Poster Session II - A251 REAL-WORLD EFFECTIVENESS, SAFETY, AND DOSE MODIFICATIONS OF UPADACITINIB IN PATIENTS WITH CROHN’S DISEASE: A CANADIAN MULTICENTER STUDY

Background Upadacitinib (UPA) is widely used for the treatment of Crohn’s disease (CD) but data on clinical outcomes in daily practice remain scarce. Aims To assess the real-world effectiveness, safety, and dose modifications of UPA in patients with CD. Methods This multicenter retrospective cohort study included adults with CD initiating UPA at 7 Canadian centers (2019/10–2025/04). Patients were followed from treatment start until database lock (June 2025), discontinuation, or loss to follow-up (censored). The primary endpoint was corticosteroid-free clinical remission (CFCR; HBI <5 off systemic corticosteroids ≥90 days). Biochemical (FCP <250 µg/g and/or CRP <10 mg/L) and endoscopic (SES-CD ≤4) remission rates were captured. We documented adverse events (AEs), analyzed drug persistence by Kaplan–Meier, and evaluated predictors of CFCR by Cox regression. Results We included 295 patients (mean age 44.9 ± 15.9 years; 46.1% female) with a median follow-up of 37.7 weeks (IQR 42.4). During the maintenance phase, 24 (8.1%) patients underwent dose escalation. Prior exposure to any and ≥3 advanced therapies was noted in 93.2% (n = 275) and 39.3% (n = 116), respectively, and 25.1% (n = 74) used corticosteroids at baseline. UPA starting doses were 15 mg in 6.4% (n = 19), 30 mg in 8.1% (n = 24), and 45 mg in 84.4% (n = 249) of patients. At the end of follow-up, 55.3% (n = 163) achieved CFCR; clinical and biochemical remission rates were 60.0% (n = 177) and 44.4% (n = 131). CFCR was achieved in 40.0% (n = 8), 63.5% (n = 47), 61.9% (n = 52), and 48.3% (n = 56) of patients exposed to 0, 1, 2, or ≥ 3 advanced therapies, respectively. Endoscopic remission occurred in 24/61 colonoscopies. There were 145 AEs reported in 104 patients, predominantly acne (n = 28), and infections (n = 29; 8 classified as serious (S)AEs). Of all AEs, 10.3% were SAEs, including thrombo-embolic events (n = 2), herpes zoster (n = 1), and malignancy (n = 1). There were 29 (9.8%) patients who experienced 40 CD-related hospitalizations, and 26 (8.8%) patients underwent 35 surgeries: 12 intestinal resections, 3 colectomies, 16 perianal, and 4 other IBD-related procedures. UPA was discontinued in 19.3% (n = 57), mainly for lack of efficacy (n = 30) or AEs (n = 13). Drug persistence at year 1 was 73.8%. No factors were significantly associated with CFCR on Cox regression. Conclusions In this large and refractory Canadian CD cohort with extensive prior exposure to advanced IBD therapies, a substantial proportion of patients experienced clinical benefit from UPA. Treatment persistence was high while SAEs were rare, and AEs resulted in treatment discontinuation in < 5%.