FRI0438 Risk factors for uveitis in patients with ankylosing spondylitis: a retrospective cohort study

Background The uveitis is one of the most common extra-articular manifestation of ankylosing spondylitis (AS), occurring in 20–30% of the patients with AS. But it is not clear which factors are associated with the occurrence of uveitis in patients with AS. Objectives To evaluate the risk factors for uveitis in patients with AS. Methods A total of 1061 Korean patients who were diagnosed with AS at Seoul National University Hospital between January 2004 and December 2015 were evaluated in this study. Using electronic medical record review, the information on the occurrence of uveitis after AS diagnosis and the other demographic and clinical characteristic were obtained, which included age, gender, disease duration, previous history of uveitis, smoking status, Bath ankylosing spondylitis activity index, HLA-B27 positivity and level of serum inflammatory markers. After dividing the patients into those who experienced uveitis after AS diagnosis (uveitis group) and those who did not (non-uveitis group), the potential risk factors for the occurrence of uveitis were evaluated using univariable and multivariable logistic regression analysis. Receiver operating curve analysis was performed. Results Among the 1061 patients, 837 (78.9%) were male and their mean (SD) age was 27.9 (6.2) years. During the observation of 4231 patient-year, uveitis occurred in 142 (13.4%) patients (uveitis group). Patients in the uveitis group had longer disease duration than those in the non-uveitis group (10.0±6.6 vs. 8.2±6.2 years, p=0.003). The proportion of the patients with prior uveitis before diagnosis of AS or HLA-B27 positivity were higher in the uveitis group than those in the non-uveitis group. (58.5% vs. 12.1%, 95.7% vs.83.7%, respectively) There were no significant differences in ESR and CRP between the two groups. In the multivariable analysis, two most significant risk factors for new-onset uveitis was the prior uveitis before the diagnosis of AS (OR=11.18 [95% confidence interval (CI): 7.09–17.63]) and HLA-B27 positivity (OR=9.15 [95% CI: 1.23–68.41]). The area under the curve (AUC) of the final multivariable model was 0.809 (95% CI: 0.77–0.85). Multivariate analysis showed that history of uveitis and the presence of HLA-B27 as the final risk factors for the occurrence of uveitis. Conclusions Prior uveitis before the diagnosis of AS and HLA-B27 positivity were associated with the occurrence of uveitis. These results can be useful to predict occurrence of uveitis in patients with AS. Disclosure of Interest None declared