iconstar paper   HIV Articles  
Back grey arrow rt.gif
Tenofovir Disoproxil Fumarate Intravaginal Ring Protects High-Dose Depot Medroxyprogesterone Acetate-Treated Macaques From Multiple SHIV Exposures
  Download the PDF here
JAIDS Journal of Acquired Immune Deficiency Syndromes: Jan 1 2015
Smith, James M. PhD*; Srinivasan, Priya MS*; Teller, Ryan S. PhD; Lo, Yungtai PhD; Dinh, Chuong T. BS; Kiser, Patrick F. PhD; Herold, Betsy C. MD
*Division of HIV/AIDS Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, GA; Department of Biomedical Engineering, Northwestern University Evanston, Evanston, IL; Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, NY; Total Solutions, Inc, Atlanta, GA; and Department of Pediatrics and Microbiology-Immunology, Albert Einstein College of Medicine, Bronx, NY.
Preclinical HIV prevention models use either a single high-dose viral challenge in depot medroxyprogesterone acetate-treated macaques or repeated viral challenges in cycling macaques. We tested the efficacy of an intravaginal tenofovir disoproxil fumarate (TDF) ring in a model combining repeated 30-mg injections of depot medroxyprogesterone acetate every 6 weeks with vaginal viral challenges weekly for 12 weeks. Twelve macaques were randomized to TDF or placebo rings. All placebo macaques became infected after a median of 2 exposures, whereas only 1 TDF macaque became infected at the eighth exposure (P = 0.0012). The TDF ring provides durable protection in a stringent challenge model.
Pre-exposure prophylaxis (PrEP) has become one of the frontline strategies to reduce HIV acquisition and transmission. Coitally dependent topical PrEP with a 1% tenofovir (TFV) gel provided modest protection in one human study, but difficulties with adherence limit the effectiveness of gel formulations.1 Intravaginal rings (IVRs) have the potential to overcome some adherence hurdles and, by providing sustained drug delivery, may prove more effective compared with vaginal gels.2 The concentrations of drug that must be delivered to protect against HIV are not known and may be affected by hormonal contraception, which could modulate drug pharmacokinetics (PK) and pharmacodynamics in ways that have yet to be fully delineated.
Epidemiological studies suggest that the use of injectable depot medroxyprogesterone (DMPA) may increase the risk of HIV acquisition.3,4 The underlying mechanisms are not yet established, although in vitro and animal studies indicate that DMPA may have pleiotropic effects, including thinning of the epithelial barrier, alterations in the vaginal microbiome, and immunomodulatory effects.5 Concerns about DMPA have fostered the design of multipurpose IVRs to deliver antiretroviral drugs in combination with levonorgestrol or other hormonal contraceptives.6,7
The nonhuman primate model provides the opportunity to evaluate the potential safety and efficacy of oral and topical PrEP as illustrated by results with oral Truvada, TFV gel, and a polyurethane reservoir IVR delivering the prodrug, tenofovir disoproxil fumarate (TDF).8,9 The vaginal exposure models include a single high-dose (500-10,000 TCID50) simian immunodeficiency virus (SIV) or simian-human immunodeficiency virus (SHIV) challenge in DMPA-treated (single 30-mg injection) rhesus macaques, and a lower dose (50 TCID50) multiple SHIV challenge model in cycling pigtail macaques.10 We hypothesized that combining DMPA treatment with repeat viral challenges may provide insights into how DMPA might impact TDF PK and PrEP efficacy. We elected to treat pigtail macaques with 30-mg DMPA every 6 weeks, a dose that is approximately 10-fold higher in milligram/kilogram than the human dose, which results in marked thinning of the epithelium5 and compared efficacy of the TDF IVR to a placebo IVR in DMPA-treated pigtail macaques challenged with SHIV weekly for 12 weeks. We also compared the findings to results obtained in a previous study in which the TDF ring completely protected 6 cycling macaques from 16 weekly challenges with SHIV.9
PK and Efficacy

Twelve sexually mature female pigtailed macaques received intramuscular injections of 30-mg DMPA every 6 weeks starting 3 weeks before insertion of a TDF (T1-T6, N = 6) or a placebo IVR (P1-P6, N = 6); details on ring design were previously described.9 DMPA doses ranged from 3.8 to 5.8 mg/kg. The first of 12 weekly vaginal exposures to SHIV162p3 (50 TCID50) was administered 1 week after insertion of the IVR. The final IVR was removed at week 14. IVRs were replaced every 4 weeks, 3 days after virus exposure. Macaques were monitored weekly for infection by real-time polymerase chain reaction with a lower limit of detection of 50 copies per milliliter of plasma.9 Resistance mutations of SHIV162p3 in plasma were assessed by polymerase chain reaction with a mean limit of detection for the K65R mutation of 0.4%.11
Vaginal secretions were collected at each IVR exchange and during the follow-up period with Ultracell surgical sponges (3.5 x 4 mm). TFV content was quantified with a lower limit of detection of 5 ng/mL.9 Residual TDF in used rings and plasma progesterone levels were assayed as previously described.12,13 All macaques were housed at the Centers for Disease Control and Prevention under the Guide for the Care and Use of Laboratory Animals (National Research Council of the National Academies, 2010).
Statistical Analysis
SHIV infection-free survival curves for the TDF + DMPA, DMPA placebo, historical TDF IVR cycling, and historical cycling, no ring groups were performed using Kaplan-Meier plots. Each treatment group was compared with its own control group using a log-rank test. Wilcoxon rank-sum tests with Bonferroni adjustment for multiple comparisons were used to compare TFV levels in vaginal secretions between DMPA and cycling macaques. Student t tests with Bonferroni adjustment for multiple comparisons were used to compare residual drug levels in used rings between DMPA and cycling macaques. Statistical analyses were performed with GraphPad Prism (version 6; GraphPad Software, La Jolla, CA) and SAS (version 9.3; SAS Institute, Inc, Cary, NC).
PK of Progesterone, MPA, and TFV

To determine whether DMPA fully suppressed progesterone production, plasma progesterone was measured weekly beginning 6 weeks before first DMPA dose and ending at the week of the last virus exposure (Fig. 1A). The plasma progesterone levels fell below 400 pg/mL within 2 weeks of the first DMPA dose and remained low throughout the experimental period (Fig. 1B). One animal (T3) had a single spike in plasma progesterone at week 10 and a second animal (T6) continued to show plasma progesterone peaks similar to cycling animals, despite having detectable MPA levels after each DMPA injection (see Table S1, Supplemental Digital Content, http://links.lww.com/QAI/A584).
Vaginal secretions were collected for TFV analysis in the TDF ring-treated animals at each IVR exchange and at weeks 13, 14, and 15 and were compared with those obtained in the previous study in cycling macaques.9 Median TFV levels in both proximal (circles) and distal (triangle) vaginal secretions were greater than 1 x 105 ng/mL throughout the 12-week exposure period and persisted at week 15, 1 week after removal of the last IVR (Fig. 2A). Proximal TFV drug levels in macaques that received DMPA were significantly higher compared with cycling macaques at 4 1/2 weeks (Wilcoxon rank-sum, Bonferroni adjustment P = 0.012), at 8 1/2 weeks (P = 0.007), and at 12 1/2 weeks (P = 0.007); distal TFV levels were also significantly higher at 4 1/2 weeks in DMPA-treated compared with cycling animals (P = 0.012). Similarly, the average daily release rate of TDF from the rings removed at 4 1/2 weeks was significantly higher in the DMPA-treated compared with the cycling macaques (t test, Bonferroni adjustment, P < 0.001) (Fig. 2B).9 Notably, plasma TFV levels, which were measured weekly, were below the limit of detection (5 ng/mL) in all animals.
Susceptibility of DMPA-Treated TDF and Placebo IVR Macaques to Repeated Challenges
All placebo IVR macaques became infected with a median of 2 exposures compared with median of 4 exposures for cycling animals (Fig. 2C). In contrast, only 1 of 6 DMPA TDF IVR-treated macaques became infected (T6 at week 8) (83% protection, log-rank test, P = 0.0012). Animals were monitored for SHIV by polymerase chain reaction through week 21 and no additional animals became infected. There was a trend toward higher plasma viral loads in the placebo IVR DMPA-treated macaques compared with historic cycling control animals (Figure 2D; see Figure S1, Supplemental Digital Content, http://links.lww.com/QAI/A584) that was not statistically significant (log-rank test, P = 0.301). Interestingly, peak viral load in the 1 breakthrough animal, T6, was ~100-fold lower than the median viral load of DMPA, placebo IVR-treated macaques (2.6 x 105 copies/mL vs. 2.4 x 107 copies/mL). Furthermore, T6 had undetectable plasma virus beginning 5 weeks after peak viremia, whereas the placebo + DMPA animals continued to have detectable plasma virus 12 weeks after peak viremia (median, 3.4 x 105 copies/mL). TDF-associated resistance (K65R mutation) was not detected in the TDF IVR-infected animal at any time where plasma virus was detected.11
This study tested the efficacy of a TDF IVR in a model that combined multiple high-dose injections of DMPA with weekly exposures to virus. We hypothesized that DMPA treatment would increase the susceptibility to SHIV by inducing an exaggerated and sustained luteal phase. Previous studies indicate that cycling animals are more susceptible in the luteal phase of the menstrual cycle and that the window of susceptibility in cycling macaques encompasses approximately one half of the menstrual cycle.13 In this study, DMPA-treated macaques in the placebo arm became infected after a median of 2 challenges compared with 4 challenges in historical control cycling animals. This trend is consistent with previous studies reporting an increased risk of acquisition of SIV with DMPA. Notably, the dose and intervals of DMPA administration were not designed to simulate the hormone levels or vaginal histology of women using DMPA for contraception, but rather to provide a "worst-case" scenario by maintaining a reduced vaginal epithelial layer throughout the exposure period.5 The small group sizes in this study did not offer sufficient power to statistically evaluate the role of DMPA in differences observed in time to infection, peak viral load, or viral clearance from the plasma.
Although no IVR has yet proven effective against HIV acquisition in women, several have provided varying degrees of protection in nonhuman primates with PK and efficacy studies informing IVR design.2,9,14 For example, results from PK studies prompted the advancement of the current TDF reservoir ring rather than a matrix IVR design.11 The reservoir TDF ring provided complete protection in cycling macaques against 16 weekly virus exposures9 and protected 5 of 6 DMPA-treated macaques from 12 weekly challenges in this study. The addition of DMPA to the repeat challenge model may provide a highly stringent model of the efficacy of drug delivery systems if infections consistently occur earlier in the control animals as observed here.
The TDF IVR produced high TFV levels in macaque vaginal fluids. We focused on measuring TFV rather than TDF levels because luminal TDF is hydrolytically converted into TFV and thus less variable than TDF levels.9 Importantly, the TFV levels exceeded those in the CAPRISA 004 study (1000 ng/mL) that correlated with protection in women receiving 1% TFV gel.15 Similar TFV levels were observed proximal and distal to the ring indicating well-dispersed tissue coverage with sustained delivery. Notably, high levels of TFV persisted 1 week after ring removal (week 15), suggesting that women may be protected for several days after ring removal. The TFV levels in vaginal secretions of DMPA-treated animals were significantly higher at several time points in this study compared with those observed in the cycling animals, which may reflect either greater drug release and/or thinning of the epithelial barrier allowing drug to more easily penetrate the tissues and more readily return to the vaginal vault. However, because we did not obtain biopsies and thus did not measure intracellular TFV-diphosphate levels, we cannot exclude the possibility that differences in drug levels reflect differences in drug metabolism.
The one TDF animal that became infected at the eighth exposure cannot be explained with the data available. The TFV levels in vaginal secretions in this animal were within the range of those in protected animals. However, this macaque did have an unusual pharmacological response to the DMPA injections; plasma MPA levels were within range of other DMPA-treated animals and consistent with levels that have been previously shown to suppress progesterone production,5 yet T6 continued to produce progesterone (Fig. 1). This aberrant response might be expected to decrease the risk of SHIV if the vaginal epithelium fluctuated in thickness with the changes in progesterone levels. Unfortunately, we were unable to sample the vaginal epithelium during the exposure period. It is also possible that lower than expected intracellular TFV-diphosphate levels or increased numbers of activated immune cells contributed to the lack of protection in this one animal.
More intensive PK studies in DMPA-treated animals are planned to assess the impact of sustained DMPA treatment on vaginal epithelium, mucosal immune cell populations, and levels of TFV-diphosphate in lymphocytes from cervicovaginal tissue and lymph nodes. These studies should include a range of DMPA dosing and intervals as the current regimen was not designed to simulate the hormone levels or vaginal histology of women using DMPA for hormonal contraception, but rather to provide a "worst-case" scenario by maintaining a reduced vaginal epithelial layer throughout the exposure period.5
In summary, we describe a stringent model combining high-dose DMPA with repeated viral exposures to evaluate the potential efficacy of a TDF IVR. The predictive value of this and other models requires testing of other PrEP options and validation with clinical studies. The findings further support the clinical advancement of this polyurethane reservoir TDF IVR, which is currently being evaluated in a phase 1 clinical study (ClinicalTrials.gov Identifier: NCT02006264) and suggest that protective levels of TDF may be observed in cycling women as well as women receiving DMPA hormonal contraception.

  iconpaperstack View Older Articles   Back to Top   www.natap.org