Adalimumab, etanercept and ustekinumab for treating plaque psoriasis in children and young people

Position of technologies in the treatment pathway and comparators

4.3 The committee was aware that the marketing authorisations were different for adalimumab, etanercept and ustekinumab (see section 2). It was aware that adalimumab could be given as an alternative to non-biological systemic therapies but heard from the clinical expert that in clinical practice all 3 drugs are used as third line after topical therapies, phototherapy and non-biological systemic agents. It heard however, that patients and clinicians would welcome the opportunity to offer biologicals earlier in the treatment pathway. The committee concluded that the most appropriate comparator for adalimumab as a second-line treatment was non-biological systemic therapy (such as methotrexate). It also concluded that the most appropriate comparators for adalimumab, etanercept and ustekinumab as third-line treatment were each other and best supportive care, and that this was the point at which biologicals would most likely be used in the NHS.

Generalisability of the clinical trials to clinical practice

4.5 The committee considered the severity of psoriasis and the way it was defined in clinical practice and in the trials. It heard from the clinical experts that clinicians use both the PASI and the Children's Dermatology Life Quality Index (CDLQI; a questionnaire designed for use in children aged 5 years to 16 years) when monitoring disease and choosing who to offer biological therapies. The committee noted that percentage reduction in PASI score was the primary end point in M04‑717 and 20030211, and a secondary end point in CADMUS. It heard from the clinical expert that a 75% reduction in PASI (PASI 75) is a broadly used assessment method in children and young people. The committee agreed that the appropriate outcomes were captured in the trials. The committee concluded that PASI was a relevant measure used in clinical practice in the NHS and that PASI 75 was a clinically relevant definition of response to treatment.

4.6 The committee discussed the baseline characteristics of the patients in the trials:

Network meta-analysis results

4.7 The committee heard from the assessment group that it was not possible to connect the interventions and comparators together using direct evidence from children and young people alone because the trials did not use a common comparator. The committee understood that the assessment group's preferred analysis included all available adult data, because of the lack of evidence in children and young people. The committee understood that the assessment group adjusted for differences in population response rates and placebo response rates because they differed between trials and between children and adults. It agreed with the assessment group that all available adult evidence should be included in the network, and that it was appropriate to adjust the data for population characteristics and placebo response rates.

4.8 The results of the network meta-analyses are presented in table 1. The results for PASI 75 showed that the effectiveness of ustekinumab and adalimumab were similar, and that ustekinumab and adalimumab were more effective than etanercept. The committee heard from the clinical expert that this reflected clinical practice because clinicians are unlikely to offer etanercept as a first biological therapy. The committee was concerned that using adult data could potentially bias the effect estimates, but agreed that this was mitigated by the assessment group having adjusted for population and placebo effects. The committee concluded that, despite the uncertainty associated with the network meta-analyses (see section 4.7), the results showed adalimumab, etanercept and ustekinumab to be more clinically effective than placebo. In addition, the committee concluded that ustekinumab and adalimumab had broadly similar effectiveness, and that both were more clinically effective than etanercept.

Table 1 Network meta-analyses results

Model structure

4.9 The committee considered the assessment group's de novo Markov model. It noted the assessment group had done analyses for 3 different populations based on the position of the technology in the treatment pathway, and the different ages specified in the marketing authorisations for each intervention:

4.10 The committee discussed the length of the time horizon used in the assessment group's model. The assessment group assumed that at the age of 18 years, NICE technology appraisal guidance on etanercept, adalimumab and ustekinumab for biologicals in adults would apply. In the model, the time horizon varied according to population: 14 years for population 1 (aged 4 to 17 years); 12 years for population 2 (aged 6 to 17 years); and 6 years for population 3 (aged 12 to 17 years). The committee heard from one of the companies that a lifetime time horizon was needed to capture the full benefits and costs of treatment because the effects of psoriasis continue into adulthood. It heard from the assessment group that, in its model, most people had withdrawn from biological treatment after 14 years. It also noted that the time horizon did not have a large effect on the incremental cost-effectiveness ratios (ICERs). The committee concluded that, although a lifelong time horizon would better reflect the treated natural history of disease, given the data available, the assessment group's approach was acceptable.

4.11 The committee considered the stopping rules used by the assessment group in its model, that is, that clinicians should assess and stop treatment in patients whose disease has not responded by week 12 for etanercept, and week 16 for adalimumab and for ustekinumab. It agreed that this was consistent with the guidance in the summary of product characteristics for etanercept and adalimumab, but not for ustekinumab, which states that response should be assessed at 28 weeks, rather than 16 weeks. The committee understood that 16 weeks was used in NICE's technology appraisal guidance for ustekinumab for treating moderate to severe plaque psoriasis in adults. The committee agreed that it was desirable to have similar stopping rules for children and adults to avoid unnecessary changes in care during the transition from children to adult services. In addition, the committee agreed it was appropriate to use PASI 75 to assess response to treatment (see section 4.5). The committee concluded that the assessment group's approach and stopping rules based on PASI 75 were appropriate.

Utilities

4.12 The committee discussed the challenges of measuring health-related quality of life in children and young people with psoriasis. The committee appreciated that the assessment group assumed that biological therapies improve quality of life but do not extend life. The committee noted that the trials did not collect data on EuroQol‑5 Dimension‑Youth (EQ‑5D‑Y, a generic preference-based measure for quality of life in people aged 8 years to 15 years), and reported only CDLQI and Pediatric Quality of Life Inventory (PedsQL, an approach to measuring health-related quality of life in healthy children and young people, and those with acute and chronic health conditions). In its model, the assessment group mapped PedsQL scores from the CADMUS trial to EQ‑5D‑Y using a mapping algorithm.

4.13 The committee noted that, when using this mapping algorithm, the quality of life in children and young people at the beginning of the trials was higher than in adults with severe plaque psoriasis (such as in NICE technology appraisal guidance on etanercept, adalimumab and ustekinumab). It also noted that the utility gain associated with an improvement in PASI response in children and young people was lower than in adults. The committee heard that it was implausible that children benefit less than adults, particularly because children experience similar physical symptoms, but some might feel more socially stigmatised than adults. The committee acknowledged that the gain in quality of life associated with an improvement in psoriasis was uncertain. It agreed that it was likely that the increase in quality of life in children and young people would be higher than estimated by the assessment group in its model. The committee concluded that it was appropriate to apply the most optimistic adult utility gains to children and young people.

4.14 The committee heard from the clinical and patient experts that carer disutility should be considered when appraising treatments for severe plaque psoriasis in children. The committee heard that children need help administering their treatments (such as applying emollients) and this can be time consuming, especially for best supportive care. The committee appreciated that it was difficult to estimate the disutility associated with psoriasis for carers and that, in the absence of quantitative estimates of these, the assessment group had not been able to incorporate carer disutilities in its analyses. The committee concluded that it would take into account the reduced disutility to carers with biological treatments in its decision-making.

Resource use and costs

4.15 The committee considered costs used by the assessment group in its model:

Cost-effectiveness results and conclusions

4.16 The committee recalled that its preferred assumptions included:

4.17 The committee agreed that the scenario analysis that most closely matched these assumptions was the assessment group's scenario analysis that combined the effect of using adult EQ‑5D data from NICE technology appraisal guidance on adalimumab, and which assumed 6.49 days in hospital per year for children and young adults having best supportive care from Fonia et al. (2010). However, the committee noted that some of its preferred assumptions were not fully reflected in the scenario analysis and took into account the potential for bias in the ICERs: