(181) Multivariable analysis of baseline variables associated with efficacy outcomes in the ELEVATE UC clinical program

Background: : Baseline (BL) variables that accurately predict responses to advanced therapies have potential to improve ulcerative colitis (UC) management. Etrasimod is an oral, once-daily, selective sphingosine 1-phosphate (S1P)1,4,5 receptor modulator for the treatment of moderately to severely active UC.

Objective: : BL data from patients in the ELEVATE UC clinical program were used to identify variables associated with etrasimod response.

Methods: : In this post hoc analysis of ELEVATE UC 52 (NCT03945188) and ELEVATE UC 12 (NCT03996369), efficacy endpoints of clinical remission and endoscopic improvement, and change from BL in modified Mayo score (MMS) and endoscopic subscore (ES), were analyzed using logistic or linear regression models, respectively, at Weeks 12 (pooled data) and 52 (ELEVATE UC 52 only). BL variables identified by univariate regression analyses (≥1 comparison p< 0.1) were further assessed using multivariable logistic/linear regression using backward selection (stay criterion p< 0.05).

Results: : At Week 12, BL corticosteroid use (no vs yes, odds ratio; 95% confidence interval: 0.61; 0.41, 0.90), a BL ES of 3 vs 2 (0.44; 0.30, 0.64), and higher high-sensitivity C-reactive protein (hsCRP) (≥0.5 to < 3 vs < 0.5 mg/L: 0.43; 0.24, 0.77) were independently associated with lower odds of achieving clinical remission. Biologic/Janus kinase inhibitor (bio/JAKi) naïve vs experienced was associated with higher odds of achieving clinical remission at Week 52 (2.02; 1.13, 3.61). Similar variables were associated with endoscopic improvement at Week 12. At Week 52, a BL MMS 4–6 vs 7–9 (2.23; 1.38, 3.60) was linked with increased odds of achieving endoscopic improvement. At Week 12, higher hsCRP (≥0.5 to < 3 vs < 0.5 mg/L, least squares mean difference: 0.62), a BL ES of 3 vs 2 (0.48), and bio/JAKi exposure (naïve vs experienced: -0.42) were associated with change from BL in MMS, and higher disease activity (hsCRP ≥0.5 to < 3 vs < 0.5 mg/L: 0.27; ES 3 vs 2: -0.43) with change from BL in ES; most associations were not observed at Week 52.

Conclusions: : In this analysis, BL variables related to higher disease activity were associated with lower odds of achieving efficacy endpoints at Week 12. These were not associated with all efficacy endpoints at Week 52, suggesting some of these patients may require more time to achieve clinical outcomes.

Disclaimer: Pfizer’s generative artificial intelligence tool MAIA was used to assist production of the abstract first draft. Authors reviewed/edited and take responsibility for the content.

Sponsorship:: Pfizer