The solubility-permeability interplay when using cosolvents for solubilization: revising the way we use solubility-enabling formulations

Abstract

We have recently reported the interplay between apparent aqueous solubility and intestinal membrane permeability, showing the trade-off between the two when using cyclodextrin- and surfactant-based systems as solubility-enabling formulations. In these cases, the decreased permeability could be attributed directly to decreased free fraction of drug due to the complexation/micellization inherent in these solubilization methods. The purpose of this study was to investigate the direct solubility-permeability interplay, using formulations in which complexation is not the mechanism for increased solubilization. The apparent aqueous solubility (S(aq)) and rat intestinal permeability (P(eff)) of the lipophilic drug progesterone were measured in systems containing various levels of the cosolvents propylene glycol and PEG-400, since this solubilization method does not involve decreased free fraction. Thermodynamic activity was maintained equivalent in all permeability studies (75% equilibrium solubility). Both cosolvents increased progesterone S(aq) in nonlinear fashion. Decreased P(eff) with increased S(aq) was observed, despite the constant thermodynamic activity, and the nonrelevance of free fraction. A mass-transport analysis was developed to describe this interplay. The model considers the effects of solubilization on the membrane permeability (P(m)) and the unstirred water layer (UWL) permeability (P(aq)), to predict the overall P(eff) dependence on S(aq). The analysis revealed that (1) the effective UWL thickness quickly decreases with ↑S(aq), such that P(aq) markedly increases with ↑S(aq); (2) the apparent membrane/aqueous partitioning decreases with ↑S(aq), thereby reducing the thermodynamic driving force for permeability such that ↓P(m) with ↑S(aq); (3) since ↑P(aq) and ↓P(m) with ↑S(aq), the UWL is shorted out and P(eff) becomes membrane control with ↑S(aq). The model enabled excellent quantitative prediction of P(eff) as a function of S(aq). This work demonstrates that a direct trade-off exists between the apparent solubility and permeability, which must be taken into account when developing solubility-enabling formulations to strike the optimal solubility-permeability balance, in order to maximize the overall oral absorption.