Background: Trials of a vaginal Tenofovir gel for pre-exposure prophylaxis (PrEP) for HIV have given conflicting results. Knowledge of concentrations of Tenofovir and its active form Tenofovir diphosphate, at putative sites of anti-HIV functioning, is central to understanding trial outcomes and design of products and dosage regimens. Topical Tenofovir delivery to the vaginal environment is complex, multivariate and non-linear; determinants relate to drug, vehicle, dosage regimen, and environment. Experimental PK methods cannot yield mechanistic understanding of this process, and have uncontrolled variability in drug sampling. Mechanistic modeling of the process could help delineate its determinants, and be a tool in design and interpretation of products and trials.
Methods and findings: We created a four-compartment mass transport model for Tenofovir delivery by a gel: gel, epithelium, stroma, blood. Transport was diffusion-driven in vaginal compartments; blood concentration was time-varying but homogeneous. Parameters for the model derived from in vitro and in vivo PK data, to which model predictions gave good agreement. Steep concentration gradients occurred in stroma ≤ 8 hours after gel release. Increasing epithelial thickness delayed initial TFV delivery to stroma and its decline: tmax increased but AUC at 24 hours was not significantly altered. At 24 and 48 hours, stromal concentrations were 6.3% and 0.2% of C(max). Concentrations in simulated biopsies overestimated stromal concentrations, as much as ∼5X, depending upon time of sampling, biopsy thickness and epithelial thickness.
Conclusions: There was reasonably good agreement of model predictions with clinical PK data. Conversion of TFV to TFV-DP was not included, but PK data suggest a linear relationship between them. Thus contrasts predicted by this model can inform design of gels and dosage regimens in clinical trials, and interpretation of PK data. This mass transport based approach can be extended to TFV conversion to TFV-DP, and to other drugs and dosage forms.