Chiara Zecchin (1), Steven Yancey (2), Angela Davis (2), Alienor Berges (1)
(1) GlaxoSmithKline, Stevenage, UK (2) GlaxoSmithKline, Collegeville, PA, USA
Introduction: Depemokimab is an anti-IL-5 monoclonal antibody with enhanced potency and half-life compared to mepolizumab. It is currently in Phase 3 development in asthma, and a life cycle innovation program is planned in additional eosinophilic conditions. PK and pharmacology data were collected in a single subcutaneous ascending dose first-time in human (FTIH) study (NCT03287310) in subjects with mild to moderate asthma.
Objectives: To select the optimal dosing regimen of depemokimab in several eosinophilic diseases, using Model Informed Drug Development (MIDD), leveraging mepolizumab knowledge, in the absence of Ph2 trials.
Methods: A PK-blood eosinophils model, developed for mepolizumab on pooled data from studies in various eosinophilic conditions, with a wide range of baseline eosinophil count values and dose regimens, was available. This model was updated, maintaining the same model structure, to reflect the longer half-life (t1/2) and enhanced potency of depemokimab compared to mepolizumab. The PK parameters, and the IC50, the between subject variability and the residual unexplained variability were re-estimated based on the data from depemokimab FTIH [1]. Other model parameters were fixed to the value of the mepolizumab model (including the impact of the baseline eosinophil levels on the maximum achievable drug inhibitory response).
Simulation of depemokimab Ph3 trials investigated the effect of dose, dosing frequency, and baseline eosinophil level (including values observed in the mepolizumab Ph3 trials, and higher values reflective of corticosteroid tapering and flares). For each scenario, 1000 trials were simulated in R [2]. Parameters were sampled from a multivariate normal distribution, centred in the parameter estimated values, with covariance estimated by NONMEM in the $COV step. Residual unexplained variability was included in the simulation.
For each indication, depemokimab dosing regimen was selected to match the blood eosinophil pharmacology observed with mepolizumab in the pivotal Ph3 trial of the respective indication, at the approved therapeutic dose, over a period of 52 weeks.
Results:
A one-compartment pharmacokinetic model with first-order absorption and first order elimination described depemokimab PK data well. Depemokimab exposure increased dose proportionally, except between 2 and 10 mg for undetermined reasons. The population PK parameters were aligned with a typical monoclonal antibody with extended t1/2. The estimated (95% CI) value of CL/F was 0.132 (0.118-0.147) L/day, V/F was 7.88 (6.82-8.94) L, the absorption rate constant was 0.383 (0.338-0.428) 1/day, the half-life was 41.3 (38.9-43.7) days, and the variance of the residual error was 0.131 (0.111-0.150). Between subject variability (BSV) was 26% for CL/F and 31% for V/F and their covariance was 27%. Bodyweight was included in the model as covariate, using allometry with fixed exponents of 0.75 and unity for CL/F and V/F, respectively. The exposure-eosinophils model was an indirect response model parameterised in terms of baseline blood eosinophils (KRO), rate of elimination (KOUT), maximum inhibitory effect (Imax), concentration resulting in 50% of maximum inhibitory drug effect (IC50), and Hill coefficient (GAMA). Between subject variability was included on KRO and Imax using an exponential model. Measured baseline blood eosinophil count was a covariate of KRO and Imax, and disease was a covariate for KRO. The model was fitted to PKPD data with the ADVAN6 TOL5 subroutine and FOCE method with interaction using NONMEM version 7.3 [3]. The estimated (95%CI) value of IC50 was 0.0960 (0.0822-0.112) ug/mL, Imax was 0.170 (0.162-0.179), KRO was 0.177 (0.160-0.197) GI/L and KOUT was 0.310 (0.259-0.371) 1/day. The between subject variability was 30% for KRO, and 60% for Imax. The standard deviation of the exponential residual error was 0.350.
Simulations identified the Phase 3 dosing regimen of depemokimab for each indication, and were key in supporting the dose rationale during regulatory interactions.
Conclusion: The use of MIDD and pharmacology principles, together with the precedented safety of the IL-5 neutralisation pathway, provided a strong rationale to transition depemokimab directly from the FTIH to confirmatory Phase 3 trials. This example is a successful MIDD application to bridge dosing across LCI indications based on key disease characteristics and robust pharmacology data predictive of efficacy.
References:
[1] Singh D., et al. “A Phase 1 study of the long-acting anti-IL-5 monoclonal antibody GSK3511294 in patients with asthma”. Br J Clin Pharmacol. 2021.
[2] The R Foundation for Statistical Computing, version 3.3.2 (2016-10-31)
[3] Beal SL., et al. “NONMEM 7.3. 0 Users Guides. (1989–2013).” ICON Development Solutions, Hanover, MD. Funding: GSK
Disclosures:
CZ, SY, AD, and AB are / were employees and shareholders of GSK.
Reference: PAGE 30 (2022) Abstr 10085 [www.page-meeting.org/?abstract=10085]
Poster: Drug/Disease Modelling - Other Topics