Long-Term Comparative Outcomes of TNF-α Antagonists vs. Vedolizumab as First-Line Biologic Therapy for Refractory Ulcerative Proctitis: A Propensity-Matched Study

Background: The optimal first-line biologic therapy for refractory ulcerative proctitis (UP) remains uncertain, largely because patients with UP are frequently excluded from biologic clinical trials, limiting evidence to guide treatment selection. This study evaluated outcomes among patients with UP treated with first-line TNF inhibitors or vedolizumab. Methods: We performed a retrospective cohort study using the TriNetX database between 1995 and 2023. Propensity score matching was applied to balance demographics, laboratory parameters, and baseline medications. Primary outcomes included corticosteroid use, all-cause emergency room (ER) visits and hospitalizations, and colectomy, assessed at 6, 12, and 24 months after initiation of TNF inhibitors or vedolizumab. Secondary outcomes described real-world biologic usage patterns in UP. Results: Among 641 patients with UP receiving advanced therapy, the most commonly used biologics were adalimumab (39%), infliximab (27%), and vedolizumab (26%). Ustekinumab was used in 12% of patients. In matched analyses, TNF inhibitor therapy was associated with reduced ER visits and hospitalizations at 6 and 12 months compared with vedolizumab (6 months: 14.3% vs. 25%, aOR 0.50, p-value 0.03; 12 months: 16.1% vs. 35.2%, aOR 0.35, p-value 0.01). By 24 months, no significant differences were observed. Corticosteroid use and colectomy rates were similar across therapies at all time points. In a subgroup comparison between adalimumab and vedolizumab, results were consistent with the primary analysis, with lower short-term ER visits and/or hospitalizations among patients receiving adalimumab. Conclusions: In this propensity-matched analysis, TNFi therapy was associated with lower short-term healthcare utilization, with no significant differences observed in corticosteroid use. These findings should be interpreted cautiously given nonspecific outcomes and potential residual confounding from unmeasured disease variables such as endoscopic activity.