Tenofovir (TFV) is a proven microbicide when administered topically as a vaginal gel. To improve its efficacy, TFV was combined with the nonnucleoside reverse-transcriptase inhibitor UC781 in a vaginal gel. Mixture design of experiments theory was used to define a range of gel compositions with varying rheological properties and to assess in vitro drug release and tissue retention. Experiments and computations led to the specification of three different gels referred to as a spreading gel (SG), an intermediate spreading gel (ISG), and a bolus gel (BG). These three gels, all containing 1.0% TFV and 0.1% micronized UC781, were evaluated for in vitro release, in vitro tissue retention and safety, and in vivo pharmacokinetics in the rabbit. There were some differences in in vitro release rates of UC781 (the higher the gel viscosity, the slower the release rate) across gels, while release of TFV was independent of gel type. In an organotypic human vaginal-ectocervical (VEC) tissue model, the amounts of tissue-associated TFV and UC781 were several orders of magnitude higher than their in vitro half-maximal inhibitory concentration. There were no differences in VEC tissue concentrations of TFV or UC781 between the SG, ISG, and BG. All three gels were well tolerated in the VEC model as assessed by tissue viability, electrical resistance, histology, and cytokine (interleukin-8 and interleukin-1 beta) release. The local vaginal tissue concentrations in rabbits following a single dose or seven once-daily doses were variable and generally lower than those found in the VEC tissue model. The approach described herein provides a rational schema to design and evaluate vaginal gels for use as microbicides.
Copyright © 2012 Wiley Periodicals, Inc.