From the Experimental Model Settings to the Physical System for Interpretable Decision Making
Axiosis sprl & Arxios sprl - Bousval - Belgium
Objectives: The presentation will emphasize the importance of putting the experimental settings in perspective to the physical studied system before taking any decision. This will be essentially done by going through different alternatives for evaluating the apparent permeability index obtained from in vitro CACO2 models. The comparison of these alternatives can be considered as an objective on its own.
Methods: The permeability index is used as part of a general screening process to study drug absorption. It is typically computed by adapting a straight line to the initial portion of the amounts in the receiver compartment; disregarding the first few points when lagging of the transfer process through the membrane is evident. Modeling the transfer process via a two-compartmental system yields an immediate analogue of the index as the initial slope of the receiver quantity, but the two-compartment model often does not match observations well. A three-compartment model, describing the cellular layer as well as donor and receiver compartments, typically better represents the kinetics, but has the disadvantage of always having zero initial flow rate to the receiver compartment!
Results: The alternative definition for the index proposed in Palumbo et al.  applicable for three-compartment models will be first assessed. That new definition will be shown to reduce to the classical formulation as the cellular layer's volume tends towards zero. We will discuss the fact that although mathematically well defined, the proposed alternative lacks of biological interpretations. We will then move to the concept of reformulating the model before taking decision. We will prove that by virtually assuming a continuous and instantaneous resetting of the donor and receiver concentrations, it is straightforward to obtain a computationally robust as well as biological interpretable general definition.
Conclusions: It is often critical to move from the model based on the experimental settings to another model representing the physical studied system in order to reach a robust conclusion. Making such clear distinction for the evaluation of the apparent permeability index has shown to be quite useful for obtaining a generic computational method that avoids any subjective interventions.
 P. Palumbo, U. Picchini, B. Beck, J. Van Gelder, N. Delbar and A. DeGaetano - A General Approach to the Apparent Permeability Index - J. PK/PD, 35, 235-248 - 2008