Golimumab for the treatment of psoriatic arthritis

4 Consideration of the evidence

4.1 The Appraisal Committee reviewed the data available on the clinical and cost effectiveness of golimumab, having considered evidence on the nature of psoriatic arthritis and the value placed on the benefits of golimumab by people with the condition, those who represent them, and clinical specialists. It also took into account the effective use of NHS resources and the impact of the patient access scheme (see section 2.4).

4.2 The Committee understood that psoriatic arthritis can cause significant distress and psychological impact on the person's life, employment and social activities. The Committee heard from a patient expert that TNF inhibitors are valued options for the treatment of psoriatic arthritis and have a positive impact on quality of life. It understood that people with the condition may prefer the option of a treatment that is self-injectable and/or has a longer retreatment interval. The Committee understood that people value having a choice of TNF inhibitors and that another treatment option will always be welcome.

4.3 The Committee considered current clinical practice for the treatment of psoriatic arthritis. It understood that TA199 recommends adalimumab, etanercept and infliximab for the treatment of psoriatic arthritis in people who have peripheral arthritis with three or more tender joints and three or more swollen joints, and when the psoriatic arthritis has not responded to adequate trials of at least two standard DMARDs (administered either individually or in combination). The Committee also noted that TA199 specifies that treatment should be with the least expensive drug, taking into account drug administration costs, required dose and product price per dose. It heard from the clinical specialists that they considered there to be little demonstrable difference between the TNF inhibitors in terms of their clinical effectiveness. The clinical experts did, however, note slight differences among the TNF inhibitors in TA199 with regard to the subjective reduction in response to treatment in the skin and joint components of the disease. The Committee heard from the commissioning expert that subtle differences in cost and administration, particularly with regard to dose escalation (as included in the marketing authorisation for infliximab) and hospitalisation, can make a big cost difference. The Committee concluded that adalimumab, etanercept and infliximab were the appropriate comparators for golimumab.

4.4 The Committee heard from the clinical specialists and the patient expert that people often prefer a less frequent dosing schedule; that is, a longer time period between treatments. However, the Committee noted that the longer retreatment interval associated with golimumab could potentially result in more discomfort because of waning efficacy before retreatment. It understood that people with psoriatic arthritis and their clinicians may therefore value the once-monthly, self-injectable administration of golimumab. The Committee concluded that golimumab could, on balance, be a valued additional treatment option for people with psoriatic arthritis.

4.5 The Committee considered the evidence on the clinical effectiveness of golimumab. It understood that the main clinical effectiveness data were derived from a single phase III RCT. The Committee noted statistically significant outcomes for the 50 mg dose compared with placebo in terms of improvements in joint disease, skin disease and functional status (see sections 3.3 and 3.4). The Committee concluded that golimumab was clinically effective compared with placebo.

4.6 The Committee discussed the 100 mg dose of golimumab, which may be considered for people who weigh more than 100 kg and whose psoriatic arthritis has not responded after three or four doses of golimumab (as stated in the SPC). It noted that neither the 100 mg arm nor dose escalation to 100 mg in the 50 mg arm in the GO-REVEAL trial was limited to people who weighed more than 100 kg, and therefore the trial population did not reflect the population in the marketing authorisation for the 100 mg dose. The Committee heard from clinical specialists that they would be more likely to select a different TNF inhibitor than to increase the dose if the 50 mg dose of golimumab failed to produce a response. The Committee concluded that it was uncertain of the extent to which the 100 mg dose would be used in clinical practice.

4.7 The Committee noted that there had been no head-to-head trials of golimumab and any of the other TNF inhibitors, and that, as a result, the manufacturer had conducted a mixed treatment comparison. The Committee recognised the limitations of mixed treatment comparisons and was aware that the associated results would need to be interpreted with caution. It also noted the manufacturer's reservations about mixed treatment comparisons and the uncertainty associated with the use of such methodologies, but noted that no alternative methods or data had been provided. Following the consultation on the Appraisal Consultation Document, the manufacturer suggested the removal of the Mease 2000 results (for etanercept) from the mixed treatment comparison, because this trial did not disaggregate HAQ scores in the same way as other trials, and showed better results for etanercept than Mease 2004 (the larger etanercept trial). The Committee had misgivings about the selective removal of individual trials, but heard from the ERG that extracting the Mease 2000 study from the mixed treatment comparison had little effect on the results. The Committee agreed it would bear this in mind when considering the results of the mixed treatment comparison.

4.8 The Committee carefully considered the results of the mixed treatment comparison. It noted that for PsARC response and absolute change in PASI from baseline, the results showed that golimumab was generally equivalent to the other TNF inhibitors. However, it also noted that golimumab had the lowest HAQ score change from baseline (both in participants whose disease responded to treatment based on PsARC score and those whose disease did not respond based on PsARC score) compared with the other TNF inhibitors.

4.9 The Committee further discussed the HAQ results from the mixed treatment comparison. The Committee understood from the clinical specialists and the patient expert that pain and disability caused by arthritis (as captured by HAQ score and reflected in the manufacturer's economic model) often have a significant impact on the person's quality of life. The Committee was concerned that, out of the four TNF inhibitors that were compared, golimumab had the lowest HAQ score change from baseline and that this might indicate inferiority of its anti-arthritic activity (see section 3.12); however, the Committee was also aware of the limitations of the mixed treatment comparison methodology. Therefore the Committee also considered the radiographic progression data, which, together with the change in HAQ score, could be used to assess the effect of a treatment on disease progression. The Committee noted the statistically significant reduction from baseline in vdH-S score (a measure of radiographic progression) for golimumab compared with placebo at 24 weeks (−0.16 for 50 mg golimumab compared with 0.27 for placebo, p = 0.01; see section 3.4). It noted that the change from baseline in vdH-S score at week 24 for golimumab was less than that for infliximab (−0.70 for 5 mg/kg infliximab compared with 0.82 for placebo, p < 0.001), which was the other TNF inhibitor for which radiographic progression was measured by vdH-S score in the trials included in the mixed treatment comparison. The Committee was aware, however, that this difference may be due to differences in the trial populations, as reflected by the respective changes from baseline with placebo. The Committee also understood that the absolute differences between the two changes from baseline were small. Although the evidence suggested that golimumab may be less effective in its anti-arthritic activity (based on the HAQ score results from the mixed treatment comparison and the data for radiographic progression), on balance the Committee concluded that the evidence was not robust enough to confirm clinically important differences in the effectiveness of golimumab compared with the other TNF inhibitors.

4.10 The Committee considered the evidence on the adverse event rates associated with the use of golimumab. It noted a number of reported 'serious' adverse events, but understood that GO-REVEAL was not powered to detect statistically significant differences in adverse event outcomes. The Committee considered the additional evidence submitted by the manufacturer on the long-term adverse event data for golimumab in people with psoriatic arthritis, and also for people with rheumatoid arthritis and ankylosing spondylitis. It concluded that although there remains uncertainty about golimumab's long-term adverse event profile, it had not been shown to be different from that of other TNF inhibitors.

4.11 The Committee considered the economic model presented by the manufacturer. The Committee noted that the model assumed people continuing on therapy maintained their initial improvement in HAQ score. The Committee considered the utility estimates incorporated in the model, and noted that the utility formula was derived from the HAQ score change and the PASI response. The HAQ score change had a greater effect on utility than the PASI response did, indicating that the calculated utility benefit was driven more by the reduction in joint symptoms than the reduction in skin disease. The Committee concluded that this was appropriate (see section 4.9).

4.12 The Committee considered the results of the manufacturer's base-case analysis, which compared each of the TNF inhibitors (including golimumab) with palliative care. The Committee heard from the ERG that the pair-wise comparisons with palliative care needed to be reworked into an incremental analysis comparing each treatment with the next most effective alternative. The ERG re-presented these results. The Committee was aware that the acquisition costs of adalimumab, etanercept and golimumab (50 mg) were similar, and that the acquisition cost of infliximab was dependent on the person's weight and the amount of the drug required, with additional administration costs for infliximab (related to intravenous infusion). The Committee noted that all alternatives to etanercept were either dominated (infliximab was more expensive but no better than etanercept) or extendedly dominated (adalimumab and golimumab were, in effect, less cost effective than etanercept; see section 3.20). The Committee agreed that golimumab was, in effect, less cost effective than etanercept.

4.13 The Committee was aware, however, that TA199 recommends adalimumab and infliximab alongside etanercept. The Committee therefore also considered whether golimumab was at least as cost effective as adalimumab and infliximab. The Committee was aware that in the incremental analysis, both adalimumab and golimumab were extendedly dominated by etanercept. However, the Committee noted that the pairwise ICER of golimumab compared with adalimumab alone would be approximately £24,000 per QALY gained. The Committee similarly noted that the pairwise ICER for golimumab compared with infliximab would be approximately £45,000, aware that in this instance the ICER would represent a 'savings per QALY lost', as golimumab was associated with both lower costs and fewer QALYs compared with infliximab (see section 3.19). Given the weaknesses of the evidence suggesting lesser clinical effectiveness of golimumab compared with the other TNF inhibitors, and the estimates of golimumab's cost effectiveness compared with adalimumab and infliximab, the Committee concluded that the 50 mg dose of golimumab was acceptable when the criteria in TA199 are met; that is, the person has peripheral arthritis with three or more tender joints and three or more swollen joints, and the psoriatic arthritis has not responded to adequate trials of at least two standard disease-modifying antirheumatic drugs (DMARDs), administered either individually or in combination.

4.14 The Committee considered the 100 mg dose of golimumab. The Committee was aware that the SPC for golimumab states that for people who weigh more than 100 kg whose disease does not show an adequate clinical response after three or four doses, the dose of golimumab may be increased to 100 mg once a month. The Committee heard two different opinions about the proportion of people who would be eligible for the higher dose. The Committee agreed that this proportion was uncertain, but that it could be substantial. The Committee noted that the 100 mg dose of golimumab was not considered in the economic model, but that, because of the patient access scheme (as described in section 2.4), the cost of the 100 mg dose would be equal to that of the 50 mg dose. In addition, the Committee acknowledged the comments from the clinical specialists that, in clinical practice, people would be more likely to be switched to a different TNF inhibitor if no response was observed with the 50 mg dose, than to have the dose increased (see section 4.6). The Committee also noted TA199 states that treatment choice should be based on cost (taking into account drug administration costs, required dose and product price per dose), with treatment initiated with the least expensive drug. Therefore the Committee concluded that with the incorporation of the patient access scheme, golimumab would be considered an acceptable option for the treatment of psoriatic arthritis if used as described for other TNF inhibitors in TA199.

4.15 The Committee discussed the discontinuation of treatment with etanercept, infliximab and adalimumab in TA199. The Committee considered that the recommendation to discontinue treatment based on an inadequate PsARC response at 12 weeks included in TA199 was also appropriate for golimumab. The Committee was aware that no evidence had been provided by the manufacturer for the use of golimumab after the failure of other TNF inhibitors. The Committee was therefore unable to make recommendations about the use of golimumab following the failure of other TNF inhibitors.

4.16 The Committee was aware that there may be some circumstances that could affect a person's responses to components of the PsARC such as any physical, sensory or learning disabilities, or communication difficulties. The Committee concluded that in such cases, healthcare professionals should make any adjustments they consider appropriate.

4.17 The Committee was aware of registries that collect data on the long-term outcomes of treatment with TNF inhibitors for rheumatoid arthritis. The Committee noted the importance of registries in gathering data and supported the inclusion of outcomes specific to psoriatic arthritis in a suitable registry so that specific information about treatments and treatment-related adverse events in psoriatic arthritis can be collected.

4.18 In summary, the Committee considered the clinical and cost effectiveness of golimumab in the light of the submitted evidence and the comments of the clinical specialists, the commissioning expert and the patient expert. The Committee noted that although the evidence suggested that golimumab may be less effective in its anti-arthritic activity (based on the HAQ score results from the mixed treatment comparison and the data for radiographic progression), the evidence was not robust enough to confirm clinically important differences in the effectiveness of golimumab compared with the other TNF inhibitors. The Committee further noted that golimumab was, in effect, not cost effective when compared with etanercept, but may be cost effective when compared with adalimumab and infliximab. The Committee was aware that the patient access scheme (as described in section 2.4) would provide the 100 mg dose of golimumab at the same cost as the 50 mg dose. The Committee concluded that, with the incorporation of the patient access scheme and if the criteria specified in TA199 were met, golimumab should be recommended as an option for the treatment of active and progressive psoriatic arthritis in adults, as described for other TNF inhibitor treatments in TA199.