S732 Comparative Long-Term Drug Survival of First-Line Biologic Agents for Ulcerative Colitis

Abstract

Introduction: Real-world data comparing long-term survival of first-line biologics for ulcerative colitis (UC) is limited. We sought to compare long-term drug survival related to non-response (NR) and adverse effects (AEs) among biologic-naïve UC patients initiating infliximab (IFX), adalimumab (ADA), or vedolizumab (VDZ). Methods: Retrospective cohort study of adult, biologic-naïve UC patients initiating IFX, ADA, or VDZ 6/1/14-11/30/20 at a large academic center. Electronic health records were reviewed for clinical data. The primary outcome was time to biologic discontinuation for primary/secondary NR (including colectomy), confirmed by documentation of inadequate therapeutic effect. The secondary outcome was time to discontinuation for AEs. Kaplan-Meier analysis (KMA) and Cox regression were used to compare outcomes. Patients were censored at loss of follow-up, discontinuation for reasons unrelated to outcome, colectomy for dysplasia, or death, whichever occurred first. Confounders for Cox analyses were chosen a priori from the literature; p< 0.025 was considered significant. The proportional hazards assumption was tested using Shoenfeld residuals. Results: 503 UC patients initiated IFX (n=194), ADA (n=169), or VDZ (n=140) 6/1/14-11/30/20. Baseline characteristics compared by biologic are presented in Table 1. Biologic discontinuation for NR occurred in 52/194 (26.7%) for IFX, 77/169 (45.6%) ADA, and 30/140 (21.4%) VDZ (p< 0.01). Discontinuation for AEs occurred in 41/194 (21.1%) for IFX, 22/169 (13.0%) ADA, and 7/140 (5.0%) VDZ (p< 0.01). KMA demonstrated separation in drug survival curves for NR (Fig 1A) and AEs (Fig 1C). For the NR Cox model (Fig 1B), ADA but not VDZ was associated with greater hazard for discontinuation vs IFX (HR 2.4, 95% CI 1.5-3.6); higher serum albumin was associated with lower hazard for discontinuation (HR 0.7, 95% CI 0.5-0.9). For the AE Cox model (Fig 1D), VDZ but not ADA was associated with lower hazard for discontinuation vs IFX (HR 0.2, 95% CI 0.1-0.5); dose escalation was associated with lower hazard (HR 0.4, 95% CI 0.2-0.6) and female sex was associated with greater hazard for discontinuation (HR 2.0, 95% CI 1.1-3.7). Common AEs involved joints, rashes, infections, and infusion reactions (Fig 1E). Conclusion: After adjusting for confounders, ADA but not VDZ had a higher rate of NR compared to IFX. VDZ but not ADA had a lower rate of AEs compared to IFX. Prospective studies comparing long-term outcomes of biologics in UC are needed to personalize treatment decisions.Table 1.: Disease characteristics of the patient cohort.Figure 1.: Modified pouchitis disease activity index (mPDAI) among pouch phenotypes.