From Quantum Chemistry to Prediction of Drug Solubility in Glycerides

Abstract

Lipid-based delivery is a key technology for dealing with the challenges of poorly soluble drugs. Therefore, prediction of drug solubility in lipid-based excipients and their mixtures is an important research goal in computational pharmaceutics. This study is based on the conductor-like screening model for real solvents (COSMO-RS), which combines quantum chemical surface calculations with fluid phase thermodynamics. An experimental dataset of 51 drugs was collected with measured thermochemical data and solubility results in medium and long-chain tri- and monoglycerides. For the theoretical model, the excipients were represented by a single structure in a simplified glyceride approach. COSMO-RS was able to capture the solubility trends in the different excipients. Only a few compounds showed rather poor predictions and these outliers were often comparatively larger molecules. The present study also evaluated the effects of individual fatty acid hydrolysis on glycerides’ solubilization capability. In conclusion, the application of COSMO-RS modeling for drug solubility prediction in lipid-based formulations is highly promising, in particular for rank-ordering excipients in an early development phase. In future, this in silico approach may also address solubilization effects of minor components in excipients or in excipient mixtures, which is interesting from a product quality perspective so that it can further advance this field of molecular pharmaceutics.

Synergistic Computational Modeling Approaches as Team Players in the Game of Solubility Predictions

“…Another study used COSMO-RS theory to predict solubility in glycerides. 12 An experimental dataset of 51 drugs was presented and the assumed simplified lipids provided reasonable predictions of the observed drug solubility. Effects of the number of ester bonds, chain length or saturation can be modeled, which gives valuable information to formulation scientists to rank excipients for their drug load capacity.…”

“…Importantly, transparency allows users to understand and challenge the model development facilitating future fine-tuning of models and improvement of predictions. Transparency also allows them to be used for solubility in complex solvents such as lipids 12 ; the underlying assumptions of the theoretical model can be used or adapted as needed for the specific solvent system of interest. This is a clear difference to entirely data-driven models in which statistical relationships are targeted between a molecular descriptor space and the solubility in a particular solvent system.…”

“…These include viewing solvents as being completely pure, and in the examples of lipidic solvents, typically the main component is used to describe the complete solvent. 12 For these reasons a theoryguided approach is an interesting alternative for solubility predictions. An example of this approach for solubility prediction is theory-guided selection of molecular descriptors for a statistical correlation model.…”

Artificial Neural Networks to Predict the Apparent Degree of Supersaturation in Supersaturated Lipid-Based Formulations: A Pilot Study

“…Over the last decade, interest regarding the use of ML algorithms across diverse disciplines in pharmaceutical design and development has grown [ 11 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 ]. While ML models have been produced to optimise lipid-based formulation (LBF) development [ 3 , 22 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ], the application of more novel ML approaches for bio-enabling formulations currently focuses on solid dispersions (SDs) [ 21 , 34 , 35 ]. However, their application to LBFs, particularly supersaturated LBFs (sLBFs), remains unexplored.…”

In-Silico Screening of Lipid-Based Drug Delivery Systems

“…The API solubility in the liquid excipient (mixture) is a key property for the development of LBDDS. [8] Therefore recent works focused on API-solubility measurements in either TGs [9][10][11], natural edible oils [11][12][13], or other commercial excipients [14,15]. The most-common experimental methods to measure API solubilities are differential scanning calorimetry (DSC) [9,13,16], Raman spectroscopy [9,13], UV-vis [14,17], and high-performance liquid chromatography (HPLC) [9,13,15].…”