Draft guidance consultation

First-line treatment for chronic hand eczema includes topical corticosteroids with or without topical calcineurin inhibitors (treatments applied to the skin). Second-line treatment includes phototherapy and oral alitretinoin (treatment taken by mouth). Subsequent treatments may include immunosuppressants, oral corticosteroids, biological medicines and oral Janus kinase (JAK) inhibitors. Phototherapy includes psoralen (treatment that makes the skin sensitive to sunlight) and ultraviolet light A (collectively known as PUVA). The patient group submission highlighted that current treatments are not effective for some people. They also noted that alitretinoin can cause birth defects, so a pregnancy prevention programme is necessary for people who could become pregnant. The clinical experts described how phototherapy is inconvenient for most people because it involves regular visits to hospital (20 to 30 sessions). The committee concluded that people with chronic hand eczema would welcome an additional treatment option.

In the company's submission, delgocitinib was positioned as a second-line treatment option for adults with moderate to severe chronic hand eczema when topical corticosteroids have not worked or are not suitable. The company later proposed to optimise this positioning, so that in people with severe disease delgocitinib would only be used if alitretinoin has not worked or is not suitable. For people with moderate disease, the company did not consider alitretinoin a relevant treatment option, as NICE's technology appraisal guidance on alitretinoin for the treatment of severe chronic hand eczema does not include people with moderate chronic hand eczema. The clinical experts explained that it is often difficult to differentiate between moderate and severe disease because severity can be subjective. They said that PUVA would typically be used for people with moderate chronic hand eczema but accepted that there may be a proportion of people with moderate disease who have alitretinoin 'off label'. They also noted that treatment choice is guided by method of administration, as well as safety and efficacy. For example, some people may prefer the convenience of a topical cream, even if it means a slower or less complete therapeutic response. The committee thought that delgocitinib would be prescribed in secondary care by a dermatologist with experience in chronic hand eczema, in line with the summary of product characteristics. It thought that the company's original positioning of delgocitinib reflected how it would likely be used in clinical practice. Based on this positioning, it concluded that PUVA and alitretinoin are both the most relevant comparators to delgocitinib for both moderate and severe chronic hand eczema.

The key clinical trial evidence for delgocitinib came from the DELTA trials. These include DELTA 1 (n=487) and DELTA 2 (n=473), which were phase 3, randomised double-blind trials comparing delgocitinib with vehicle cream (no active treatment). The population included adults with moderate or severe chronic hand eczema for whom topical corticosteroids had not worked or were not suitable. The Investigator's Global Assessment for chronic hand eczema (IGA-CHE) scale was used to define moderate (IGA-CHE score of 3) and severe (IGA-CHE score of 4) disease. People were randomised to delgocitinib cream or vehicle cream twice daily for up to 16 weeks. People who completed 16 weeks of treatment in DELTA 1 or 2 were able to enrol in DELTA 3 (n=801), which was a phase 3, open-label extension study of 36 weeks' duration. In DELTA 3, everyone had delgocitinib twice daily as needed (based on IGA-CHE score). The company also presented evidence from DELTA FORCE (n=513), which was a phase 3 randomised assessor-blinded trial comparing delgocitinib with alitretinoin. The population included adults with severe chronic hand eczema (IGA-CHE score of 4) for whom topical corticosteroids had not worked or were not suitable. People were randomised to delgocitinib cream twice daily for 16 weeks or alitretinoin daily for 12 weeks. After these time points, each treatment was used as needed (based on IGA-CHE score). The EAG considered the DELTA trials to be at low risk of bias. They noted that there was no trial data that compared delgocitinib with alitretinoin in people with moderate chronic hand eczema. The committee acknowledged this but concluded that the DELTA trials were appropriate for decision making.

The primary outcome in DELTA 1 and 2 was IGA-CHE treatment success at week 16, defined as an IGA-CHE score of 0 (clear) or 1 (almost clear) and an improvement from baseline of at least 2 points. In the full-analysis set, the proportion with IGA-CHE treatment success at week 16 was significantly higher for the delgocitinib arm than for the vehicle-cream arm in DELTA 1 (mean difference 9.8%, p=0.006) and DELTA 2 (mean difference 22.2%, p<0.001). The primary outcome in DELTA FORCE was the mean change in Hand Eczema Severity Index (HECSI) score from baseline to week 12, with higher scores indicating greater disease severity. In the full-analysis set, the percentage reduction in mean HECSI score at week 12 was significantly larger in the delgocitinib arm than in the alitretinoin arm (mean difference ‑16.1%, 95% confidence interval ‑23.28% to ‑8.86%, p<0.001). The committee concluded that delgocitinib is an effective treatment for improving chronic hand eczema symptoms.

Because of a lack of direct evidence comparing delgocitinib with PUVA, the company did a network meta-analysis (NMA). The analysis included data from 5 randomised controlled trials to compare the efficacy of delgocitinib with vehicle cream, alitretinoin and PUVA. These included DELTA 1 and 2, DELTA FORCE, Worm 2022 and ALPHA (alitretinoin compared with PUVA in people with severe chronic hand eczema). For all analyses, the company selected the fixed-effects models. The results showed that people who had delgocitinib cream were statistically significantly more likely than those having PUVA or alitretinoin to have clear or almost clear skin. The company said the fixed-effects model results reflected the direct evidence for delgocitinib and alitretinoin, because DELTA FORCE was the only study in the NMA informing this comparison. The EAG noted that there was substantial clinical and methodological heterogeneity among the studies. For example, there were differences in the baseline characteristics of the included populations and the endpoints used to determine treatment success. It said that using a fixed-effects model was not appropriate because of this heterogeneity. Instead, it preferred to use direct evidence from DELTA FORCE to estimate the treatment effects for delgocitinib compared with alitretinoin. To estimate the treatment effects for delgocitinib compared with PUVA, the EAG asked the company to perform unanchored matching-adjusted indirect comparisons (MAICs) comparing delgocitinib (pooled population from DELTA 1 and 2) with PUVA (population from ALPHA). The EAG's preferred MAIC compared delgocitinib with PUVA in people with severe disease and based on hyperkeratotic status. The company considers the results of the MAIC to be confidential, so they cannot be reported here. The committee noted that the effective sample size for the matched population in the MAIC was small (n=39). It concluded that the company's NMA was more appropriate because it pooled the relative treatment effects using methods that preserve within-trial randomisation. So, the company's fixed-effects NMA was preferred over the EAG's approach to estimate the relative treatment effects for delgocitinib, vehicle cream, alitretinoin and PUVA.

The company used the worst observation carried forward (WOCF) approach to impute missing trial data to estimate treatment effects. The EAG thought that using the WOCF approach could bias comparisons of delgocitinib with vehicle cream or alitretinoin in favour of delgocitinib. This is because it thought the WOCF approach generated bias against the treatment arm with the highest dropout rate. The committee noted that dropout rates were greater in the vehicle-cream and alitretinoin arms of the DELTA studies. The EAG explained that this could underestimate the effectiveness of the comparator arms in the model. Instead, it suggested that multiple imputation, with a 'missing not at random' assumption, should be used because it is likely to be associated with a lower risk of bias. The company explained that, in the DELTA trials, most people stopped vehicle cream or alitretinoin because of adverse events or a lack of efficacy. Because of this, the company thought that a multiple-imputation approach could introduce bias and potentially underestimate the relative treatment effect of delgocitinib. The committee recognised the limitations associated with both the WOCF and multiple-imputation approaches. It thought it was unclear what the impact would be on the cost-effectiveness results if an alternative approach to WOCF was used to impute missing data across the DELTA trials. It concluded that analyses using the first and last observation carried forward method may help to reduce some of this uncertainty.

The company presented a Markov cohort model. The population included adults with moderate (IGA-CHE score of 3) or severe (IGA-CHE score of 4) chronic hand eczema for whom topical corticosteroids had not worked or were not suitable. The comparators modelled included PUVA for people with moderate and severe disease and alitretinoin for people with severe disease. The time horizon was 10 years and the model included a 4-week cycle with half-cycle correction. In the model, everyone had 12 weeks of initial treatment with delgocitinib, alitretinoin or PUVA. The company included stopping rules based on response to initial treatment, continued treatment and retreatment. Response to treatment was based on IGA-CHE score and included full, partial, low and insufficient response health states. The EAG thought the model structure was appropriate for addressing the decision problem. The committee recalled that although alitretinoin is licensed for severe disease, it may be used off label for some people with moderate disease (see section 3.3). The committee concluded that the company's model was acceptable for decision making. It also concluded that the model should consider alitretinoin as a comparator for people with moderate disease.

There was no evidence for the safety or efficacy of alitretinoin or PUVA in moderate disease. It recalled that both DELTA FORCE and ALPHA included only people with severe chronic hand eczema. In the model, the company assumed that the relative treatment effect of delgocitinib to alitretinoin and delgocitinib to PUVA in moderate chronic hand eczema was consistent with that seen in severe chronic hand eczema. So, although there may be a difference in the treatment effect between moderate and severe chronic hand eczema, this difference was consistent across treatment comparisons. The company did subgroup analyses from the DELTA 1 and 2 trials, which it said supported its assumption. The EAG thought that the company's subgroup analyses were insufficient to confirm whether the relative treatment effects between moderate and severe disease were consistent. So, the EAG's base case compared delgocitinib with alitretinoin or PUVA in only the severe population. The clinical experts said that all treatments may be slightly more effective in moderate disease, but agreed that the relative effectiveness of treatments would likely be consistent across moderate and severe populations. The committee also noted that the EAG's scenario analyses in moderate disease had a minimal to small impact on the cost-effectiveness results. It concluded that the company's assumption of equivalence in the relative treatment effects between moderate and severe chronic hand eczema was reasonable.

Some people with chronic hand eczema experience thickening of the outermost layer of the skin (hyperkeratosis). The EAG's clinical experts said the presence of hyperkeratosis influences how well treatments work and can guide treatment decisions. DELTA FORCE was the only trial for delgocitinib that was stratified by hyperkeratotic status. The committee discussed the subgroup results from the trial, which showed that the relative treatment effects for delgocitinib compared with alitretinoin differed by hyperkeratotic status. The company considers the results of the analyses to be confidential, so they cannot be reported here. The clinical experts said that hyperkeratotic status would guide treatment decisions only when hyperkeratosis is very pronounced. In this case, they would typically recommend alitretinoin because it is more effective than delgocitinib in this population. But they would also consider other factors, such as convenience and safety (see section 3.3). The committee noted that using the treatment effects based on hyperkeratotic status had a large impact on the cost-effectiveness results for delgocitinib compared with alitretinoin. The company explained that there was a lack of hyperkeratotic and non-hyperkeratotic subgroup data in the ALPHA trial, which meant that an indirect comparison between delgocitinib and PUVA was not possible. The committee concluded that the treatment effects in the model should be informed by the full population from DELTA FORCE rather than based on hyperkeratotic status.

The company's model used discontinuation rates from week 12 to week 24 in DELTA FORCE to estimate the discontinuation rate for people having retreatment. So, the model used discontinuation rates informed by data from people whose eczema did not have a full response to inform discontinuation rates for people whose eczema previously did have a full response. Because of this, the EAG said the discontinuation rates used for people having retreatment were too high, and time on treatment was likely underestimated. The EAG's clinical experts stated that around 25% of people on alitretinoin are still having treatment after 2 years in clinical practice. This was much higher than the proportion of people having alitretinoin in the model at 2 years (the company considers the exact figures to be confidential so they cannot be reported here). The clinical experts said it was difficult to predict how long people remain on treatment. But they said the proportion of people remaining on treatment in the model at 1, 2 and 3 years appeared too low for all treatments. They thought it was reasonable to assume that 25% of people would have long-term intermittent courses of alitretinoin to manage their chronic hand eczema. They thought it would be easier for people to restart treatment with delgocitinib than the other treatments when their symptoms flare up. This is because delgocitinib is a topical treatment with a year-long shelf life once opened, in contrast to the other treatments for which a new hospital visit would be needed to restart treatment each time. The committee concluded that modelled time on treatment did not reflect clinical practice. It would like to see analyses that include 25% of people on alitretinoin remaining on treatment at 2 years in the model. The committee thought the proportion of people remaining on delgocitinib at 2 years in the model should be at least 25%, but the proportion on PUVA may be lower than 25%. It discussed that the analyses should appropriately adjust the modelled time on treatment at years 1 and 3 in the model. It understood that the incremental cost-effectiveness ratio (ICER) was likely to be highly sensitive to the difference in proportions of people remaining on treatment.

The EAG made additional amendments to the model to inform its base case. These included:

• using delgocitinib dosing data from DELTA FORCE to inform the comparison between delgocitinib and alitretinoin

• using the weighted average of week 12 and week 24 delgocitinib dosing data from DELTA 1, DELTA 2 and DELTA FORCE to inform the comparison between delgocitinib and PUVA

• using the proportion of people who discontinue to next-line treatment and best supportive care based on the ALPHA trial

• removing alitretinoin from the next-line treatment basket and reducing the efficacy of the basket to 25.6%

• adjusting frequencies of dermatologist visits in each health state to align with the EAG's clinical expert opinion

• not including adverse events from DELTA FORCE.

NICE's manual on health technology evaluations 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. But it will also take into account other aspects, including uncaptured health benefits. The committee discussed the high level of uncertainty, specifically:

• the use of the WOCF approach to impute missing data for estimating treatment effects

• time on treatment in the model being underestimated for delgocitinib, alitretinoin and PUVA and not reflecting clinical practice.

The committee concluded that its preferred modelling assumptions included:

• using the company's fixed-effects NMA results to inform treatment effects in the model (see section 3.6)

• using health state-specific utility values (see section 3.12)

• including the EAG's preferred amendments to the model, except for adverse events (see section 3.13)

• including adverse events from DELTA FORCE (see section 3.13).

The exact ICERs are confidential and cannot be reported here because they include confidential discounts for alitretinoin and other treatments in the pathway. The committee noted it did not have ICERs that reflected all its preferred assumptions. The most plausible ICERs for delgocitinib compared with PUVA were below the committee's preferred threshold (£20,000 per QALY gained). The most plausible ICERs for delgocitinib compared with alitretinoin were above £30,000 per QALY gained. The committee considered whether delgocitinib could be recommended for people for whom alitretinoin isn't suitable. It recalled that it can be difficult to identify distinct populations based on subtype or severity (see section 3.3 and section 3.10). And also that treatment choice is guided by method of administration, as well as safety and efficacy (see section 3.3). The committee concluded that is not possible to clearly define different distinct patient populations who would be offered one treatment over the other. So, it could not recommend delgocitinib because it was not a cost-effective use of NHS resources.

The committee requested the following additional analyses from the company to address the areas of outstanding uncertainty:

• using the first and last observation carried forward method to impute missing data in the DELTA trials (see section 3.7)

• updated cost-effectiveness estimates that reflect the proportion of people remaining on delgocitinib, alitretinoin and PUVA in clinical practice at 1, 2 and 3 years (see section 3.11).

The committee acknowledged that chronic hand eczema disproportionally affects people from certain groups. It recognised that current treatments or services for chronic hand eczema may not be suitable for all people and their availability may depend on where a person lives. The committee agreed that these issues could not be addressed by its recommendations.