Purpose: Develop a preclinical in vitro algorithm enabling de novo design of semisolid vaginal drug delivery gels, by using biomechanical modeling of gel spreading in the vaginal canal and empirically relating gel composition to mechanical properties and predicted performance.
Methods: Gel performance was defined through a multivariate objective function constructed from gels' mechanical properties and selected performance criteria for gel spreading within the vaginal canal. Mixture design of experiment was used to establish a semi-empirical relationship linking composition-property and property-performance relationships for gels with varying concentrations of hydroxyethylcellulose and Carbopol 974P. This permits definition of a local optimum for gel composition and volume of administration, within a defined gel composition space.
Results: Rheological behavior and, consequently, the value of the objective function varied broadly with composition. The algorithm indicated a 3.0 wt% HEC gel as the near optimal composition for a 3.5 mL applied volume for gels designed to spread throughout the vagina.
Conclusions: The algorithm introduced herein is a novel tool that facilitates an understanding of the composition-property-performance relationship for vaginal semisolid drug delivery gels. This approach has promise as a scientific methodology for evaluation and optimization of vaginal gels prior to in vivo investigations.