S2000 Safety of COX-2 Inhibitor NSAIDs in Inflammatory Bowel Disease: A Propensity-Matched Cohort Study Using Real-World Data

Introduction: Nonsteroidal anti-inflammatory drugs (NSAIDs) are widely used for pain and inflammation but are often avoided in inflammatory bowel disease (IBD) due to concerns about worsening disease activity through COX-1 inhibition and disruption of mucosal integrity. Selective and preferential COX-2 inhibitors may mitigate this risk, but real-world safety data in ulcerative colitis and Crohn’s disease are limited. Methods: We conducted a retrospective cohort study using the TriNetX platform to evaluate the association between prescribed NSAIDs and outcomes in ulcerative colitis (UC) and Crohn’s disease (CD). Patients were grouped as: (1) COX-2 inhibitors vs no NSAID, and (2) celecoxib vs meloxicam. Selective COX-2: celecoxib; preferential COX-2: meloxicam, nabumetone, etodolac. Propensity score matching was performed on demographics, comorbidities, recent ER/inpatient encounters, and markers of IBD severity, including recent use of steroids or IBD-specific therapies. The primary outcome was a composite of new intravenous steroids, oral steroids, new biologic/Janus kinase inhibitors (JAK)/S1P initiation and/or fecal calprotectin ≥200 µg/g. Secondary outcomes included the individual components, IBD-related surgery, International Classification of Diseases-based documentation of UC or CD complications, ER/inpatient encounters, mortality, and—for CD—bowel obstruction or intervention for stricturing/perianal disease. Patients who experienced the study outcomes prior to the index prescription were excluded. Results: Across 4 matched cohorts (2 in UC and 2 in CD), sample sizes ranged from 805 to 3,009 per group. In UC, the primary composite outcome occurred in 5.2%–5.8% of patients across comparisons, with no significant differences observed. In CD, the primary outcome ranged from 5.9%–7.0%, also without significant associations. No statistically significant differences were observed in steroid or biologic/JAK/S1P use, fecal calprotectin, surgery, documented complications, hospitalization, or mortality. Conclusion: In this large real-world analysis, prescribed COX-2 inhibitors were not associated with increased risk of IBD activity. These findings suggest COX-2 inhibitors may be a reasonable option in appropriately selected patients with UC or CD.