[Professor Nick Levell, Dr Jane Ravenscroft, Professor Carsten Flohr, Professor Mike Cork, Professor Anthony Bewley, Dr Ser-Ling Chua, Dr Jui Vyas, Mr Will Price on behalf of the BAD’s Therapy & Guidelines sub-committee]

1) Has all of the relevant evidence been taken into account?

We agree with the committee that all the available evidence was not included in the original document but needs to be considered. This includes data from studies that were not available in the original analysis, e.g. JADE-DARE, AD-UP, RISING-UP.

The committee may wish to consider the results of the TREAT trial (ciclosporin vs. methotrexate in adolescents), although we realise these treatments are not currently licensed in adolescents with eczema. The investigators would be happy to share the results confidentially with NICE in a few months’ time when they have been accepted for publication.

There is also additional published evidence regarding methotrexate in adults which would ideally be considered because methotrexate is the most commonly used first-line treatment for eczema and is much cheaper than ciclosporin or the new drugs.

Evidence from trial situations is limited because trial data necessarily exclude some patients, e.g. significant mental health difficulties, patients who have not responded to other treatments, patients with very severe disease who it would be unethical to randomise to placebo or who would drop out of the standard run-in period due to disease flares when their current systemic therapy needs to be stopped. Hence, real-world effectiveness data, such as that from the A-STAR registry, are likely to be more representative of the patient populations treated within NHS clinics than cohorts enrolled in trials.

There are many patients who have been treated with abrocitinib under the Early Access to Medicines Scheme (EAMS) and with tralokinumab and upadacitinib via free-of-charge (FOC) or nearly FOC schemes from their respective manufacturers. Such patients would have not only failed treatment with at least one conventional systemic but also a novel agent (most likely dupilumab). This shows the need for additional systemic therapies and the companies may have data from enrolment into these schemes that could be additionally used for the NICE technology assessment.

It is essential that patients commencing new treatments for eczema are entered into an independent research data platform, such as A-STAR, to allow evidence of comparative effectiveness, safety and treatment effect waning to be taken into account for future appraisals. A-STAR also collects health resource use data that could be used for the health economic evaluation that forms part of the NICE technology assessments.

2) Are the summaries of clinical and cost effectiveness reasonable interpretations of the evidence?

The summary statement that the drugs in question ‘have been indirectly compared with some standard treatments, but the results are highly uncertain’ seems disingenuous in the light of overwhelming data showing clinical benefit, including the living NMA (Drucker et al.) just published in JAMA Dermatology. This NMA has conducted indirect comparisons between the novel systemics using much of the same data as NICE, and demonstrated evidence of efficacy against conventional and other novel drugs, in particular dupilumab., Drucker states, ‘’This systematic review with meta-analysis found that compared with dupilumab, abrocitinib, 200 mg daily, and upadacitinib, 30 mg daily, were associated with reductions in EASI scores; upadacitinib, 15 mg daily, was associated with similar reductions, and tralokinumab, 300 mg every other week, and baricitinib, 2 and 4 mg daily, were associated with fewer reductions’’.

The uncertainty identified by NICE principally surrounds the meta-analysis approach taken, which essentially favours head-to-head clinical trial data with IMP vs. best supportive care. Unfortunately, such data largely does not exist, and therefore, it seems unreasonable to derive a model that favours it. However, the efficacy of the medications under review against placebo was undoubted and agreed in the report. Indeed, it was a drug vs. placebo technology appraisal model that has been applied for novel AD therapies until now. The uncertainty of effect size due to a circuitous indirect comparison with ciclosporin, risks loss of sensitivity. This results in a negative appraisal which would exclude highly effective treatments from clinical use for our patients. In the single area where head-to-head data is available (albeit not as primary endpoint) abrocitinib vs. dupilumab data, the clinical efficacy of abrocitinib is demonstrated, supporting the argument that the indirect comparison model used in the NMA by Drucker et al. is robust.

Another possible reason for difference in analysis results between NICE and Drucker include the categorical/dichotomous outcomes used in the NICE analysis vs. continuous outcomes in Drucker.

The cost economic modelling is problematic for several reasons highlighted by the committee in the consultation document. The committee have suggested a number of areas where changes are needed, which we broadly agree with.

We are very concerned that different methodology used in the appraisals for novel eczema treatments to date (dupilumab, baricitinib and this one) has led to different conclusions regarding cost effectiveness. For example, the utility value of best supportive care in this model was associated with higher quality of life and lower costs compared with the dupilumab appraisal. The consequence of this is that dupilumab, which was previously recommended, now appears less cost effective. We feel strongly that this needs to be addressed because of the implications for future NICE appraisals of new eczema treatments.

It was stated in the ACD document that “The limitations in the clinical evidence mean the results from the economic model are very uncertain. Because of this it is not possible to determine a suitable cost-effectiveness estimate for abrocitinib, tralokinumab and upadacitinib. So, they cannot be recommended.” However, when dupilumab and baricitinib were being appraised by NICE, the evidence available was more limited.

‘Best supportive care’ (BSC) is defined in this model as a single health care state. Costs of BSC are calculated by the weighted average of responders and non-responders at 16 weeks (as guided by the NMA of clinical effectiveness). This is likely to be an underestimate of true costs of BSC (see below). We do not feel it is appropriate to have a single BSC state or that this state should be assumed to be stable for the duration of modelling (5 years). In the current modelling, BSC is associated with high quality of life and low costs. This is incorrect and leads to favouring of ineffective treatments.

Patients in BSC are likely to require high-cost, ongoing care, including:

Ongoing hospital outpatient attendances and investigations.
Alternative systemic drugs requiring hospital follow-ups, blood tests, other monitoring and management of drug side effects and toxicity.
GP support.
Hospital admission for severe uncontrolled disease.
Mental health input due to impact of severe disease.
High social costs due to inability to work, costs of carers and impact on family.

Costs of being on different classes of drugs is not considered in the model (apart from the actual drug costs). Conventional systemic drugs and JAK inhibitors require frequent blood tests. Biologics including dupilumab and tralokinumab require minimal blood tests, and less frequent hospital attendances once stable.

3) Are the recommendations sound and a suitable basis for guidance to the NHS?

The BAD does not consider these recommendations, as they stand, to be sound or a suitable basis for guidance to the NHS, for reasons outlined above and below.

The most important group of patients who need these drugs are those who have tried all existing treatments for eczema, and none have worked. These patients are desperately in need of new treatments to allow them to return to productive living, as evidenced by the uptake of JAK inhibitors on the EAMS and FOC schemes, where these drugs have proved life-changing for some.

There is a need for drugs of different classes to be available because eczema is a long-term disease affecting patients of all ages and with differing co-morbidities and preferences. Clinical judgement will influence treatment choice for individual patients, based on efficacy, adverse effects profile, pre-existing co-morbidities and cost. For this reason, it is not appropriate to dictate a rigid sequence of treatments.

Currently, the free of charge schemes run until October 2022, after which patients may not be able to access drugs which are currently working for them.

4) Are there any aspects of the recommendations that need particular consideration to ensure we avoid unlawful discrimination against any group of people on the grounds of race, gender, disability, religion or belief, sexual orientation, age, gender reassignment, pregnancy and maternity.

We agree with the committee decision to recommend that adolescents are treated the same as adults to avoid age discrimination.

We note that the Scottish Medicines Consortium has approved tralokinumab (https://www.scottishmedicines.org.uk/medicines-advice/tralokinumab-adtralza-full-smc2403/) and upadacitinib (https://www.scottishmedicines.org.uk/medicines-advice/upadacitinib-rinvoq-full-smc2417/), and our understanding is that abrocitinib may be approved shortly. The current recommendation would therefore create a situation where patients in Scotland have access to the medications under assessment, while patients in England might not.

Download the response as a PDF here.