New methods to understand the physical chemistry and improve the formulation of topical drug products.

Abstract

The compositions of semisolids tend to be modified without quantitative measurements to assess whether the changes improve or hinder their function. A more scientific approach to their formulation is needed. Consequently, three methods to obtain the diffusion coefficient and the solubility of drugs in semisolids have been devised and were used to establish if changes in diffusivity, the kinetic determinant of release, or solubility, the thermodynamic determinant of release, or a combination of the two are responsible for altered release arising from changes in formula composition or processing. The first method involves direct equilibrium measurement of solubility in a specifically constructed apparatus. Equilibrium is achieved between an initially blank layer of semisolid and a polymer matrix containing suspended drug. In the second method, release data, gathered using a <italic>in vitro </italic> release test prescribed by the FDA, are plotted on a log-log scale, rate of release versus total concentration. The solubility is taken as the intersection of lines representing the respective release dependencies of solution and suspension systems. The third method involves estimating the diffusivity by following the square root of time release from a solution of the drug in the test matrix. One then follows the release from a concentrated suspension of the drug in the same matrix, yielding the product of the diffusivity and the solubility. Solubility is calculated using the previously determined diffusivity. The methods were used on two different matrices. Hydrocortisone's solubility in creams ranged from 0.012% to 0.064%, and diffusivity ranged from 1.6 x 10<super>-4</super> to 7.4 x 10<super>-4</super> cm<super> 2</super>/h, depending on compositions. Benzocaine's solubility in a gel formulation was found to be 0.31% and its diffusivity 1.0 x 10<super>-2</super> cm<super>2</super>/h. The agreement between the equilibrium and kinetic methods was excellent and the extrapolation also produced a fair result; the results were successfully used to predict and understand excipient-induced trends in drug release. The methods are believed to be useful and valid for most semisolids. Finally, a conceptual model for the influence of particle size on release was developed, implemented in a numerical simulator, and found to explain the major trends seen in the experimental data.