P677 The Oxford Anti-TNF Study of Immunogenicity and Sustained Response (OASIS-IBD): Analysis of 40,000 Patient Years of Anti-TNF Use

Abstract

Abstract Background Anti-TNF therapy is established as a mainstay in IBD management. Recent data raise controversies regarding efficacy of these agents, sustainability of response, and identification of factors that may influence outcomes, including dose escalation and need for long-term concurrent immunosuppression (Lancet Gastro.Hep.2019;4:341). Methods We report a retrospective cohort study of 699 IBD patients on anti-TNF at our centre between 2005-2022. Clinical data were obtained from patient records and pharmacy database (TABLE 1). Primary outcomes were overall loss of response (LOR) at any timepoint; and secondary LOR (after an initial clinical response to therapy). Cox regression analysis was performed to identify variables independently affecting outcome. Kaplan-Meyer survival curves were constructed to demonstrate overall response rates (persistence on drug) and secondary LOR (Figure 1a-d). Dose adjustments were made proactively in a Virtual Clinic. Results 204 patients had a diagnosis of ulcerative colitis (UC), 481 Crohn’s Disease (CD) and 14 IBD-unclassified. 126 patients received both infliximab (IFX) and adalimumab (ADA). Median duration on IFX was 42 months (range 0-204); median duration on ADA was 48.5 months (range 0-180). 48% of patients on IFX underwent dose escalation, as did 40% on ADA. Overall, 58 patients (7.6%) remained on treatment at 120 months. Overall survival rates for IFX at 12 months were 83% (82%CD vs. 82%UC); and at 36 months were 71% (69%CD vs. 74%UC). Overall survival rates for ADA at 12 months were 84% (89%CD vs. 59%UC); and at 36 months were 72% (78% CD vs. 42% UC) (Figure 1a-b). Secondary LOR rates for IFX at 12 months were 6% (5%CD vs. 9%UC); and at 36 months were 16% (15%CD vs.19%UC). Secondary LOR rates for ADA at 12 months were 7% (5%CD vs. 19%UC); and at 36 months were 16% (12% CD vs. 42% UC) (Figure 1c-d). There was no difference in survival analysis for IFX between CD and UC. On Cox regression analysis, for IFX, immunomodulation was associated with reduction in overall LOR (HR=0.48, p=1.07e-05) but not secondary LOR (HR=0.78, p=0.31) (Figure 1e-f). For ADA, outcomes were better for CD vs. UC for both overall LOR (HR=0.35, p=3.29e-07) and secondary LOR (HR=0.35, p=9.99e-05). Conclusion We demonstrate encouraging response rates in the first three years of therapy, with a small proportion remaining on therapy beyond 120 months. Survival rates on IFX were similar for UC and CD, but for ADA were significantly lower in UC. Secondary LOR was not affected by immunomodulation. Further analysis of relation to drug levels, dose escalation and genotyping is underway. This work was supported by The Leona M & Harry B Helmsley Charitable Trust.