P519 Mirikizumab pharmacokinetics and exposure-response relationships in patients with moderately to severely active ulcerative colitis: results from randomised phase 2 and phase 3 induction and maintenance trials

Abstract

Abstract Background Mirikizumab (MIRI) is a p19-targeted anti–interleukin-23 developed for ulcerative colitis (UC), and has demonstrated efficacy in Phase 2 (AMAC, NCT02589665) and Phase 3 (LUCENT-1, NCT03518086; LUCENT-2, NCT03524092) trials in patients with moderately to severely active UC. We characterised MIRI pharmacokinetics (PK) and the exposure-response (E-R) relationship between MIRI exposure and clinical outcomes in these 3 trials. Methods Patients received 12-week MIRI induction (50, 200 or 600 mg [AMAC] or 300 mg [LUCENT-1] intravenous [IV] every 4 weeks [Q4W]), and induction responders then received MIRI maintenance (200 mg subcutaneous [SC] Q4W or every 12 weeks for 92 weeks in AMAC; 200 mg SC Q4W for 40 weeks in LUCENT-2). Serum PK parameters were analysed by non-linear mixed-effects modelling, and covariate effects on MIRI exposure were evaluated using simulation-based estimations. Relationships between MIRI exposure and clinical remission, clinical response, endoscopic remission and symptomatic remission, based on the Mayo scoring system, were assessed using logistic regression models. Results MIRI PK (clearance, 0.022 L/hr (95% CI 0.02,0.03); volume of distribution for central and peripheral compartments, 3.11 L (95% CI 3.07, 3.15) and 1.69 L (95% CI 1.03, 2.57); half-life, 9.5 days) were best described by a 2-compartment model with first-order absorption. Clearance increased with body weight and decreased with albumin concentration, and bioavailability decreased with body mass index, but impacts of covariates were small relative to random variability. MIRI SC bioavailability was 48%. Modelling of E-R relationships during MIRI induction in AMAC revealed statistically significant positive associations of MIRI exposure with clinical response and with change in Modified Mayo Score. Maximum effect was predicted at 300 mg, with no significant improvement at higher doses. Steep positive relationships between exposure and efficacy in LUCENT-1 were not explained by patient covariates, and analyses suggest confounding by unknown variables. MIRI exposure during maintenance was not a significant predictor of efficacy in AMAC and LUCENT-2. Conclusion MIRI PK characteristics were typical of an immunoglobulin G monoclonal antibody. The MIRI E-R relationship revealed an apparent lack of sensitivity of efficacy to individual patient exposures at the dose levels studied in Phase 3. Overall, the PK and E-R data suggest that dose adjustments due to patient factors are not justified.