P-040 Assessing Risk Factors Predicting Loss of Response to Vedolizumab in Ulcerative Colitis and Crohn's Disease: Outcomes from the VICTORY Consortium

identified and followed. Incidence rates of EIMs were reported during lines of therapy and compared between the VDZ and anti-TNF cohorts. Incidence rates per 100 person-years were calculated as the number of patients with newly diagnosed EIMs during lines of therapy divided by sum of person-time which was the time from the lines of therapy start date to the event date for those with an event of interest or the entire duration of lines of therapy for those without an event. Incidence rate ratios (IRR) were across cohorts. Descriptive analyses were performed for all lines combined and stratified by lines of therapy. Cox proportional hazards models were employed to estimate the hazard ratio for EIMs adjusting for patient demographic characteristics and baseline proxy IBD severity. Results: A total of 597 and 16,058 unique CD patients and 417 and 7083 unique UC patients on VDZ and anti-TNFs respectively were identified. Compared to anti-TNFs, CD patients treated with first or second line VDZ were more likely to develop “any EIMs” (IRR: 1.73; 95% CI, 1.33, 2.24). This finding remained after adjusting for disease severity (HR: 1.37; 95% CI, 1.06, 1.78). CD patients on VDZ were specifically more likely to develop erythema nodosum (IRR: 5.10; 95% CI, 1.85, 14.05), aphthous stomatitis (IRR: 4.54; 95% CI, 1.66, 12.47), arthropathy (IRR: 1.69; 95% CI, 1.31, 2.20), and primary sclerosing cholangitis (IRR: 5.47; 95% CI, 1.70, 17.68) compared to patients treated with anti-TNFs. UC patients on VDZ did not appear to have a statistically significant increase in “any EIMs” incidence compared to patients on anti-TNFs (IRR: 1.22; 95% CI, 0.87, 1.71) but were more likely to develop episcleritis (IRR: 4.90; 95% CI, 1.48, 16.24), and pyoderma gangrenosum (IRR: 7.45; 95% CI, 1.65, 33.61) specifically. Conclusions: This analysis provides evidence suggesting that patients on VDZ may be more likely to develop EIMs compared to patients on anti-TNF therapies. A potential confounding factor is that patients on VDZ may have more severe disease due to their increased use of steroids and immunomodulators before initiating therapy. However, the VDZ cohort still experienced higher risk of developing EIMs compared to the anti-TNF cohort after adjusting for proxy disease severity. Similar results were observed when stratifying by line of therapy. Thus, it is important to continue monitoring the impact of novel therapies on patient safety and outcomes as more real-world data become available.