Protection Against Rectal Chimeric Simian/Human Immunodeficiency Virus Transmission in Macaques by Rectal-Specific Gel Formulations of Maraviroc and Tenofovir

Abstract

Background. Rectal human immunodeficiency virus (HIV) transmission is an important driver of the HIV epidemic. Optimally formulated gels of antiretroviral drugs are under development for preventing rectally acquired HIV. We investigated in a macaque model the pharmacokinetics and efficacy of 3 rectal gel formulations

Methods. Single-dose pharmacokinetics of low-osmolar 1% maraviroc (MVC), 1% tenofovir (TFV), or 1% MVC/1% TFV combination gel were evaluated in blood, rectal fluids, colorectal biopsy specimens, and rectal lymphocytes. Efficacy was evaluated over 10 twice-weekly rectal SHIV162p3 challenges in rhesus macaques that received either placebo (n = 7), MVC (n = 6), TFV (n = 6), or MVC/TFV (n = 6) gel 30 minutes before each challenge.

Results. MVC and TFV were detected in plasma 30 minutes after gel application and remained above 95% inhibitory concentrations in rectal fluids at 24 hours. MVC, TFV, and TFV diphosphate (TFV-DP) concentrations in colorectal tissues collected up to 30 cm from the anal margin were all high at 2 hours, demonstrating rapid and extended tissue dosing. TFV-DP concentrations in tissue homogenates and rectal lymphocytes were highly correlated (r² = 0.82). All 3 gel formulations were highly protective (82% efficacy; P ≤ .02 by the log-rank test).

Conclusions. Desirable pharmacokinetic profiles and high efficacy in this macaque model support the clinical development of these gel formulations for preventing rectal HIV infection.

Human immunodeficiency virus type 1 (HIV) continues to spread globally, with an estimated 2.1 million new infections in 2013 acquired primarily through heterosexual sex [1]. Women bear the greatest burden of infections, which led to early efforts to develop prevention products specifically for use by women. These products included microbicide gels containing antiretroviral drugs that women could apply vaginally, before and after sex, to prevent acquisition of HIV [2–4]. Tenofovir (TFV), a nucleotide reverse transcriptase inhibitor, has been formulated as a 1% aqueous vaginal gel and evaluated in clinical trials in women [2–5]. Studies found this gel formulation to be safe and effective in reducing HIV acquisition when used pericoitally, thus providing the first proof-of-concept of efficacy of this prevention modality [2].

In developed countries, among the drivers of the HIV epidemic are persons who engage in receptive anal intercourse, particularly men who have sex with men (MSM) [6]. In certain areas in the United States, the incidence rates among black MSM are similar to those in HIV-endemic sub-Saharan African regions and in MSM populations in Asia [1, 7, 8]. Because anal sex is also practiced among heterosexual couples worldwide, it is reasonable to assume that such practices may also contribute to HIV spread in many populations [9–12]. Therefore, there is increasing interest in extending the development of microbicides to include rectal formulations that can afford protection against rectal HIV acquisition [13–15]. Success of the vaginal 1% TFV gel formulation prompted its clinical evaluation for rectal safety [13]. However, many participants in this study reported gastrointestinal symptoms, including abdominal bloating, cramps, and defecation urgency [13]. Similar effects are produced by hyperosmolar rectal enemas, suggesting that these symptoms may have been linked to the high osmolality of the vaginal gel formulation (3111 mOsmol/kg) [16, 17]. Consequently, TFV was reformulated with a low-glycerin (5% w/w mg rather than 20%) gel that reduced the osmolality to 836 mOsmol/kg. This reduced-glycerin formulation was found to be well tolerated rectally, underscoring the importance of low osmolality for rectal gel formulations [15].

The recognition of functional and structural differences between the rectum and vagina has led to efforts in developing low-osmolar gels suitable for rectal use [14, 15]. Here, we evaluated in a macaque model 3 rectum-specific gel formulations containing 1% maraviroc (MVC) or 1% TFV alone or in combination. All gels were formulated with high drug concentrations and designed to have similar osmolality (approximately 816–940 mOsmol/kg), neutral pH, and favorable rheological profiles. The formulations included MVC and TFV, as both are Food and Drug Administration–approved oral drugs for the treatment of HIV-infected persons and lead candidates for oral and topical prophylaxis [2, 18–24].

Macaque models of mucosal chimeric simian/human immunodeficiency virus (SHIV) transmission can serve as invaluable preclinical tools for assessing the efficacy of rectal gels against rectal infection. Previously, vaginal MVC and TFV gels both protected macaques from vaginal acquisition of simian immunodeficiency virus (SIV) or SHIV, supporting evaluation of rectal protection [19–21, 25]. Here, we used a repeat-exposure macaque model to assess the pharmacokinetic (PK) profile and efficacy of all 3 gel formulations against repeated rectal challenges with an R5-tropic SHIV. We demonstrate in vivo evidence of rapid drug dosing and high efficacy against rectal SHIV transmission, supporting the clinical development of these rectum-specific gel formulations as a promising strategy for prevention of rectal HIV infection.

METHODS

Gel Formulations

Pure MVC and TFV were sourced from ViiV and Gilead Sciences, respectively. Aqueous low-osmolar gels containing 1% MVC, 1% TFV, and 1% MVC/1% TFV were prepared as described in the Supplementary Methods.

Virus Stock

SHIV_162P3 was generated as described in the Supplementary Methods.

Single-Dose PK Studies

A crossover study design was used to evaluate the PK of all 3 rectal gel formulations in 6 rhesus macaques, with a 1-week washout period between each gel product. Briefly, macaques received rectal washes (3–4) with sterile saline to remove feces, followed by a single 4-mL gel dose into the rectum (8 cm) via a sterile gastric feeding tube. At 0.5, 2, 6, and 24 hours after the dose, samples of blood plasma and rectal fluid (obtained by Weck-cel spear) were collected and analyzed for MVC and TFV concentrations. Tissue drug (MVC, TFV, TFV diphosphate [TFV-DP]) levels were investigated in another group of SHIV-infected macaques that were euthanized 2 hours (n = 3) and 24 hours (n = 3) after receiving a single 4-mL rectal dose of 1% MVC/1% TFV gel. Intact rectum and colonic tissue specimens (approximately 30 cm) excised at the time of necropsy were gently rinsed with saline to remove feces and dissected into 5-cm tissue sections collected 5, 10, 15, and 30 cm from the anal margin. Biopsy specimens were collected from each section using 3.7-mm biopsy forceps (Radial JawTM 3 Biopsy Forceps, Boston Scientific) and immediately homogenized in 1 mL of ice-cold 80% methanol. Rectal lymphocytes were isolated from tissue sections 5, 10, and 15 cm from the anal margin by established enzymatic extraction methods previously described [26].

Sample Analysis

MVC, TFV, and TFV-DP concentrations were measured using validated high-performance liquid chromatography–tandem mass spectrometry methods as previously described [26–28]. MVC and TFV concentrations in plasma and rectal secretions were expressed as nanograms of drug per milliliter of biological matrix. Limits of detection (LODs) for MVC and TFV were 1.5 ng/mL and 3 ng/mL, respectively. Area under the curve (AUC), maximum drug concentration (C_max), and time of C_max (T_max) calculations were done using GraphPad PRISM (version 5.04) software. MVC and TFV concentrations in tissues were expressed as nanograms of drug per milligram of tissue. For intracellular TFV-DP analysis, tissue homogenates and isolated cell lysates were analyzed using an indirect assay as described elsewhere [28]. TFV-DP levels were expressed as femtomoles per milligram of tissue and as femtomoles per million cells (lower limit of quantification, 2.5 fmol/sample).

Rectal Efficacy Studies

Efficacies of rectal gels were evaluated in Chinese-origin Rhesus macaques. Macaques were administered 4 mL of 1% MVC (n = 6), 1% TFV (n = 6), 1% MVC/1% TFV (n = 6), or placebo (n = 7) gel in the rectum 30 minutes before each rectal challenge with SHIV_SF162P3 inoculum (500 50% tissue culture infective doses). Both the gel and virus inoculum were delivered 8 cm into the rectum, using a gastric feeding tube of adjusted length. Gel dosing and rectal challenges were performed twice weekly for up to 5 weeks (10 challenges). Blood specimens were collected 30 minutes after gel dosing to monitor for drug levels and systemic SHIV infection. Virus challenges were stopped when a macaque tested positive for SHIV RNA in plasma obtained on 2 consecutive occasions. All SHIV-infected macaques continued to receive gel applications twice weekly for up to 10 weeks after infection to monitor for drug resistance. All experiments were done under highly controlled conditions by the same personnel, using the same virus stock, inoculum dose, and challenge procedures. These studies adhered to the Guide for the Care and Use of Laboratory Animals; all procedures were approved by the institutional animal care and use committees of the Centers for Disease Control and Prevention.

SHIV RNA Testing

SHIV RNA in plasma was quantified by reverse transcription–polymerase chain reaction (PCR; sensitivity, 50 RNA copies/mL) as previously described [26]. The time of infection was estimated as 1 week (2 challenges) before the first confirmed detection of SHIV RNA. The correction takes into account the lag between virus infection and detection of SHIV RNA in plasma.

Statistical Methods

Assuming constant risk per SHIV challenge, we estimated efficacy from the number of infections per total exposures in each group relative to controls, and log-rank tests were used to compare survival curves. Peak RNA viremias were compared using a 2-tailed Wilcoxon rank sum test. Differences in repeatedly measured drug absorption values (log₁₀ transformed) were assessed using linear mixed effects models, with adjustment for changes over time (ie, models included independent variables for time).

RESULTS

Plasma Pharmacokinetics

The mean concentrations of MVC and TFV in plasma and rectal fluids over 24 hours are shown in Figure 1A and 1B, respectively; the T_max, C_max, and AUCs are summarized in Table 1. MVC and TFV concentrations in plasma were highest at 2 and 0.5 hours, respectively, and gradually declined to undetectable levels by 5 and 24 hours, respectively (Figure 1A). Plasma TFV exposures after receipt of MVC/TFV gel were similar those for TFV gel (P = .41), indicating similar TFV release from either TFV gel and MVC/TFV gel. In contrast, plasma MVC levels were significantly less (1.7-fold; P = .007) following rectal dosing with MVC/TFV gel. Interestingly, despite the same concentration in gel, the TFV plasma AUC_{0–24 hours} was consistently higher than that of MVC for the single-drug (2.3-fold; P = .007) and combined-drug (5-fold; P < .0001) gels.

Figure 1.

Pharmacokinetic parameters of rectum-specific gels containing maraviroc (MVC) and tenofovir (TFV). MVC and TFV concentrations over 24 hours following rectal gel dosing. A, The time of maximum concentration (T_max) and subsequent decay of MVC and TFV concentrations in plasma after rectal dosing with single-drug and combined-drug gel are shown. B, MVC and TFV concentrations in rectal fluids after gel dosing.

Table 1.

Pharmacokinetic Parameters of Maraviroc (MVC) and Tenofovir (TFV) in Macaques Following Rectal Administration of 1% MVC Gel, 1% TFV Gel, and 1% MVC/1% TFV Gel

Variable	Plasma			Rectal Fluid
	Cmax, ng/mL, Mean	Tmax, h	AUC, ng h/mL, Mean	Cmax, ng/mL, Mean	Tmax, h	AUC, ng h/mL, Mean
TFV
Alone	42.7	0.5a	121.1	3.5 × 106	2.0a	29.3 × 106
Combined with MVC	40.2	0.5a	126.0	1.3 × 106	2.0a	8.6 × 106
MVC
Alone	7.3	2.0	54.2	0.7 × 106	2.0a	4.9 × 106
Combined with TFV	5.3	0.5a	15.8	0.69 × 106	2.0a	4.1 × 106

Abbreviations: AUC, area under the drug concentration-time curve; C_max, maximum drug concentration; T_max, time of C_max.

^a First time point at which specimens were collected.

Rectal Fluid PK

MVC and TFV levels in rectal fluids were highest at 2 hours and remained detectable 24 hours after receipt of the gel dose, although they were 2–3 logs lower than the highest concentrations (Figure 1B). Rectal exposures to MVC and TFV were similar following dosing with single-drug and combined-drug gels (P = .47 and P = .30, respectively). However, the AUC_{0–24 hours} of TFV in rectal fluids was higher than that of MVC following receipt of single-drug (3.7-fold; P = .001) and combined-drug (1.9-fold; P = .002) gel. As expected, C_max values for MVC and TFV in rectal fluids (0.7 × 10⁵ ng/mL and 3.5 × 10⁶ ng/mL, respectively) were much higher than in plasma and orders of magnitude greater (2–3 log₁₀) than those previously reported in macaques administered human-equivalent doses of oral MVC (44 mg/kg) and tenofovir disoproxil fumarate (TDF; 22 mg/kg), respectively [27, 29].

Rectal Tissue PK

Following a single rectal dose of MVC/TFV gel, the median MVC and TFV concentrations were high at 2 hours in biopsy specimens from tissue sections collected 5 cm (46 and 43 ng/mg), 10 cm (39 and 169 ng/mg), 15 cm (243 and 246 ng/mg), and 30 cm (152 and 160 ng/mg) from the anal margin, respectively (Table 2). Likewise, median TFV-DP concentrations in biopsy specimens collected at 5 cm (415 fmol/mg), 10 cm (429 fmol/mg), 15 cm (1697 fmol/mg), and 30 cm (868 fmol/mg) were also high at 2 hours, suggesting rapid intracellular delivery and phosphorylation of TFV to TFV-DP. Notably, median intracellular TFV-DP concentrations in lymphocytes purified from tissues collected at 5 cm (1389 fmol/10⁶ cells), 10 cm (3588 fmol/10⁶ cells), and 15 cm (3448 fmol/10⁶ cells) at 2 hours were high and exceeded the in vitro 95% inhibitory concentration (1373 fmol/10⁶ cells) [26]. When comparing drug concentrations at 24 hours, extracellular MVC and TFV exposures in all tissues sections (5, 10, 15, and 30 cm) were approximately 1 log₁₀ lower than at 2 hours. In contrast, TFV-DP levels remained high at 24 hours in biopsy specimens and lymphocytes collected at 5 cm (446 fmol/mg and 2391 fmol/10⁶ cells, respectively) and were consistently detected in distal tissues located 15 cm (33 fmol/mg and 529 fmol/10⁶ cells, respectively) and 30 cm (19 fmol/mg and 262 fmol/10⁶ cells, respectively) from the anal margin. Taken together, these findings demonstrate rapid dosing and high drug exposures in rectal tissues that extend beyond (by 20–30 cm) the site of gel application.

Table 2.

Drug Concentrations in Rectal Tissue Specimens Collected 2 and 24 Hours After Intrarectal Administration of 1% Maraviroc (MVC)/1% Tenofovir (TFV) Gel

Time, Drug	5 cm, Median (Range)	10 cm, Median (Range)	15 cm, Median (Range)	30 cm, Median (Range)
2 h
MVC, ng/mg	46.1 (14.9–193.7)	139.6 (81.8–203.3)	243.8 (16.7–291.3)	152.9
TFV, ng/mg	43.0 (6.9–172)	169.6 (8.2–181)	246.3 (2.6–527.6)	160.4
TFV-DP, fmol/mg	415 (68–788)	429 (61–798)	1697 (13–2083)	868
TFV-DP, fmol/106 cells	1389 (375–11 407)	3588 (497–30 262)	3448 (252–14 398)	ND
24 h
MVC, ng/mg	0.6 (0.4–3.9)	0.3 (0.1–0.5)	0.8 (0.3–2.5)	0.8 (0.3–0.9)
TFV, ng/mg	6.5 (0.6–9.7)	1.7 (0.6–1.8)	1.0 (1.0–4.1)	0.7 (0.01–0.9)
TFV-DP, fmol/mg	446 (31–534)	ND	33 (25.1–64.0)	19 (6–24.1)
TFV-DP, fmol/106 cells	2391 (515–6770)	529 (53–1233)	262 (58–576)	ND

Abbreviations: DP, diphosphate; ND, not determined.

Covariate Analysis of TFV-DP in Rectal Tissue Specimens and Purified Lymphocytes

A total of 14 paired rectal biopsy specimens and purified lymphocytes were available from tissue sections collected from the 6 SHIV-infected macaques euthanized 2 and 24 hours after MVC/TFV gel dosing. Linear regression analysis demonstrated a strong, positive correlation (r² = 0.82; P < .001) between log₁₀ TFV-DP concentrations in tissue homogenates and cell lysates. Measurements of TFV-DP in biopsy specimens collected at 2 and 24 hours were highly predictive of intracellular TFV-DP levels in purified lymphocytes (8.6-fold higher TFV-DP in lymphocytes; Figure 2).

Figure 2.

Pharmacokinetic correlation of tenofovir diphosphate (TFV-DP) concentration in rectal tissues and lymphocytes. The relationship between TFV-DP concentrations in tissue homogenates (log fmol/mg tissue) and isolated lymphocytes (log fmol/10⁶ cells) is shown for paired samples (n = 14) collected at necropsy from proximal, medial, and distal tissues 2 and 24 hours after administration of a 1% maraviroc/1% TFV gel dose.

Rectal Gel Containing MVC or TFV Alone or in Combination Protects Against Rectal SHIV Infection

The cumulative proportion of macaques that remained uninfected over 10 rectal SHIV exposures are shown in Figure 3. All controls were infected after 10 SHIV challenges (median required for infection, 6 challenges). In contrast, 4 of 6 macaques in each treatment group were protected after 10 exposures and remained negative for SHIV by PCR and seronegative throughout the 10-week washout period. Infections in the MVC, TFV, and MVC/TFV groups occurred at challenge (2 and 9), (7 and 9), and (3 and 10), respectively. Differences in risk for infection between placebo and treatment arms were statistically significant (P = .02 for MVC gel, P = .003 for TFV gel, and P = .01 for MVC/TFV gel; all P values were calculated by the log-rank test), with estimated efficacies of 81.5% for MVC gel, 83.2% for TFV gel, and 82.0% for MVC/TFV gel.

Figure 3.

Rectal gels containing maraviroc (MVC) and tenofovir (TFV) provide high protection against rectal chimeric simian/human immunodeficiency virus 162P3 (SHIV_162P3) infection. Survival curves of the cumulative percentage of uninfected macaques as a function of the number of weeks to SHIV positivity. Small black arrows indicate times of SHIV exposure (2 per week). Animals received gel twice weekly 1 week before, during, and 1 week after the challenge period (shaded area). After 5 weeks (10 SHIV_162p3 exposures), challenges were stopped, and animals were monitored for 10 additional weeks in the absence of virus challenge. Animals that became infected continued to receive gel up twice-weekly to 10 weeks after infection.

Plasma Drug Concentrations Following Gel Dosing

We analyzed systemic drug exposures longitudinally in protected and infected animals in plasma specimens collected 30 minutes after gel dosing during the 5-week challenge period. Figure 4A shows plasma MVC concentrations in animals receiving MVC alone were similar in the 4 protected animals (mean, 4.5 ng/mL; range, 0–19 ng/mL) and the 2 with breakthrough infections (mean, 3.3 ng/mL; range, 0–11 ng/mL) during the challenge period (P = .63). Likewise, TFV levels in protected animals (mean, 11.9 ng/mL; range, 0–44 ng/mL) and those with breakthrough infections (mean, 9.3 ng/mL; range, 0–34 ng/mL) were comparable (P = .82) in animals treated with TFV alone (Figure 4B). While MVC exposures in infected animals treated with MVC/TFV gel were comparable to those that were protected (P = .63), TFV levels were significantly lower (by approximately 0.50 log₁₀) among animals that became infected (P = .03). However, the mean TFV level in these animals (11.0 ng/mL) was similar to that detected in animals that were protected by TFV gel. We additionally assessed plasma drug exposures in all 6 breakthrough infections around the estimated time of infection (7 days before SHIV RNA detection) and found no differences in plasma drug exposures during the infection window, relative to earlier challenges (data not shown). Taken together, these findings suggest that failures may not be related to inadequate drug exposure in plasma or to inability of animals to absorb drug.

Figure 4.

Drug absorption was similar in protected and breakthrough infections. Assessment of plasma drug levels in macaques at 30 minutes after gel dosing for each chimeric simian/human immunodeficiency virus exposure during the 5-week challenge period. Protected and infected macaques were treated with 1% maraviroc (MVC; A), 1% tenofovir (TFV) gel (B), and 1% MVC/1% TFV (C) gel. Solid black line indicates mean drug levels.

Virus Loads and Drug Resistance Testing in Breakthrough Infections

All 6 infected macaques continued to receive rectal dosing of gel twice weekly for an additional 6–10 weeks after infection. Figure 5 shows that the virus load kinetics in breakthrough infections were similar across all treatment groups and controls.

Figure 5.

Viral replication in breakthrough infections is similar to that in controls and shows no evidence of K65R emergence. Individual virus load kinetics of chimeric simian/human immunodeficiency virus (SHIV)–infected controls (black dotted) and mean viral load kinetics of the 2 breakthrough infections under continued gel dosing twice weekly with 1% maraviroc (MVC; red), 1% tenofovir (TFV; blue), and 1% MVC/1% TFV (green) are shown (standard error bars shown in same color). Time 0 indicates SHIV RNA levels at first detection. Wild-type SHIV detected at time 0, at week 2, and among all samples tested during follow-up (up to 10 weeks after infection).

We additionally tested for the TFV-associated K65R mutation in animals exposed to TFV alone (4235 and RQ4979) or TFV/MVC gel (41 803 and KCV). Sensitive testing by allele-specific PCR for the K65R mutation (LOD, 0.4%) revealed wild-type SHIV at the time of infection, at peak viremia, and during the following 6–10 weeks after infection, indicating the absence of low-frequency K65R mutants in plasma.

DISCUSSION

In this study, we evaluated, in a rhesus macaque model, 3 rectum-specific gel formulations consisting of MVC and TFV alone or in combination. We demonstrated that all 3 gels were highly effective in preventing rectal infection and related efficacy with favorable PK properties, including rapid drug absorption and extensive coverage of colorectal tissues far beyond the site of dosing. Our study provides proof of concept of efficacy and supports the clinical development of these rectum-specific gel formulations as a promising strategy to prevent rectal HIV infection. The protection against rectal transmission by these gels mirrors that of a fixed-dose combination of oral TDF plus emtricitabine (FTC) in the same model, which predicted the clinical efficacy of oral TDF/FTC in MSM [30, 31].

Rectal gels were formulated as aqueous viscous hydrogels with neutral pH and similar osmolality (816–940 mOsmol/kg). The detection of MVC or TFV in plasma 30 minutes after gel application confirms rapid drug release from all 3 gels. The rapid movement through tissue ensures a short pharmacologic lag. This is particularly important for MVC's ability to block virus entry, the earliest step in virus infection. The approximately 3–5-fold lower levels of MVC in plasma relative to TFV despite similar drug concentrations in the gel is consistent with previous findings following vaginal MVC gel dosing, which inferred that the ionization state of MVC at pH 7.0 may affect its tissue permeability [19]. Alternatively, the release rate of MVC may be slightly lower than that of TFV, which may explain the lower MVC levels in both the rectal fluid and plasma. Despite the disparity, MVC is approximately 1000 times more potent than TFV in vitro and should, in principle, confer more in vivo activity even at lower concentrations [26, 27, 32, 33]. The finding of high TFV-DP concentrations in colorectal tissue biopsy specimens and purified lymphocytes 2 hours after application demonstrates that TFV is rapidly delivered and phosphorylated in tissues and HIV target cells, respectively. Importantly, we noted high drug exposures in colorectal tissues 20–30 cm beyond the point of gel application, which is desirable, given the requirement of rectal gels to protect a large surface area. We also found that these TFV-DP concentrations were within the range of those seen in humans who received a rectal dose of 1% TFV gel, highlighting the clinical relevance of the model [34]. We further documented a strong correlation between log₁₀ TFV-DP concentrations in tissues and purified lymphocytes. The identification of this relationship supports the usefulness of obtaining tissue biopsy specimens for analyzing drug concentrations and measuring correlates of protection and, thus, circumventing the need to collect large tissue samples for isolating purified lymphocytes.

Similar to the established macaque models found to predict clinical trial results of antiretroviral prophylaxis products, our model included rectal challenges twice weekly (up to 10) to mimic high-risk human exposures [20, 30, 35]. In this model, all controls become infected with SHIV after a median of 4 exposures, which allows measurement of protection over multiple transmission events per animal and, thus, provides a consequential increase in statistical power relative to a single high-dose model [36]. Here, we showed that all 3 gels were highly protective and exhibited a protective efficacy of approximately 82%. The protection afforded by 1% MVC gel is noteworthy because in this model oral MVC administered at a clinically relevant dose failed to protect macaques from rectal SHIV infection [27]. More recent data from humans who received oral MVC also did not demonstrate protection in ex vivo HIV infection of rectal explants [37]. The observed efficacy of MVC gel is likely related to the >1000-fold higher concentration of MVC found in rectal tissues following gel application, compared with concentrations after oral dosing [27]. Furthermore, the high mucosal MVC concentrations and protection are consistent with the vaginal efficacy of MVC gel previously reported in macaques [19, 21]. The findings also suggest that the MVC concentrations in rectal tissues were sufficient to overcome the higher frequency of activated lymphocytes and higher CCR5 expression on target cells in the rectal mucosa [38, 39]. Notably, the protection afforded by 1% TFV gel was similar to that of 1% MVC gel despite the far lower (approximately 1000-fold) potency of TFV. However, these findings are in agreement with previous data showing approximately 62% protection by a hyperosmolar vaginal TFV formulation in a single high-dose rectal SIV challenge model, which may have underestimated the efficacy, compared with the repeat low-dose challenge model [18].

The finding that MVC/TFV gel was unable to boost efficacy is intriguing because drug combinations often confer higher in vivo activity in both treatment and oral prophylaxis [30, 40]. It is possible that the single-drug gels reflect maximal mucosal protection attainable for the rectal compartment and that combining MVC and TFV provided no additional activity in blocking pathways responsible for the breakthrough infections. While it is clear that breakthrough infections were established and expanded in cells not fully protected by drug, the location and source of these cells, and the mechanism by which they were exposed to SHIV, remain unknown. Despite the long intracellular half-life and high TFV-DP levels in colorectal tissues and lymphocytes shortly after dosing, the sharp decay of TFV-DP by 24 hours was much faster than predicted, which may have shortened the durability of antiviral activity and possibly contributed to breakthrough infections.

Although our findings in this study support the clinical development of all 3 gel formulations, factors other than efficacy, such as cost, ease of formulation, and activity against circulating drug-resistant viruses, should be considered. For example, MVC/TFV gel may be more costly to formulate but might have the advantage of providing higher activity against drug-resistant viruses. While biological efficacy demonstrated in this model is promising, it may not necessarily translate into efficacy in humans if compliance and/or desirability of the product are poor. Therefore, assessment of desirability of rectal gels in end users is critical. There are also limitations to our study, as the macaque model cannot fully reproduce conditions seen in humans. These include a high virus inoculum and the absence of semen or coital activity, all of which could influence drug PK and efficacy.

Drug resistance emergence is a concern for antiretroviral-based topical microbicides, particularly with ARVs commonly used for treatment. We found no evidence of TFV resistance in the 4 macaques who did not respond to TFV or TFV/MVC gels despite continued dosing for several weeks after infection. In addition, plasma virus loads during acute viremia in animals with breakthrough infections were similar to those in untreated controls, which differs from findings associated with oral TFV dosing, which reduces SHIV viremia and selects for TFV resistance [30]. These findings are consistent with previous studies and likely reflect the low systemic antiviral activity from gel dosing [26]. While we have not tested for shifts in MVC susceptibility, previous data from breakthrough SHIV infections of oral or vaginal MVC found no evidence of selecting for MVC-resistant variants, suggesting a low likelihood of MVC resistance in this study [27, 41].

In conclusion, the macaque model described here provided both PK and efficacy data that support the development of rectum-specific gel formulations as a prevention strategy against rectal acquisition of HIV.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.