(317b) Modeling Drug Precipitation in the Human Gastrointestinal Tract Using gCOAS

Over the last decade there has been a significant increase in the number of poorly soluble drugs in the pharmaceutical industry.¹ Poor solubility, especially for weakly basic compounds, can lead to crystallization in vivoand consequently to poor bioavailability.

Upon oral administration, weak bases can experience major pH swings when transiting from the acidic environment of the stomach to the alkaline milieu of the small intestine. This can lead to crystallization and precipitation of the compound reducing its bioavailability and extent of absorption.² Several computational oral absorption models have been developed over the last decade to predict the extent of oral absorption in vivo. However, while these models have successfully incorporated most of the pharmacokinetic and pharmacodynamic processes taking place in the GI tract,³they lack a scientific rationale to properly model drug precipitation. The onset of precipitation is often dictated by a static precipitation time rather than a supersaturation driving force.

The present work aims to use a novel in silico computational oral absorption model, gCOAS (Process Systems Enterprise Ltd, London, UK), to simulate drug precipitation in vivo with a strong emphasis on first principles modeling of crystallization kinetics. gCOAS is inspired by the Sugano framework⁴ and offers a flexible modeling platform to model the different processes taking place in vivo. A key feature of the software is the ability to customize the segmentation and physiological parameters of the GI tract. Each compartment in the model incorporates a population balance model, and kinetic expressions for transit, dissolution, nucleation, growth, and permeation. The model also includes mass and charge balances of both drug species (ionic, non-ionic and micellar) and physiological ions in solution in the GI tract. In addition, to properly model drug transit and fluid flow, the volume of fluids in each compartment is dynamically modeled and is considered to be dependent on the volume of fluid uptake and the volume of endogenous secretions such as saliva, pancreas, mucous, and bile secretions.

In this study, dipyridamole, a weakly basic BCS class II compound, is chosen as the model active pharmaceutical ingredient. Solubility and nucleation threshold experiments are conducted under varying pH and bile salt concentrations to develop a solubility-supersolubility phase diagram for dipyridamole to serve as a design space for the nucleation and growth experiments.

Kinetic expressions using the experimental nucleation and growth data are then integrated in gCOAS. The flexible platform of the latter allows for the implementation of custom nucleation and growth kinetic models to account for the effect of bile salts on the crystallization kinetics.

  References

1.        Hauss, D. J. Oral lipid-based formulations. Advanced Drug Delivery Reviews 59, 667–76 (2007).

2.        Modern Pharmaceutics. 864 (CRC Press: 2002).doi:10.1201/9780824744694

3.        Agoram, B., Woltosz, W. S. & Bolger, M. B. Predicting the impact of physiological and biochemical processes on oral drug bioavailability. Advanced Drug Delivery Reviews 50, S41–67 (2001).

4.        Sugano, K. Introduction to computational oral absorption simulation. Expert opinion on drug metabolism & toxicology 5, 259–293 (2009).

5.        Yu, L. X., Crison, J. R. & Amidon, G. L. intemational journal of Compartmental transit and dispersion model analysis of small intestinal transit flow in humans. International journal of pharmaceutics 140, 111–118 (1996).