Comparative effectiveness and safety of upadacitinib versus vedolizumab in anti-TNF-exposed patients with ulcerative colitis.

BACKGROUND & AIMS: Head-to-head comparative evidence between upadacitinib and vedolizumab for anti-tumor necrosis factor (TNF)-exposed ulcerative colitis (UC) patients is lacking. This study compared their real-world effectiveness and safety. METHODS: This multicenter retrospective cohort included anti-TNF-exposed UC patients with a partial adapted Mayo score ≥ 2 who received upadacitinib or vedolizumab between March 2021 and February 2026 and had a minimum of 16 weeks of follow-up. Clinical outcomes were assessed at weeks 16 and 26, with endoscopic evaluation at week 26. The primary outcome was steroid-free symptomatic remission (SFSR) at week 16, defined as rectal bleeding subscore = 0 and stool frequency subscore ≤ 1, with no corticosteroid use for at least 2 weeks prior to assessment. Propensity score-based inverse probability of treatment weighting (IPTW) was used to adjust for imbalance of baseline characteristics. RESULTS: A total of 242 patients were analyzed (120 vedolizumab, 122 upadacitinib). After IPTW, upadacitinib was superior in achieving week-16 SFSR (71.7% vs 54.6%; adjusted odds ratio [aOR] 2.11, 95% CI, 1.67-2.65, P < .001). At week 26, upadacitinib remained superior in SFSR (82.6% vs 61.8%; aOR 2.93, 95% CI, 1.85-4.66, P < .001) and also demonstrated higher rates of endoscopic improvement (75.3% vs 47.1%; aOR 3.42, 95% CI, 2.57-4.56, P < .001), and endoscopic remission (41.5% vs 22.2%; aOR 2.49, 95% CI, 1.71-3.63, P < .001). There was no significant difference regarding the risk of drug discontinuation between the two groups (aHR 0.46, 95% CI, 0.20-1.04, P = .058). However, the overall incidence of adverse events was higher with upadacitinib (0.475 vs 0.213 per person-year, P < .001), while the rates of serious adverse events were not significantly different between groups. CONCLUSIONS: Upadacitinib demonstrates superior effectiveness over vedolizumab in anti-TNF-exposed UC but carries a higher burden of adverse events, highlighting the need for personalized benefit-risk assessment to guide treatment selection.