icon star paper   HIV Articles  
Back grey_arrow_rt.gif
Protease Inhibitor and NNRTI Concentrations in the Genital Tract of HIV+ Women
  JAIDS Journal of Acquired Immune Deficiency Syndromes: Volume 37(5) 15 December 2004
Min, Sherene S. MD*; Corbett, Amanda H. PharmD†; Rezk, Naser MS†; Cu-Uvin, Susan MD¶; Fiscus, Susan A. PhD*‡; Petch, Leslie PhD§; Cohen, Myron S. MD*‡; Kashuba, Angela D. M. PharmD†
From the Schools of *Medicine and †Pharmacy and Department of ‡Microbiology and Immunology, and §Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill; and ¶Brown University School of Medicine, Providence, RI.
The pharmacokinetics of antiretrovirals (ARVs) in the female genital tract (FGT) are likely to influence vertical and sexual transmission of HIV, the development of viral resistance, and post-exposure prophylaxis regimens. This study is the first to compare ARV concentrations in direct aspirates of cervicovaginal fluid (CVF) and blood plasma (BP). This unique method provides direct assessment of concentrations without the confounding of cervicovaginal lavage dilution. Of 8 ARVs, CVF concentrations ranged from <10% to >100% of BP concentrations. These large differences in CVF penetration suggest that further research into ARV pharmacokinetics and drug efficacy in the FGT is necessary.
ARV concentrations in the FGT are extremely critical for the treatment and prevention of HIV transmission. However, only a single study has been undertaken to date, in which ARV concentrations were detected in only 16 of 43 samples (37%).6 Ours is the first study to investigate the feasibility of measuring ARV drug concentrations in direct CVF aspirates. Volumetric vaginal aspirators, previously used for family planning methods, were successfully used to reliably and directly collect CVF. Patients could perform this method of collection with ease, making frequent sampling for pharmacology studies feasible. Using this method, we found measurable drug concentrations in CVF aspirates for all ARVs tested with the exception of ritonavir and saquinavir. The alternative method, CV lavage, requires study personnel intervention and introduces an imprecise dilution factor. Our method should be used in future evaluations of FGT PK/PD relationships.
A number of physiochemical properties may contribute to the passage of ARV compounds across membranes.10 In our study, compounds with greater lipophilicity and lower protein binding achieved the highest concentrations in the female GT. The large protein binding influence is similar to what has been seen with drug concentrations in the male GT11 and in breast milk.
Because this was primarily a feasibility study, with pharmacology data generated in a small number of subjects, there are several limitations to extrapolating these findings. Drug concentrations were compared at only 2 time points. Differential CVF/BP concentration ratios of the ARV agents between peak and trough sampling suggest that full pharmacokinetic profiles of ARV agents will be required before conclusive statements can be made about the specific degree of GT exposure. Also, our subjects were not sampled at standardized times during the menstrual cycle. Because the volume of CVF varies during the menstrual cycle, it will be important to determine whether menstrual cycle stage affects drug concentrations. Finally, the degree of protein binding in CVF is unknown. One study determined that concentrations of albumin and α-1-acid glycoprotein in cervical mucus were <1% that of blood. Therefore, although lower total drug concentrations may be present in this fluid, low protein content may make these local concentrations as effective as higher drug concentrations found in BP. Consequently, future evaluations of ARV concentration-efficacy relationships in the FGT (using HIV RNA or DNA) will be very important.
Due to sample collection limitations and less well-defined PK/PD relationships, NRTIs were not measured in this initial evaluation. Investigations of this ARV class should include assessing intracellular triphosphate concentrations, the active form of these compounds. However, intracellular evaluation requires collection of cervical mononuclear cells, and a robust method for collecting sufficient numbers of cells had not been developed at the time of this study. A method is currently under study in our laboratory.
This investigation was not powered to detect statistically significant differences between BP and GT ARV concentrations or between different ARV therapies and their effect on viral persistence and evolution. Timings of HIV-1 RNA samples with respect to length of treatment may not have been adequate to detect a difference in viral replication and evolution between the BP and GT. The main purpose of the study was to assess our method of collection and to determine whether drug concentrations could be reliably measured by directly sampling CVF.
Our observations raise the possibility that different ARV drugs may have different antiviral effects in the FGT, although definitive conclusions could not be made in this exploratory study. These findings have implications for vertical and sexual transmission of the virus from HIV-infected women and for prevention of sexual transmission to uninfected women. Further research into ARV pharmacology in the FGT and the effects on viral replication and evolution is clearly required to assess its impact on sexual and vertical transmission of HIV.
UNAIDS and the World Health Organization estimate that women represent approximately 50% of people living with HIVand AIDS and 50% of new infections worldwide. It is important to study the behavior of HIV in the female genital tract (FGT), as it may be a separate biologic compartment for HIV replication. Differential penetration of antiretrovirals (ARV) into the FGT could alter the rates of vertical and heterosexual transmission, in addition to efficacy of post-exposure prophylaxis regimens. Decreased ARV concentrations in the FGT may predispose a woman to develop HIV mutations that could impact transmission to others as well as the development of ARV failure in the individual.
To date, only a single investigation of ARV concentrations in the FGT has been undertaken. However, the samples collected were cervicovaginal lavage fluid (CVF), and the extent of sample dilution could not be determined. Additionally, sampling was performed inconsistently with respect to drug administration.
We use a novel direct collection strategy for determining protease inhibitor (PI) and nonnucleoside reverse transcriptase inhibitor (NNRTI) concentrations in the FGT to avoid any confounding obtained with lavage.
Patient Characteristics and Visits

Plasma and genital samples were collected from 7 women (4 African American, 3 white) with a median age of 40 years (range 26-60 years) who completed a total of 24 study visits. One patient was excluded after her first visit due to anemia. The study participants were taking a mean of 3 (range 2-4) ARV medications at the time of study enrollment.
Concentration of Antiretroviral Drugs
The extracellular fractions of 68 paired BP and CVF samples were evaluated for PI and NNRTI concentrations (table 1). BP concentrations for all ARV drugs measured were within expected steady-state concentrations.
Among PIs, indinavir had the highest concentrations in the FGT at trough and peak compared with BP. Amprenavir CVF concentrations were approximately 50% that of BP. Lopinavir, saquinavir, and ritonavir had the lowest concentrations in the FGT. Ritonavir was below quantifiable limits in 9 of 11 C12h and 8 of 10 C3-4h CVF samples. We chose to assign these nonquantifiable results 25 ng/mL (the lower limit of quantification). Therefore, the true geometric mean and 95% CIs will be less than what is reported in Table 1. Among NNRTIs, nevirapine had the highest concentration in CVF.
HIV RNA Measurements and Genotyping
HIV-1 RNA was measured in 23 paired BP and GT samples from 7 subjects while on their current ART and at least 1 month and up to 5 months after initiation of new ART. No significant differences between BP and GT HIV-1 RNA were observed in paired samples obtained on current or new ART. Reverse transcriptase and protease genotyping were performed on paired BP and GT samples from 2 subjects on current ART and from 2 subjects on new ART. On all paired samples, major and minor reverse transcriptase and PI mutations were similar in the 2 compartments, regardless of drug therapy.
Study Population

HIV-1-infected, nonpregnant women were recruited from the University of North Carolina Infectious Diseases Clinic and the Immunology Center of the Miriam Hospital between August 2001-July 2002. Subjects were considered for enrollment if a screening HIV-1 RNA was >=1000 copies/mL, the primary provider planned to change ARV therapy (ART), and the current regimen contained at least 1 PI or 1 NNRTI. Women with a history of anemia, hysterectomy, abnormal vaginal bleeding or discharge, and active sexually transmitted diseases were excluded from the study. Participants refrained from intercourse or douching 48 hours prior to each visit. Participants completed an ARV administration log for 72 hours prior to each study visit to evaluate adherence. Samples were not collected during menstruation. Each patient was expected to complete 2 study visits while taking her current ART and 2 study visits after changing her ART.
Informed consent was obtained from all study participants. Human experimentation guidelines of the US Department of Health and Human Services and those of the University of North Carolina at Chapel Hill and Brown University School of Medicine were followed in the conduct of this research study.
Sample Collection and Processing
At each study visit, blood plasma (BP) and CVF were obtained simultaneously around directly observed morning drug doses of ART at the subject's usual dosing time. Trough concentrations were collected at the end of the dosing interval (8 or 12 hours after documented dosing), and peak concentrations were collected 3-4 hours after an observed dose. Efavirenz concentrations were obtained 12 hours after an evening dose and 3-4 hours after an observed morning dose. Blood was collected in tubes containing EDTA; plasma was separated by centrifugation, aliquoted into cryovials, and stored at -80°C. CVF was obtained from the posterior fornix of the vagina via direct aspiration with a volumetric vaginal aspirator (Rovumeter; Recipe Pharmaceuticals, Munich, Germany) by the patient or study coordinator. CVF was stored in preweighed cryovials at -80°C. A pelvic examination was performed 3-4 hours after an observed morning ARV dose. Five Sno*strips (Akron, Abita Springs, LA) were placed in the endocervical canal for 3-5 minutes and then stored in nucleic acid sequence-based amplification buffer at -80°C for future HIV RNA determinations. All samples were processed within 1 hour of collection.
NNRTI and PI Concentrations
All testing was performed by the University of North Carolina Center for AIDS Research. Plasma PI and NNRTI concentrations were determined by validated high-performance liquid chromatography UV methods after solid-phase extraction, as modified from Rezk et al8 The lower limit of quantification was 25 ng/mL for all drugs. Inter- and intraday coefficients of variation were <8%.
HIV-1 RNA Determinations
All testing was performed by the University of North Carolina Center for AIDS Research. BP HIV-1 RNA was quantified with the Roche Amplicor Monitor standard assay, version 1.0 (Roche Molecular Systems, Branchburg, NJ). Cervical fluid HIV-1 RNA was extracted using the NucliSens method (bioMérieux, Durham, NC) and amplified and quantified with the Roche Amplicor Monitor standard assay or the Organon NucliSens HIV QT 2.0 (bioMérieux). The lower limit of detection was 200 copies/mL.
HIV-1 RNA Genotypes
Viral RNA was extracted from 140 μL of BP using the QIAamp Viral RNA isolation kit from Qiagen (Venlo, The Netherlands) or 24 μL of cervical fluid as described above. Reverse transcription polymerase chain reaction and sequencing were performed using the TRUGENETM HIV-1 Genotyping kit (Bayer Diagnostics, Tarrytown, NY). Genotypes were analyzed using the OpenGene system, version 3.1.5 software (Bayer Diagnostics). Genotypes were only obtained for paired BP and cervical specimens with a cervical HIV-1 RNA >400 copies/mL.
Statistical Analysis
ARV drug concentrations are expressed as geometric mean (μg/mL) with corresponding 95% CI. CIs were calculated using repeated-measures analysis of variance testing. This pilot study was not powered for specific statistical testing but rather designed to provide data on which to base future sample size calculations for evaluating pharmacokinetic-pharmacodynamic (PK/PD) relationships in the FGT. Statistical analysis was performed using SAS v8.0 (SAS Corp., Cary, NC).
  icon paper stack View Older Articles   Back to Top   www.natap.org