DOP58 Non-white race is associated with decreased efficacy of tumor necrosis factor antagonist therapy in Ulcerative Colitis: A post-hoc analysis of individual level data from golimumab clinical trials

Abstract

Abstract Background Observational studies suggest non-white patients with inflammatory bowel disease (IBD) have worse clinical outcomes.1 There are limited data on whether race impacts response to biologic therapy. We therefore aimed to evaluate the efficacy of the tumor necrosis factor (TNF) antagonist golimumab comparing white to non-white participants, using individual participant level data from phase 2/3 randomized clinical trials of TNF antagonist therapy in ulcerative colitis (UC). Methods We conducted a pooled analysis of individual-level data from the induction and maintenance trials of golimumab in UC accessible through Yale University Open Data Access Project (YODA). There were insufficient non-white participants in infliximab studies accessible through YODA, precluding meaningful analysis. We analyzed patients in the placebo and treatment arms separately. Our primary outcome was clinical response and secondary outcomes were clinical remission and endoscopic healing according to clinical trial definitions. We compared white and non-white (defined as Black, Asian, or Other race) participants using multivariable logistic regression a priori adjusting for age, sex, treatment group, baseline Mayo score, immunomodulator and corticosteroid use. Effect estimates were expressed as adjusted odds ratios (aOR) and 95% confidence intervals (95% CI). Results A total of 1,006 participants were included in the induction trial (PURSUIT-SC; 18% non-white) and 783 participants in the maintenance trial (PURSUIT-M; 17% non-white). Non-white participants had significantly lower odds of week 6 clinical response (aOR 0.43, 95%CI 0.28–0.66), clinical remission (aOR 0.41, 95%CI 0.22–0.77) and endoscopic remission (aOR 0.48, 95%CI 0.30–0.74) compared to white participants (Figure). Non-white participants also had a lower adjusted odds of week 30 clinical response (aOR 0.64, 95%CI 0.40–1.01), clinical remission (aOR 0.45, 95%CI 0.28–0.74), and endoscopic remission (aOR 0.62, 95%CI 0.41–0.96). By week 54, the lower odds of these outcomes among non-whites were no longer statistically significant (Figure). Conclusion Non-white UC patients were less likely to achieve clinical response, clinical remission, and endoscopic healing with golimumab compared to white patients in these clinical trials. Further studies are needed to understand these differences and whether they are observed with other drugs or outside the context of clinical trials. Reference