G. Baneyx (1), C. Meille (1), S. Belli (2), A. Iliadis (3) and T. Lavé (1)
F. Hoffmann-La Roche Ltd, Non-Clinical Safety, (1) Modeling and Simulation, (2) Drug Metabolism and Pharmacokinetics (DMPK), Basel, Switzerland; (3) Dpt. of Pharmacokinetics, INSERM U911 CRO2, Aix-Marseille University, Marseille, France
Objective: Penetration of anticancer drug in tumor tissue is an important factor to achieve exposure of cancer cells to efficacious concentrations and thus a therapeutic effect [1]. The objective of this study was to develop a model based approach to characterize and differentiate the in vitro penetration of Erlotinib and Gefitinib for 2 cancer cell lines.
Methods: Erlotinib and Gefitinib are selective and potent inhibitors of the epidermal growth factor (EGFR) in human [3]. EGFR is expressed in the selected epidermoid (A431) and colorectal (DiFi) human carcinoma cells [4]. A multicellular layers (MCL) system [2] was used to generate the drug penetration kinetics through a Teflon membrane (control) and through 3 or 6 layers of A431 or DiFi cells. Drug concentrations were measured by LC-MS/MS in donor and receiver chambers and cell lysates. The model structure consisted in a set of compartments corresponding to donor, each cell layer and the receiver chamber. Drug exchanges were driven by passive diffusion between compartments, unspecific binding to MCL system and specific binding to cells. Main model parameters (permeability, binding and volumes) were estimated using population approach with the software Monolix 4.1.3 [5] considering the well as the statistical unit.
Results: MCL data exploration clearly showed much more limited penetration for Gefitinib than Erlotinib especially with the DiFi cell line. Model parameters were accurately estimated with RSE from 4 to 39%. Permeability values through A431 and DiFi cells for Gefitinib were 12.6 and 2.7 nm/min respectively and 2.5 and 10 fold higher for Erlotinib. In the model, the low mass balance of Gefitinib with DIFI cells was explained by unspecific and specific binding whereas no binding was required for Erlotinib. After 6 and 24h, fraction of dose reaching receiver chamber was respectively 8.3 and 3.7 fold lower for Gefitinib than Erlotinib.
Conclusion: The developed PK model was able to capture the dynamics of drug penetration for both cell lines. In this multicellular layers system, Gefitinib showed much more limited penetration than Erlotinib, especially with DiFi cells. Model based analysis allowed characterization and differentiation of drugs with similar chemical structures. The model can be used to provide mechanistic insights for tumor drug distribution and to explore, by changing experimental conditions, impact of pH on drug penetration.
References:
[1] Minchinton et al. Nat Rev Cancer 2006; 6:583-92.
[2] Tannock et al. Clin Cancer Res 2002; 8:878-884.
[3] Scheffler et al. Clin Pharmacokinet 2011; 50:371-403.
[4] Untawale et al. Cancer Res 1993; 53:1630-1636.
[5] Kuhn et al. Computational Statistics and Data Analysis 2005; 49:1020-1038.
Reference: PAGE 22 () Abstr 2867 [www.page-meeting.org/?abstract=2867]
Poster: Oncology