Comparing the efficacy and tolerability of biologic therapies in psoriasis: an updated network meta-analysis

BACKGROUND/OBJECTIVES The rapid expansion of psoriasis biologics has led to an urgent need to understand their relative efficacy and tolerability to better inform treatment decisions and specifically, to inform guideline development. METHODS We searched MEDLINE, PubMed, EMBASE and Cochrane databases for randomized controlled trials (RCTs),published up to 7th September 2018, of 11 licensed, NICE-approved biologics targeting TNF(adalimumab, etanercept, infliximab, certolizumab pegol), IL-12/IL-23p40 (ustekinumab), IL-17A (secukinumab, ixekizumab), IL-17RA (brodalumab) and IL-23p19 (guselkumab, tildrakizumab, risankizumab). Data were extracted and synthesised using RevMan 5.3. A frequentist network meta-analysis, conducted using Stata13 (StataCorp), ascertained direct/indirect evidence comparing biologics with one another, methotrexate, or placebo. This was combined with hierarchical cluster analyses to consider efficacy (PASI90 or PGA0/1; PASI75; DLQI improvement) and tolerability (drug withdrawal due to adverse events) outcomes at 10-16 weeks, followed by study quality, heterogeneity, and inconsistency assessments. RESULTS We identified 62 RCTs presenting data on direct comparisons (n=31,899 participants). All biologics were efficacious compared with placebo or methotrexate at 10-16 weeks. Hierarchical cluster analyses revealed that adalimumab, brodalumab, certolizumab pegol, guselkumab, risankizumab, secukinumab, tildrakizumab and ustekinumab were comparable with respect to high short-term efficacy and tolerability. Infliximab and ixekizumab clustered together, with high short-term efficacy but relatively lower tolerability compared with other agents, although the number of drug withdrawal events across the network was low, so these findings should be treated with caution. CONCLUSIONS Using our methodology, we find most biologics cluster together with respect to short-term efficacy and tolerability, and we do not identify any single agent as 'best'. These data need to be interpreted in the context of longer-term efficacy, effectiveness data, safety, posology, and drug acquisition costs when making treatment decisions.