ONE-YEAR COMPARATIVE EFFECTIVENESS OF UPADACITINIB VERSUS TOFACITINIB FOR ULCERATIVE COLITIS: A MULTICENTER COHORT STUDY

Background: We compared real-world outcomes of upadacitinib (upa) vs tofacitinib (tofa) for ulcerative colitis (UC) through 52 weeks.

Methods: In this retrospective cohort study, adults initiated upa or tofa for UC between 1/1/2020-2/1/2023 at two US academic centers. Electronic records were manually reviewed for clinical data. Patients with prior colectomy and those initiating therapy primarily for non-UC indications were excluded. The primary outcome was steroid-free clinical remission at 52 (+/- 4) weeks, defined using the following tiered criteria based on available documentation: A. Simple clinical colitis activity index (SCCAI) <2 points B. Partial Mayo score <2 points C. Provider global assessment of clinical remission and no use of oral corticosteroids at assessment. Secondary outcomes were SFCR at 12 (+/- 4) weeks, endoscopic response (improvement in Mayo endoscopic subscore [MES] by >1 point), and endoscopic remission (MES=0) within 52 weeks. Other outcomes were normalization of fecal calprotectin (<250 ug/g; only assessed among those with baseline elevation) within 52 weeks and treatment persistence at 52 weeks. Drug-related adverse events (AEs) within 52 weeks were reported descriptively. Inverse probability of treatment-weighted (IPTW) logistic regression was performed to compare upa vs tofa for primary/secondary outcomes.

Results: 155 unique patients initiated upa (n=81) and tofa (n=74). Baseline characteristics (Table 1) were similar, however upa patients more commonly had prior vedolizumab and ustekinumab failures. A higher proportion of upa patients met all outcomes except for endoscopic response (similar between groups) (Fig 1A). AEs for upa (n=14 among 11 patients) included peripheral edema, shingles, rash (n=2), pneumonia, diverticulitis, streptococcal pharyngitis, cellulitis, acne, angina (due to anemia), bacteremia, elevated liver enzymes, nausea, and neutropenia and for tofa (n= 12 among 10 patients) included Clostridioides difficile infection, pneumonia, chemical pregnancy / miscarriage (n=2), skin abscess, dental abscess, elevated liver enzymes, ankle swelling, norovirus infection, shingles (n=2), and DVT. After IPTW, which successfully balanced covariates (Fig 1B), upa was associated with significantly higher odds of SFCR at 12 weeks (OR 2.3) and 52 weeks (OR 3.0) and non-significantly higher odds of endoscopic response (OR 1.2) and endoscopic remission (OR 2.2) vs tofa (Fig 1C). Results were similar in a sensitivity analysis that excluded upa patients with prior tofa exposure (n=24) (Fig 1C).

Conclusion: Upa was associated with significantly higher odds of SFCR at 12 and 52 weeks and non-significantly higher odds of endoscopic response and remission vs tofa for UC. AEs were consistent with known safety profiles. These findings should be confirmed prospectively.