Paediatric non-infectious uveitis in Cape Town, South Africa: a retrospective review of disease characteristics and outcomes on immunomodulating treatment

BACKGROUND: Non-infectious uveitis is a well-reported cause of blindness in more developed countries, however data from sub-Saharan Africa is lacking. Here we aim to describe the diseases associated with paediatric non-infectious uveitis and the effect of currently available treatment in this setting. METHODS: A retrospective observational analysis of children with non-infectious uveitis from January 2010 to December 2017, attending the tertiary paediatric rheumatology and ophthalmology referral units in Cape Town was conducted. Statistical analysis utilising STATA13 software was performed with p < 0.05 considered significant. RESULTS: Twenty-nine children were identified: median age at first visit of 74 months (IQR 49–86 months), female to male ratio of 0.9:1, predominantly of mixed ancestry (72.4%). Juvenile idiopathic arthritis associated uveitis (JIAU) (48.3%), idiopathic uveitis (41.4%), sarcoidosis (6.9%) and Behcet’s disease (3.5%) were diagnosed. Chronic anterior uveitis (72.4%) was the most frequent finding. Fifty-five percent had complications at presentation and all children with idiopathic uveitis presented with cataracts. Only 6.5% of the JIA cohort had JIAU. All JIA children had chronic anterior uveitis. There were no differences between JIA children with uveitis and those without uveitis, for sex (p = 0.68) and race (p = 0.58). Significantly, children with uveitis presented at an overall younger age (p = 0.008), had oligo-articular JIA (p = 0.01) and were antinuclear antibody positive (p < 0.001). Children with idiopathic uveitis were predominantly male (66.6%) with chronic anterior uveitis (41.7%). Nineteen children (65.5%) in the cohort had inactive disease on treatment at 12 months from diagnosis, which included 10 on topical corticosteroid therapy. At the last clinical visit 17 (58.6%) on standard initial therapy, 8 (27.6%) on tumour necrosis factor inhibitors and 2 on additional DMARDs were in remission. Five of these children still required topical corticosteroids. Surgery was performed in 41.4%, primarily in the idiopathic group. Visual acuity improved or was maintained on treatment. CONCLUSION: Current practice seems to detect children with potentially sight-threatening disease but the high rate of complications and the low percentage of children with JIAU raises concerns of delayed healthcare intervention. Tumour necrosis factor inhibitors have improved outcomes in refractory cases in this cohort, however further studies are needed.