This is the first study to comprehensively evaluate antiretroviral drug exposure in the GT after single and multiple dosing. In this study of 27 HIV-infected women, GT concentrations of all 11 antiretroviral drugs reported here were detected rapidly after single doses. Differing GTexposures between the drugs relative to BP, however, suggest that only certain drugs may be preferable for PrEP or PEP regimens.
Of the nucleoside analogs evaluated, 3TC, ZDV, FTC, and TDF achieved GT concentrations similar to, or higher than, BP, and therefore would be excellent PEP and PrEP candidates for further study. Mean 3TC and ZDV concentrations in the GT were near or above extracellular concentrations suggested for clinical efficacy [20
]. The median GT AUC of FTC at steady state was approximately three times higher than the suggested BP AUC of 10 ug*h/ml for clinical efficacy [23
Of the other compounds studied, LPVand ATVachieved low to moderate GT concentrations. Since these may be similar to the recommended BP trough concentrations necessary for efficacy in patients with wild-type HIV-1 virus [22
], further study of these agents is warranted for use in PrEP and PEP either given alone or in combination with NRTIs. EFV had very low GT concentrations relative to BP, and is unlikely to be useful in PEP/PrEP regimens.
Low drug exposures in the GT may contribute to higher viral shedding in the GT and the development of viral resistance. Recently, receipt of NNRTI-containing therapy was found to be independently associated with HIV shedding at the cervix (odds ratio = 2.24 compared to a PI-based regimen) in 107 women (31 on NVP and 76 on EFV) [24
]. Additionally, data presented from the ACTG study A5077 suggested an increased rate of NNRTI mutations in the female GT [25
] in individuals who were highly antiretroviral drug-experienced. Additionally, a recent study of 14 women found three with partial or complete compartmentalization of HIV in the GT compared to BP [26
]. In our study, the low exposures of EFV (0.5% GT: BP AUC) observed in this study suggest NNRTI drug resistance to be biologically plausible [27
]. As drug concentrations in the GT are seldom measured, however, a direct relationship between drug exposure, viral burden, and viral resistance patterns in the GT has not been established.
It is currently unclear which specific mechanisms and physicochemical properties of drugs dictate passage into the female GT. However, drug penetration appears to be related to the protein-binding capacity of each drug. Highly protein-bound drugs such as LPV (98–99% protein bound) [29
] and ATV (86% protein bound) [30
] had lower GT exposures (8 and 18%, respectively, at steady state), whereas drugs with low protein-binding such as the nucleoside/tide analogs ZDV, 3TC, and FTC (<4% to <38% protein bound) [31
], generally had higher GTexposures (ranging from 235–411%). d4Tand ddI were exceptions to this trend, as GT exposures were low (5 and 21%, respectively) despite <5% protein binding [34
]. For d4T, peak plasma concentrations occur at approximately 2 h post-dose. Since our first sampling time was 2 h post-dose, it is possible that the true peak concentration was not captured, and complete exposure information was not obtained.
To further this initial pharmacokinetic evaluation, a number of investigations are desirable. Linking drug exposure to biologic response (e.g., decrease in HIV RNA) in the female GT is critical to confirm that higher extracellular drug concentrations are predictive of a more potent response and guide drug selection for PrEP and PEP. Pharmacokinetic/pharmacodynamic modeling analyses are ongoing in our laboratory to evaluate the correlation between differential GT drug exposure and virologic response. Additionally, our previous NRTI work in male GT mononuclear cells suggests similar or lower concentrations of intracellular triphosphate active metabolites compared to BP despite higher GT extracellular drug exposures [36
]. Therefore, research characterizing the intracellular active triphosphorylated metabolites for NRTIs in the female GT is necessary. Our study observed an extracellular half-life of TDF in the GT secretions twice as long as the extracellular half-life of TDF in BP, suggesting that TDF might be dosed at extended intervals for PEP or PrEP. Under first dose conditions, the median half-life was 14.5 h (95% CI: 3.6–29.2 h) in the GT, and 7.4 h (95% CI: 4.8–12.4 h) in BP (data not shown). However, characterizing intracellular concentrations of tenofovir diphosphate in GT mononuclear cells would be most important to move this application forward.
Although for most drugs, no differences in plasma drug exposure have been found between HIV-infected subjects and healthy volunteers [28
], it may also be beneficial to evaluate the GT exposures of these drugs in healthy women, as they will be the primary recipients of PrEP and PEP. Additionally, no data exist on the influence of local infection and inflammation on drug concentrations in the female GT. As 130–1800% increases in drug concentrations have been documented with the inflammation present in meningitis and prostatitis, [37
], GT infections may also increase antiretroviral drug concentrations in addition to increasing susceptibility to HIV infection [40
Finally, understanding antiretroviral protein binding in the GT is necessary. This investigation measured total (protein unbound drug + protein bound drug) drug concentrations in both BP and GT. However, it has been suggested that concentrations of drug binding proteins (specifically albumin and alpha-1 acid glycoprotein) in the female GT may be lower than in BP [43
]. If this is true, despite differences in total drug concentrations between BP and GT, free drug concentrations in the GT may be similar to, or greater than, BP. Hence, free drug concentrations may be a more important predictor of efficacy. Although these types of investigations are technically difficult and lend themselves better to in-vitro study, this work will be required for a more thorough understanding of antiretroviral behavior in the GT.
In conclusion, this investigation is the first to comprehensively evaluate antiretroviral drug exposure in the female GT. Regardless of drug, GT concentrations were detected 2–4 h after a single antiretroviral dose. Additionally, GT exposures were similar after a first dose and at steady state. The results of this investigation support the use of 3TC, ZDV, TDF, and potentially FTC as excellent PrEP/PEP candidates. ATVand LPV/r may prove useful agents due to favorable GT concentrations in relation to HIV-1 wild-type susceptibility. We believe agents that achieve less than 10% of BP exposure, such as EFV and d4T, are less likely PrEP/PEP candidates.