Ofatumumab for treating relapsing multiple sclerosis

3.1 The clinical and patient experts said that multiple sclerosis is a chronic, disabling neurological condition. The patient experts explained that symptoms of relapsing multiple sclerosis and the adverse effects from treatment can limit people's ability to work, and to engage in social and family life. The dosing frequency and monitoring needs of some treatments can disrupt people's lives and careers. The committee noted that ofatumumab is self-administered at home. It heard that a treatment that could be self-administered monthly is less disruptive to people's lives than treatments administered by intravenous infusions in hospital, so would be valued by people with multiple sclerosis.

3.2 The clinical experts said that ofatumumab would be offered as a first-line therapy or after other treatments. The committee agreed that, because ofatumumab is an anti‑CD20 monoclonal antibody, it should be in a similar position in the treatment pathway to ocrelizumab. Ocrelizumab is also an anti‑CD20 monoclonal antibody recommended first line or after other treatments. The clinical experts noted that there are no clear rules for sequencing of treatments. However, in practice, clinicians generally offer a different disease-modifying treatment for relapsing–remitting multiple sclerosis when the patient's current treatment no longer prevents disease relapses. The clinical experts said that pregnancy planning can also be an important factor in choosing to change treatments for relapsing–remitting multiple sclerosis. They also recommend stopping all treatments when people can no longer walk or when they develop secondary progressive multiple sclerosis. The committee concluded that ofatumumab was likely to be used first line or after other treatments in people who have active relapsing–remitting multiple sclerosis.

3.3 The company limited its submission to relapsing–remitting multiple sclerosis rather than all relapsing forms of multiple sclerosis, as specified in its proposed marketing authorisation. The clinical experts agreed that this was reasonable.

3.4 Various treatment options are available, with different methods and schedules of administration, and the committee noted that people have different preferences. The patient expert confirmed that ofatumumab could be self‑administered relatively easily monthly at home, after initial training from a health professional. The committee concluded that differences in dosing schedule, adverse effects and monitoring requirements between ofatumumab and other treatments may influence patient choice. It agreed that it is important to take these into account when making decisions about treatment.

3.5 The main evidence for the clinical effectiveness of ofatumumab compared with teriflunomide came from 2 trials, ASCLEPIOS I (n=927) and ASCLEPIOS II (n=955). These were phase 3 randomised controlled trials in adults with relapsing multiple sclerosis. The main purpose of these trials was to consider if patients who had ofatumumab had fewer relapses and slower disease progression compared with patients who had teriflunomide. Participants had had at least 1 relapse in the past year, 2 relapses in the last 2 years, or a positive gadolinium-enhancing MRI scan in the last year. Few patients in each trial were from the UK, but the clinical experts and ERG noted that there were no major concerns about the generalisability of the evidence. The committee accepted that the baseline characteristics of the patients in ASCLEPIOS I and II reflected people with relapsing–remitting multiple sclerosis having treatment in the NHS. It concluded that the results of the clinical trials were generalisable to NHS clinical practice.

3.6 The company carried out 2 post-hoc analyses of the ASCLEPIOS trials data for the 2 subgroups: patients with highly active disease or with rapidly evolving severe disease. These subgroups had been defined in the NICE scope for this appraisal. The highly active subgroup was defined as people in the ASCLEPIOS relapsing–remitting multiple sclerosis population who had previously had any disease-modifying treatment and stopped their last treatment because of lack of efficacy. The rapidly evolving severe subgroup were people with relapsing–remitting multiple sclerosis, who had had at least 2 relapses in the last year and at least one T1 gadolinium-enhancing lesion on baseline brain MRI. The company was not able to carry out a network meta-analysis for these subgroups because of a lack of published comparator trial data. The committee noted that there was limited evidence for these subgroups and that the company had included them because previous appraisals had considered them. The clinical experts said that clinicians are more likely to use categories that describe treatment and relapse history and, in practice, these subgroups would not be used. The committee concluded that the evidence was not robust enough for them to be considered separately, and the relapsing–remitting multiple sclerosis population would be considered as a whole in this appraisal.

3.7 The ASCLEPIOS trials showed that ofatumumab is more effective than teriflunomide for all main clinical outcomes and had no unexpected safety concerns. Ofatumumab reduced annualised relapse rate compared with teriflunomide with an annual relapse rate ratio of 0.50 (95% confidence interval 0.37 to 0.65) in ASCLEPIOS I and 0.42 (95% confidence interval 0.31 to 0.56) in ASCLEPIOS II. Fewer patients had confirmed disability worsening at 3 months and 6 months for ofatumumab compared with teriflunomide. The hazard ratio for ofatumumab compared with teriflunomide was 0.68 (95% confidence interval 0.50 to 0.92) for confirmed disability worsening at 6 months for the prespecified pooled ASCLEPIOS population (because both trials had the same design and were carried out at the same time). The committee concluded that ofatumumab was clinically effective compared with teriflunomide.

3.8 Because the company had direct comparative evidence only for ofatumumab and teriflunomide, it provided a network meta-analysis to estimate ofatumumab's effectiveness compared with the other comparators in the scope. The company chose 31 studies to inform its mixed treatment comparison for annualised relapse rates in the relapsing–remitting multiple sclerosis population. There was uncertainty in the results because most comparisons were informed by a single trial, and many of the comparators were indirectly compared with ofatumumab by 1 or more intermediate comparator. However, the committee concluded that there was a lower annualised relapse rate for ofatumumab in the whole population compared with most comparators, except the monoclonal antibodies (alemtuzumab, natalizumab and ocrelizumab) and cladribine.

3.9 The company chose 21 studies for confirmed disability worsening at 3 months and 20 studies for confirmed disability worsening at 6 months in its base-case analyses. The ASCLEPIOS trials used a different definition of disability worsening from the one commonly used in other trials, based on Expanded Disability Status Scale (EDSS) changes from baseline. The EDSS is used to measure how much someone is affected by their multiple sclerosis. To account for using a different definition, the company carried out additional analyses on the 3‑month and 6‑month data from the ASCLEPIOS trials using 'aligned criteria'. The company also carried out another analysis of the ASCLEPIOS trial data according to the methods set out in the protocol of OPERA trials, which were the pivotal trials for ocrelizumab in people with relapsing multiple sclerosis. The committee considered the evidence for confirmed disability progression at 3 months and 6 months. It noted that the point estimates for hazard ratios measuring ofatumumab against comparators for confirmed disability progression (aligned criteria) at 6 months were below 1 for the comparators except alemtuzumab, natalizumab and ocrelizumab, and that the 95% credible intervals crossed 1 for all the comparators except teriflunomide. This suggested that there was a statistically significant difference between treatment with ofatumumab compared with teriflunomide for most comparators except the other monoclonal antibodies. The committee concluded that ofatumumab may slow confirmed disability progression more than most comparators except the monoclonal antibodies alemtuzumab, natalizumab and ocrelizumab.

3.10 The company's model structure was similar to that of models used in previous appraisals of multiple sclerosis technologies. The model was a Markov transition model consisting of 21 health states based on the EDSS. The EDSS has 10 functional states, with higher numbers reflecting a greater functional impact. The company's model consisted of 10 EDSS states for relapsing–remitting multiple sclerosis, 10 states for secondary progressive multiple sclerosis and death. The model used natural history data from the British Columbia multiple sclerosis registry for transitions between relapsing–remitting multiple sclerosis health states. It used the London Ontario multiple sclerosis registry (as used in NICE's technology appraisal guidance on fingolimod for highly active relapsing–remitting multiple sclerosis) to model transitions from relapsing–remitting multiple sclerosis to secondary progressive multiple sclerosis health states. And it used data from the London Ontario registry supplemented by data from EXPAND (as used in NICE's technology appraisal guidance on siponimod for secondary progressive multiple sclerosis) as sources of natural history data between secondary progressive multiple sclerosis health states. The company sourced treatment effects for ofatumumab and all comparators from the company's network meta-analysis and applied them as:

3.11 The company's original base case included direct medical disease management costs for each EDSS status score (the system used to classify the severity and progression of multiple sclerosis). But it used the same disease management costs for each EDSS health state for both relapsing–remitting multiple sclerosis and secondary progressive multiple sclerosis. The ERG commented that the economic analysis should include different disease management costs for people with secondary progressive multiple sclerosis and relapsing–remitting multiple sclerosis to better reflect that the costs of management and care changes over time. After technical engagement, the company agreed and updated its base case in line with the ERG's preferred assumptions. The committee noted that including disease management costs specifically for people with secondary progressive multiple sclerosis had a minor impact on the incremental cost-effectiveness ratio (ICER). But it was satisfied that the company's updated base case better reflected the natural history of multiple sclerosis.

3.12 The values the company chose in its original base case show that for relapsing–remitting multiple sclerosis, there is a steady decrease in annual relapse rates. For secondary progressive multiple sclerosis, relapse happened more often at some higher EDSS scores than at some lower EDSS scores. The ERG preferred an alternative approach, decreasing annual relapse rates as EDSS levels increase. After technical engagement, the company agreed with the approach taken by the ERG and updated its base case accordingly. The committee heard from the clinical experts that, because of the natural course of multiple sclerosis, relapse rates were unlikely to increase as EDSS score increased. The committee concluded that the approach taken in the company's updated analyses was appropriate.

3.13 The company's economic model used the EDSS scale to show how people move between the different health states: from no disability from their multiple sclerosis to mild, moderate and severe disability. There were no placebo-arm data from ASCLEPIOS I and II to inform the probability of progressing from one health state to another. So the company used transition matrices from the British Columbia longitudinal multiple sclerosis dataset to model transitions between relapsing–remitting multiple sclerosis health states. The company and ERG used different transition matrices for progressing from relapsing–remitting multiple sclerosis to secondary progressive multiple sclerosis. The ERG preferred to use transition matrices from 1 analysis of the London Ontario multiple sclerosis dataset from NICE's technology appraisal guidance on peginterferon beta-1a for relapsing–remitting multiple sclerosis (from now on referred to as TA624). The company explained that it used a different analysis of the London Ontario dataset from NICE's technology appraisal guidance on fingolimod for highly active relapsing–remitting multiple sclerosis (from now on referred to as TA254) because the transition probabilities used by the London Ontario dataset from TA624 had not adjusted for active or benign forms of relapsing multiple sclerosis. The London Ontario dataset from TA254 excluded people with less progressive relapsing multiple sclerosis and therefore fully represented eligible people. The committee heard from the clinical experts that it was uncommon for people with secondary progressive multiple sclerosis disease symptoms to improve, but noted that the alternative sources had a minimal impact on cost effectiveness. The committee concluded that in this case, the probability of progressing to secondary progressive multiple sclerosis should take account of only active forms of relapsing multiple sclerosis.

3.14 If direct trial data were not available for health state utility values, the company used data from alternative sources. In the secondary progressive multiple sclerosis health states, data from the EXPAND trial (a trial including people with secondary progressive multiple sclerosis) were supplemented by data from Orme et al. (2007) to inform health state utility values for an EDSS status of 7 or more. The company's rationale for this was that values taken from the EXPAND trial consistently decrease with each progressive EDSS state, and this aligns with a clinical expectation of reduced quality of life with disability progression. The ERG preferred to use the data taken directly from Orme et al. (2007). But the company noted that utility values at EDSS health state 3 were lower than at EDSS health state 4, suggesting a better quality of life in people with greater disability. The committee noted that there were negative utility values at higher EDSS states. The clinical experts said that this can be because of the limitations people experience as their multiple sclerosis disability progresses. But overall the committee concluded that as disability progresses, quality of life will reduce.

3.15 The company assumed in its base case that the treatment effect of ofatumumab and its comparators did not wane over time, but that any waning in the model would be captured by all‑cause discontinuation (stopping for any reason, including perceived lack of efficacy). In the company's response to technical engagement, it analysed the outcomes of disease worsening and rates of relapse considered in the ASCLEPIOS trials. This showed that over the 27‑month data period, there was no evidence of waning of treatment effect. The ERG was satisfied that the analyses carried out by the company showed that there was no evidence of treatment effect waning in the ASCLEPIOS trials. But because the data were short term, the ERG suggested it would be appropriate to assume a waning of treatment effect for all disease-modifying multiple sclerosis treatments, as seen in previous multiple sclerosis appraisals. The company provided 2 scenario analyses considering the waning of treatment effect:

3.16 In its response to technical engagement, the company maintained that it did not think it was valid to include treatment effect waning in the base case. The ERG viewed including treatment effect waning as a precaution, with the expectation that the effect of all multiple sclerosis treatments was likely to wane over time eventually. This was consistent with some other multiple sclerosis appraisals. For this reason, the ERG preferred to use a conservative assumption of waning, with a 25% reduction in effectiveness after 5 years, and a 50% reduction after 8 years. The committee considered both approaches as well as considering the approaches taken in other technology appraisals for disease-modifying multiple sclerosis treatments. It noted that there had been no clear consistent approach to waning in the technology appraisals of other disease-modifying multiple sclerosis treatments.

3.17 The network meta-analyses showed that ofatumumab had similar efficacy to other monoclonal antibodies. The clinical experts said that because monoclonal antibodies generally have a higher efficacy than other drugs for relapsing multiple sclerosis, they would be expected to have less waning of treatment effect over time. The committee noted that discontinuation and waning were connected, but also noted that the ASCLEPIOS trials were too short to predict long-term discontinuation rates. The clinical experts explained that other monoclonal antibodies such as natalizumab had maintained efficacy over a long period of time. They said that all-cause discontinuation showed that people may choose to discontinue for reasons other than efficacy waning (for example, treatment fatigue or pregnancy). The committee recognised that, because ofatumumab and ocrelizumab were both anti‑CD20 monoclonal antibodies, it was reasonable to consider ocrelizumab the closest comparator to ofatumumab. The committee noted that in NICE's technology appraisal guidance on ocrelizumab for relapsing–remitting multiple sclerosis, all-cause treatment discontinuation was accepted as a proxy for treatment waning in the absence of evidence for a waning effect. But it also noted that scenarios that included waning were also relevant for consideration. The committee recognised that this was a difficult area with limited data and concluded that in this case treatment discontinuation could be considered a proxy for waning.

3.18 NICE's guide to the methods of technology appraisal notes that, above a most plausible ICER of £20,000 per quality-adjusted life year (QALY) gained, judgements about the acceptability of a technology as an effective use of NHS resources will take into account the degree of certainty around the ICER. The committee will be more cautious about recommending a technology if it is less certain about the ICERs presented. The committee considered the probabilistic and deterministic cost-effectiveness estimates presented by the company and ERG. Its preferred assumptions included: