AMSTERDAM – Etanercept does not appear to increase the risk of uveitis relative to methotrexate in children with juvenile idiopathic arthritis (JIA), according to a large retrospective cohort study presented at the EULAR 2018 Congress.
“When we added in a fully adjusted hazard ratio, it showed a lower risk in the development of uveitis in patients on etanercept [compared with methotrexate],” reported Rebecca Davies, a research assistant at the University of Manchester (England).
These data were characterized as reassuring for JIA patients being considered for etanercept, but Ms. Davies was cautious about suggesting that etanercept has a protective effect. Even though hazard ratios were calculated with propensity-adjusted Cox regression analyses, Ms. Davies believes the best interpretation of these data is that etanercept therapy is not likely to contribute significantly to the risk of uveitis.
The substantial differences between the comparator groups provide one reason to refrain from speculating that etanercept is protective against uveitis. The risk of uveitis is greater both in younger patients and in the first year after diagnosis. Patients in the etanercept group were older than patients in the methotrexate group and they started etanercept a longer time after the JIA diagnosis.
“The initiation of etanercept later in the disease may mean that more of these patients had already passed through the window of greatest risk,” Ms. Davies explained.
The data for this study were drawn from a British Society for Pediatric and Adolescent Rheumatology cohort registry. Confined to patients first starting as opposed to restarting therapy, 1,009 patients initiating etanercept were compared to 508 patients initiating methotrexate.
In addition to an older age (11 vs. 9 years) and disease duration at start of therapy (3 vs. 1 years), a lower proportion of patients in the etanercept group had a persistent oligoarthritis subtype (5% vs. 17%). During follow-up, there were 15 cases (0.15%) of uveitis in the etanercept group and 18 (3.5%) in the methotrexate group.
The crude incidence of uveitis was 0.6 per 100 patient-years for etanercept versus 2.4 per 100 patient-years for methotrexate, according to Ms. Davies. After adjustment for a broad number of variables, including age, gender, disease scores, disease duration, baseline steroid use, and the presence of comorbidities, there was still a 70% lower risk of uveitis among those treated with etanercept (hazard ratio 0.30; 95% confidence interval, 0.1-0.9).
The low relative rate of uveitis after starting etanercept is discordant with several previous studies, according to Ms. Davies. In two retrospective studies conducted in the United States and one in Canada, etanercept treatment was associated with higher rates of uveitis than other tumor necrosis factor inhibitors. In a German study, etanercept was associated with a higher risk of uveitis than that of methotrexate.
Although a highly effective anti-inflammatory agent such as etanercept might be expected to have a protective effect against uveitis, at least relative to methotrexate, Ms. Davies suggested that the previous reports of potential causal association and the limitations of this retrospective analysis require a more cautious interpretation.
“It is possible that those considered to be at high risk of developing uveitis were kept away from etanercept,” said Ms. Davis, providing one of several explanations why skepticism is needed in regard to assuming uveitis protection from etanercept.