Thermodynamic Perspectives on the Impact of a Second Drug on Amorphous Drug Solubility

Drug amorphous solubility can be changed in the presence of other compounds, making it essential to elucidate the underlying mechanisms for designing supersaturated formulations. In this study, we experimentally determined how a second drug affects the amorphous solubility of ritonavir (RTV) and analyzed these effects from a thermodynamic perspective. Lopinavir (LPV), cilnidipine (CND), and probucol (PBC) were used as second drugs. The coexistence of each second drug in an aqueous solution reduced the amorphous solubility of RTV. In the presence of LPV and CND, the experimentally determined RTV amorphous solubility was close to the value predicted under the assumption of ideal mixing of RTV and second drugs. In contrast, in the presence of PBC, the experimentally determined RTV amorphous solubility exceeded the predicted value. Dynamic vapor sorption (DVS) measurements revealed that the RTV/LPV coamorphous absorbed water similarly to amorphous RTV. Conversely, water absorption in the RTV/CND and RTV/PBC coamorphous decreased compared with amorphous RTV. Using the experimentally determined amorphous solubilities and water absorption data, the interaction parameters between RTV and each second drug within the water-saturated drug-rich phase were calculated. The absolute value of the interaction parameter in the RTV/LPV system is relatively small, suggesting that incorporating LPV into the RTV-rich phase had minimal impact on water absorption and drug-drug interaction strength in the RTV-rich phase, resulting in experimentally determined solubility values that align closely with those predicted by ideal mixing of RTV and LPV. Meanwhile, the interaction parameters of the RTV/CND and RTV/PBC systems were negative, indicating relatively strong drug-drug interactions that can further reduce RTV amorphous solubility. However, for these two systems, the mixing of second drugs also decreased the water content in the RTV-rich phase, which would mitigate the extent of the solubility reduction. In the RTV/CND system, the strong drug-drug interaction and reduced water content largely offset each other. This results in an experimentally determined RTV amorphous solubility similar to the value predicted by the ideal mixing of RTV and CND. In contrast, in the RTV/PBC system, the water content of the RTV-rich phase was more substantially decreased, leading to a higher experimentally determined value of RTV amorphous solubility than that predicted by ideal mixing of RTV and PBC. Overall, this study elucidates the impact of a second drug on the amorphous solubility of a primary drug and provides valuable insights for the design of supersaturated formulations containing multiple drugs.