Efficacy and Safety of Nemolizumab for Treatment of Adult Atopic Dermatitis.

Abstract

To the Editor: We were interested to read the meta-analysis by Xuemin et al [1] on the efficacy and safety of nemolizumab for treatment of adult atopic dermatitis. This is an important area, and an overview of the available data is timely. We concur with the conclusions of the authors that nemolizumab is a promising treatment, and Galderma is currently undertaking further phase 3 clinical trials to gather new data. On one important point, however, we struggle to reconcile the results and conclusions of the review with our knowledge of the published data. The authors comment that the 60-mg Q4w regimen is likely to be optimal. The only complete phase 3 trial that has compared that dose [2] on a proper likefor-like basis with a lower dose currently in late development (30 mg Q4w) is actually in favor of the lower dose. The percent change in the Eczema Area and Severity Index (EASI) (from baseline to week 16) for the 30-mg Q4w dose vs placebo (difference, –23.7; 95%CI, –8.9 to –38.5; P=.002) was reported in the phase 2b trial of Silverberg et al [3]. The treatment effect observed with EASI was greater than that observed with the 60-mg Q4w dose in the phase 3 trial by Kabashima et al [2] (difference, –12.6; 95%CI, –1.3 to –24.0). Using the method of Bucher et al [4] to undertake a formal indirect comparison, we observed a numerically greater benefit for 30 mg Q4w (difference, –11.1; 95%CI, 7.6 to –29.8) (Figure). The 30-mg dose is under evaluation in 2 pivotal phase 3 trials (NCT03989349 and NCT03985943), and analysis of the outcomes of both will help us resolve this important question. We do not recognize some of the data included by the reviewers in their work; for example, the results above for Silverberg et al [3] derived from the clinical study report are not reflected in the report by Xuemin et al [1]. This is troubling. While the overall conclusions are supported by individual trials, the dosage conclusions do not reflect current data.