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

Intestinal drug solubility is a key parameter controlling oral absorption but varies both intra and inter individuals and between the fasted and fed states, with food intake known to alter the bioavailability of many compounds. Intestinal solubility can be measured in vitro either using sampled fed human intestinal fluid (FeHIF) or simulated fed intestinal fluid (SIF) but neither approach is optimal. FeHIF is difficult to obtain and variable, whilst for fed SIF multiple recipes are available with no consensus on the ideal version. A recent study characterised FeHIF aspirates using a multidimensional approach and calculated nine simulated media recipes that covered over ninety percent of FeHIF compositional variability. In this study the equilibrium solubility of thirteen drugs have been measured using the nine simulated media recipes and compared to multiple previous design of experiment (DoE) studies, which have examined the impact of fed SIF media components on solubility. The measured nine media solubility data set is only statistically equivalent to the large scale 92 media DoE in 4 out of 13 drug comparisons, but has improved equivalence against small scale DoEs (9 or 10 media) with 6 out of 9 or 10 out of 12 (9 and 10 media respectively) equivalent. Selective removal of non-biorelevant compositions from the 92 media DoE improves statistical equivalence to 9 out of 13 comparisons. The results indicate that solubility equivalence is linked to media component concentrations and compositions, the nine media system is measuring a similar solubility space to previous systems, with a narrower solubility range than the 92 point DoE but equivalent to smaller DoE systems. Phenytoin and tadalafil display a narrow solubility range, a behaviour consistent with previous studies in fed and fasted states and only revealed through the multiple media approach. Custom DoE analysis of the nine media results to determine the most statistically significant component influencing solubility does not detect significant components. Indicating that the approach has a low statistical resolution and is not appropriate if determination of media component significance is required. This study demonstrates that it is possible to assess the fed intestinal equilibrium solubility envelope using the nine media recipes obtained from a multi-dimensional analysis of fed HIF. The derivation of the nine media compositions coupled with the results in this study indicate that the solubility results are more likely to reflect the fed intestinal solubility envelope than previous DoE studies and highlight that the system is worthy of further investigation.