FormalPara Key Summary Points

Why carry out this study?

There are multiple treatment options for atopic dermatitis, but relatively few head-to-head randomized controlled trials have compared treatments with one another.

Prior head-to-head randomized controlled trials have not assessed more stringent endpoints, which have been found to be associated with improved patient outcomes and have been highlighted as optimal targets for treat-to-target recommendations.

The study used population-adjusted indirect comparison to estimate how upadacitinib 15 mg would have performed if included in a head-to-head trial that had assessed efficacy of upadacitinib 30 mg against dupilumab 300 mg.

What was learned from the study?

Across all outcomes and timepoints assessed, the estimated absolute response rates were greatest for upadacitinib 30 mg, followed by upadacitinib 15 mg and then dupilumab.

Upadacitinib 30 mg appears to be the most efficacious treatment in attaining independent and simultaneous achievement of more stringent skin and itch outcomes for adults with moderate-to-severe atopic dermatitis, followed by upadacitinib 15 mg and then dupilumab 300 mg.

Introduction

Randomized controlled trials represent the gold standard of establishing efficacy of treatments, but for diseases where there are multiple treatment options, such as atopic dermatitis, randomized controlled trials comparing treatments with one another may be few. Indirect treatment comparisons may be used to evaluate therapeutic options in these situations.

Upadacitinib and dupilumab are two treatment options for atopic dermatitis. Upadacitinib is an oral, small-molecule Janus kinase inhibitor that has greater inhibitory potency for JAK1 than for JAK2, JAK3, or tyrosine kinase 2 [1]. Upadacitinib is metabolized in the liver, primarily via the CYP3A4 pathway and excreted predominantly unchanged in the urine and feces [2]. Dupilumab is a subcutaneously administered human monoclonal antibody that inhibits interleukin-4 and interleukin-13 signaling [3]. Dupilumab is metabolized into peptides and amino acids through catabolism [4]. Common adverse events for upadacitinib include acne, upper respiratory tract infection, nasopharyngitis, headache, plasma creatine phosphokinase elevation, atopic dermatitis (AD) worsening, nausea, and oral herpes; for dupilumab, common adverse events include conjunctivitis, allergic conjunctivitis, exacerbation of AD, nasopharyngitis, headache, fatigue, allergic rhinitis, cough, diarrhea, vascular disorders, injection-site reactions, non-skin infections, herpes viral infections, and upper respiratory tract infection [5, 6]. Both upadacitinib and dupilumab have been found to be well tolerated with a favorable benefit–risk profile compared with placebo [1, 3]. Efficacy of upadacitinib in treating atopic dermatitis has been assessed in trials, including Measure Up 1 (NCT03569293), Measure Up 2 (NCT03607422), and Heads Up (NCT03738397). Measure Up 1 and Measure Up 2 assessed efficacy of upadacitinib 30 mg and upadacitinib 15 mg against placebo, while Heads Up assessed efficacy of upadacitinib 30 mg in a head-to-head trial against dupilumab 300 mg. A head-to-head trial of upadacitinib 15 mg against dupilumab 300 mg, though, has not been conducted. Network meta-analysis has shown that upadacitinib 30 mg and upadacitinib 15 mg are among the most efficacious targeted systemic therapies, [7, 8] but prior indirect comparisons have not looked at stricter outcomes such as complete skin clearance and itch resolution.

This analysis builds on prior indirect comparisons in that it uses individual patient-level data for all treatment arms assessed across multiple trials to compare upadacitinib and dupilumab on more stringent skin and itch outcomes. This analysis aims to use indirect comparison to compare upadacitinib and dupilumab on stringent and composite measures of skin clearance and itch resolution to estimate how upadacitinib 15 mg would perform compared to dupilumab 300 mg.

Methods

This analysis is a post hoc analysis of clinical trial data. The analysis used individual patient data from three clinical trials: Measure Up 1 (NCT03569293), Measure Up 2 (NCT03607422), and Heads Up (NCT03738397). Measure Up 1 and Measure Up 2 were replicate multicenter, randomized, double-blind, placebo-controlled phase 3 trials with co-primary endpoints of achievement of at least 75% improvement in Eczema Area and Severity Index (EASI) score from baseline and achievement of a validated Investigator’s Global Assessment for Atopic Dermatitis (vIGA-AD) response (defined as a vIGA-AD score of 0, clear or 1, almost clear with ≥ 2 grades of reduction from baseline) at week 16. Patients were eligible on the basis of age (adolescents aged 12–17 years or adults aged 18–75 years) and disease criteria (moderate to severe AD [≥10% of body surface affected by AD, EASI score of ≥ 16, vIGA-AD score of ≥ 3, Worst Pruritus Numerical Rating Scale (WP-NRS) score of ≥ 4]) [1]. Heads Up was a head-to-head, multicenter, randomized, double-blinded, double-dummy, active-controlled phase 3b clinical trial which compared the safety and efficacy of upadacitinib 30 mg with dupilumab 300 mg. Patients were eligible on the basis of age (adults ages 18–75) and disease criteria (moderate-to-severe AD defined as ≥ 10% of body surface area affected by AD, EASI ≥ 16, vIGA-AD ≥ 3, and WP-NRS ≥ 4 at baseline) [9].

This analysis uses a population-adjusted indirect comparison, using covariate adjustment as a treatment effect calibration to estimate how upadacitinib 15 mg would have performed in a head-to-head comparison with dupilumab 300 mg. This study uses an “anchored” population adjustment because both studies have upadacitinib 30 mg as a common comparator. The theoretical basis of this analysis is laid out by Zhang et al. (2009) [10]; the covariate-adjustment method, an individual patient data calibration method mentioned in National Health and Care Excellence (NICE) guidance, was applied [11]. A logistic regression model was fit to the Measure Up 1 and Measure Up 2 pooled data to understand the relationship between upadacitinib 30 mg and upadacitinib 15 mg for each outcome, adjusting for patient-level baseline covariates (age, gender, race, country, previous use of systemic therapy, disease duration, vIGA-AD, and EASI). Using this estimated relationship, the differential between upadacitinib 30 mg and upadacitinib 15 mg was applied to the Heads Up data to estimate how upadacitinib 15 mg would have performed if included in Heads Up.

Prespecified efficacy outcomes included the achievement of no/minimal itch (WP-NRS 0/1), EASI score of ≤ 3 (EASI ≤ 3), 100% improvement in EASI (EASI 100), both EASI ≤ 3 and WP-NRS 0/1 simultaneously (EASI ≤ 3 & WP-NRS 0/1), both ≥ 90% improvement in EASI (EASI 90) and WP-NRS 0/1 simultaneously (EASI 90 & WP-NRS 0/1), and both EASI 100 and WP-NRS 0/1 simultaneously (EASI 100 & WP-NRS 0/1). Predicted differences were used to estimate absolute response rates at weeks 4, 8, 12, 16, and 24. The analysis included adult patients across all studies and used non-responder imputation.

This study did not require institutional review board approval as it did not access identifiable patient information or have direct interaction with patients. This post hoc analysis used data from previously conducted studies and did not contain data from any new studies with human participants or animals performed by any of the authors.

Results

Across all outcomes assessed at week 4, the estimated absolute response rates were greatest for upadacitinib 30 mg, followed by upadacitinib 15 mg and then dupilumab (Fig. 1). Numerical differences between upadacitinib 15 mg and dupilumab 300 mg were most pronounced on the achievement of EASI ≤ 3, WP-NRS 0/1, and EASI 90 & WP-NRS 0/1; differences favoring upadacitinib 15 mg over dupilumab 300 mg were statistically significant on these outcomes (Fig. 2).

Fig. 1
figure 1

Response rates at week 4 by endpoint. *Upadacitinib 15 mg response rates are predicted using the characteristics of the HEADS Up data and these values are compared with actual dupilumab results. DUPI dupilumab, EASI Eczema Area and Severity Index, UPA upadacitinib, WP-NRS Worst Pruritus Numerical Rating Scale

Full size image
Fig. 2
figure 2

Predicted difference of upadacitinib 15 mg versus dupilumab 300 mg at week 4 by endpoint. EASI Eczema Area and Severity Index, WP-NRS Worst Pruritus Numerical Rating Scale

Full size image

Estimated absolute response rates increased for all treatment arms over time, though the rank order remained consistent throughout: upadacitinib 30 mg, followed by upadacitinib 15 mg, and then dupilumab (see Supplementary Table 1). At week 16, the primary endpoint time point of these trials, a > 15 percentage-point difference was observed with upadacitinib 30 mg over dupilumab across all outcomes (Fig. 3). Statistically significant differences favoring upadacitinib 15 mg over dupilumab were observed for EASI ≤ 3, EASI 100, and EASI 100 & WP-NRS 0/1 (Fig. 4). These patterns also held at week 24, with statistically significant differences between upadacitinib 15 mg and dupilumab observed for EASI 100 and EASI 100 & WP-NRS 0/1 (Figs. 5 and 6).

Fig. 3
figure 3

Response rates at week 16 by endpoint. *Upadacitinib 15 mg response rates are predicted using the characteristics of the HEADS Up data and these values are compared with actual dupilumab results. DUPI dupilumab, EASI Eczema Area and Severity Index, UPA upadacitinib, WP-NRS Worst Pruritus Numerical Rating Scale

Full size image
Fig. 4
figure 4

Predicted difference of upadacitinib 15 mg versus dupilumab 300 mg at week 16 by endpoint. EASI Eczema Area and Severity Index, WP-NRS Worst Pruritus Numerical Rating Scale

Full size image
Fig. 5
figure 5

Response rates at week 24 by endpoint. *Upadacitinib 15 mg response rates are predicted using the characteristics of the HEADS Up data and these values are compared with actual dupilumab results. DUPI dupilumab, EASI Eczema Area and Severity Index, UPA upadacitinib, WP-NRS Worst Pruritus Numerical Rating Scale

Full size image
Fig. 6
figure 6

Predicted difference of upadacitinib 15 mg versus dupilumab 300 mg at week 24 by endpoint. EASI Eczema Area and Severity Index, WP-NRS Worst Pruritus Numerical Rating Scale

Full size image

Discussion

For diseases where there are multiple treatment options, as is the case with AD, head-to-head randomized controlled trials comparing treatments with one another may be few. Indirect treatment comparisons are valuable to provide evidence in the absence of multiple head-to-head trials. However, because they have not been frequently reported, prior indirect comparisons have not been conducted on the more stringent endpoints assessed here.

More stringent outcomes have been found to be associated with improved patient outcomes, including the achievement of both skin and itch resolution [12]. Developing treat-to-target recommendations supports stringent outcomes as optimal targets, highlighting the importance of both clinician- and patient-reported measures such as skin and itch assessments to inform disease management [13]. These indirect comparisons can enrich shared decision-making discussions between patients and physicians with evidence on these optimal treatment targets.

Indirect comparisons of treatments for moderate-to-severe AD have been conducted using network meta-analysis [7, 8, 14, 15].Previous analyses evaluated skin and itch improvement separately across various immunomodulatory treatments. Due to limited data availability, prior assessments have focused primarily on stringent skin improvement outcomes, specifically EASI 90 and vIGA-AD 0/1. Consequently, the indirect comparison of simultaneous achievement of both skin and itch improvement was not compared indirectly. This analysis partially fills this evidence gap by providing indirect comparison results on these novel stringent endpoints. Although the analysis presented here is limited to upadacitinib (15 mg and 30 mg) and dupilumab, the rank order and temporal dynamics of these treatments are generally consistent with prior indirect comparisons, with greater degrees of skin and itch efficacy observed with upadacitinib compared with dupilumab, and with the highest degrees of efficacy observed with upadacitinib 30 mg.

Limitations

Outside of a head-to-head clinical trial comparing upadacitinib 15 mg with dupilumab 300 mg, having individual patient data for all treatment options in an indirect treatment comparison is the next best option. This analysis was limited to select upadacitinib and dupilumab trials because individual patient-level data were not available for all AD immunomodulatory treatment trials. Covariate adjustment was used to account for imbalances across the included trials, and while this approach has both published theoretical underpinnings and applications, it assumes potential imbalances are sufficiently addressed by the baseline covariates included in the statistical model. Other indirect comparison approaches (such as network meta-analysis) were infeasible because the outcomes of interest are not commonly or consistently reported across AD immunomodulatory treatment trials. Pending future reporting of these stringent outcomes, other indirect comparison approaches may be employed and can complement these findings with the inclusion of additional treatments.

Conclusions

For adults with moderate-to-severe AD, upadacitinib 30 mg appears to be the most efficacious treatment in attaining independent and simultaneous achievement of more stringent skin and itch outcomes across multiple timepoints over 24 weeks, followed by upadacitinib 15 mg and then dupilumab 300 mg.