Background/Purpose: Uveitis associated with JIA (JIA-U) is the most common extra-articular complication of JIA. As JIA-U is associated with significant risk of visual loss and impairment, achieving disease control as early as possible is crucial. Treatment with TNF-α inhibitors (TNFi) has significantly improved outcomes of JIA-U, yet studies show 25% of patients with JIA-U may fail treatment and remain at high risk for vision loss. In this study, we examined children with JIA-U within the CARRA II registry treated with TNFi in order to describe medication use, percentage of patients successfully treated with TNFi, characteristics associated with TNFi success, and medications used for treatment after TNFi failure.

Methods: Retrospective cohort study utilized data from the CARRA II JIA registry. We included all patients with diagnosis of JIA-U, 0-25 yrs., with uveitis onset before age 19 yrs. and identified a subset of patients who ever-used TNFi. We collected data about patient demographics, uveitis course and treatment. We defined: “TNFi failure” as discontinuing TNFi due to active uveitis or side effects or remaining on TNFi with active uveitis at end of study; “TNFi success” as stopping TNFi with quiet uveitis or remaining on TNFi with quiet uveitis at end of study. Medication use post-TNFi failure was summarized. Data was summarized using standard paired t-tests or chi-squared tests as appropriate to compare characteristics between the groups. Logistic regression was used to estimate the effects of baseline characteristics on TNFi success.

Results: 874 patients were diagnosed with JIA-U with uveitis onset before age 19 yrs. Of these, 617 were at some time treated with TNFi; in 58 of these patients, data were inadequate to determine if TNFi was successful in treating uveitis. Of the remaining 559 patients (77.46% female; 82.73% white race; 61.72% ANA positive; and 50.09% oligo-articular JIA), 418 (74.77%) were successfully treated with TNFi and 141 (25.22%) were determined to have failed TNFi. In a multivariate logistic regression model, race (white/non-white) and JIA subtype (oligoarticular/non-oligoarticular) were significant factors for TNFi success with whites and non-oligoarticular JIA more likely to succeed (OR=1.92, p=0.0104 and OR=1.54, p=0.0470, respectively). Age at uveitis diagnosis, sex and ANA status were not significantly associated with success. Of the 141 patients who failed TNFi, 117 (83%) had active uveitis while remaining on TNFi at the end of study and 24 were determined to be failures after halting TNFi. Of the later, 12 patients were started on the following medications within the year (mean gap 33.75 days): abatacept (n=1), tocilizumab (n=6), cyclosporine (n=1), methotrexate (n=2), tofacitinib (n=1), anakinra (n=1).

Conclusion: This first analysis of JIA-U patients in the CARRA II registry includes a large cohort of patients who used a TNFi for JIA-U. The percentage of patients successfully treated with TNFi is consistent with the current literature. White race and non-oligoarticular JIA were associated with successful response to TNFi. A variety of biologic and non-biologic medications were used after TNFi failure with tocilizumab the most frequent.

« Back to Pediatric Rheumatology Symposium 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/uveitis-treatment-in-the-carra-ii-registry-tumor-necrosis-factor-inhibitors-and-beyond/