Natural edible oils (NEOs) are common excipients for lipid-based formulations. Many of them are complex mixtures comprising hundreds of different triglycerides (TGs). One major challenge in developing lipid-based formulations is the variety in NEO compositions affecting the solubility of active pharmaceutical ingredients. In this work, solubilities of indomethacin (IND), ibuprofen (IBU), and fenofibrate (FFB) in soybean oil and in coconut oil were measured via differential scanning calorimetry, high-performance liquid chromatography, and Raman spectroscopy. Furthermore, this work proposes an approach that mimics NEOs using one key TG and models the API solubilities in these NEOs based on perturbed-chain statistical associating fluid theory (PC-SAFT). Key TGs were determined using the 1,2,3-random hypothesis, and PC-SAFT parameters were estimated via a group-contribution method. Using the proposed approach, the solubility of IBU and FFB was modeled in soybean oil and coconut oil. Furthermore, the solubilities of five more APIs (IND, cinnarizine, naproxen, griseofulvin, and felodipine) were modeled in soybean oil. All modeling results were found in very good agreement with the experimental data. The influence of different NEO kinds on API solubility was examined by comparing FFB and IBU solubilities in soybean oil and refined coconut oil. PC-SAFT was thus found to allow assessing the batch-to-batch consistency of NEO batches in silico.
Keywords: PC-SAFT; lipid-based drug delivery systems; oils; solubility.