Cost-Effectiveness of Fingolimod in Treating Patients With Relapsing-Remitting Multiple Sclerosis

Multiple sclerosis is a chronic, neurodegenerative, immune-mediated disease of the central nervous system that affects approximately 400,000 individuals in the United States and 2.1 million people worldwide.^1,2 It is a lifelong condition with an onset generally occurring in early adulthood (20-40 years of age)^3,4 characterized by inflammation, demyelination, and axonal degeneration,^5-7 resulting in progressive disability.⁸

Approximately 85% of patients have relapsing-remitting multiple sclerosis (RRMS) at diagnosis, making it the most common form of MS.⁹ The clinical pattern of RRMS includes repeated episodes of inflammation and significant symptom worsening (“relapses” or “exacerbations”) interspersed between periods of partial or full recovery (“remission”). The number of early relapses, along with the time between first and second relapse (interattack interval), has been found to be predictive of long-term disability outcomes. In a study of RRMS patients in the London, Ontario, natural history cohort, patients with a higher number of relapses and a shorter interattack interval early in the disease were more likely to become more highly disabled, as measured by the Expanded Disability Status Scale.¹⁰

Overall, MS has a significant and devastating personal and economic burden due to the early age of onset and accumulation of physical disability.^11,12 In managed care organizations, the accumulated medical costs for MS patients are more than 2 to 3 times greater than those for all enrollees.¹³ MS costs result from relapses and natural progression of disability and neurologic compromise.¹⁴ Kobelt and colleagues, who surveyed 1909 MS patients in the United States in 2003-2004 to investigate the cost of MS treatment, estimated the total average cost to be $47,215 per patient per year (2004 costs).¹⁵ The number and severity of relapses are significant predictors of total cost of care for managed care organizations.¹⁶ O’Brien et al determined the cost of healthcare services required for relapses of low, medium, or high intensity using data from a number of sources including inpatient resource utilization databases in California, Florida, Maryland, Massachusetts, and Washington, and the 2002 Medicare national physician fee schedules.¹⁷ The authors found that the cost of treating relapses ranged from $333 for low-intensity relapses to $17,617 for high-intensity relapses.¹⁷ Therefore, reducing both severity and frequency of relapses may decrease overall healthcare costs of MS.

Although there is no cure for MS, the goals of treatment are to decrease relapses and slow the progression of disability.³ An ideal agent should also provide maximal efficacy, safety, tolerability, and convenience.¹⁸ Until recently, the 5 Food and Drug Administration (FDA)-approved disease-modifying therapies (DMTs) generally used first line included 4 interferon-beta (IFNß) type-1 products: intramuscular (IM) IFNß-1a (Avonex), subcutaneous (SC) IFNß-1a (Rebif), SC IFNß-1b (Betaseron), and SC IFNß-1b (Extavia); and glatiramer acetate (Copaxone). These immunomodulatory agents have been demonstrated in randomized controlled trials and long-term extension studies to improve disease outcomes in patients with MS, through reduced relapse rates and slowed disability progression.

However, there remain unmet needs in the treatment of RRMS. Current first-line DMTs have only moderate efficacy in patients with RRMS, reducing relapse rates by up to 34% versus placebo.¹⁹ As shown in

Table 1

, in placebo-controlled studies, IM IFNb-1a reduced relapse rates by 18%. The SC DMTs have improved efficacy in reducing relapse, with a 32% relapse rate reduction for SC IFNß-1a and a 34% relapse rate reduction with SC IFNß-1b (Table 1). Nonadherence with DMTs also contributes to gaps with current therapy (discontinuation ranging from 5% to 50% in clinical trials).^19,20,30,31 In the first 6 months of therapy, one-third of patients discontinue therapy due to perceived lack of efficacy, injection-site reactions, and other side effects including flu-like reactions, depression, and fatigue.^32,33 As the IFNs and glatiramer are administered parenterally, injection burden and tolerability may also contribute to poor adherence.

Fingolimod (Gilenya), a once-daily oral DMT indicated for treatment of patients with relapsing forms of MS to reduce the frequency of clinical relapses and delay accumulation of physical disability, was approved by the FDA in September 2010. The efficacy and safety of fingolimod were established in placebo-controlled and active-controlled 2- and 1-year trials. In the FTY720 Research Evaluating Effects of Daily Oral therapy in Multiple Sclerosis (FREEDOMS) study, clinical results showed fingolimod 0.5 mg had a significantly lower annualized relapse rate (ARR) than placebo (0.18 vs 0.40; P <.001), for a relative reduction of 54%. Results also indicated that a significantly higher percentage of patients treated with fingolimod 0.5 mg were without relapse after 24 months of therapy compared with placebo (70% vs 46%; P <.001), and patients had a lower risk of disability progression (88% vs 81%; P = .03).³⁴ Results of the multicenter, randomized, double-blind, 1-year Trial Assessing Injectable Interferon versus FTY720 Oral in Relapsing-Remitting Multiple Sclerosis (TRANSFORMS) study demonstrated a 52% relative reduction in ARR for fingolimod 0.5 mg compared with weekly IM IFNβ-1a (P <.001).²¹

Fingolimod was generally well tolerated in its clinical trial program. The most frequent adverse reactions in clinical trials (incidence >10% and greater than placebo) were headache, influenza, diarrhea, back pain, liver enzyme elevations, and cough. Bradycardia has been observed with fingolimod administration, necessitating a baseline electrocardiogram in patients at higher risk for bradyarrhythmia and in-office monitoring for all patients receiving their first dose of fingolimod.²²

When evaluating innovative therapies where resources may be limited, it is important to consider cost-effectiveness of treatment. However, no cost-effectiveness data are currently published on the use of fingolimod in patients with RRMS. The objective of this analysis was to estimate the cost per relapse avoided with oral fingolimod 0.5 mg compared with other first-line DMTs in a hypothetical managed care plan from a US commercial payer perspective.

METHODSModel Design

An economic model was developed in Microsoft Excel to estimate costs of treating MS with DMTs in a managed care plan. The model population included prevalent and incident RRMS patients; comparators were first-line DMTs: fingolimod, SC IFNβ-1a, SC IFNβ-1b (Extavia and Betaseron), IM IFNβ-1a, and glatiramer. Because the frequency of relapses has a high impact on patient burden, medical care, and resources, cost per relapse avoided for each DMT was selected as the relevant outcome of interest.

Model Inputs

Patient Population. The patient population was RRMS patients who were candidates for DMT and enrolled in or covered by a hypothetical health plan of 1,000,000 members.

Relapses. The expected number of relapses over a 2-year period in treated patients was determined by comparing the rate of relapse in untreated patients with the rate in patients treated with each of the DMTs. The 2-year time period was chosen because most clinical trials of DMTs reported relapse rates over 2 years. The relapse rate for untreated patients was based on a weighted average of placebo-treated patients in each of the respective phase 3 studies (2.58 relapses per 2 years).²⁰ The relative risk reduction in relapses was determined for each comparator from respective clinical trials. Using these inputs, the number of relapses avoided was calculated as the relative risk reduction multiplied by 2.58 relapses in an untreated patient; the number of relapses that could be expected to occur for each DMT over 2 years was 2.58 minus the number of relapses avoided (Table 1).

The cost of each relapse was based on the cost of managing the relapse as well as the intensity of the relapse. The direct cost of low-, moderate-, and high-intensity relapse management was obtained from O’Brien et al.¹⁷ Low-intensity relapses were defined as those requiring a physician office visit, treatment with symptom-related medication, and a follow-up visit. Medium-intensity relapses were defined as those requiring a physician office or emergency department visit, treatment with intravenous methyl prednisone and symptom-related medications, a follow-up office visit, or consults with a physical, occupational, or speech therapist. High-intensity relapses were defined as those requiring a physician office or emergency department visit, treatment requiring a full hospital admission, and postdischarge services including outpatient follow-up, rehabilitation, home healthcare, skilled nursing facility care, and short-stay nursing home care, or hospital readmission within 30 days. The relative incidence of each relapse severity was obtained from the Evidence for Interferon Dose-response: European North American Comparative Efficacy (EVIDENCE) trial (Table 1).²⁰ The weighted average cost of a direct relapse was estimated to be $4972.41, ranging from $332.63 for low-intensity relapses to $17,617.27 for high-intensity relapses. All costs were inflated to 2010 dollars using the consumer price index for the first half of 2010. The model did not discount costs, nor did it include indirect costs.

Treatment Costs. Acquisition costs for DMT were obtained from AnalySource, a Web-based drug pricing software tool (August 2011 wholesale acquisition costs).³⁵ In order to obtain the net drug acquisition cost for each comparator, the prescribing information was used to determine the number of doses per package and number of packages necessary per year per treated patient. This model assumed no product rebates, discounts, or patient copays or coinsurances. The model assumed that all patients were 100% adherent to treatment. The net yearly cost of therapy was calculated to be $48,081.45 for fingolimod, $36,755.55 for SC IFNβ-1b (Extavia), $40,311.24 for SC IFNβ-1b (Betaseron), $41,708.55 for glatiramer, $39,806.00 for IM IFNβ-1a, and $40,184.00 for SC IFNβ-1a. Annual treatment costs were doubled to account for the 2-year time horizon.

Monitoring Costs. Because monitoring requirements differ based on whether a patient is a newly diagnosed (incident) or already treated (prevalent) patient, monitoring costs for each patient population were calculated separately (annual rates and costs of monitoring are shown in

Table 2

). Monitoring requirements based on the agents’ respective product prescribing information included complete blood counts, thyroid panels, and liver function tests. The frequency of neurologist visits for an established patient was obtained from a cross-sectional survey conducted in 121 patients with MS.³⁸ Based on the recommendation from prescribing information, all patients new to fingolimod were assumed to undergo a 6-hour in-office observation for signs or symptoms of bradycardia after the first dose.²² The model assumed that all patients at high risk for bradycardia or atrioventricular block who were new to fingolimod had an electrocardiogram in the last 6 months; therefore, the cost of this test was not included in the model.²² As the prescribing information also recommends that a recent (ie, within 6 months) complete blood count and transaminase and bilirubin levels be available before initiating therapy,²² the model assumed all patients new to fingolimod require 1 complete blood count and 1 liver function test.

The unit cost for each parameter was obtained from the 2010 Physician’s Fee and Coding Guide³⁶; total monitoring costs were calculated as the unit cost of the monitoring parameter multiplied by the annual test frequency. The cost of other potential monitoring requirements for high-risk patients, such as ophthalmologist visits, dermatology visits, and pregnancy tests, were not included in the model, nor were tests commonly required for MS diagnosis. The monitoring costs were calculated for the 2-year time horizon, accounting for patients who were new to therapy and on existing therapy each year.

TRANSFORMS Subanalysis

The TRANSFORMS study was a 12-month, phase 3, multicenter, randomized, double-blind, double-dummy, parallel group study of 1292 RRMS patients comparing fingolimod with IM IFNβ-1a.²¹ Data from this study were used to compare the head-to-head cost per relapse avoided for patients treated with each agent. The annual number of relapses expected for patients prior to treatment with fingolimod or IM IFNβ-1a, 0.77 per year, was obtained from a real-world observational retrospective analysis.²³ Using the results from TRANSFORMS, the ARR was 0.16 for patients treated with fingolimod and 0.33 for patients treated with IM IFNβ-1a (P <.001).²¹ Thus, over a 1-year period, treatment with fingolimod avoided 0.61 relapses and treatment with IM IFNβ-1a avoided 0.44 relapses.

A post hoc analysis of TRANSFORMS evaluated the intensity of relapses with fingolimod and IM IFNβ-1a.²⁴ Relapses were categorized based on associated resource utilization: no steroid treatment or hospitalization (low intensity), steroid treatment but no hospitalization (moderate intensity), and hospitalization required (high intensity). The proportion of patients who experienced high-intensity relapses was lower with fingolimod than with IM IFNβ-1a. In addition, the proportion of patients who experienced relapses requiring steroid treatment but not hospitalization was also lower with fingolimod compared with IM IFNβ-1a (Table 1).²⁴ The cost of managing relapses of varying severity was obtained from O’Brien et al (Table 1).¹⁷

Sensitivity Analyses

In order to determine the impact of varying model inputs on cost per relapse avoided with fingolimod, univariate analyses were performed. Each model input was varied ±10% around the base-case value while keeping other inputs constant to determine the effect on cost per relapse avoided. A sensitivity analysis of the TRANSFORMS subanalysis was also performed. As in the general analysis, a univariate analysis was performed; each input was varied by ±10% around the base-case value while keeping the other inputs constant.

RESULTS

The model determined the cost per relapse avoided for each comparator. Although the 2-year total cost of therapy was highest for fingolimod at $104,271, the cost per relapse avoided was lowest at $74,843. The DMT with the next lowest cost per relapse avoided was SC IFNβ-1b (Extavia) at $94,423 per relapse avoided (total 2-year cost of therapy: $82,827.57), or $19,580 more per relapse avoided than fingolimod. The highest cost per relapse avoided was more than $197,000 per relapse avoided for IM IFNβ-1a (total 2-year cost of therapy: $91,520.63). The costs per relapse avoided are shown in

Table 3

and

Figure 1

.

TRANSFORMS Subanalysis

Results from the TRANSFORMS subanalysis indicate that based on the ARR of fingolimod and IM IFNβ-1a, and the differential intensity of relapses, average annual cost of a relapse was $582 for patients on fingolimod and $1864 for patients on IM IFNβ-1a. Although total cost of therapy was higher for fingolimod than for IM IFNβ-1a, total cost per relapse avoided was lower for fingolimod ($82,016) than for IM IFNβ-1a ($96,282) (Figure 1). Intramuscular IFNβ-1a cost $14,266 more per relapse avoided than fingolimod.

Sensitivity Analyses

Sensitivity analyses indicate that the model inputs with greatest impact on cost per relapse avoided were the fingolimod drug acquisition costs and the average number of relapses in untreated patients. For example, by decreasing the average number of relapses in untreated patients by 10% (from 2.58 to 2.32), the average cost per relapse avoided over 2 years would increase by 10.6%, or from $74,843 to $82,756 per relapse avoided. The total cost of managing a relapse had a negligible effect on the cost per relapse avoided: increasing the weighted cost of managing a relapse by 10% (from $4972 to $5469) would result in an increase of $424 to the cost per relapse avoided (from $74,843 to $75,267). Results of the sensitivity analysis are displayed in

Figure 2

.

To determine the sensitivity of the cost-effectiveness of fingolimod and IM IFNβ-1a in the TRANSFORMS subanalysis, a univariate analysis was conducted on the cost per relapse avoided for fingolimod. Results indicated that inputs with the greatest impact were the number of annual relapses before treatment and the ARR for fingolimod (Figure 2). Decreasing the number of annual relapses prior to treatment by 10% (from 0.77 to 0.69 relapses per year) increased the cost per relapse avoided by 15.1% to $94,396.

DISCUSSION

Multiple sclerosis poses a significant physical and social burden on patients, due to the early age of onset of this lifelong disease. The economic burden imposed by this disease is substantial: for managed care organizations, MS patients incur high medical costs¹³ largely due to the management of relapses, accumulation of disability, and the neurologic compromise they face.¹⁴ Two key predictors for total cost of care for MS patients are the level of disability and number of relapses.^15,16 In a study of MS patients treated with DMTs, Kobelt and colleagues found that the mean annual cost of an MS patient with severe disability was $64,492 (2004 dollars), nearly twice that of the mean annual cost for an MS patient with mild disability ($32,297).¹⁵ Reducing the number of relapses a patient experiences may result in a subsequent reduction in total cost of care for that patient.

Making effective, safe, and well-tolerated oral medications available to MS patients has the potential to reduce the annual number of relapses a patient experiences while also improving adherence with therapy. Oral fingolimod is a novel agent with a unique dual mechanism of action that has the potential to fulfill the currently unmet need for an effective yet convenient treatment for MS. In patients with RRMS, oral fingolimod has demonstrated superior efficacy compared with placebo and IM IFNβ-1a, a current first-line DMT.^21,34 The results of this economic model suggest that fingolimod is also cost-effective as a first-line treatment for RRMS. The average 2-year cost per relapse avoided in patients treated with fingolimod was $74,843, approximately $20,000 less than the next closest comparator. Using the head-to-head TRANSFORMS data, the 1-year cost per relapse avoided was $82,016 for fingolimod compared with $96,282 for IM IFNβ-1a. The difference in the cost-effectiveness of fingolimod changes substantially when using the head-to-head TRANSFORMS data, largely due to differences in the IM IFNβ-1a relapse rates between the placebo-controlled trial and TRANSFORMS. In the fingolimod placebo-controlled trial (FREEDOMS), the annual relapse rates were 0.18 for fingolimod and 0.40 for placebo.³⁴ In the IM IFNβ-1a placebo-controlled trial, the annual relapse rates were 0.67 for IM IFNβ-1a and 0.82 for placebo.²⁸ The relapse rate differences are thought to be due to differences in the MS patient population and disease severity (ie, Expanded Disability Status Scale scores) between the placebo-controlled and active-controlled trials. The IM IFNβ-1a placebo-controlled trial was conducted in the 1990s; since then, substantial changes in the patient population and treatment patterns have occurred in comparison with the more recently conducted fingolimod trials. This may explain the substantial differences in annual relapse rates versus placebo from the IM IFNβ-1a placebo-controlled trial to the fingolimod placebo-controlled trial, which drive the large cost per relapse difference in the placebo model comparisons.

Results of the sensitivity analyses indicate that the inputs that had the highest impact on the cost per relapse avoided were those that differentially impacted the treatment arms, including drug acquisition costs and average number of relapses expected for an untreated patient.

This model had a number of limitations. In general, modeling requires a variety of assumptions regarding the disease of interest, treatment patterns, and costs, and the results presented here are simply an estimate of the cost per relapse avoided of using fingolimod to treat RRMS. This model assumed 100% adherence to DMTs—much higher than rates reported in the real world.³⁹ While such real-world rates of adherence and persistence are not likely, no data are currently available on adherence or persistence with each DMT, including fingolimod. The impact of poor adherence or persistence on the effectiveness of therapy in MS is also not well known. As the IFNs and glatiramer commonly used for treating MS are administered parenterally 1 to 7 times weekly, injection burden and tolerability may contribute to a number of potential barriers to adherence. Because fingolimod is the only oral DMT currently available for treatment of RRMS, it is possible that adherence to this treatment will be improved compared with other treatments. Studies have shown that one-third of MS patients discontinue therapy within the first 6 months due to perceived lack of efficacy, injection-site reactions, and other side effects of therapy including flu-like reactions, depression, and fatigue.^32,33 To avoid biasing the results, a conservative estimate of complete adherence to all DMTs was assumed.

The model does not consider the impact or cost of treating adverse events due to DMTs because most adverse events seen in clinical trials were generally mild and required minimal or no medical interventions. Serious events that would potentially have a financial impact on managed care organizations, such as hepatotoxicity, serious infections, progressive multifocal leukoencephalopathy, and arrhythmias (due to bradycardia or atrioventricular block), are rare.

Although the level of disability increases in RRMS as the disease progresses and is also a significant contributor to total cost of care, the impact of disability progression was not included in the model. This is because the model evaluated the costs of care for an MS patient over a short 2-year horizon. Health plan costs of disability progression are generally incurred over a long time period (10+ years). As such, the true impact of DMT on disease progression and associated costs would not be captured accurately during the model time frame. Finally, the model is largely based on placebo comparators. With the exception of the head-to-head TRANSFORMS trial comparing fingolimod to IM IFNβ-1a, the relapse rates in the model are based on placebo-controlled trials.

The results of this model indicate that fingolimod is a cost-effective DMT, with the lowest cost per relapse avoided compared with other first-line DMTs. Although the total cost of fingolimod is the highest, its greater efficacy in terms of ARR results in the lowest total cost per relapse avoided of the 6 first-line DMTs included in the model. In addition to having a positive impact on the cost of treating RRMS patients, decreasing the number of relapses a patient experiences may also slow the rate of disability progression. As such, fingolimod as a first-line therapy for the treatment of RRMS has the potential to have a positive benefit for both payers and patients.