Background/Purpose: Filgotinib is an oral, selective Janus kinase 1 (JAK1) inhibitor showing good efficacy and safety in Phase 2B studies in rheumatoid arthritis (RA) patients. The exposure-response modeling and simulation to support the filgotinib dose selection for Phase 3 studies in RA is presented here.

Methods: Population predicted and individual responses to treatment were investigated on the basis of simulated exposures to filgotinib and its active metabolite. Non-linear mixed-effects E-R models using either direct response regression (DAS28) or logistic regression (ACR 20/50/70) models were built to describe improvement in clinical response from baseline in RA patients treated for 24 weeks. The drug effect was modeled using linear or E_max model and a composite effective exposure (AUC_Eff) defined as sum of individual-predicted exposure for filgotinib and its metabolite taking into account an estimated 10-fold relative potency for filgotinib versus the metabolite. The observed DAS28 change from baseline and ACR responder rates were described as a sum of individual placebo and drug effects. Continuous covariates were evaluated in the models as power functions while binary covariates were tested as factors. The analysis was performed pooling all Phase 2 data available over the doses of 30 to 300 mg as twice or once daily regimen. Simulations of clinical responses (DAS28, ACR and Low Disease Activity [LDA]) were investigated over the 50 to 200 mg daily dose range.

Results: The PK of filgotinib and its metabolite were described by two independent population PK models. Renal function (CLcr) and race were included as statistically significant covariates on the metabolite clearance while no significant covariates were identified in the PK model for filgotinib. Based on the distribution of random effects, filgotinib and its metabolite follow dose-linear PK across the studied dose range. Region was identified as the only significant covariate on the placebo response; placebo effect was predicted to be higher in Western Europe, USA and South America than in the other countries, and included in the exposure-response model for both endpoints. Clinical response at week 12 and week 24 (DAS28 change from baseline and ACR responder rate) increased with dose. A rapid onset in clinical response was also observed during the first 12 weeks followed by continued increase in effect until week 24. Time to reach a given response was shorter at higher doses. Dosing frequency (once or twice daily) did not influence PK of filgotinib and its metabolite as well as clinical response. Simulated clinical response at week 24 following 50 to 200 mg daily dose is reported as mean (95%CI) in the table below:

Conclusion: Modeling and simulation on the basis of Phase 2 clinical data show a dose-related response with a maximum efficacy achieved at a daily dose of 200 mg filgotinib, the maximum dose currently being tested in the Phase 3 program.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/dose-selection-of-filgotinib-a-selective-jak1-inhibitor-for-rheumatoid-arthritis-phase-3-studies-exposure-das28-and-acr-modeling-approach/