(036) Serious Infection Risk Associated With Biologics and Systemic Nonbiologics in Patients With Psoriasis: A Systematic Review and Network Meta-Analysis of Randomised Controlled Trials

Background: Psoriasis is a common inflammatory skin condition affecting people of all age groups. Biologic treatments are effective in the treatment of moderate to severe psoriasis. However, biologic-induced serious infection is a concern due to their immunomodulatory properties.

Objectives: To determine the serious infection risk associated with biologics used for psoriasis from available evidence from randomised controlled trials (RCTs).

Methods: Medline, Embase, CINAHL, Cochrane register and ClinicalTrials.gov were searched for RCTs comparing biologics and systemic nonbiologics with each other or placebo in all patients with moderate to severe psoriasis from inception through July 17, 2023. Studies with less than 50 patients, predominantly psoriatic arthritis patients or if published in non-English languages were excluded. Infections that led to hospitalisation, intravenous antibiotics or death were considered serious. The number of patients with serious infection, and number of patients randomised in individual trials during the induction period of 8 to 24 weeks were used to compute relative effects at the drug and class levels using a fixed effect model. We conducted pairwise and network meta-analysis (NMA), and generated network plots, league tables and surface under the cumulative ranking (SUCRA) curve values using the Netmeta package in R. Side splitting was used to assess coherence in the networks.

Results: Our search resulted in 1115 records after deduplication. Of these, 109 RCTs were eligible and 101 involving 43653 patients were included in the quantitative analysis after excluding those comparing different doses of the same drug. Patients were predominantly male (up to 85%) with a mean age ranging from 14 years to 52 years and received one of the 18 drugs or placebo. Side splitting indicated network coherence (p>0.05). In the pairwise meta-analysis, risankizumab had a lower serious infection risk than ustekinumab (OR 0.16; 95% CI 0.04 to 0.60). In the drug level NMA, none of the drugs were associated with serious infection risk when compared to placebo or each other. At the class level, interleukin 23 inhibitors (IL-23i) showed a lower serious infection risk compared to placebo (OR 0.46; 95% CI 0.22 to 0.98). Ustekinumab showed an increased risk compared to IL-23i (OR 2.48; 95% CI 1.03 to 6.00). Bimekizumab ranked safest in the SUCRA rankings followed by guselkumab, risankizumab and tildrakizumab. Serious infection associated deaths were rare (n=3).

Conclusions: This study shows a lower serious infection risk with IL-23i compared to placebo and ustekinumab. IL-17i, TNFi and nonbiologics showed no difference compared to placebo or any other drug classes. Observational studies of more than 24 weeks assessment duration are needed to evaluate longer term risk.