Uveitis occurrence in patients with ankylosing spondylitis according to the type of tumour necrosis factor inhibitor: a cohort study of 175 patients.

To compare the occurrence of non-infectious uveitis based on the type of tumour necrosis factor (TNF) inhibitor used to manage spondyloarthritis in ankylosing spondylitis (AS) patients.The occurrence (new-onset and recurrence) of uveitis was reviewed retrospectively in AS patients receiving different TNF inhibitor therapies (adalimumab [ADA], infliximab [IFX], etanercept [ETN], and golimumab [GOL]) for the management of spondyloarthritis from 2005 to 2018. Kaplan-Meier analysis was performed to calculate the cumulative occurrence rates of uveitis during TNF inhibitor therapy, and a log-rank test was used to analyse differences between the survival curves. Multivariable Cox proportional-hazards models were used to compute hazard ratios (HRs) of different TNF agents for uveitis occurrence after adjusting for concurrent confounding factors.The three-year cumulative occurrence rates of uveitis were significantly different according to the type of anti-TNFs used (23.1% in IFX, 18.5% in ETN, and 11.9% in ADA+GOL group) (p=0.020). The risk of new-onset uveitis was similar for different drugs. However, the IFX group showed a 5.4 times higher risk of recurrence than the ADA+GOL group (p=0.022). After adjusting for other confounding factors, IFX use was independently associated with a more frequent occurrence of uveitis in AS patients (HR=2.01; p=0.011).A significant number of AS patients who received anti-TNF therapy developed uveitis. Different types of anti-TNF drugs were associated with uveitis recurrence. Particularly, chimeric mouse-human monoclonal antibody (IFX) was found to increase the risk of uveitis occurrence compared to humanised monoclonal antibody (ADA or GOL).