Mangas-Sanjuan,V.(1), Gonzalez-GarcÃa, I.(2), Gonzalez-Alvarez, I.(1), Bermejo, M.(1), Casabo, VG.(2)
(1) Pharmacy and Pharmaceutical Technology Area, Miguel Hernández University; (2)Pharmacy and Pharmaceutical Technology Department, University of Valencia.
Introduction: The permeability calculation is derived from the Fick’s Law, which its first approximation is a lineal regression model where sink conditions and constant donor concentrations are assumed[1] (Sink model). Artursson et al.[2] published a linear regression model to estimate the permeability under sink conditions, where the donor concentration changes in time, using the accumulated transported fractions (Corrected Sink model). Tavelin et al.[2] proposed a non linear regression model (No Sink model), but could not correctly predict the permeability when there is a lag time and when it is an increased permeation at initial phases of the experiment.
Objectives: The aim was to develop a new non linear model which could estimate the permeability in all types of profiles, simulating different scenarios of variability.
Methods: The non linear model proposed in this work is a modification of Tavelin’s non linear model. The comparison between four models was done simulating 1000 experiments with 3 well for each experiment with exponential interindividual and residual errors. Three types of profiles were simulated under sink and non sink conditions and four types of interindividual and residual variabilities combinations (H-H, H-L, L-H, L-L). Permeability was calculated using the four models described. Mean estimation error, intraassay estimation error and individual estimation error were calculated using simulated and estimated values. ANOVA was performed with the mean estimation error using SPSS 20.0.
Results: Modified non sink model presented the lowest estimation error in most of the cases. Linear regression models estimation error were similar to modified non sink model under sink conditions, but higher under non sink conditions with statistical significance. Non sink model obtained higher estimation errors in profiles 1 and 3, compared to modified non sink model with statistical significance.
Conclusions: The modified non sink equation predicts under the standards of accuracy and precision in all types of profiles and variability scenarios.
References:
[1] Dawson, D., Principles of membrane transport, in Handbook of physiology, R. B, Editor 1991, American Physiological Society: Bethesda. p. 1-45.
[2] Tavelin, S., et al., Applications of epithelial cell culture in studies of drug transport. Methods Mol Biol, 2002. 188: p. 233-72.
Reference: PAGE 22 () Abstr 2746 [www.page-meeting.org/?abstract=2746]
Poster: Absorption and Physiology-Based PK