Abstract 4510: Model-based phase II dose selection of c-Met inhibitor MSC2156119J

Abstract

Abstract Objectives: To evaluate the dose/exposure/target inhibition/tumor growth relationship of MSC2156119J, an orally administered, reversible, ATP-competitive, highly potent and selective c-Met receptor tyrosine kinase inhibitor, in order to determine the recommended phase II dose (RP2D) for the program. Methods: Interim pharmacokinetic (PK) and biomarker (PD) data, i.e. plasma concentrations and c-Met inhibition in tumor biopsies from an ongoing first-in-human (FIM) solid tumor trial of MSC2156119J, were analyzed using the population modeling approach. The exposure/target inhibition relationship established in rich sampled KP-4 xenografted mice was utilized as prior information to develop the human model. Model-based simulations of the c-Met inhibition profiles in humans, aiming for a level that achieved tumor regression in mice, helped to determine the biologically active dose of MSC2156119J. Results: During the FIM study, the maximum tolerated dose of MSC2156119J was not reached at a dose of 1,400 mg/day. Instead, a biologically active dose was selected as the RP2D, based on a translational modeling approach that integrated the quantitative relationship between dose, exposure, and target as well as tumor inhibition in humans and in a preclinical model. In KP-4 xenografted mice, tumor c-Met inhibition (quantified as normalized phosphorylated c-Met expression [pMET]) was described by a turnover full inhibitory Imax model, and tumor growth inhibition was best described by the Simeoni model. Simulations demonstrated that nearly complete pMET inhibition (≥95%) is required for tumor stasis or even regression. In the FIM study, a one-compartment linear model with an absorption transit compartment best described the PK of MSC2156119J at doses of 30-700 mg. The turnover model developed from mice data was applied to evaluate the level of c-Met inhibition in paired human tumor biopsies (taken pre- and on-treatment). System turnover parameters for pMET were set equal to those estimated in mice, while the treatment potency (1/IC50 of the inhibition of build-up) of MSC2156119J in humans was found to be 1.7-fold higher than that in mice. Assuming a 30% inter-individual variability in IC50, human simulations suggested that a 500-mg once-daily dose regimen of MSC2156119J could achieve continuous pMET inhibition of ≥95% in 90% of the population. The model and simulation will be confirmed using additional PK data and further tumor biopsies. Conclusions: c-Met inhibition in human tumor lesions was described by a turnover model developed in KP-4 xenografted mice. Using this translational modeling and simulation approach and human PK and target inhibition data from paired tumor biopsies, a biologically active dose of 500 mg was proposed as the RP2D for MSC2156119J. This dose achieves continuous pMET inhibition of ≥95% in 90% of the population and has been adopted for the current development program for MSC2156119J. Citation Format: Wenyuan Xiong, Samer El Bawab, Friedhelm Bladt, Michael Meyring, Manfred Klevesath, Gerald Falchook, David Hong, Andreas Johne, Pascal Girard. Model-based phase II dose selection of c-Met inhibitor MSC2156119J. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4510. doi:10.1158/1538-7445.AM2015-4510