Jong Bong Lee(1), Atheer Zgair(1), Tae Hwan Kim(2), Hwi-yeol Yun(3), Peter Fischer(1), Pavel Gershkovich(1)
(1)School of Pharmacy, University of Nottingham, UK; (2)School of Pharmacy, Sungkyunkwan University, South Korea; (3)College of Pharmacy, Chungnam National University, South Korea
Objectives: Firstly, to obtain experimental permeability data for a list of compounds using simple and accessible Caco-2 cell permeability assay and compare with permeability predicted by GastroPlus™. Secondly, to assess the effect of difference between the two permeability data in simulation of pharmacokinetic profiles.
Methods:
1) Caco-2 cell permeability assay was performed using Caco-2 cell monolayer maintained for 21 days. Monolayer integrity was monitored by measurement of transepithelial electric resistance (TEER) before and after the permeability assay. 17 compounds with structural diversity (antipyrine, atenolol, cetirizine, cimetidine, desipramine, dexamethasone, furosemide, hydrochlorothiazide, ketoprofen, metoprolol, naproxen, piroxicam, propranolol, ranitidine, sildenafil, terbutaline and verapamil) were tested and apparent permeability coefficient (Papp) values were obtained. Papp values of 14 compounds were used to develop a user-defined permeability model in GastroPlus™. This model was applied to all compounds to obtain human effective permeability (Peff) from in vitro experimental data.
2) GastroPlus™ and its built-in ADMET Predictor module were used for in silico prediction of Peff values based on chemical structures of the compounds.
Results: User-defined permeability model was developed in GastroPlus™ by correlating the experimental Papp values of the 14 compounds with built-in Peff values in GastroPlus™ and a log-linear model was obtained (R2=0.8799). The Peff values predicted from this model and Peff values from previously reported literature had a correlation of R2=0.7375. Peff values predicted in silico by GastroPlus™ showed a correlation of R2=0.6917 with the literature values. Peff values obtained from experimental model and Peff values obtained from in silico prediction resulted in R2=0.6990, indicating reasonable prediction power of the software. The influence that the difference between two dataset has on outcomes of simulated pharmacokinetic profiles were assessed for dexamethasone, sildenafil and cetirizine.
Conclusions: Although GastroPlus™ was able to predict permeability of compounds with reasonable accuracy, experimentally obtained permeability data were more accurate when used in simulation of reliable pharmacokinetic profiles. In silico permeability predictions could be utilised to facilitate decision making processes but permeability remains as a value that experimentation is desired whenever possible.
Reference: PAGE 25 (2016) Abstr 5890 [www.page-meeting.org/?abstract=5890]
Poster: Drug/Disease modeling - Absorption & PBPK