(A-315) Risk of Incident Inflammatory Bowel Disease in New Users of Dipeptidyl Peptidase 4 Inhibitors, Sodium Glucose Cotransporter 2 Inhibitors, or Sulfonylurea in the U.S. Food and Drug Administration’s Sentinel System

Background: Previous study results were inconclusive regarding the potential risk of inflammatory bowel disease (IBD: Crohn’s disease [CD], ulcerative colitis [UC], and indeterminate colitis) after dipeptidyl peptidase-4 inhibitors (DPP4i) exposure.

Objectives: Compare the risk of incident IBD between new users of DPP4i and sodium-glucose co-transporter 2 inhibitors (SGLT2i) and between new users of DPP4i and sulfonylureas (SU).

Methods: In Sentinel from January 2008 to May 2024, we identified new users of DPP4i, SGLT2i, or SU aged ≥18 years who had type 2 diabetes and no evidence of IBD, IBD treatment, diverticulitis, colitis, intestine or colon surgery, endoscopy, or use of glucagon-like peptide-1 receptor agonists (GLP1RA) in the year before index dispensing. Incident IBD events were identified using diagnosis codes, preceded by endoscopy and biopsy, and followed by IBD treatment. Follow-up began on the index date and ended at the first occurrence of IBD, exposure to comparator drug or GLP1RA, treatment cessation, death, disenrollment, or study end. We excluded IBD events in the first 180 days of follow-up to rule out latent cases. Average treatment effect on the treated weights were calculated for SGLT2i and SU to balance observed baseline covariates. In weighted cohorts, we report Kaplan-Meier (KM) plots and adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for IBD and for CD and UC subgroups.

Results: In weighted DPP4i (n=588,233) and SGLT2i (n=574,152) cohorts, baseline variables were balanced (mean age 66 years and 50% female in both cohorts). Mean at-risk person-year (PY) was 1.22 and 0.93 PY, and 159 and 99 IBD cases were identified in the DPP4i and SGLT2i cohorts, respectively, yielding incidence rates of 22 and 19 cases/100,000 PY. Comparing DPP4i with SGLT2i cohort, HR was 1.18 (95% CI: 0.87, 1.59) for IBD, 1.28 (95% CI: 0.63, 2.59) for CD, and 1.01 (95% CI: 0.72, 1.40) for UC. KM plots suggested a potential violation of proportionality assumption.
In weighted DPP4i (n=394,487) and SU (n=391,339) cohorts, baseline variables were balanced (mean age 65 years and 51% female in both cohorts). Mean at-risk PY was 1.28 and 1.36 PY, and 162 and 161 IBD cases were identified in the DPP4i and SU cohorts, respectively, yielding incidence rates of 32 and 30 cases/100,000 PY. Comparing DPP4i with SU cohort, HR was 1.05 (95% CI: 0.86, 1.28) for IBD, 0.70 (95% CI: 0.45, 1.09) for CD, and 1.09 (95% CI: 0.85, 1.41) for UC.

Conclusions: Overall risk of IBD, UC, and CD did not appear to differ substantially between DPP4i and SGLT2i cohorts or DPP4i and SU cohorts in our study. For the DPP4i/SGLT2i comparison, the potential violation of proportionality assumption suggests the hazard ratio for IBD varied over time and merits additional analyses.