Vaginal drug distribution modeling

Abstract

This review presents and applies fundamental mass transport theory describing the diffusion and convection driven mass transport of drugs to the vaginal environment. It considers sources of variability in the predictions of the models. It illustrates use of model predictions of microbicide drug concentration distribution (pharmacokinetics) to gain insights about drug effectiveness in preventing HIV infection (pharmacodynamics). The modeling compares vaginal drug distributions after different gel dosage regimens, and it evaluates consequences of changes in gel viscosity due to aging. It compares vaginal mucosal concentration distributions of drugs delivered by gels vs. intravaginal rings. Finally, the modeling approach is used to compare vaginal drug distributions across species with differing vaginal dimensions. Deterministic models of drug mass transport into and throughout the vaginal environment can provide critical insights about the mechanisms and determinants of such transport. This knowledge, and the methodology that obtains it, can be applied and translated to multiple applications, involving the scientific underpinnings of vaginal drug distribution and the performance evaluation and design of products, and their dosage regimens, that achieve it.

Keywords: Gel; HIV; Microbicides; Modeling; Mucosa; Ring; Vagina.

Figure 1

Drawing of human vaginal canal and mucosal tissue (not to scale). The canal contains gel that partially coats the mucosal surface (anti-pathogenic microbicide molecules are depicted as red dots), and there is semen in the lumenal space not occupied by gel. On the left is a column that represents the tissue collected by a punch biopsy, typically used to collect tissue for measuring drug concentration in pharmacokinetic studies.

Figure 2

Schematic drawing of the gel spreading and drug mass transport problem. Gel flow and drug mass transport are symmetrical in the half planes about y = 0. The problem is therefore solved for y > 0.

Figure 3

Concentration distributions at 1, 4, 12 and 24 hours after insertion of 4 mL of 1% TFV gel to the vaginal fornix. Gel spreads from right-to-left. This view shows the upper half plane of the domain of the problem (cf. Fig 2).

Figure 4

Compartment-averaged concentrations vs. time of TFV and TFV-DP from model predictions for the 1% TFV clinical gel inserted to the fornix in a human vaginal canal of average size (cf. Figure 3). Two clinically relevant gel volumes, 2 mL and 4 mL, are considered.

Figure 5

Values of percent protected in stroma vs. time for three log-spaced values of the target EC₅₀ value.

Figure 6

Model simulations of concentrations of TFV measured in biopsies and of TFV-DP in stroma vs. time after initial gel insertion. A second gel insertion is performed at times given in the figure.

Figure 7

Model predictions of distances that gel has spread along the vaginal canal for fresh gel vs. 12 months of accelerated aging. Data for the prototype microbicide gel inserted to the fornix is a vaginal canal of average dimensions (see above; the length of the canal is 13 cm). The line for 4 mL of aged gel becomes horizontal at the moment the gel has completed coating the entire length of the canal (subsequent to which it begins to leak).

Figure 8

Model predictions of average TFV concentration in a biopsy vs. time after insertion for fresh vs. aged gel (as in Figure 7).

Figure 9

Model predictions of average TFV-DP concentration in the stroma vs. time after insertion for fresh vs. aged gel (as in Figure 7).

Figure 10

Heat map plot of Tenofovir concentration distribution at steady state for an intravaginal ring. Comparison of TFV concentration distributions achieved by the gel and IVR can be made by plotting the average concentration in each compartment as a function of time, cf. Figure 11.

Figure 11

Comparison of volume averaged TFV concentrations vs. time for the 1% TFV gel vs. an intravaginal TFV releasing ring. For details see text.

Figure 12

Fractional vaginal surface area coated vs. time, predicted by the model for the 1% Tenofovir gel for human, macaque and murine sized vaginal canals. Gels are inserted to either the fornix or mid canal. Further details are given in text.