Ixekizumab for treating moderate to severe plaque psoriasis

Generalisability of the trial populations

4.4 The committee noted that the evidence for ixekizumab mostly came from 3 trials: UNCOVER‑1, -2 and -3, which were double-blinded, randomised controlled trials that included 3,866 patients. UNCOVER‑1 compared ixekizumab with placebo, and UNCOVER‑2 and ‑3 compared ixekizumab with placebo and with etanercept. The primary outcome was Psoriasis Area and Severity Index (PASI) 75, which is a 75% reduction in the PASI score from when treatment started, measured at 12 weeks (the end of the induction period).

4.5 The committee was aware that the trials included patients with a PASI score of 12 or more and that in previous appraisals of technologies for treating psoriasis, a PASI score of 10 or more had been defined as severe disease. The committee heard from the clinical experts that the PASI is a composite measure of disease severity that combines the extent of the body surface area involved and the severity of the redness, thickness and scaling in each area. It understood that higher values represent increased severity. The committee also understood that the trials included patients with Dermatology Life Quality Index (DLQI) scores ranging from 0 to 30, and that in previous NICE appraisals of technologies for treating psoriasis, a DLQI score of more than 10 had been defined as severe disease. It was aware that the DLQI is a questionnaire that aims to measure how much psoriasis affects the life of people who have it. It heard from the company that the trial eligibility criteria had also included a static Physician Global Assessment score of 3 or more, where scores of 3, 4 and 5 are defined as moderate, severe and very severe disease, respectively. The committee heard from the clinical experts that in clinical practice disease severity is rarely defined in this way. The committee heard from the clinical experts that patients being considered for biological treatment would tend to have a PASI score of 10 or 12 or above, and that the patients in the UNCOVER trials were representative of patients seen in the NHS who would be considered for biological treatment.

4.6 The committee noted that the population in the trials and the marketing authorisation for ixekizumab included people who are candidates for systemic therapy, which includes both non-biological and biological treatments. It noted that the UNCOVER trials included patients who had never had systemic treatment, had had systemic treatment, or had already had biological treatment. Overall, the committee concluded that although previous treatment for trial participants varied, the populations of the UNCOVER trials were likely to be generalisable to patients in the NHS who would be considered for biological treatment, and were appropriate for decision-making on the clinical effectiveness of ixekizumab.

Ixekizumab compared with placebo and etanercept

4.7 The committee noted that patients randomised to ixekizumab had clinically and statistically significantly higher PASI 75 response rates at week 12 than placebo and etanercept. The odds ratios for ixekizumab producing a higher PASI 75 response rate at week 12 are detailed in table 1.

Table 1 Odds ratios for ixekizumab producing higher PASI 75 response rate at week 12

The committee noted that the dose of etanercept given in the trials was double that which is recommended in NICE's technology appraisal guidance on etanercept and efalizumab for the treatment of adults with psoriasis, and was considered by the clinicians to be more effective than lower doses of etanercept. Therefore the committee considered that the odds ratios could underestimate the treatment effect of ixekizumab compared with etanercept. The committee concluded that ixekizumab was more clinically effective than placebo and etanercept.

Ixekizumab in patients who have already had biological treatment

4.8 The committee noted that the benefit of treatment with ixekizumab compared with placebo and etanercept could be seen in the subgroup of patients who had already had biological treatment, as well as those who had not previously had biological treatment. It heard that the company did a test for interaction that showed little difference between the results of the trials across subgroups defined by baseline disease severity and previous biological treatment. The committee noted that the trials were not powered to detect statistically significant differences between subgroups, and so it could not be certain that the treatment effect did not differ across subgroups, however it agreed there was no evidence of a subgroup effect. The committee heard from the clinical experts that biological treatments generally work less well in patients who have already had a biological treatment, but that it depends on the particular biological treatment, and factors such as disease severity. The committee therefore concluded that, based on the data available, ixekizumab was more effective than etanercept or placebo both for patients who had previously had biological treatment and for those who had not.

Network meta-analysis results

4.9 The committee discussed the company's network meta-analysis that compared ixekizumab with adalimumab, ustekinumab, secukinumab and infliximab indirectly. The results showed that the relative risk of ixekizumab achieving a PASI 75 response at 12 weeks was significantly higher than that of all the biological treatment comparators except infliximab.

Adverse events

4.10 The committee was aware that the rates of serious adverse events including non-melanoma skin cancer, malignancies other than non-melanoma skin cancer, and severe infection, were very low, and that most of the adverse events related to treatment were mild to moderately severe and did not lead to stopping treatment. It heard from the clinical experts that serious infection was the main concern with biologicals, but that treatment was generally well tolerated. The committee concluded that the tolerability of ixekizumab was similar to that for other biological treatments approved for treating psoriasis.

Model structure

4.11 The committee considered the Markov state transition model the company used to model the cost effectiveness of ixekizumab. It modelled 7 biological treatment sequences and contained 4 health states:

4.12 The committee noted that the ERG considered the treatment sequencing approach to be better than comparing individual treatments because it more closely reflected clinical practice. It noted that the model represented both patients who had never had systemic treatments, and those who had already had biological treatment. The committee concluded that the company's model structure reflected clinical practice.

4.13 The committee understood that the company had used market share information to determine the most commonly used treatment sequences in the NHS. Each of the 7 biological treatments was modelled first in a sequence of 3 biologicals, as follows:

4.14 It heard from the clinical experts that the sequences of treatment included in the company's economic model mostly reflected current practice in the NHS, except that etanercept and infliximab were not used as a first biological treatment. Also, depending on the weight of the patient, ustekinumab 90 mg may be used as a second biological treatment if there is an inadequate response to ustekinumab 45 mg. This sequence was not included in the company's economic model. The committee also considered whether ixekizumab would replace an existing biological treatment, or extend a sequence of biological treatments. It heard from clinical experts that although ixekizumab could initially extend an existing biological treatment sequence, over time as more data become available it may replace an older, less efficient biological treatment. The committee recognised that the treatment sequences presented did not cover all possible sequences but concluded that the sequences included by the company in its economic model reasonably represented current NHS practice.

Modelling utility benefit

4.15 The committee understood that the company had used the subgroup of patients with a DLQI of more than 10 from the UNCOVER trials to estimate utility benefit. It noted that this differed from the intention-to-treat populations (that is, those with a DLQI ranging from 0 to 30) from the UNCOVER trials and from other trials included in the company's network meta-analysis, that were used to model treatment effectiveness. The committee acknowledged that this subgroup could not be used to model treatment effectiveness because not all the trials included in the network meta-analysis reported subgroup data. It accepted that because the subgroup population with a DLQI of more than 10 represented the patients seen in clinical practice, it was appropriate to analyse this subgroup where possible. It concluded that it was therefore appropriate to use this subgroup of patients to estimate utility benefit.

4.16 The committee considered when patients would get the benefit of treatment, and when to apply the utility gains from treatment in the model. It was aware that the company had applied utility gains in the maintenance period of treatment, but not the induction period. It heard from both the ERG and the company that ixekizumab is associated with a rapid response and therefore utility gains during the induction period were likely. The committee concluded that it would be appropriate to include utility gains for ixekizumab in the induction period, and that excluding this underestimates the quality-adjusted life year (QALY) gain associated with ixekizumab.

4.17 The committee noted that the company did not include the disutility of adverse events in its model. It considered that this did not reflect the importance of manageable side effects (see section 4.1). It heard from the company that there were little data available on severe adverse events that led to stopping treatment. The committee heard from the clinical experts that biological treatments were generally well tolerated (see section 4.10) and that their side effects profiles were similar, and concluded that it was therefore acceptable to exclude the adverse events in the model.

Costs of adverse events

4.18 The committee acknowledged that serious adverse events including non-melanoma skin cancer, malignancies other than non-melanoma skin cancer and severe infection did not occur very often (see section 4.10). However, the committee considered that it was appropriate to capture all the benefits and costs, including the costs of adverse events over the time horizon of the model. The committee concluded that the company should have included the costs of adverse events in its economic model, particularly given that the quality-of-life data were likely to already include any disutility from adverse events.

Costs of best supportive care

4.19 The committee noted that the costs of best supportive care had been estimated used the Fonia et al. (2010) study. It was aware from previous appraisals for psoriasis that when this study had been used, the conclusion was that it was likely to overestimate the costs of best supportive care. The committee understood from the ERG's analysis that if the costs of best supportive care were lower, the incremental cost-effectiveness ratios (ICERs) for ixekizumab compared with other biological treatments would increase. It considered that the estimates from Fonia et al. do not represent best supportive care after many biological treatments have not worked because they include costs of systemic treatments that are unlikely to be given after 3 biological treatments. The committee concluded that the costs of best supportive care remained uncertain, but given that data are lacking in this area, considered that the Fonia et al. estimates were appropriate for the company to use in the economic model.

Results of cost-effectiveness analysis

4.20 The committee noted that the company had given deterministic results in its base case. The committee preferred to use a probabilistic base-case analysis for decision-making. It agreed that the cost of adverse events should be included and that utility gains should be applied in the induction period of treatment, which the ERG had done in its base-case analysis. The committee noted that the ERG had also fixed errors in the company's model for adverse event rates and costs, calculating the standard error for NHS reference costs, and calculating the number of doses of secukinumab in the maintenance period. The committee therefore preferred to consider the results of the ERG's base-case analysis in its decision-making.

4.21 The committee noted that when validating the company's model, the ERG's ICERs for each comparator alone (that is, not in a sequence with other treatments) compared with best supportive care were more than £30,000 per QALY gained. The committee considered that including potentially non-cost-effective comparators, especially within sequences of treatments could result in misleading ICERs. The committee therefore also took into account comparisons against best supportive care in its decision-making. It noted that the ERG had presented analyses with ixekizumab at different positions within the treatment pathway (as the first biological treatment or as the second), and considered each in turn.

4.22 The committee considered the cost-effectiveness analyses for ixekizumab as the first biological treatment in a treatment sequence, taking into account the patient access schemes associated with ixekizumab and secukinumab. It noted:

4.23 The committee considered the cost-effectiveness analyses for ixekizumab as the second biological treatment in a treatment sequence, taking into account the patient access schemes associated with ixekizumab and secukinumab. It noted:

4.24 The committee considered the cost-effectiveness analyses for ixekizumab (not in a sequence) compared with best supportive care, and the ICERs for each comparator (not in a sequence) compared with best supportive care, which were used by the ERG to validate the model (see section 4.21). The committee recognised these analyses used the company's assumptions. It noted:

4.25 The committee was aware that the company had not explored the full range of treatment sequences that might be offered in current NHS practice (see section 4.14). The committee recognised that best supportive care was not a relevant comparator given the position of ixekizumab in the treatment pathway, but it considered the comparison in its decision-making to account for potential bias from including non-cost-effective comparators within all other analyses. The committee considered the cost effectiveness of ixekizumab in the light of previous appraisals in this disease area and concluded that the most plausible ICER was likely to be in line with the other biological treatments already recommended in previous NICE guidance. The committee concluded that the ICER was within the range that could be considered a cost-effective use of NHS resources.

Stopping rule

4.26 The committee was aware that previous appraisals for treating psoriasis recommended stopping treatment if there was an inadequate response; an adequate response was defined as either a 75% reduction in the PASI score from when treatment started, or a 50% reduction in the PASI score and a 5‑point reduction in DLQI from when treatment started. The latter stopping rule was originally included based on clinical advice, rather than specific evidence on the clinical- and cost effectiveness of this treatment strategy. The committee agreed that if there was no response to ixekizumab, the patient should not continue treatment. The committee noted that PASI 75 was the primary outcome in the trial data used to model the cost effectiveness of ixekizumab. It also considered the cost-effectiveness evidence for a PASI 50 stopping rule submitted by the company as a scenario analysis during consultation. It noted that the cost effectiveness of ixekizumab improved when a PASI 50 stopping rule was applied. The committee thought this may be counter-intuitive because there would be a smaller benefit for those with a PASI 50 response for the same cost. It heard from the company that, when the PASI 50 stopping rule is applied, patients remain on ixekizumab for longer, which increases the benefit accrued from having active treatment. The committee acknowledged that the sequencing model made it difficult to ascertain what was driving the results. The committee appreciated that, in patients with psoriasis affecting their hands and feet, a PASI 75 may be difficult to achieve because gains in quality of life from having treatment could represent clinical improvement but not be accounted for in the PASI score. The committee therefore concluded that, for consistency with previous appraisals for biological treatments in psoriasis, ixekizumab should be stopped if there is an inadequate response at 12 weeks, with an adequate response defined as a 75% reduction in the PASI score from when treatment started or a 50% reduction in the PASI score and a 5‑point reduction in DLQI from when treatment started.