Effect of composition of simulated intestinal media on the solubility of poorly soluble compounds investigated by design of experiments.

The composition of the human intestinal fluids varies both intra- and inter-individually. This will influence the solubility of orally administered drug compounds, and hence, the absorption and efficacy of compounds displaying solubility limited absorption. The purpose of this study was to assess the influence of simulated intestinal fluid (SIF) composition on the solubility of poorly soluble compounds. Using a Design of Experiments (DoE) approach, a set of 24 SIF was defined within the known compositions of human fasted state intestinal fluid. The SIF were composed of phospholipid, bile salt, and different pH, buffer capacities and osmolarities. On a small scale semi-robotic system, the solubility of 6 compounds (aprepitant, carvedilol, felodipine, fenofibrate, probucol, and zafirlukast) was determined in the 24 SIF. Compound specific models, describing key factors influencing the solubility of each compound, were identified. Although all models were different, the level of phospholipid and bile salt, the pH, and the interactions between these, had the biggest influences on solubility overall. Thus, a reduction of the DoE from five to three factors was possible (11-13 media), making DoE solubility studies feasible compared to single SIF solubility studies. Applying this DoE approach will lead to a better understanding of the impact of intestinal fluid composition on the solubility of a given drug compound.