When does economic model type become a decisive factor in health technology appraisals? Learning from the expanding treatment options for relapsing–remitting multiple sclerosis

Figures & data

Table 1. Summary of scenarios presented in the Markov and DES models.

Intervention	Comparator	Model	Scenario
Fingolimod	DMF	Markov	Markov HA RRMS base case
		DES	DES HA RRMS base case, which includes PML as a long-term AE for fingolimod
	Natalizumab	Markov	Markov RES RRMS base case
		DES	DES RES RRMS base case, which includes PML as a long-term AE for fingolimod and natalizumab
		DES	PML removed from fingolimod and natalizumab in DES
	Alemtuzumab	Markov	Markov HA RRMS base case
		DES	DES HA RRMS base case, with inclusion of alemtuzumab relapse-triggered retreatment, and long-term AEs for both fingolimod and alemtuzumab
		DES	Retreatment of alemtuzumab removed in DES
		DES	Retreatment of alemtuzumab removed, and PML removed for fingolimod, and renal disorders and long-term thyroid function removed for alemtuzumab; length of all AEs set to match Markov

Abbreviations. AEs, adverse events; DES, discrete event simulation; DMF, dimethyl fumarate; HA, highly active; PML, progressive multifocal leukoencephalopathy; RES, rapidly evolving severe; RRMS, relapsing–remitting multiple sclerosis.

Table 2. Cohort level characteristics of the IPD used.

Characteristics	Value HA RRMS				Value RES RRMS
	DES		Markov		DES		Markov
Size of the population	1381		1000		528		1000
No. of cohort iterations	2000		–		2000		–
Number of patients in initial EDSS scores
0	82	5.94%	59.4	5.94%	39	7.39%	73.9	7.39%
1	118	8.54%	85.4	8.54%	55	10.42%	104.2	10.42%
2	456	33.02%	330.2	33.02%	203	38.45%	384.5	38.45%
3	309	22.38%	223.8	22.38%	105	19.89%	198.9	19.89%
4	268	19.41%	194.1	19.41%	82	15.53%	155.3	15.53%
5	101	7.31%	73.1	7.31%	26	4.92%	49.2	4.92%
6	47	3.40%	34.0	3.40%	18	3.41%	34.1	3.41%
7	0	0%	0	0%	0	0%	0	0%
8	0	0%	0	0%	0	0%	0	0%
9	0	0%	0	0%	0	0%	0	0%
Female to male ratio	2.8:1		2.8:1		2:1		2:1
Age of cohort (mean)	38.23		38.23		34.85		34.85
Years since diagnosis (mean)	6.42		6.42		4.61		4.61
Source	HA RRMS subgroup from FREEDOMS I, FREEDOMS II and TRANSFORMS		HA RRMS subgroup from FREEDOMS I, FREEDOMS II and TRANSFORMS		RES RRMS subgroup from FREEDOMS I, FREEDOMS II and TRANSFORMS		RES RRMS subgroup from FREEDOMS I, FREEDOMS II and TRANSFORMS

Abbreviations. DES, discrete event simulation; EDSS, expanded disability status scale; HA, highly active; IPD, individual patient data; RES, rapidly evolving severe.

Figure 1. The structures of the Markov and DES models. In the Markov model when patients are transferred to SPMS, their EDSS score increases by 1; this assumption was applied in the DES analyses to match the approach taken in both models. Note that in the models, it is possible for a patient to move between states that are more than one EDSS point apart. For example, a transfer from RRMS with EDSS score of 1 to EDSS score of 7 is possible, without going through all intermediate steps. The omission of these transition arrows is for the sake of clarity. Moreover, from each EDSS state patients can transfer to the death state.

Abbreviations. DES, discrete event simulation; EDSS, expanded disability status scale; QALYs, quality-adjusted life years; RRMS, relapsing-remitting multiple sclerosis; SPMS, secondary progressive multiple sclerosis.

Table 3. Efficacy inputs for the Markov and DES models.

DMT	Subgroup	Outcome vs. placebo (over 24 months)		Source
DMF	HA RRMS	ARR ratio	0.57	Pooled DEFINE and CONFIRM HA subgroups^40
		HR for 3 month confirmed disability progression	1.19
Fingolimod	HA RRMS	ARR ratio	0.52	In line with the previous Markov model in HA RRMS^22; this analysis has also been published in Derfuss et al.^41
		HR for 3 month confirmed disability progression	0.66
	RES RRMS	ARR ratio	0.43	FREEDOMS + FREEDOMS II pooled RES subgroups^23
		HR for 3 month confirmed disability progression (definition as per AFFIRM trial)	0.61
Natalizumab	RES RRMS	ARR ratio	0.19	AFFIRM treatment-naïve RES subgroup (Hutchinson et al.)^42
		HR for 3 month confirmed disability progression	0.47
Alemtuzumab	HA RRMS	ARR ratio	0.54	Calculated using Bucher indirect comparison (see Supplementary Table 1)
		HR for 3 month confirmed disability progression	0.58

Abbreviations. ARR, annualized relapse rate; DES, discrete event simulation; DMF, dimethyl fumarate; HA, highly active; HR, hazard ratio; RES, rapidly evolving severe; RRMS, relapsing–remitting multiple sclerosis.

Figure 2. Cost-effectiveness plane of deterministic results from comparison of DES and Markov models. All analyses were carried out with the intervention and comparators at their list price. A confidential PAS discount is available for fingolimod and DMF, this was not considered in the model. Values for each point are the incremental QALY and incremental cost for each comparison. Note: ‘No AEs’ indicates the removal of serious long-term AEs from the DES model which were not included in the Markov model (e.g. PML). ‘No retreatment’ indicates the removal of relapse-triggered retreatment for alemtuzumab in the DES model.

Abbreviations. AE, adverse events; ALE, alemtuzumab; DES, discrete event simulation; DMF, dimethyl fumarate; FIN, fingolimod; HA, highly active; ICER, incremental cost-effectiveness ratio; NAT, natalizumab; PML: progressive multifocal leukoencephalopathy; QALY, quality-adjusted life year; RES, rapidly evolving severe; RRMS, relapsing-remitting multiple sclerosis; WTP, willingness-to-pay.

European Medicines Agency. Tecfidera: EPAR – Public assessment report. Available at: http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Public_assessment_report/human/002601/WC500162070.pdf [Last accessed 12 March 2018]

 Google Scholar

Maruszczak MJ, Montgomery SM, Griffiths MJ, et al. Cost-utility of fingolimod compared with dimethyl fumarate in highly active relapsing–remitting multiple sclerosis (RRMS) in England. J Med Econ 2015;18:874-85

 PubMed Web of Science ®Google Scholar

Derfuss T, Bergvall NK, Sfikas N, et al. Efficacy of fingolimod in patients with highly active relapsing–remitting multiple sclerosis. Curr Med Res Opin 2015;31:1687-91

 PubMed Web of Science ®Google Scholar

Montgomery SM, Maruszczak MJ, Slater D, et al. A discrete event simulation to model the cost-utility of fingolimod and natalizumab in rapidly evolving severe relapsing–remitting multiple sclerosis in the UK. J Med Econ 2017;20:474-82

 PubMed Web of Science ®Google Scholar

Hutchinson M, Kappos L, Calabresi PA, et al. The efficacy of natalizumab in patients with relapsing multiple sclerosis: subgroup analyses of AFFIRM and SENTINEL. J Neurol 2009;256:405-15

 PubMed Web of Science ®Google Scholar