Network meta-analysis results of safety outcome.

Backgrounds and aims
There are currently no studies comparing histologic remission of FDA-approved biologics for moderate to severe ulcerative colitis (UC), except for one head-to-head VARSITY trial. The current study employs a network meta-analysis to compare the efficacy, including histologic remission and safety of biologic agents for UC.
Methods
Using four electronic databases, including Pubmed, EMBASE, The Cochrane Library, and ClinicalTrials.gov, a search was conducted of all literature published until September 2022. Included were studies of randomized controlled trials with adult patients with moderate to severe UC using biologics approved by the FDA. An odd ratio with a 95 percent credible interval and ranking information was calculated for each endpoint.
Results
The results of the network meta-analysis did not reveal statistically significant differences among biological agents. However, the ranking information for each biological agent exhibited the following patterns. Vedolizumab was ranked first for overall efficacy endpoints in the maintenance phase, including histologic remission. Except for histologic remission, Ustekinumab was identified as the top-ranked drug for induction phase efficacy endpoints other than histologic remission. Adalimumab was identified as the top-ranked drug for maintenance phase corticosteroid-free remission. Vedolizumab was identified as the top-ranked drug in the induction phase for Treatment Emergent Adverse Events (TEAE). Adalimumab was identified as the top-ranked drug in the induction phase for infection. For TEAE and infection in the maintenance phase and Treatment Emergent Severe Adverse Events (TESAE) in both the induction and maintenance phases, Ustekinumab was determined to be the top-ranked medication.
Conclusions
Including histologic remission, for the overall efficacy endpoints in the maintenance phase, VDZ was identified as the first rank drug, but there was no statistically significant difference between biologics. Therefore, the generalization of the results of this study is bounded due to the intrinsic limitations of the study provided.