Models for Predicting Drug Absorption From Oral Lipid-Based Formulations

Abstract

In this review, we describe the in vitro tools currently used to identify when a lipid-based formulation has the potential to deliver a poorly water-soluble drug via the oral route. We describe the extent to which these tools reflect the in vivo performance of the formulation and, more importantly, we present strategies that we foresee will improve the in vitro-in vivo correlations. We also present emerging computational methods that are likely to allow large parts of the formulation development to be carried out in the computer rather than in the test tube. We suggest that these computational tools will also improve the mechanistic understanding of in vivo formulation performance in the complex and dynamic environment of the gut.

Keywords: Formulation performance; In vitro models; Lipid-based formulations; Molecular dynamics simulations; Multivariate data analysis; Prediction.

Fig. 1

Current and future assessment of the loading capacity (maximum amount of drug that can be dissolved) and in vivo performance of LBFs. ( a) The solubility of the drug in key excipients and the loading capacity of the potential LBFs are currently assessed experimentally, typically in a 96-well titer plate, for a large number of excipients and formulations. This exercise requires a large amount of the compound to be synthesized and may result in sub-optimal formulations to be selected, as only a standard selection of formulations is studied. (b) Recently, computational multivariate data analysis models have been developed to allow the drug solubility in excipients and the loading capacity of the formulations to be predicted from calculated molecular descriptors and information obtained from the solid state. (c) The use of MD simulations allows molecular interactions between drug and excipients/LBFs to be identified and the free energy of solvation in the formulation to be calculated. These computational simulations are likely to result in more accurate prediction of the solubility and loading capacity and will also increase the mechanistic understanding of the solvation process of the API in complex formulations. (d) Novel in vitro and in silico tools to predict the dynamic gut. The lipolysis as performed today is shown on the left hand side with the reaction vessel to which the LBF is added. A future goal is to connect this in vitro model with an absorption chamber mimicking the intestinal wall which will allow absorption of API and LBF components to occur simultaneously with the digestion. Furthermore, protocols for MD simulations that assess the impact of dispersion and digestion of the LBF on the solvation capacity of the intestinal fluid need to be developed. These simulations should capture the restructuring of the solubilizing nanoaggregates present in the intestinal fluid to better predict in vivo performance of LBF dosed drugs. (e) The end goal is the accurate optimization of loading capacity and in vivo performance using in silico tools.