Efficacy of Dupilumab in Different Racial Subgroups of Adults With Moderate-to-Severe Atopic Dermatitis in Three Randomized, Placebo-Controlled Phase 3 Trials

August 2019 | Volume 18 | Issue 8 | Original Article | 804 | Copyright © August 2019

Andrew F. Alexis MD MPH,a Marta Rendon MD,B Jonathan I. Silverberg MD,c David M. Pariser MD,D Benjamin Lockshin MD,e Christopher E.M. Griffiths MD,f Jamie Weisman MD,g Andreas Wollenberg MD,h Zhen Chen PhD,i John D. Davis PhD,i Meng Li PhD,j Laurent Eckert PhD,k Abhijit Gadkari PhD,i Brad Shumel MD,i Ana B. Rossi MD,l Neil M.H. Graham MD,i Marius Ardeleanu MDi

ªDepartment of Dermatology, Mount Sinai West, New York, NY

BRendon Center for Dermatology and Aesthetic Medicine, Boca Raton, FL

cDepartment of Dermatology, Preventive Medicine and Medical Social Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL

DDepartment of Dermatology, Eastern Virginia Medical School and Virginia Clinical Research, Inc., Norfolk, VA

EU.S. Dermatology Partners, Rockville, MD

fDermatology Centre NIHR Manchester Biomedical Research Centre, University of Manchester, Manchester, UK

gAdvanced Medical Research, PC, Atlanta, GA

hDepartment of Dermatology and Allergy, Ludwig Maximilian University, Munich, Germany

IRegeneron Pharmaceuticals, Inc., Tarrytown, NY

jSanofi, Bridgewater, NJ

kSanofi, Chilly-Mazarin, France

lSanofi Genzyme, Cambridge, MA

deaths; and conjunctivitis from baseline to week 16. The generic term “conjunctivitis” was used to summarize a cluster of Medical Dictionary for Regulatory Activities Preferred Terms including conjunctivitis, allergic conjunctivitis, viral conjunctivitis, bacterial conjunctivitis, and atopic keratoconjunctivitis. Individual TEAEs by racial subgroup were not analyzed due to low numbers of patients in each subgroup.

The PK profile among different racial subgroups was assessed as functional dupilumab levels in serum.26 The lower limit of quantification of functional dupilumab concentration was 0.078 mg/L in undiluted human serum.

Statistical Analysis

Baseline, efficacy, and safety data for all three studies were pooled. For CHRONOS, efficacy and safety data were used through week 16, with week 52 data not reported due to insufficient numbers of patients in some racial subgroups to enable statistical analyses at that time point.

Efficacy analyses were performed on the full analysis set, which included all randomized patients. Safety analyses by study drug were performed using the safety analysis set, which included all randomized patients who received at least one dose of any study drug. PK analyses were performed on the PK analysis set, which included all patients who had a PK measurement for a particular time point.

All efficacy endpoints except IGA and EQ-5D-3L were analyzed using an ANCOVA model with baseline measurement as covariate, and treatment, region, and baseline IGA strata as fixed factors for all studies, and study identifier. Values after first rescue treatment were set to missing (censored) and imputed using the multiple imputation method. P values (nominal) were derived from an ANCOVA model using baseline as covariate, and treatment, subgroup, treatment-by-subgroup interaction, region, and baseline IGA strata as fixed factors. The ANCOVA model was further adjusted with baseline weight (kg) as an additional factor to generate weight-adjusted results for continuous endpoints. IGA and EQ-5D-3L responder analyses were conducted using a Cochran–Mantel–Haenszel (CMH) test stratified by region, baseline disease severity (IGA=3 vs IGA=4), and study identifier. CMH models are not amenable to adjustment with continuous variables, therefore, body weight adjustment analyses were not conducted for responder analyses.

Safety outcomes were summarized using descriptive statistics. Statistical significance of differences in efficacy between the dupilumab dose groups was not investigated.

All analyses were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC, USA).



The pooled analysis population included 1,429 White patients, 501 Asian patients, and 128 Black/African American patients (Table 1). Patient race was self-identified. White patients were enrolled in the United States of America (USA, 29.2%), Poland (16.9%), Germany (14.4%), Canada (12.0%), Spain (4.1%), Estonia (3.8%), the United Kingdom (UK, 3.1%), Italy (2.5%), the Netherlands (2.5%), Australia (2.4%), Hungary (1.9%), France (1.5%), the Czech Republic (1.4%), Lithuania (1.2%), Bulgaria (1.0%), Denmark (1.0%), Finland (0.6%), Romania (0.3%), and New Zealand (0.2%). Asian patients were enrolled in Japan (44.5%), Republic of Korea (21.2%), Canada (15.6%), the USA (11.8%), Singapore (2.8%), Hong Kong (1.0%), the Netherlands (1.0%), the UK (0.8%), Australia (0.6%), New Zealand (0.4%), Germany (0.2%), and Spain (0.2%). Black/African American patients were enrolled in the USA (92.2%) and Canada (7.8%).

Baseline demographics and disease characteristics were generally balanced among treatment groups21,22 and among racial subgroups in each of the studies (Table 1). Most patients had a median AD duration of >20 years, mean baseline EASI >30, and mean weekly Peak Pruritus NRS >7, showing a high disease burden at baseline.


In the White and Asian subgroups, both dupilumab regimens significantly improved AD signs (EASI, IGA), symptoms (POEM), including itch (Peak Pruritus NRS) and pain/discomfort (EQ- 5D-3L), and QoL (DLQI) vs placebo at week 16 (Figure 1A-J, P<0.0001). Weight-adjusted analyses support these findings (Figure 2A-H). In Black/African American patients, both dupilumab regimens led to significant improvement vs placebo in the EASI endpoints (Figure 1A-B) and mean change in Peak Pruritus NRS (Figure 1C) and DLQI (Figure 1E), while only dupilumab 300 mg qw showed significant improvement vs placebo in mean percent change in Peak Pruritus NRS (Figure 1D), POEM endpoints (Figure 1G-H), and proportions of patients achieving the IGA success endpoint (Figure 1I) (P<0.05). Weight-adjusted analyses are consistent with these findings (Figure 2).


Dupilumab was generally well tolerated, with an acceptable safety profile in all three trials.21,22 TEAEs occurred at similar rates across treatment groups. Conjunctivitis and injection-site reactions were more frequent in the dupilumab-treated groups in all studies, but were mild-to-moderate and rarely led to treatment discontinuation. AD exacerbations and skin infections occurred more frequently in the placebo groups.