M.I. Titze (1), O. Schaaf (2), M. H. Hofmann (2), M. Sanderson (2), S. Zahn, (2), J. Quant (2), T. Lehr (1)
(1) Clinical Pharmacy, Saarland University, Germany, (2) Boehringer Ingelheim RCV GmbH, Vienna, Austria
Objectives: The insulin-like growth factor 1 receptor (IGF-1R) plays an important role in tumor growth. BI 893923 is a novel and selective ATP-competitive IGF1R/INSR inhibitor. Aim of this work was to develop a population PK/PD model for human GEO colon carcinoma mice xenografts and to simulate the tumor growth inhibition (TGI) based on this model for different doses and schedules of BI 893923. The PK/PD model comprises the characterization of the relationship between BI 893923 plasma concentration and the biomarker phosphorylated IGF-1R (pIGF-1R) and further the evaluation of the association between pIGF-1R modulation and tumor growth.
Methods: Tumor bearing mice were treated with different doses (0-120 mg/kg p.o.) and dosing schedules (single dose (sd), multiple daily dosing (qd), or twice daily dosing (bid)) of BI 893923. Modeling was done in a sequential manner by first fitting plasma concentrations to a PK model using non-linear mixed-effects modeling implemented in NONMEM V7.3.0 [1]. The developed PK model was linked to the pIGF-1R model and finally integrated in the tumor growth model. Model assessment was guided by common evaluation tools. Simulations of the final model with various dosing scenarios (sd, bid with intervals of 1, 4, 6, 12 hours) were done to determine the dosing regimen leading to the highest TGI.
Results: A three-compartment model with two absorption compartments for sequential fast/slow absorption and linear elimination from the central compartment best described BI 893923 PK. A turnover model was selected to describe pIGF-1R. The BI 893923 concentration in the effect compartment was linked to the inhibition of IGF-1R phosphorylation by an indirect response model. Control tumor growth was described by the Simeoni growth model [2] with an exponential growth rate of 0.0049 h-1 and linear growth rate of 1.24 mm3/h. In contrast to the Simeoni model no cell death was induced but both growth rates were inhibited by the decreased pIGF-1R level as a combination of direct and time-delayed inhibition. Simulations revealed that with increasing interval for bid dosing the TGI increased.
Conclusions: The final PK/PD model described the relationship between BI 893923 plasma concentration, IGF-1R phosphorylation and tumor growth very good. Simulations suggest a bid dosing with an interval of 12 hours as the most effective one. The developed model is considered to be a predictive tool for the human therapeutic dose estimation.
References:
[1] Beal SL, Sheiner LB, Boeckmann AJ & Bauer RJ (Eds.) NONMEM Users Guides. 1989-2011. Icon Development Solutions, Ellicott City, Maryland, USA.
[2] Simeoni, M, Magni P, Cammia C. Predictive Pharmacokinetic-Pharmacodynamic Modeling of Tumor Growth Kinetics in Xenograft Models after Administration of Anticancer Agents. Cancer Res. (2004) 64: 1094–1101.
Reference: PAGE 24 (2015) Abstr 3320 [www.page-meeting.org/?abstract=3320]
Poster: Drug/Disease modeling - Oncology