Efficacy and safety of treatments for atopic dermatitis: a model based meta-analysis of randomized controlled trials

We aimed to quantitatively compare the efficacy and safety of currently marketed Atopic Dermatitis (AD) therapies and placebo in AD patients, providing evidence-based guidance for clinical treatment selection.

Randomized controlled trials (RCTs), drug labels, and regulatory review documents related to AD treatments and placebo were retrieved from PubMed, FDA, EMA, NMPA, and PMDA databases from inception to July 11, 2023. The response rates of Eczema Area and Severity Index (EASI) 75, clear (0) or almost clear (1) with a ≥ 2-point improvement in Investigator's Global Assessment (IGA) including validated IGA (vIGA) (hereinafter referred to as IGA), EASI 90 and ≥ 4-point improvement in Pruritus Numerical Rating Scale (PP-NRS4) were used as the efficacy index. Time-effect models were subsequently established using a model-based meta-analysis (MBMA). For treatments not suitable for MBMA or subgroup analyses stratified by disease severity and age groups, meta-analyses were conducted. Safety outcomes were compared via pooled analysis.

A total of 274 articles and 4 FDA review documents were identified, with 77 studies and 2 FDA reviews ultimately included (total sample size: 20,262). Results showed that all topical and systemic medications demonstrated superior efficacy (EASI 75 response rate) to placebo. Among topical treatments, methylprednisolone showed the highest efficacy. For non-steroidal topicals, ruxolitinib cream exhibited the best efficacy, followed by tacrolimus ointment and difamilast ointment. Among systemic treatments, upadacitinib demonstrated the highest efficacy, followed by abrocitinib, with dupilumab ranking third as the most effective biologic. Baricitinib showed relatively lower efficacy. Biologics exhibited slower onset compared to other treatments. Safety analysis revealed higher overall adverse event incidence with systemic treatments versus topical treatments.

MBMA was conducted for the time courses of several efficacy index. In conclusion, results showed that there is still a lack of treatment methods that balance efficacy, safety and speed of onset. The results of this MBMA can be used as a reference for optimizing clinical medication and provide a basis for decision-making in the future development of new AD drugs.