) were based on regional data (median prices for physician visits and monitoring sessions). Natalizumab infusion costs were obtained from the national cost-per-patient database. Drug costs for both therapies were 15,651 SEK/28 days.

Results:

After 3 years, fingolimod use was associated with savings of 124,823 SEK/patient compared with natalizumab (total cost/patient: 566,718 SEK vs 691,542 SEK). Cost savings with fingolimod were 40,402 SEK/patient after 1 year and 301,730 SEK/patient after 10 years. Treatment with natalizumab was 18% more expensive than fingolimod therapy after 1 year and 23% more expensive after 10 years.

Limitations:

Based on the CHMP assessment, it was assumed that fingolimod and natalizumab have similar efficacy. The analysis was conducted for Sweden, and caution is needed in extrapolating the results to other countries.

Conclusion:

Fingolimod is cost-saving compared with natalizumab for the treatment of relapsing–remitting multiple sclerosis in Sweden.

Key words::

• Multiple sclerosis
• Fingolimod
• Natalizumab
• Cost-minimization
• Sweden
• Costs

Introduction

MS is a leading cause of disability in young people1. This chronic, progressive, inflammatory, and degenerative disease of the central nervous system (CNS)2 has a particularly high prevalence in Sweden, where it is estimated to affect 188.9 per 100,000 people3. Most patients (80–85%) present with the relapsing–remitting form of MS (RRMS), which is characterized by self-limiting acute attacks of neurological dysfunction (relapses) followed by periods of remission with partial or full recovery2. The course of MS is highly unpredictable, but in approximately two-thirds of patients with RRMS the disease progressively worsens to result in irreversible disability4, leading to reduced quality-of-life (QoL)5 and premature death6. QoL is diminished by the physical and emotional symptoms of MS, such as impaired mobility, relapses, fatigue, pain, and depression, and the consequences of these symptoms on a patient’s ability to work and perform daily tasks7,8. As MS progresses, QoL decreases9–11, and healthcare costs and lost productivity escalate9,10. A cross-sectional study assessing the cost of MS in Sweden in 1998 revealed that the total costs to society were ∼4.9 billion Swedish krona (SEK), with major cost items being personal assistance, and long-term sickness or early retirement12. Direct and indirect costs both rose significantly with increased disability: costs were almost 5-times higher for a patient with severe MS than for an individual with mild disease12. Another cross-sectional study published in 2006 also found that costs were significantly correlated with MS severity, with total costs increasing 7-fold from early (Expanded Disability Status Scale (EDSS) scores 0–1) to late (EDSS scores 8–9) disease13.

Disease-modifying therapies (DMTs) form the mainstay of RRMS management, and aim to reduce the frequency and severity of relapses, and to slow disease progression14. Interferon beta and glatiramer acetate are recommended as first-line DMTs for RRMS, but approximately one-third of patients continue to experience high disease activity despite receiving these conventional therapies15,16. Fingolimod, a sphingosine 1-phosphate receptor modulator, is the first oral agent approved for treating RRMS17; it is licensed for use in patients with high disease activity, despite treatment with interferon beta and for individuals with rapidly evolving severe (RES) RRMS in the European Union (EU)18. In phase 3 trials, fingolimod reduced the annualized relapse rate and risk of disability progression compared with placebo in FREEDOMS (FTY720 Research Evaluating Effects of Daily Oral therapy in MS)19 and compared with interferon beta-1a in TRANSFORMS (Trial Assessing Injectable Interferon vs FTY720 Oral in RRMS)20. In Sweden, six DMTs are included in the drug benefit scheme; these include interferon beta-1a (Avonex, Biogen Idec, Weston, Massachusetts, USA, and Rebif, EMD Serono, Rockland, Massachusetts, USA), interferon beta-1b (Betaferon, Bayer Healthcare, Berlin, Germany, and Extavia, Novartis, Basel, Switzerland), glatiramer acetate (Copaxone, Teva Pharmaceuticals, Petah Tikva, Israel) and natalizumab (Tysabri, Biogen Idec)21. Natalizumab is a recombinant anti-alpha4 monoclonal antibody administered by 4-weekly infusions; it is approved as a second-line therapy for RRMS, with the same EU licensed indication as fingolimod22.

There is much interest in the comparative efficacy of fingolimod and natalizumab, particularly as they are licensed for the same EU patient population; however, no head-to-head comparisons have been carried out. Crude comparisons between the primary outcomes of the phase 3, placebo-controlled trials for fingolimod and natalizumab have suggested differences in efficacy, although they do not adjust for differences in patient baseline characteristics23. A recent indirect treatment comparison indicates that these differences may be largely explained by differences in patient populations and definitions of confirmed disability progression between the trials24. Independent analyses by the EU Committee for Medicinal Products for Human Use (CHMP) and the Scientific Advisory Group Neurology have concluded that fingolimod has broadly similar efficacy to natalizumab in patients with high disease activity, despite previous treatment or with RES RRMS25. Based on this conclusion, a cost-minimization analysis (CMA), which formed part of a reimbursement submission to the Swedish Dental & Pharmaceutical Benefits Board (TLV), was performed to compare differences in treatment costs between fingolimod and natalizumab from a societal perspective in Sweden.

Methods

As fingolimod is considered to have broadly similar efficacy to natalizumab in patients with RRMS25, a CMA was performed. The CMA was carried out according to the TLV guidelines for health economic assessment, which stipulate that (i) a societal perspective should be taken, including all relevant costs irrespective of the payee; (ii) production loss for treatment and sickness should be included; and (iii) costs should be discounted at 3% per annum26. All costs in SEK were for 2010, and unit costs were based on Swedish regional data. The analysis used a 10-year time horizon, but the base-case scenario reports cost estimates after 3 years. The CMA includes costs associated with: (i) initiating therapy; (ii) continuing therapy; (iii) drug acquisition and administration; and (iv) lost patient productivity and leisure time.

Costs associated with initiating therapy

This component includes costs associated with monitoring, laboratory tests and examinations, and physician visits carried out before or during the first treatment with each drug. Table 118,27–30 lists the procedures associated with initiating therapy; these were obtained from the Summary of Product Characteristics (SmPC) for fingolimod18 and natalizumab27, and the Swedish MS Society’s guidelines28,29. Costs were based on unit costs (Table 2) from Swedish regional data. The cost of pregnancy testing was excluded because it is inexpensive and performed before initiating both treatments. A recent review of the risks and benefits of fingolimod by the CHMP has recommended that all patients starting treatment with fingolimod should have their heart activity monitored using an electrocardiogram (ECG) before receiving the first dose of the medicine and continuously for at least 6 h after this first dose, and that blood pressure and heart rate should be measured before treatment and then hourly for the first 6 h after the first dose30. Monitoring should be extended for at least 2 h in patients whose heart rate is lowest 6 h after receiving the first dose of fingolimod30. In patients who develop clinically-significant heart problems, such as brachycardia or atrioventricular block, monitoring should continue at least overnight and until the problems have resolved30. However, findings from a recent open-label study of fingolimod suggest that these events are very rare31. The costs associated with heart activity monitoring for fingolimod were included in the analysis and based on unit costs from southern Sweden. The proportion of patients requiring overnight hospitalization following treatment initiation, according to the SmPC, will ultimately depend on the patient population starting fingolimod therapy in Sweden; as overnight extended monitoring is required in pre-defined patient populations where treatment with fingolimod should be considered only if the anticipated benefits outweigh the potential risks18. It was, therefore, conservatively assumed that 10% would require overnight hospitalization, and alternative assumptions (±5%) were tested in sensitivity analyses to assess the impact of the extended monitoring.

Table 1.  Procedures associated with initiating fingolimod or natalizumab therapy (before or during the first treatment with each drug).^a

Fingolimod	Natalizumab
• Neurologist visit	• Neurologist visit
• Blood status	• Blood count with differential
• Assessment of levels of transaminases and bilirubin • Testing for VZV in patients who have not had chickenpox or have not been vaccinated against VZV • Blood pressure and heart rate monitoring after the first dose and every hour for 6 h after the first dose • ECG monitoring before initiating treatment, and then continuously for the first 6 h after the first dose. Extended ECG monitoring for at least 2 h or until resolution in patients with heart rate at its lowest since taking fingolimod at the end of the 6-h period. Extended ECG monitoring overnight or until resolution in patients with heart problems such as bradycardia or atrioventricular block	• Assessment of liver status • Testing for neutralizing antibodies to natalizumab • MRI • Testing for antibodies to JC virus to identify patients at risk of developing progressive multifocal leukoencephalopathy^b

^aProcedures associated with initiating therapy were obtained from the Summary of Product Characteristics for fingolimod18 and natalizumab27, and the Swedish MS Society’s guidelines28,29. Frequencies for blood pressure, heart rate, and ECG monitoring were obtained from the European Medicines Agency April 2012 questions and answers document30.

^bIn Sweden this test is funded by the manufacturers of natalizumab (Biogen Idec). Therefore, the costs of this test are not included in the analysis.

ECG, electrocardiogram; JC, John Cunningham; VZV, varicella zoster virus.

Table 2.  Unit costs in Sweden at the time of the study in 2010.

Resource	Cost (SEK)	Source
Medical staff (per visit)
Neurologist	2819	Median of 10 regional prices for neurologist visits
Nurse	731	Regional rates and allowances for southern healthcare region in 2010
Ophthalmologist	1385	Median of 11 regional prices for eye doctor visits
Treatment (each time)
Infusion	3750	See section on natalizumab infusions
Examinations
MRI (each time)	4843	Median of eight listed prices for MRI
Observation period, including blood pressure and  heart rate measurements (per 6 h)	2185	Calculated based on average salaries for nurses from the SCB, and a 20% mark-up to cover materials, administration, etc.
ECG (12 lead)	801	Regional rates and allowance for southern healthcare region in 2010
Continuous ECG	2211	Regional rates and allowance for southern healthcare region in 2010
Hospitalization, per day on neurology ward	3672	Regional rates and allowance for southern healthcare region in 2010
Samples (per sample)
Blood count, including differential	31	Laboratory Medicine in Västerbotten, 2010
Antibodies to VZV	147	Regional rates and allowances for southern healthcare region in 2010
Antibodies to JC virus	147	Assume the same cost as for antibodies to VZV
ALT or AST, hepatic status	12	Regional rates and allowances for southern healthcare region in 2010
Bilirubin	9	Regional rates and allowances for southern healthcare region in 2010
Time required (per hour)
Loss of production (work)	228	Average hourly earnings for private employees in December 2010 (SCB), including social security contributions (31.42%)
Loss of leisure time	57	Estimates from the Swedish Transport Administration Agency 2008

ALT, alanine aminotransferase; AST, aspartate transaminase; ECG, electrocardiogram; JC, John Cunningham; SCB, Swedish Central Statistical Office; SEK, Swedish krona; VZV, varicella zoster virus.

Costs associated with continuing treatment

This component includes costs associated with monitoring, laboratory tests and examinations, and physician visits carried out following treatment with each drug. The base-case analysis includes costs associated with follow-up procedures specified in the SmPC and Swedish MS Society guidelines for fingolimod18,28, and natalizumab27,29 (Table 3). Costs for these procedures were based on unit costs from Swedish regional data (Table 2).

Table 3.  Follow-up procedures for treatment with fingolimod and natalizumab according to the product Summary of Product Characteristics and the Swedish MS Society guidelines.^a

Procedure	Frequency
Fingolimod
Neurologist visit	No specific requirements, but adopt the same frequency as for natalizumab (i.e., biannually)
Eye examination (for macular edema)^b	Within 3–4 months from start of treatment
MRI	Before initiation of treatment and annually
Blood count, including differentials	At 1 and 3 months and then every 3 months
Blood pressure	At 1, 3, and 6 months, and then every 6 months
Liver status (transaminases)	At 1, 3, 6, 9, and 12 months, and then every 6 months
Natalizumab
Neurologist visit	Biannually
MRI	Before initiation of treatment and annually
Blood count, including differential	At first 3 infusions, then every 6 months
Liver status	At first 3 infusions, then every 6 months
Antibody monitoring^c	At 3, 6, 12, and 24 months

^aDetails of follow-up procedures were obtained from the Summary of Product Characteristics for fingolimod18 and natalizumab27, and the Swedish MS Society’s guidelines28,29.

^bPatients with uveitis or diabetes mellitus are recommended to undergo more frequent monitoring, but the prevalence of these conditions is such that the cost at the population level is negligible. Sensitivity analysis is performed on these costs instead.

^cMonitoring for antibodies against John Cunningham virus (JCV) are recommended as anti-JCV antibodies are associated with an increased risk of progression to progressive multifocal leukoencephalopathy27.

Costs of drug acquisition and administration

This component includes drug acquisition and administration costs. Drug costs for fingolimod and natalizumab were identical in Sweden at the time of this study. The pharmacy retail price of both treatments was 15,651 SEK/28 days and 204,022 SEK/year. No costs are associated with oral administration of fingolimod; natalizumab is administered by intravenous infusion every 4 weeks by medical staff. The costs associated with natalizumab infusion were estimated using the following sources: fixed-price contracts for natalizumab infusions in Swedish counties; published prices for the diagnosis-related group (DRG) codes for outpatient intravenous drug delivery (907O); and prices listed in the national cost-per-patient database. The national cost-per-patient database was used to estimate the cost of natalizumab infusions in the base-case scenario. It contained information on the cost of intravenous drug delivery in nine hospitals and the data could be selected to include only patients with MS and exclude the costs of drug acquisition. The weighted mean cost of infusion used in the base-case analysis was 3750 SEK/visit (range: 1059–12,038 SEK). Costs from the other sources were not used in the base-case analysis because the fixed-price contracts (estimated to be 5886–12,649 SEK/visit) incorporated other costs, such as those of physician visits and MRI, and the DRG code prices (estimated to be 2469–7152 SEK/visit) included patients without MS.

Costs associated with lost patient productivity and leisure time

Hours of lost patient productivity were estimated based on the time taken for treatment and follow-up sessions during working hours. Time at work was estimated using data from a Swedish cost-of-illness study in which 12.4% and 21.7% of patients with MS worked full-time or part-time, respectively13. It was estimated that 23% of the time for visits within the entire MS population occurs during working hours and the remaining 77% represents lost leisure time. The numbers of hours taken to attend treatment and follow-up sessions for fingolimod and natalizumab were determined by estimating the number of trips needed per year and the time taken to attend these sessions (Table 4). The time taken for procedures performed for both treatments was not included. Unit costs per hour of lost productivity were calculated based on average hourly earnings for private employees in December 2010, including social security contributions (228 SEK/h). Unit costs per hour of lost leisure time were calculated based on values used in cost–benefit analyses by the Swedish Transport Administration for 2008 (57 SEK).

Table 4.  Time taken to attend treatment and follow-up sessions for fingolimod and natalizumab.

Treatment and follow-up sessions	Hours per session	Times per year		Hours per year
		Fingolimod	Natalizumab	Fingolimod	Natalizumab
First year of treatment
Observation at treatment initiation^a	7.8	1	0	8	0
MRI at treatment initiation	1	1	1	1	1
Ophthalmologist visit	1	1	0	1	0
Infusion^b	2	0	13	0	26
Annual MRI	1	1	1	1	1
Travel to hospital^c,d	1.5	7	13	11	20
Total				21	48
Following years of treatment
Infusion	2	0	13	0	26
Annual MRI	1	1	1	1	1
Travel to hospital^c,d	1.5	4	13	6	20
Total				7	47

Hours may not add up to totals due to rounding.

^aFingolimod Summary of Product Characteristics recommends that patients are initially observed for 6 h18. In this analysis it is further assumed that 10% of patients will require monitoring for 24 h.

^bAccording to the Swedish MS Society guidelines infusions take 1 h, after which patients are observed for 1 h28.

^cAssuming a trip to hospital takes 45 min on average.

^dPatients treated with natalizumab require 13 trips to hospital per year to perform the infusions; all other examinations are assumed to coincide with these visits. Patients treated with fingolimod require seven hospital visits in the first year to perform all the examinations and tests. From year 2 onwards, patients need to make four visits a year.

Costs associated with treating adverse events

For both treatments, most adverse events (AEs) were assumed to be mild and of short duration, and therefore not associated with significant costs. The most serious AEs associated with using fingolimod (increased risk of bradycardia or atrioventricular block, macular edema, and infection) and natalizumab (progressive multifocal leukoencephalopathy) are rare, and, although the costs of managing these serious conditions may be high on an individual basis, the population-level costs were assumed to be very low and were therefore not included. The total cost of treating AEs in patients receiving natalizumab has been estimated at ∼500 SEK for the first year and 300 SEK for the second year32. These costs are low compared with other factors such as the cost of infusions every 4 weeks. Since differences in the costs of treating AEs for fingolimod and natalizumab were calculated to be negligible, the costs of treating AEs did not form part of the analysis for either drug.

Assumption of mortality and adherence to treatment

Patients were assumed to be 37 years old at the start of treatment based on data from the fingolimod phase 3 FREEDOMS study19. The relative risk of mortality was based on a Norwegian study that followed patients with MS for 50 years and estimated a 2.7-fold increase in the risk of death compared with the general population33. Age-specific mortality in the general population was obtained from the Swedish Central Statistical Office34. A withdrawal rate of 8.4% was used for both treatments, based on the weighted average values from FREEDOMS19 and TRANSFORMS20. The same withdrawal rate was used for both treatments because in the respective phase 3, placebo-controlled trials there were lower withdrawal rates in the treatment groups than in the placebo groups19,35.

Sensitivity analyses

Sensitivity analyses were performed in which the costs of monitoring, infusion, lost patient productivity and leisure time, discontinuation rates, and ophthalmologist visits, were varied. As recommended by the TLV guidelines26, the effect of changing the discount rate to 0 or 5% was also investigated. Costs of infusion were also investigated based on estimates from fixed-price contracts and DRG prices, and on the lowest value from the national cost-per-patient database.

Results

The results of the analysis show that fingolimod is cost saving compared with natalizumab for the treatment of RRMS in Sweden. After 3 years, fingolimod was associated with total savings of 124,823 SEK/patient compared with natalizumab (total cost/patient: 566,718 SEK vs 691,542 SEK). Savings for the four different cost components over 3 years are shown in Table 5. Natalizumab was associated with higher costs than fingolimod for two of the components. The additional expense of natalizumab over fingolimod was mainly due to the costs of infusions (124,845 SEK) and lost patient productivity and leisure time (8518 SEK). The increased costs associated with lost patient productivity and leisure time reflect the fact that patients treated with natalizumab require 13 trips to hospital per year, whereas those receiving fingolimod need seven trips in the first year and four visits per year thereafter. Therapy initiation costs were higher for fingolimod than for natalizumab; the additional expense was due primarily to the need for ECG monitoring before the first treatment (6365 SEK) and an ophthalmologist visit.

Table 5.  Cost-minimization results for treatment of patients with relapsing–remitting MS in Sweden over a 3-year period: fingolimod vs natalizumab^a (base-case analysis).

Cost category	Fingolimod	Natalizumab	Difference^b
Investigations/tests before and during first treatment
Medical staff	2819	2819	0
Samples	199	190	9
MRI	4843	4843	0
Observation	6365	0	6365
Total	14,226	7852	6374
Continuing therapy
Drug acquisition	522,486	522,486	0
Drug administration  (e.g., infusions)	0	124,845	−124,845
Total	522,486	647,331	−124,845
Treatment follow-up
Medical staff	17,164	14,439	2725
Samples	440	1000	–559
MRI	12,403	12,403	0
Total	30,006	27,841	2166
Time (difference)
Loss of productivity  (work time)	–	4637	−4637
Loss of leisure time	–	3881	−3881
Total	–	8518	−8518
Overall total	566,718	691,542	−124,823

Costs may not add up to the total sums due to rounding.

^aAll costs are in Swedish krona (SEK) per patient.

^bCost for fingolimod minus cost for natalizumab.

A year-by-year analysis of cumulative cost savings with fingolimod is shown in Table 6. Cost savings with fingolimod were 40,402 SEK/patient after 1 year and 301,730 SEK/patient after 10 years. Cost savings are maintained but gradually decrease in magnitude each year as the number of treated patients reduces due to treatment withdrawal or mortality. Treatment with natalizumab was 18% more expensive than that with fingolimod after 1 year and 23% more expensive after 10 years.

Table 6.  Cumulative cost savings with fingolimod over 10 years.^a

Year	Proportion treated (%)^b	Cost saving per year^c	Cumulative cost saving
1	95.7	40,402	40,402
2	87.5	44,727	85,128
3	80.0	39,695	124,823
4	73.1	35,222	160,045
5	66.8	31,246	191,291
6	61.0	27,713	219,004
7	55.7	24,572	243,575
8	50.9	21,777	265,352
9	46.4	19,292	284,644
10	42.4	17,086	301,730

^aAll costs are in Swedish krona (SEK) per patient.

^bThe proportion of patients who are still receiving treatment, given factors such as predicted mortality and adherence rates. In addition, following on from the initial treatment (where the rate is 100%), we used the proportion of patients expected to be treated by the middle of the first year for the proportion treated in year 1.

^cCost saving with fingolimod over natalizumab.

Sensitivity analyses varying costs of infusions, monitoring sessions, ophthalmologist visits, and lost patient productivity and leisure time showed that fingolimod was cost saving vs natalizumab in all cost scenarios evaluated (Table 7). When the cost of infusion was varied, by using cost estimates from other sources (average fixed-price contracts, 9675 SEK and average DRG code price, 5022 SEK), fingolimod was associated with even greater cost savings over natalizumab after 3 years (322,079 and 167,171 SEK/patient, respectively, vs 124,823 SEK/patient in the base-case analysis). Even with the most conservative estimates for natalizumab infusions (1059 SEK based on the lowest price in the cost-per-patient database), fingolimod was associated with savings of 35,234 SEK/patient over 3 years. These results show that fingolimod and natalizumab would be cost neutral only if natalizumab infusions were not associated with any costs (i.e., 0 SEK instead of 3750 SEK per infusion).

Table 7.  Sensitivity analyses varying the costs of initiating and following treatment with fingolimod or natalizumab over 3 years.

Resource	Estimate	Cost saving (SEK/patient)^a
Basis analysis		124,823
Percentage of patients stopping treatment annually	0%	142,798
Percentage of patients stopping treatment annually	10%	121,591
Discounting	0%	128,583
Discounting	5%	122,474
Cost of ophthalmologist visit	1000 SEK	125,192
Cost of ophthalmologist visit	5000 SEK	121,363
Cost of monitoring	1000 SEK	130,189
Cost of monitoring	10,000 SEK	121,189
Percentage of patients needing overnight hospitalization at treatment initiation	5%	125,007
Percentage of patients needing overnight hospitalization at treatment initiation	15%	124,640
Cost of infusion, lowest price for cost-per-patient database	1059 SEK	35,234
Cost of infusion, average price from fixed-price contracts	9675 SEK	322,079
Cost of infusion, average DRG price	5022 SEK	167,171
Cost of lost patient productivity and leisure time	0 SEK	116,305
Cost of lost patient productivity and leisure time	+100%	133,341

^aCost saving with fingolimod over natalizumab.

DRG, diagnosis-related groups; SEK, Swedish krona.

Discussion

The CHMP has concluded that the efficacy of fingolimod is broadly similar to that of natalizumab25, and has given both drugs the same EU licensed indication18,27. Therefore, a CMA rather than a cost-effectiveness analysis was performed in order to compare the two treatments. The TLV has previously determined that natalizumab is a cost-effective treatment in Sweden36; therefore, it was of interest to see how the costs associated with once-daily oral fingolimod treatment compare with those for natalizumab, which is administered by intravenous infusion every 4 weeks. The results of this analysis suggest that fingolimod is cost saving compared with natalizumab in a Swedish healthcare setting. Fingolimod was associated with savings of over 40,000 SEK/patient/year. Extrapolation of these cost savings per person to the total number of patients receiving second-line treatment with natalizumab in Sweden, estimated to be 1152 individuals in a recent post-marketing surveillance publication22, results in total savings of ∼47 million SEK/year. Moreover, according to the results for the 10-year analysis, the potential cost savings associated with fingolimod over natalizumab for the first year’s therapy are maintained, and thus year-on-year savings can be expected. Based on this analysis, the TLV has concluded that fingolimod is a cost-effective therapy and has granted reimbursement in Sweden.

Several procedures are recommended when starting these treatments. Initiation of fingolimod therapy is associated with transient decreases in heart rate; these symptoms are reported in 0.5% of patients and typically resolve within 1 month of treatment37. Close observation of patients for signs and symptoms of bradycardia for the first 6 h following the first dose is therefore recommended. A recent review of the risks and benefits of fingolimod by the CHMP has recommended that ECG monitoring should be performed before treatment and then continuously for the first 6 h after the first dose, and that blood pressure and heart rate should be measured before treatment and then hourly for the first 6 h after the first dose30. Monitoring should be extended for at least 2 h in patients whose heart rate is lowest 6 h after receiving the first dose of fingolimod30. In patients who develop clinically-significant heart problems such as brachycardia or atrioventricular, block monitoring should continue at least overnight and until the problems have resolved30. The costs associated with ECG surveillance are included in this analysis. Therefore, this CMA includes all relevant costs associated with both therapies. The costs of monitoring after the first fingolimod dose (6365 SEK) are small in comparison with the total cost saving of fingolimod over natalizumab treatment.

The route of drug administration accounts for the main difference in costs between fingolimod and natalizumab therapy, with the costs associated with infusions making up almost all of the additional costs of natalizumab over 3 years. Cost estimates for natalizumab infusions therefore have the greatest potential to influence the results of the CMA. As there were no published Swedish studies examining the cost of natalizumab infusions in isolation, the costs were calculated after exploring three different information sources: fixed-priced contracts; DRG code rates for outpatient intravenous drug delivery; and the cost of intravenous drug delivery in patients with MS included in the national cost-per-patient database. The first two strategies were not chosen for the analysis because the cost of infusion could not be distinguished from other (i.e., drug) costs and it was impossible to confirm the patient population in which the infusion was used. The third method of estimating infusion costs was used as it excludes drug costs and was based on a population confirmed to have MS. Moreover, this estimate was the most conservative of the three. Even using the lowest estimate for the cost of infusions in the national cost-per-patient database (1059 SEK), as in the sensitivity analysis, fingolimod treatment shows a cost saving over natalizumab therapy. This infusion cost may include the cost for treatment with mitoxantrone and human immunoglobulin, other infusional treatments for MS. However, these treatments are rarely used, so the effect of including them is likely to be negligible.

In addition to the costs of infusion, the other main cost associated with natalizumab over fingolimod is lost patient productivity and leisure time. Sensitivity analyses found that the cost saving with fingolimod remains robust to changes in estimates of unit costs for lost patient productivity and leisure time.

This study has some limitations. First, both drugs were assumed to have equivalent efficacy. This has not been demonstrated in a head-to-head clinical trial. However, both drugs have identical indications, which suggests that their efficacies are similar; based on the reimbursement submission, the TLV has also concluded that fingolimod and natalizumab have similar efficacies in the licensed EU patient population, and is thus in agreement with the CHMP. Secondly, costs of managing AEs for both drugs were assumed to be negligible. The main rationale for this assumption is that most serious AEs for both drugs are rare19,20,35. Therefore, even though costs may be high on an individual level, those for managing serious AEs were considered to be negligible on a population level in comparison to other expenses (such as the cost of infusions). Finally, the costs and resource utilization employed in this analysis are unlikely to be representative of all countries. Caution is therefore needed in extrapolating these results, but it is likely that the conclusions will be valid in countries other than Sweden which follow the same administration procedures and have similar cost structures.

Conclusion

This CMA showed that fingolimod is associated with savings of ∼40,000 SEK/patient/year vs natalizumab for the treatment of RRMS in Sweden. Based on these results, the TLV has concluded that fingolimod is a cost-effective therapy and has granted reimbursement.

Transparency

Declaration of funding

This study was funded by Novartis Sweden AB. The study was designed by Niklas Bergvall, Magnus Tambour, and Freddie Henriksson, and data were gathered and analyzed by Novartis. All authors contributed to the interpretation of the data, were involved in drafting and reviewing the manuscript, and approved the submitted paper.

Declaration of financial/other relationships

Professor Fredrikson has received honoraria for lectures, consultancy, and educational activities from Allergan, Bayer, Biogen Idec, Genzyme, Merck-Serono, Novartis, Sanofi-Aventis, and Teva. Dr Tambour is an employee of Novartis and reports having equity interests in Novartis. Drs Bergvall and Henriksson are employees of Novartis.

Acknowledgments

Dr Gemma Carter (Oxford PharmaGenesis, Ltd) provided medical writing support, editorial assistance, and collation and incorporation of comments from all authors. Such editorial help was funded by Novartis Pharma AG, Basel, Switzerland.

Previous presentation

The results of this analysis were presented as a poster at the 22nd meeting of the European Neurological Society (ENS) on 9–12 June 2012 in Prague, Czech Republic.