Pauline Bambury1, Giuseppe Pasculli1, Fianne Sips2, Niccolò Totis2, Maud Beneton1, Jane Knöchel1, Paolo Messina1, Mario Torchia1, Daniel Röshammar1, Sandrine Martins3, Georges Rawadi3, Karine Aguéra3
1InSilicoTrials Technologies S.p.A, 2InSilicoTrials Technologies B.V., 3StromaCare SAS
Introduction: STRO-501 is a humanized murine monoclonal Immunoglobulin 1 (IgG1) antibody (mAb) under pre-clinical development by StromaCare. It targets the soluble stromal component and more specifically, the transforming growth factor beta-induced (TGFßi/BIGH3) stromal protein. The STRO-501 mechanism of action consists in increasing the stroma stiffness resulting in cytotoxic T-cell infiltration and activation, consequently aiming to restore anti-tumor immunity. Objectives: To characterize the kinetics of STRO-501 and BIGH3 across multiple dose levels in non-human primates (NHPs), using pharmacokinetic-pharmacodynamic (PK/PD) modelling. The modelling framework was used to explore the impact of key model parameter variations on BIGH3 levels and to simulate alternative dosing regimens to further and better control the variation of BIGH3 levels, on a cloud-based study simulation platform. Methods: STRO-501 was administered according to dose escalation principles in two treatment groups and one placebo group. Group 1 received 5 mg/kg on day 1, followed by 20 mg/kg on days 57 and 64 while Group 2 received 50 mg/kg on day 1, followed by 200 mg/kg on days 57 and 64. Longitudinal STRO-501 plasma concentrations and BIGH3 levels in sera were measured for all NHPs. To characterize the binding between STRO-501 and BIGH3 and to account for the non-linear elimination observed in the PK profiles, a two-compartment target-mediated drug disposition (TMDD) model was developed using a non-linear mixed effect approach in Monolix. In this model, STRO-501 in both the central and peripheral compartments binds to BIGH3, which was represented by a single compartment. Simulations were performed using the InSilicoTrials’ platform to assess the impact of STRO-501 clearance (CL) and the binding affinity (KD) on STRO-501 PK, as well as on free and total BIGH3 levels. Specifically, the TMDD model was incorporated in our study simulator, allowing the exploration of various parameter values as well as alternative dosing regimens, with the objective to limit variation of BIGH3. A user-friendly study input interface and a simulation result dashboard were set up, facilitating non-modellers to explore the impact of several mAb PK and PD properties on various study designs. Results: Visual predictive checks confirmed a good agreement between observations and simulations, indicating that the TMDD model accurately captured the variability and trends in the observed STRO-501 and BIGH3 levels. Simulations of alternative dosing schedules with the current mAb version showed that administering STRO-501 twice a month leads to its accumulation, particularly at dose levels above 50 mg/kg. This results in a potent and sustained suppression of free BIGH3 levels, but without limiting variation in the total BIGH3 levels. An optimized mAb version was explored, featuring a decreased clearance (scaled by 0.1) and an increase KD value (scaled by 1000). This optimized mAb demonstrated the potential to efficiently suppress free BIGH3 levels while better controlling the variation in total BIGH3 levels. Specifically, administering 20 mg/kg twice a month may result in a slower yet, more sustained reduction in free BIGH3 levels compared to the current version while effectively controlling the level of total BIGH3 levels, providing a relevant compromise between efficacy, potential safety and dosing frequency. Conclusion: A mechanistic TMDD model was used to successfully characterize the PK and target binding of STRO-501 to BIGH3. Simulations conducted in the InSilicoTrials’ simulation platform provided key insights into ligand-target interactions, guiding improvements in PK/PD properties for STRO-501 optimization and dosing strategies of the mAb for the next in vivo NHPs studies. The simulation platform empowers both technical modellers (using their current tools) and non-technical users from the wider clinical team to refine mAb properties, assess the impact of key parameters such as clearance and binding affinity, and plan nonclinical trials to maximize the likelihood of success.
Reference: PAGE 33 (2025) Abstr 11335 [www.page-meeting.org/?abstract=11335]
Poster: Drug/Disease Modelling - Oncology