A Unifying Approach to Drug-in-Polymer Solubility Prediction: Streamlining Experimental Workflow and Analysis

This method paper describes currently used experimental methods to predict the drug-in-polymer solubility of amorphous solid dispersions and offers a combined approach for applying the Melting-point-depression method, the Recrystallization method, and the Melting-and-mixing method. It aims to describe and expand on the theoretical basis as well as the analytical methodology of the recently published Melting-and-mixing method. This solubility method relies on determining the relationship between drug loads and the enthalpy of melting and mixing of a crystalline drug in the presence of an amorphous polymer. This relationship is used to determine the soluble drug load of an amorphous solid dispersion from the recorded enthalpy of melting and mixing of the crystalline drug portion in a drug-polymer sample at equilibrium solubility. Due to the complex analytical methodology of the Melting-and-mixing method, a software solution called the Glass Solution Companion app was developed. Using this new tool, it is possible to calculate the predicted drug-in-polymer solubility and Flory-Huggins interaction parameter from experimental samples, as well as to generate the resulting solubility-temperature curve. This software can be used for calculations for all three experimental methods, which would be useful for comparing the applicability of the methods on a given drug-polymer system. Since it is difficult to predict the suitability of these drug-in-polymer solubility methods for a specific drug-polymer system in silico, some experimental investigation is necessary. By optimizing the experimental protocol, it is possible to collect data for the three experimental methods simultaneously for a specific drug-polymer system. These results can then be readily analyzed using the Glass Solution Companion app to find the most appropriate method for the drug-polymer system, and therefore, the most reliable drug-in-polymer solubility prediction.