Abstract
Introduction: Novel therapeutic agents demonstrated superior efficacy over placebo in psoriatic arthritis (PsA) patients. However, there is no direct comparison between different novel therapeutic agents. Therefore, we compared the efficacy of these therapeutic agents by network meta-analysis.
Methods: We searched for randomised controlled trials using ISI Web of Science, Scopus, Medline, Cochrane library, Clinicaltrials.gov and EMBase. Studies reporting the proportion of PsA patients that achieved ACR20 response were included. They were stratified into two groups: (1) biologic naive; and (2) TNFi intolerance. Results were analysed by using R statistics version 3.3.3 with "netmeta" version 0.9-4. Summary odds ratios and 95% confidence intervals were estimated using random effects model. P-rank scores were used to estimate the best therapeutic option for a specific group of PsA patients.
Results: In this network meta-analysis, 27 studies were included; 19 were in biologic-naive and 8 were in TNFi-intolerant populations. Results are summarised in the Table. Apremilast was the best option in TNFi-intolerant patients (P-rank score: 0.8437), but the differences in efficacy between apremilast and other therapeutic agents were not statistically significant. Etanercept was the best treatment option in biologic-naive patients (P-rank score: 0.8846), superior to adalimumab, apremilast, tofacitinib and ustekinumab. However, etanercept was not significantly more efficacious than certolizumab, golimumab, infliximab, ixekizumab or secukinumab.
Conclusions: All novel therapeutic agents show similar efficacy in PsA patients with TNFi failure. Etanercept is more efficacious than adalimumab, apremilast, tofacitinib and ustekinumab in biologic-naive patients.
