Safety and distribution of cellulose acetate 1,2-benzenedicarboxylate (CAP), a candidate anti-HIV microbicide in rhesus macaques

Objectives: To assess the safety and distribution of a cellulose acetate 1,2-benzenedicarboxylate (CAP) gel formulation in rhesus macaques as part of the development process for its use as a vaginally administered product in humans.

Design: The similarities between the reproductive physiology, anatomy and vaginal microflora of human and non-human primates makes non-human primates a relevant animal model to assess the safety and distribution of candidate anti-HIV microbicides.

Methods: CAP gel was instilled once or once daily for 4 days into the vaginal vault of rhesus macaques. Colposcopy and magnetic resonance imaging were performed to detect adverse effects and spread of CAP, respectively. Additionally, vaginal pH and composition of the vaginal micorflora in macaques before, during and after CAP instillations were determined, and vaginal biopsies obtained following repeated CAP exposures were examined to further document its safety.

Results: CAP is safe for repeated use and exhibits a favorable distribution profile, showing no evidence of penetration into cells that line the vaginal epithelium. Further, the presence of CAP has no adverse effect on vaginal pH or the composition of the vaginal microflora, and does not induce vaginal epithelial thinning or inflammation.

Conclusions: CAP gel shows minimal toxicity in vivo, supporting its use as a candidate vaginal microbicide in humans.

Introduction

Sexual transmission is the most common route of spread of HIV ^[1,2]. While condom use remains the most effective protection for women against sexual transmission ^[3], such protection requires partner consent that is unlikely to be successfully negotiated by women in the developing countries under the current cultural environments. Thus, there is an urgent need for novel methods to block HIV transmission that put the power of use in the hands of women ^[4-6]. Potent and affordable gels or creams that can be topically applied to the vagina before sex to protect women against HIV infection are therefore highly desirable ^[7,8].

Cellulose acetate 1,2-benzenedicarboxylate (CAP), an 'inactive' pharmaceutical excipient commonly used in the coating of enteric tablets and capsules, has a past history of extensive use, is inexpensive, and is already manufactured on industrial scale ^[9]. CAP exhibits inhibitory activity against viral and non-viral sexually transmitted disease (STD) pathogens, and importantly, also against HIV-1 in laboratory tests ^[10-12]. When administered in the form of a topical gel, CAP protected four of six rhesus macaques from challenge with SIV in rhesus macaques ^[13]. Nevertheless, besides being effective, a candidate microbicide should not damage or inflame the vagina or cervix with repeated use, conditions that could enhance HIV transmission ^[14,15]. Additionally, the microbicide should not alter the acidity or the normal microflora of the vagina, allowing unwanted bacteria to outgrow and increasing the risk of infection ^[16,17]. Lastly, the microbicide should have a desirable distribution profile in the vaginal vault following application. Towards the development of CAP as a topical microbicide for use in humans, we conducted a detail assessment of the safety and distribution of a gel formulation of CAP in rhesus macaques, using colposcopy, histology and microbiology findings as endpoints.

Methods

Animals, formulation and administration

Female rhesus monkeys (Macaca mulatta) housed at the Tulane National Primate Research Center (TNRPC) in compliance with the Guide for the Care and Use of Laboratory Animals were used. CAP gel formulation was prepared and supplied by Dow Pharmaceutical Sciences (Petaluma, California, USA). For all procedures, the monkeys were anesthetized with an intramuscular injection of Telazol (8 mg/kg) or ketamine hydrochloride (10 mg/kg), and buprenorphine (0.01 mg/kg) was administered for post-biopsy analgesia. CAP gel was instilled into the vaginal vault with the use of a 5-ml syringe and the animals were kept with the hind end elevated for an additional 30 min. Complete blood count (CBC), serum chemistry (CHEM), vaginal pH, vaginal culture and colposcopy were performed at the indicated times pre- and post-instillation.

Irritation study design

A short- and long-term irritation protocol was used. For the short-term protocol, a single dose of 5 ml CAP was instilled into the vaginal vault of four macaques. CBC, CHEM, wick and colposcopy were performed at pre-, 24 and 48 h post-instillation. For the long-term protocol, 5 ml CAP was instilled into each of eight macaques daily for 4 days. Pre-instillation CBC, CHEM, wick and colposcopy were performed, and were repeated on days 1-5, 8 and 14. Vaginal biopsies were obtained at the end of the four daily instillations.

Colposcopy and microbiologic characterization

Vaginal and ectocervical mucosal surfaces were evaluated by colposcopy using a 5-mm Hopkins rigid telescope passed into the vaginal vault through a vaginal speculum. Images were collected and stored for analyses of any gross epithelial pathology. Vaginal secretions for bacterial cultures were collected by inserting a speculum into the vagina and passing a swab into the vault. Growth of aerobic and anaerobic microorganisms were evaluated according to standard assays ^[18]. For pH measurements, standard pH paper was inserted through the speculum, allowed a 1-2 s contact with the vaginal secretions and was then read.

Magnetic resonance imaging: dose and regimen

All magnetic resonance imaging (MRI) examinations (3D SPGR T1 weighted gradient echo with fat saturation) were performed at 20 min, 24 and 48 h post 1, 2 or 5 ml CAP instillation using Telazol anesthesia. Sagittal and transverse images were used to assess spread of the gel throughout the vagina vault.

Results

CAP is safe for repeated use in vivo

Colposcopic examination of the vagina and cervix is a commonly used procedure to detect epithelial changes that may be associated with vaginal product use ^[19]. Although the significance and predictive value of colposcopic findings on the predisposition of a woman to HIV infection or inflammatory responses are not yet well established ^[20,21], it is nevertheless recommended as part of the development of any vaginal microbicide product. Colposcopic examinations were therefore performed to assess the safety of a gel formulation containing 13% micronized CAP, hydroxypropylmethyl cellulose, glycerin and other proprietary ingredients.

Five milliliters of CAP was used in a short (1 day) and a long-term (daily for 4 days) dosing regimen in order to measure potential untoward effects at the maximum volume that can be tested. Colposcopy findings are shown in Fig. 1a (short term) and 1b (long term) for representative animals. There was no evidence of erythema, edema, petechial hemorrhage, nodules or epithelial disruption with either regimen. A normal cornified epithelium was present at pre-instillation (Fig. 1a). At 24 and 48 h post-instillation, the epithelium was covered with a mixture of mucus and a small amount of CAP, with more mucus present at the later time point. Similar observations were made in animals receiving repeated dosing of CAP on days 1, 2 and 3 (Fig. 1b), with a normal cornified epithelium seen 4 days after the last instillation (day 8).

Fig. 1

Repeated use of CAP does not result in thinning of the epithelial lining, or alter the vaginal pH and composition of the vaginal microflora

To further establish the safety of CAP with repeated use, the thickness of the epithelial lining, vaginal pH and composition of the vaginal microflora were measured at baseline and during and after completion of the long-term dosing regimen. There was no significant change (P < 0.59453, ANOVA) in the pH associated with prolonged CAP administration, and importantly, no significant change in Lactobacillus colonization during and following the administration of CAP (data not shown). Histological evaluation of vaginal biopsies showed no evidence of pathology associated with repeated dosing. The extent of keratin and epithelial vacuolation appears normal, and the mild inflammation in these vaginal tissues is all within normal limits based on a comparison to histologic examination of biopsy and necropsy samples from several thousand female rhesus monkeys at TNPRC as well as published reports ^[22-24].

CAP has a favorable distribution profile

To assess the actual or optimal initial distribution and subsequent spread of the CAP formulation when placed in the vagina, MRI was used. 1, 2 or 5 ml CAP gel was instilled into the vaginal vault, and MRI performed at 20 min and 24 h post-application. Results are summarized in Fig. 2. Spread of 5 ml CAP gel instilled into the vaginal vault was clearly demonstrated within 20 min post-instillation, with no evidence of CAP migrating through the vaginal wall or into the endocervix. At 24 h post-instillation, CAP was not detectable by MRI (Fig. 2). Images of 1- and 2-ml volumes of CAP administration suggested similar distribution kinetics, but incomplete coverage of the vaginal vault. The latter could be attributed to presence of air bubbles, thickness of the MRI slice, and/or thickness of the CAP coating in some areas.

Fig. 2

Discussion

In this study, repeated application of 5 ml of a gel formulation containing 13% CAP in a micronized form was found to be safe. No irritation was observed by colposopy (Fig. 1). Application of this large volume of CAP did not adversely affect the vaginal ecosystem or alter the vaginal pH. Additionally, histological analyses of vaginal biopsies showed no evidence of increased inflammation. The effects of soluble CAP on cell viability and tissue integrity, epithelial barrier and induction of proinflammatory responses had been extensively analyzed using various tissue models ^[25,26]. Although these in vitro systems are limited in that they do not measure the full spectrum of inflammatory responses, such as the recruitment of immune cells, they nevertheless showed that CAP did not induce epithelial proinflammatory cytokines or transcription factors governing cytokine production during or after exposure. Further, CAP did not damage the membrane of epithelial cells or interfere with epithelial responses to pathogenic stimulation, but attenuated some tumor necrosis factor-α induced responses. Our findings in the macaque model reported here confirm and extend these in vitro safety assessments, demonstrating through colposcopic, histologic and microbiologic measurements that CAP is safe for use in vivo.

Given that the sites of HIV-1 infection within the female reproductive tract are largely unknown, a topic microbicide, in addition to being safe, should optimally cover the entire vaginal and ectocervical epithelium to efficiently block transmission of the virus. We demonstrated by MRI even and contiguous spread throughout the vagina as well as spread to the ectocervix within 20 min of instillation of 5 ml of the CAP gel. There was no evidence of systemic absorption of CAP gel across the vaginal mucosa with this high dose, consistent with the relatively high molecular weight of CAP ^[9] and its presence in a micronized form in the formulation used. When 1 or 2 ml of the CAP gel was used, however, there appeared to be incomplete coverage of the genital tract mucosa. Since the vagina of macaques is approximately one-quarter the size of that of humans ^[27], the findings with the 1 ml dose in rhesus monkeys would be more similar to what one might observe in women receiving a 3 ml dosing regimen.

There are obvious limitations to the use of non-human primate models for testing microbicides. The cost and shortage of primates limit the size of experimental groups and hence, may restrict statistically meaningful analysis. Further, the relevance of these models in which virus is deposited atraumatically onto the vaginal mucosa to the process of heterosexual HIV transmission in humans remains unclear. Nevertheless, non-human primates provide useful tools for the assessment of irritation potential or effects on normal flora of topical microbicides since the female genital tract of macaques has many characteristics in common with that of humans ^[27-29]. Studies in macaques have predictive values, as illustrated in several reports where data from non-human primates are consistent with phase I and II safety and acceptability trials in women ^[30-35], and serve to help in the selection of products to advance to clinical study. Our finding that CAP shows minimal toxicity in rhesus macaques, therefore, supports its testing as a candidate microbicide in humans.

Acknowledgements

This work was supported by a grant from the NIH/NICHD (P01 HD041761). We thank Nurjehan Jivani and Gordon J. Dow at Dow Pharmaceutical Sciences for the preparation and provision of CAP gel formulations used in this study, Elena Serbinova (Dow Pharmaceutical Sciences) for regulatory advice, and Wendy Chen for help with the graphics.