In vitro to in vivo extrapolation using Cloe HIA to understand site specific drug absorption and solve absorption issues

Cloe HIA data requirements for human intestinal absorption simulation

A simple set of early ADME and physicochemical data is required to run a Cloe HIA prediction as detailed below. Either measured or predicted values can be used.

Cloe HIA data delivery for understanding dose dependent drug absorption issues

Predictions are generated in a downloadable pdf report and Excel spreadsheet with interactive sensitivity analysis capability.

The following parameters are reported.

• Predicted total absorption at several user-specified dose levels.
• Identification of factors (solubility and/or permeability) limiting absorption at any poorly absorbed dose level(s).
• Predicted dependence of the solubility on pH, over the range of the GI tract contents (pH 2-7.5).
• Predicted absorption from the different segments of the GI tract at each dose level.
• Predicted time course data representing absorption over time.
• Predicted solubility over the pH range of the gastrointestinal tract, and the peak concentration of compound achieved in the different parts of the tract.

Options for using Cloe HIA to predict human intestinal absorption

Data from Cloe HIA for predicting human intestinal absorption

Cloe HIA uses physiological modelling to integrate Caco-2 permeability and phys-chem data. It models transit of fluid (containing dissolved and undissolved compound), gastric, biliary and pancreatic secretion and absorption in five segments of the GI tract.

Figure 1

Schematic diagram of the intestinal absorption model.

 

Nine compartments representing the five segments of the GI tract – stomach, duodenum, jejunum, ileum and colon – are represented. The fluid content, carrying dissolved and undissolved compound passes from one compartment to the next, simulating the action of peristaltic motion. Within each compartment the dynamic interconversion between dissolved and undissolved compound is modelled. Dissolved compound can be absorbed across the GI tract epithelium. The volume of each compartment – which represents the fluid content – is modelled dynamically, simulating the following processes:

• Transit of the fluid with characteristic rate constants through each compartment.
• Gastric secretion into the stomach, and biliary and pancreatic secretions into the duodenum.
• Absorption of fluid from duodenum, jejunum, ileum and large intestine.

Figure 2

Prediction of human intestinal absorption using Caco-2 permeability, turbidimetric solubility, pK_a and logP data combined with physiological modelling.

 

The solid line represents the line of unity and the dotted line represent two fold of the reported fraction absorbed values. Data are for 122 combinations of dose and compound for 82 compounds. Reported values are from Thomas et al (2008). Cloe HIA provides a reliable quantitative prediction of human intestinal absorption from Caco-2 P_app, solubility, pK_a and log P data. Greater than 84% of the predictions are within two-fold of the reported in vivo value for an independent test set (Figure 2) and the Spearman rank correlation coefficient between predicted and reported fraction absorbed is 0.76. It is a generic model which is applicable to a diverse set of chemistry. The model addresses dose-dependent absorption and the potential limitations of solubility on human intestinal absorption.

back to top