Efficacy and Safety of Ruxolitinib, Crisaborole, and Tapinarof for Mild-to-moderate Atopic Dermatitis: a Bayesian Network Analysis of RCTs

Given the lack of head-to-head studies of novel non-steroidal molecule topical therapies in mild-to-moderate atopic dermatitis (AD), network meta-analyses (NMAs) can provide comparative efficacy and safety data for clinical decision-making. In this NMA, we performed a literature search until 01 March 2023 for eligible studies written in English using databases, including PubMed, EMBASE, Cochrane Library, and ClinicalTrials.gov. Only double-blind randomized clinical trials (RCTs) with topical Ruxolitinib, Crisaborole, or Tapinarof versus vehicle for patients with mild-to-moderate AD were included. Baseline and follow-up data were extracted. Efficacy was evaluated using Investigator's Global Assessment (IGA) achieving "clear" or "almost clear," with 2 points or more improvement from baseline at the end of treatment, referred to as "IGA success." For binary outcomes, we analyzed in random-effects Bayesian NMA consistency models to compare the efficacy of these 3 topical therapies by odds ratio (OR) with 95% credibility interval (CrI). Overall, 10 phase 2 or phase 3 RCTs were identified, which included 4010 patients with mild to moderate AD. Compared with the topical vehicle control, all these 3 treatments had higher response rate of "IGA success" at the end of trial (Ruxolitinib 1.5% b.i.d: OR, 11.94; 95%CrI, 6.28-23.15; Crisaborole 2% b.i.d: OR, 2.08; 95%CrI, 1.46-3.52; Tapinarof 1% b.i.d: OR, 2.64; 95%CrI, 0.75-9.70). Notably, Ruxolitinib 1.5% b.i.d. had the highest probability of achieving "IGA success" in ranking analysis (Rank 1, SUCRA = 0.75) and lower risk of AE (Rank 8, SUCRA = 0.22). Besides, there was no difference in treatment-related adverse events between 3 therapies. Heterogeneity was not significant across studies.