Small scale in vitro method to determine a bioequivalent equilibrium solubility range for fasted human intestinal fluid

Abstract

Drug solubility is a key parameter controlling oral absorption, but intestinal solubility is difficult to assess in vitro. Human intestinal fluid (HIF) aspirates can be applied but they are variable, difficult to obtain and expensive. Simulated intestinal fluids (SIF) are a useful surrogate but multiple recipes are available and the optimum is unknown. A recent study characterised fasted HIF aspirates using a multi-dimensional approach and determined nine bioequivalent SIF media recipes that represented over ninety percent of HIF compositional variability. In this study these recipes have been applied to determine the equilibrium solubility of twelve drugs (naproxen, indomethacin, phenytoin, piroxicam, aprepitant, carvedilol, zafirlukast, tadalafil, fenofibrate, griseofulvin, felodipine, probucol) previously investigated using a statistical design of experiment (DoE) approach. The bioequivalent solubility measurements are statistically equivalent to the previous DoE, enclose literature solubility values in both fasted HIF and SIF, and the solubility range is less than the previous DoE. These results indicate that the system is measuring the same solubility space as literature systems with the lower overall range suggesting improved equivalence to in vivo solubility, when compared to DoEs. Three drugs (phenytoin, tadalafil and griseofulvin) display a comparatively narrow solubility range, a behaviour that is consistent with previous studies and related to the drugs’ molecular structure and properties. This solubility behaviour would not be evident with single point solubility measurements. The solubility results can be analysed using a custom DoE to determine the most statistically significant factor within the media influencing solubility. This approach has a lower statistical resolution than a formal DoE and is not appropriate if determination of media factor significance for solubilisation is required. This study demonstrates that it is possible to assess the fasted intestinal equilibrium solubility envelope using a small number of bioequivalent media recipes obtained from a multi-dimensional analysis of fasted HIF. The derivation of the nine bioequivalent SIF media coupled with the lower measured solubility range indicate that the solubility results are more likely to reflect the fasted intestinal solubility envelope than previous DoE studies and highlight that intestinal solubility is a range and not a single value.