A fully integrated PK/IVTI/IVE model in mouse to help design the FHD trial for a cell cycle inhibitor X
Damien Cronier (2), Lawrence Gelbert(1), Graham Wishart (2), Alfonso De Dios (1)
(1) Eli Lilly and Company, Indidanapolis, USA; (2) Eli lilly and Company, Windlesham, UK
Objectives: The primary objective of this modelling study was to describe the effect of compound X on the cell cycle of cancer cells by means of a multi-biomarker approach allowing full monitoring of the cell cycle kinetics. The secondary objective of this study was to estimate a pharmacologically effective dose range for compound X in human, in preparation of the First Human Dose (FHD) study.
Methods: An integrated PK/IVTI/IVE model was developed in mouse using a seauential approach. The PK of X were connected to the level of in vivo target inhibition (IVTI) in Colo-205 xenograft tumours by means of a PK/IVTI moel describing the kintics of the different phases of the cell cycle using a series of transit compartments. Cell cycle progression was assumed to be inhibited by means of an indirect resposne mechanism. In vivo efficacy (IVE) of compound X in Colo-205 was then described by means of a fully integrated PK/IVTI/IVE model consisting of a modified Gompertz model connected to the PK/IVTI model using both a cytostatic and a cytotoxix component. Finally, a humanised PK/IVTI/IVE model was obtained by connecting the mosue IVTI/IVE model to a projected human PK model obtained by allometric scaling. This humanised PK/IVTI/IVE model was used to predict an efficacious dose range in human.
Results: the fully integrated PK/IVTI/IVE model could account for the dose-dependency of IVE accross a dose range of 25 to 100 mpk and over a dosing period of 21 days. The 3 biomarkers collected along the cell cycle of colo-205 cancer cells wre connected in a mechanistic manner and the time shift for the peak of inhibition observed in the 3 compartments as well as the rebound effect were wel accounted for, suggesting a block of the cell cycle associated with cell synchronisation. The model also made it possible to correalte IVE with a minimum of 30-50% maintained thoughout the whole treament period. This IVTI threshold was subsequently used to derive an efficacious dose range in human.
Conclusions: The PK/PD relationship of compound X in Colo-205 tumours was modelled by means of a fully integrated PK/IVTI/IVE model. This model made it possible to understand the determinants of IVE and to correlate the latter to the maintaining of a minimum of 30-50% IVTI throughout the whole treatment period. This model was successfully applied to project an efficaicous dose range in human to support the design of the FHD trial.
