Micellar chromatographic partition coefficients and their application in predicting skin permeability

The major goal for physicochemical screening of pharmaceuticals is to predict human drug absorption, distribution, elimination, excretion and toxicity. These are all dependent on the lipophilicity of the drug, which is expressed as a partition coefficient i.e. a measure of a drug’s preference for the lipophilic or hydrophilic phases. The most common method of determining a partition coefficient is the shake flask method using octanol and water as partitioning media. However, this system has many limitations when modeling the interaction of ionised compounds with membranes, therefore, unreliable partitioning data for many solutes has been reported. In addition to these concerns, the procedure is tedious and time consuming and requires a high level of solute and solvent purity. Micellar liquid chromatography (MLC) has been proposed as an alternative technique for measuring partition coefficients utilising surfactant aggregates, known as micelles.

This thesis investigates the application of MLC in determining micelle-water partition coefficients (logPMW) of pharmaceutical compounds of varying physicochemical properties. The effect of mobile phase pH and column temperature on the partitioning of compounds was evaluated. Results revealed that partitioning of drugs solely into the micellar core was influenced by the interaction of charged and neutral species with the surface of the micelle. Furthermore, the pH of the mobile phase significantly influenced the partitioning behaviour and a good correlation of logPMW was observed with calculated distribution coefficient (logD) values. More interestingly, a significant change in partitioning was observed near the dissociation constant of each drug indicating an influence of ionised species on the association with the micelle and retention on the stationary phase. Elevated column temperatures confirmed partitioning of drugs considered in this study was enthalpically driven with a small change in the entropy of the system because of the change in the nature of hydrogen bonding. Finally, a quantitative structure property relationship was developed to evaluate biological relevance in terms of predicting skin permeability of the newly developed partition coefficient values. This study provides a better surrogate for predicting skin permeability based on an easy, fast and cheap experimental methodology, and the method holds the predictive capability for a wider population of drugs.

In summary, it can be concluded that MLC has the ability to generate partition

coefficient values in a shorter time with higher accuracy, and has the potential to replace the octanol-water system for pharmaceutical compounds.