Can Pure Predictions of Activity Coefficients from PC-SAFT Assist Drug-Polymer Compatibility Screening?

Area of Science:

• Pharmaceutical Science
• Thermodynamics
• Computational Chemistry

Background:

• Amorphous solid dispersions (ASDs) enhance bioavailability of poorly water-soluble drugs.
• Experimental determination of drug-polymer compatibility is labor- and cost-intensive.
• Thermodynamic models like PC-SAFT offer a computational alternative for predicting compatibility.

Purpose of the Study:

• To evaluate the predictive performance of the PC-SAFT equation of state for drug-polymer compatibility in ASDs.
• To assess PC-SAFT's ability to predict compatibility without fitted binary interaction parameters (k=0).
• To investigate the impact of API parametrization on prediction accuracy.

Main Methods:

• Utilized the PC-SAFT equation of state to predict API-polymer activity coefficients.
• Employed experimental fusion properties of APIs and a k=0 approach.
• Evaluated predictions against experimental data for nearly 40 API-polymer systems.

Main Results:

• Achieved an average quantitative error of approximately 50% in weight fraction solubility.
• Observed significant variability in prediction accuracy across different API-polymer systems.
• Poorest predictions were for self-associating polymers like poly(vinyl alcohol) due to unmodeled hydrogen bonding.

Conclusions:

• PC-SAFT with k=0 provides a useful, albeit quantitatively imprecise, tool for initial screening of drug-polymer compatibility in ASD development.
• Qualitative ranking of polymer compatibility was reasonably predicted, identifying generally better or worse polymer candidates.
• Future work should focus on refining PC-SAFT parametrization to improve quantitative accuracy and cost-performance ratio.