Filgotinib for treating moderate to severe rheumatoid arthritis

3.1 The patient expert explained that rheumatoid arthritis is a lifetime condition that has a large effect on mental and physical health and emotional wellbeing, causing fear, anxiety, stress, pain and fatigue. It can severely reduce quality of life and affect ability to work, everyday activities and relationships with children and other family members. The clinical expert stated that conventional disease-modifying antirheumatic drugs (DMARDs) such as methotrexate are inadequate for many people with active rheumatoid arthritis. Although a range of biological and targeted synthetic DMARDs are available for severe rheumatoid arthritis (see section 3.2), none of these treatments are currently available for people with moderate disease activity. Patient experts explained that currently people with moderate disease activity that has not responded adequately to conventional DMARDs have no effective treatment options. They feel that their disease needs to get worse before they can be offered effective treatments. They explained that progression in rheumatoid arthritis is relentless if not adequately treated. The clinical expert also added that for a significant proportion of people with severe disease who are eligible for treatment with biological DMARDs, their disease responds inadequately to these treatments, or they cannot tolerate the treatment. Both the clinical and patient experts said it would be helpful to have new treatments for various points in the treatment pathway. Clinical and patient experts also said that an oral drug taken daily may be preferable, especially for patients who are needle phobic or who have a significant hand disability. The committee concluded that a range of treatment options was important in rheumatoid arthritis and that filgotinib would be a welcome additional option, especially for moderate disease.

3.2 Disease severity is assessed using the disease activity score (DAS28). A DAS28 of more than 5.1 indicates severe disease, between 3.2 and 5.1 indicates moderate disease, between 2.6 and 3.2 indicates mild disease, and 2.6 or less indicates disease remission. NICE technology appraisal guidance recommends the following biological and targeted synthetic DMARDs, all with methotrexate, for severe active rheumatoid arthritis that has responded inadequately to:

3.3 Filgotinib's marketing authorisation covers its use in people with moderate rheumatoid arthritis whose disease has responded inadequately to 1 or more conventional DMARDs. However, the company's submission covers filgotinib's use in moderate rheumatoid arthritis for people whose disease has responded inadequately to 2 or more conventional DMARDs. The committee agreed with the company's positioning of filgotinib in moderate disease. It noted such positioning is aligned with the use of biologic and targeted synthetic DMARDs in severe disease. The clinical expert explained that people whose disease has an inadequate response to 2 or more conventional DMARDs are usually offered continued treatment with the same conventional DMARDs. Corticosteroids can be used to manage disease flares. The committee concluded that the appropriate position for filgotinib in moderate disease is after inadequate response to 2 or more conventional DMARDs. It also agreed that the relevant comparator for this population is best supportive care, consisting of previously used conventional DMARDs with optional corticosteroids.

3.4 Filgotinib's marketing authorisation and the company's submission cover its use at all 4 points in the severe disease treatment pathway for which other biologic and targeted synthetic DMARDs are recommended (see section 3.2). The committee agreed with this positioning. It noted that the marketing authorisation includes the use of filgotinib alone or with methotrexate. The committee agreed that all treatments listed in section 3.2, all used with methotrexate, are relevant comparators for filgotinib with methotrexate, when used at the same position in the treatment pathway. For people who cannot have methotrexate, relevant comparators for filgotinib monotherapy are adalimumab, baricitinib, certolizumab pegol, etanercept, tofacitinib, sarilumab and tocilizumab, depending on the position in the treatment pathway.

3.5 The company's clinical evidence came from 2 randomised controlled trials in people with moderate to severe rheumatoid arthritis:

3.6 In FINCH1, filgotinib with methotrexate showed a statistically significant improvement in the primary end point, American College of Rheumatology responses (ACR20) at 12 weeks, compared with adalimumab with methotrexate or placebo with methotrexate (76.6% compared with 70.5% and 49.9%, respectively, p<0.05 for both comparisons). Filgotinib also showed improvement in key secondary endpoints at both 12 and 24 weeks, including ACR50, ACR70 or EULAR responses. The committee concluded that filgotinib with methotrexate was more clinically effective than adalimumab with methotrexate or placebo with methotrexate in people with moderate to severe disease that has responded inadequately to conventional DMARDs.

3.7 In FINCH2, filgotinib with conventional DMARDs showed a statistically significant improvement in the primary outcome, ACR20 at 12 weeks, compared with placebo with conventional DMARDs (66.0% compared with 31.1%, p<0.05). Filgotinib also showed improvement in key secondary endpoints at both 12 and 24 weeks, including ACR50, ACR70 or EULAR responses. The committee concluded that filgotinib with conventional DMARDs was more clinically effective than placebo with conventional DMARDs in people with moderate to severe disease that has responded inadequately to biological DMARDs.

3.8 FINCH1 and FINCH2 trials included filgotinib only with methotrexate or with conventional DMARDs, respectively. Therefore, no clinical efficacy data are available for filgotinib monotherapy in people with moderate to severe disease that has responded inadequately to conventional or biological DMARDs. The clinical expert explained that in the FINCH2 trial, which enrolled people who had not previously had methotrexate (that is, methotrexate-naive population), filgotinib monotherapy showed improved clinical outcomes compared with placebo. The committee noted that all biological and targeted synthetic DMARDs are recommended with methotrexate, unless methotrexate is contraindicated. This is because combination therapy is thought to be more clinically effective than monotherapy. The committee concluded that the clinical efficacy of filgotinib monotherapy is uncertain because there is no clinical trial data in the target population.

3.9 A direct comparison was only possible with adalimumab and placebo, informed by FINCH1 and FINCH2 trials. To compare with other biological and targeted synthetic DMARDs, the company did 2 network meta-analyses for:

3.10 Although the network meta-analysis was used for decision making for people with severe disease (see section 3.8), the technical team noted that for moderate disease it may be more appropriate to use FINCH1 trial data because:

3.11 The ERG explained that the FINCH1 trial enrolled people who had had 1 or more conventional DMARDs, and that about half the patients with moderate disease had only had 1 previous conventional DMARD. Therefore, FINCH1 data may not be generalisable to the target population (that is, after 2 or more previous conventional DMARDs). In response to technical engagement, the company provided pairwise comparisons of clinical efficacy data for patients with moderate disease who had had 1 previous conventional DMARD compared with those who had had 2 or more previous conventional DMARDs. The company highlighted that these are exploratory post hoc analyses based on small patient numbers, and FINCH1 was not powered for such a comparison. However, the number of previous conventional DMARDs did not appear to have any notable effect on clinical efficacy estimates. The company also provided exploratory cost-effectiveness analyses for the population who had had 2 or more previous DMARDs. The committee considered all evidence provided by the company and concluded that using the overall moderate population from FINCH1 is more appropriate for decision making. It noted that this is preferred to using small post hoc subgroup data.

3.12 The revised company submission used direct head-to-head trial data to model the efficacy of filgotinib (see section 3.10). The FINCH1 trial collected EULAR response data. However, the company mapped the EULAR responses from ACR responses. The ERG explained this approach was aligned with the approach taken for the severe population, but noted it preferred to use the EULAR responses from FINCH1 directly in the model. This is because using direct data gives more precise estimates of clinical efficacy than using mapped values. The committee agreed with the ERG that EULAR response should be used directly in the model, instead of the mapped values.

3.13 The revised company base case modelled the efficacy of best supportive care based on the response rates seen in the placebo plus methotrexate arm of the FINCH1 trial. The clinical expert explained that best supportive care is not expected to give a EULAR response in clinical practice. However, the committee noted that a considerable response rate was seen in the placebo plus methotrexate arm of the FINCH1 trial, as well as in other clinical trials in rheumatoid arthritis. It noted that this response could have been caused by several factors, including a placebo effect, disease resolving naturally over time, regression to the mean, response bias and variation in symptoms. Some of these factors might have also contributed to the response to filgotinib in the FINCH1 trial. Therefore, the committee agreed it would not be appropriate to assume full clinical efficacy for filgotinib while assuming no response to best supportive care. It agreed with revised company analyses, which used FINCH1 response rates for both filgotinib with methotrexate and placebo plus methotrexate (a proxy for best supportive care). However, it acknowledged that these analyses had limitations because they did not fully reflect what is expected to happen in clinical practice.

3.14 Rheumatoid arthritis is a heterogenous disease and treatment choices are influenced by many factors (see section 3.1). Because of the large number of possible treatment sequences, it was not practical to model them all. However, the clinical expert confirmed that the company model included the most relevant comparators and treatment sequences that are used in NHS clinical practice. One exception to that, noted by both clinical and patient experts, was that further advanced therapies would be used instead of best supportive care in clinical practice. The committee acknowledged this as a limitation but noted that this approach was aligned with previous NICE technology appraisals. It also noted that this is likely to have a limited effect on the cost-effectiveness estimates in severe disease, but could be important to consider for the moderate population in the treatment sequence upon progression to severe disease (see section 3.16).

3.15 In the revised company base case, the company used patients' mean baseline DAS28 and expected DAS28 trajectory, to estimate patients' progression from moderate to severe disease. Using this approach, the modelled progression rate with best supportive care was 11% at 2 years and 39% at 5 years. The clinical expert mentioned 1 study that reported 5% progression rate at 1 year. Another study (ERAN database) reported that 19% of people with moderate disease activity 1 year after diagnosis had severe disease activity at a 2‑year visit. The committee noted this estimate may be uncertain because of small patient numbers in the registry and single assessment of disease activity at both timepoints (so results could be subject to temporary fluctuation in disease activity, including flares). It also noted no data were available to inform long-term progression rates. The clinical expert highlighted that although published data on the progression rates are lacking, the rates modelled by the company seem reasonable. The committee discussed that some patients could start treatment for severe disease when they have a flare that temporarily increases their disease activity to a severe level. This could mean that the initiation of severe treatment sequences in NHS clinical practice is higher than modelled by the company. The patient and clinical experts explained that a single flare would usually trigger a change of treatment (start of severe treatment sequence) only for patients with their usual disease activity in the higher end of the moderate disease activity range (that is, close to the severe disease activity level). However, such a change after a single flare was unlikely for patients with disease activity in the lower end of disease activity range. The committee agreed the rate of progression in NHS clinical practice is uncertain but noted that higher progression rates would result in lower incremental cost-effectiveness ratios (ICERs) for filgotinib compared with best supportive care. This was because with higher progression rates, quality-adjusted life years (QALYs) and costs are increasing in both treatment arms, but to a higher degree in the best supportive care arm. The committee concluded that although the rates of progression from moderate to severe disease in NHS clinical practice is uncertain, the company approach to model this is reasonable. It also noted that if the true rates of progression are higher than those estimated in the model, the cost-effectiveness estimates for filgotinib would improve.

3.16 The committee recalled that rheumatoid arthritis is a heterogenous disease and it was not practical to model all possible treatment sequences (see section 3.1 and section 3.14). The clinical expert explained that generally, they would follow the standard treatment sequence for severe disease once patients' disease progresses to severe disease activity. This would generally be:

3.17 In the company's base case, health-related quality-of-life data was mapped from patients' long-term HAQ-DI score trajectory using a published mapping algorithm. In addition to HAQ-DI, the algorithm used patients' current age, sex, and visual analogue scale pain scores to determine a utility value at any point in the model. In the company's base case, the visual analogue scale pain scores were mapped from HAQ-DI as per NICE's technology appraisal guidance on adalimumab, etanercept, infliximab, certolizumab pegol, golimumab, tocilizumab and abatacept for rheumatoid arthritis not previously treated with DMARDs or after conventional DMARDs only have failed. The ERG explained their initial concerns about the mapping algorithm, which seemed to provide distinct utility values rather than those based on EQ‑5D data collected in the trial. However, in response to technical engagement, the company provided corrected validation of their mapping algorithm, using individual patient data. The ERG was satisfied that the QALY outputs are fairly similar using the 2 methods. Therefore, it agreed with the company's approach and followed it in the revised ERG base case. The committee noted this approach is consistent with a number of previous appraisals. It concluded that the company's approach may have limitations but is appropriate for decision making.

3.18 NICE's guide to the methods of technology appraisal notes that above a most plausible ICER of £20,000 per 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 concluded that the cost-effectiveness results for moderate disease were uncertain because:

3.19 The committee noted that the revised company analyses applied the following committee preferences:

3.20 The ERG did analyses for people with severe rheumatoid arthritis whose disease had responded inadequately to 2 or more conventional DMARDs, applying confidential discounts for filgotinib, comparators and subsequent treatment options. Filgotinib with methotrexate provided a higher net health benefit (that is, was more cost effective) than alternative therapies used with methotrexate. Therefore, the committee concluded that it could recommend filgotinib with methotrexate as a cost‑effective use of NHS resources for people with severe rheumatoid arthritis whose disease had responded inadequately to 2 or more conventional DMARDs.

3.21 The ERG did analyses for people with severe disease whose disease had responded inadequately to 1 or more biological DMARDs, applying confidential discounts for filgotinib, rituximab and subsequent treatments. Filgotinib with methotrexate was dominated by rituximab with methotrexate (that is, filgotinib with methotrexate was more costly and less effective than rituximab with methotrexate). Therefore, the committee concluded that it could not recommend filgotinib with methotrexate as a cost‑effective use of NHS resources for people with severe rheumatoid arthritis whose disease had responded inadequately to 1 or more biological DMARDs if rituximab is a treatment option.

3.22 The ERG did analyses for people with severe disease whose disease had responded inadequately to 1 or more biological DMARDs and rituximab is not a treatment option, applying confidential discounts for filgotinib, comparators and subsequent treatment options. Filgotinib with methotrexate provided a higher net health benefit (that is, was more cost effective) than alternative therapies used with methotrexate. Therefore, the committee concluded that it could recommend filgotinib with methotrexate as a cost-effective use of NHS resources for people with severe rheumatoid arthritis whose disease had responded inadequately to 1 or more biological DMARDs, if rituximab is not a treatment option.

3.23 The ERG did analyses for people with severe disease whose disease had responded inadequately to 1 or more biological DMARDs and rituximab, applying confidential discounts for filgotinib, comparators and subsequent treatment options. Filgotinib with methotrexate provided a higher net health benefit (that is, was more cost effective) than alternative therapies used with methotrexate. Therefore, the committee concluded that it could recommend filgotinib with methotrexate as a cost-effective use of NHS resources for people with severe rheumatoid arthritis whose disease had responded inadequately to 1 or more biological DMARDs and rituximab.

3.24 The committee noted that cost-effectiveness estimates for filgotinib monotherapy were uncertain because filgotinib monotherapy has not been studied in its target population (see section 3.8). The committee also recalled that the company model assumed equal effectiveness of filgotinib monotherapy and combination therapy, which has limitations (see section 3.9). Also, for moderate disease, it recalled there was higher uncertainty related to treatment sequences after progression from moderate to severe disease (see section 3.16). However, the committee concluded that despite these limitations, filgotinib is likely to represent a good use of NHS resources for people for whom methotrexate is not suitable and so it recommended filgotinib monotherapy in the same positions as combination therapy. It also noted that this population is much smaller than the population of patients who can have methotrexate. It agreed that the small number of people who could not tolerate methotrexate should not be treated differently from other people with moderate to severe disease, as far as possible.

3.25 A potential equality issue was raised in NICE's technology appraisal guidance on upadacitinib for treating severe rheumatoid arthritis, about people with rheumatoid arthritis who have difficulty communicating. For these people, it may be more difficult to assess outcomes when using the DAS28 measure. The committee agreed that this equality issue was also important to consider for this appraisal. The committee concluded that healthcare professionals should consider any physical, psychological, sensory or learning disabilities, or communication difficulties that could affect the responses to the DAS28 and make any adjustments they consider appropriate.

3.26 The committee recalled that having a range of treatment options was important in rheumatoid arthritis (see section 3.1). It also recalled that a range of biological and targeted synthetic DMARDs are already available for severe rheumatoid arthritis (see section 3.2). It noted that a number of NICE technology appraisals are currently in development for moderate rheumatoid arthritis (NICE's ongoing technology appraisal on upadacitinib for previously treated moderate active rheumatoid arthritis and NICE's ongoing technology appraisal on adalimumab, etanercept, infliximab, certolizumab pegol, golimumab, tocilizumab and abatacept for moderate rheumatoid arthritis after conventional DMARDs only have failed). The committee concluded that healthcare professionals should choose the most appropriate treatment after discussing the advantages and disadvantages of the treatments available with the person having treatment. If more than 1 treatment is suitable, they should start treatment with the least expensive drug (taking into account administration costs, dose needed and product price per dose). This may vary from person to person because of differences in how the drugs are taken and treatment schedules.

3.27 Filgotinib is taken orally, which is valued by patients. The committee noted that for severe rheumatoid arthritis, other oral treatments with a similar mechanism of action are already available. But no biological or targeted synthetic DMARDs are currently available for people with moderate disease. The committee agreed that filgotinib is an important treatment option for these people. It concluded that all the benefits of filgotinib were adequately captured in the model.