Clinical Science: Concise Communications
HIV genital shedding and safety of Carraguard use by HIV-infected women: a crossover trial in Thailand
McLean, Catherine Aa,*; van de Wijgert, Janneke HHMb,c,*; Jones, Heidi Eb,d; Karon, John Me; McNicoll, Janet Ma,f; Whitehead, Sara Ja,f; Braunstein, Sarahb,d; Achalapong, Jullapongg; Chaikummao, Supapornf; Tappero, Jordan Wa,f; Markowitz, Lauri Ea; Kilmarx, Peter Ha
aNational Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia, USA
bPopulation Council, New York, New York, USA
cAcademic Medical Center, Amsterdam, The Netherlands
dColumbia University Medical Center, New York, New York, USA
eEmergint Corp, Louisville, Kentucky, USA
fThailand Ministry of Public Health – U.S. CDC Collaboration, Bangkok
gChiang Rai Hospital, Chiang Rai, Thailand.
*C.A.M. and J.H.H.M.vdW. contributed equally to the writing of this article.
Received 6 February, 2009
Revised 30 September, 2009
Accepted 1 October, 2009
Correspondence to Peter H. Kilmarx, MD, Centers for Disease Control and Prevention, 1600 Clifton Road, MS E-45, Atlanta, GA 30030, USA. Tel: +1 404 324 7471; e-mail: email@example.com
Objective: To evaluate the safety, including impact on genital HIV RNA shedding, of Carraguard vaginal gel in HIV-infected women.
Design: This is a randomized, controlled, crossover study of Carraguard in HIV-infected women in Thailand.
Methods: Each woman (CD4+ cell count 51–500 cells/μl and not on antiretroviral therapy) used each treatment (Carraguard, methylcellulose placebo, and no-product) once daily for 7 days during each 1-month period (3-week wash-out). Women were randomized to one of the six possible treatment sequences. Safety assessments were conducted at baseline (pregel), 15 min postgel, day 7, and day 14, and included HIV RNA measurements in cervicovaginal lavage (CVL) specimens.
Results: Sixty women were enrolled, and 99% of scheduled study visits were completed. At baseline, median age (34 years), CD4+ lymphocyte count (296 cells/μl), plasma HIV viral load (4.6 log10 copies/ml), CVL HIV viral load (3.1 log10 total copies per CVL), and sexual behaviors were similar among randomization groups. HIV viral load, leukocyte and hemoglobin levels, and epithelial cell counts in CVLs were lower 15 min after application of Carraguard or placebo compared with no product; CVL HIV viral load was still lower at day 7 but returned to baseline by day 14. Carraguard use was not associated with prevalent or incident genital findings or abnormal vaginal flora.
Conclusion: Carraguard appears to be well tolerated for once-daily vaginal use by HIV-infected women. The observed reduction in CVL HIV viral load in the gel months may be clinically relevant but could have resulted from interference with sample collection by study gels.
Microbicides are being developed for vaginal or rectal application to prevent infection with HIV and other sexually transmitted infections (STIs) . Carraguard gel (Population Council, New York, New York, USA) consists of a mixture of lambda and kappa-Carrageenan. Previous studies [2–7] indicated that Carraguard was well tolerated and potentially effective against HIV and other STIs. Although Carraguard was not shown to be effective in a recently completed phase III trial , next-generation microbicides are under development combining Carraguard with antiretroviral agents .
Microbicides may be used by HIV-infected women because they are unaware of their infection status, to protect themselves from STIs or HIV superinfection, or to protect their partners from HIV. However, HIV-positive women may be more susceptible to adverse effects of gel use and may become more infectious if gel use increases genital HIV RNA shedding . We, therefore, assessed the safety of Carraguard in HIV-positive women, including the effect of Carraguard on HIV RNA shedding.
In this randomized, controlled, three-treatment, crossover trial, each woman participated in each treatment arm (Carraguard, methylcellulose placebo, and no product). The order of treatment months was randomized to six possible treatment sequences. In each of the 3 consecutive months, women attended the clinic on day 0, day 7, day 14, and a final closing visit on day 28 of the third month.
The sample size was determined by 3 × 3 Williams design, which is a crossover design balanced for first-order carryover effects . Sixty women would provide 80% statistical power to detect a difference between a pair of treatment means of 0.43 log10 HIV RNA copies in cervicovaginal lavage (CVL) specimens or higher at the 0.025 level using a random effects linear model, if all CVL HIV results are quantifiable. Eligibility criteria included age 18–50 years, HIV infection with CD4+ cell count 51–500 cells/μl, currently not on antiretroviral therapy (ART), generally healthy, not pregnant or planning pregnancy, having a regular or no menstrual cycle, not using vaginal products (including tampons), and ability to pass an informed consent comprehension test. Women had to be sexually abstinent or have only one male partner (age ≥18 years) confirmed as HIV positive and willing to be monogamous with the participant during the study. Women were excluded if they had had a delivery, miscarriage, or abortion within the previous 6 weeks, genital surgery or nonmenstrual vaginal bleeding during sex within the previous month, abnormal cytology, STI, symptomatic bacterial vaginosis and/or yeast infection, or genital abnormality with epithelial disruption at baseline, a history of sensitivity or allergy to latex, or had participated in another clinical trial within the last month. Screening included an assessment by an HIV care specialist. All eligible women accessed ART during or shortly after the study.
Carraguard and the methylcellulose placebo are clear and odorless gels packaged in a single-use Microlax-type applicator (Norden Pac, Kalmar, Sweden), shown in previous studies to dispense about 5 ml of gel . Participants and study staff were blinded to the gel used during gel-use months. Women were instructed to insert one gel applicator vaginally every evening for 7 consecutive days, beginning 3–5 days after completion of menses (day 0). Women were asked to abstain from vaginal intercourse for 24 h prior to scheduled visits. They were instructed to use gel and condoms during sex; nonspermicidal lubricated condoms were provided. Participants were considered adherent if they used gel at least six times in 9 days.
Pelvic examinations were conducted at screening, day 0, day 7, and day 14, with two examinations on day 0: one prior to gel insertion (to assess baseline and possible carryover effects in months 2 and 3) and one 15 min after gel insertion (acute effects) or after the first examination (no-product month). During each examination, swab and lavage specimens were collected, and colposcopy was done according to the WHO manual [13,14]. Research nurses used a bulb syringe to lavage the cervix and vaginal walls with 5 ml PBS, which was aspirated after 1 min.
HIV plasma viral load (Amplicor HIV Monitor version 1.5; Roche Molecular Systems, Pleasanton, California, USA) was measured at each day 0 visit. CVL specimens were tested for HIV RNA using Amplicor HIV Monitor version 1.5, with a modified extraction step using NucliSense lysis buffer and silica-based extraction (BioMerieux, Durham, North Carolina, USA). This modification was validated on simulated HIV-positive CVL samples containing various dilutions of Carraguard or placebo gel prior to study initiation . CVL specimens were also tested for hemoglobin and leukocytes using Multistix-8 SG Urine Strip (Bayer, Pittsburgh, Pennsylvania, USA). The number of epithelial cells in the CVLs was assessed by microscopy (using a hemocytometer). Diagnostic testing was conducted as described previously [5,7]. Women with positive semen tests (Abacus Diagnostics, West Hills, California, USA) were asked to return 48 h later to provide a new CVL specimen. CVL HIV RNA levels were adjusted for HIV RNA from blood (using hemoglobin levels in CVLs), but this did not significantly alter our conclusions: unadjusted data are reported.
Tests of variation among treatments and cycles used the nonparametric Friedman chi-squared test, with each woman regarded as a ‘block.’ The Wald test based on generalized estimating equation (GEE) model results (adjusted for period and first-order carryover effects) was used to evaluate variation of prevalent and incident genital findings among treatment groups, allowing for repeated measures.
Study participants and their male partners provided written, informed consent. The protocol was approved by ethical review committees of the Population Council, Centers of Disease Control and Prevention (CDC), and Thai Ministry of Public Health.
From November 2002 to February 2004, 783 women attended recruitment sessions, 222 were screened, 64 were eligible, and 60 were randomized. Use of any contraception and palpable inguinal lymph nodes were the only covariates (out of 62) that differed significantly (P < 0.05) among the six randomization groups. Median age was 34 years. Median CD4+ cell count at baseline was 296 cells/μl, and median plasma HIV viral load was 4.6 log10 copies/ml. In the month prior to enrollment, 25% of women had had vaginal sex.
Only one participant was lost to follow-up, and 99% of visits were made on schedule. Six women started ART during the study, similarly distributed among groups. Self-reported adherence with gel use was 98%: one woman stopped gel use because her partner returned and he tested HIV-negative.
No significant variation was observed in median plasma viral load among treatments or over time. HIV RNA was detectable in the majority of CVLs at all time points except day 0, 15 min (Table 1). The proportion with quantifiable load decreased at the start of each new treatment month over time (Table 1). The median CVL viral loads were similar among treatment months at day 0 pregel use. Fifteen minutes after gel use, the median acute CVL viral loads when using Carraguard and placebo were lower (P < 0.01) than when using no product. CVL viral loads were still significantly lower when using Carraguard and placebo than when using no product at the day 7 visits, but the variation among treatments was not significant on day 14. CVL viral load was positively correlated with plasma viral load (Spearman's ρ = 0.79) and negatively correlated with CD4+ cell count (Spearman's ρ = −0.23) at baseline. CVLs from each visit were evaluated for leukocytes, hemoglobin, and epithelial cells as markers for genital inflammation and epithelial sloughing. For each analyte, there was a decrease when using placebo or Carraguard compared with using no product immediately following gel use on day 0 and day 7 (Table 2).
Genital findings were common, with 73% of women having at least one prevalent finding and 43% at least one incident finding during follow-up. Findings predominantly had intact epithelium and were located on the cervix and on the external genitalia. Findings were similarly distributed among treatments with the following exceptions: women were significantly more likely to have a finding with intact epithelium at 15 min when using no product, and any cervical finding at any visit, than when using gel. Thirty-one women had an incident cervical finding at 37 visits, distributed as five, 10, and 22 incident findings in the Carraguard, placebo, and no-product months, respectively. Of these 37 findings, 19 were petechiae, purpura, or ecchymosis. The other finding types were each identified in only two to five women.
At baseline, 85% of women were tested herpes simplex virus (HSV)-2 seropositive, 33% had vaginal yeast by Gram stain, and 7% had bacterial vaginosis by Nugent criteria. No nonviral STIs were diagnosed. Bacterial vaginosis prevalence was 20, 18, and 11% during the Carraguard, placebo, and no-product months, respectively (P = 0.21 and P = 0.60 for Carraguard and placebo versus no product, respectively). Vaginal yeast prevalence was 47, 43, and 33% in the Carraguard, placebo, and no-product months, respectively (P = 0.43 and P = 0.76 for Carraguard and placebo versus no product, respectively).
The most commonly reported symptoms were genital itching, burning and pain (reported by 58% at least once during follow-up and by 32% at baseline), abdominal pain (43 and 12%), genital rash (22 and 15%), abnormal vaginal discharge (17 and 5%), and genital ulcer (17 and 3%). Symptoms were similarly distributed among treatments except for abnormal vaginal discharge and genital ulcers, which each occurred in 8% of women's months while using Carraguard, 2% while using placebo, and 12% while using no product (P = 0.03 for the three-way comparison).
In this crossover study of 60 HIV-positive women, Carraguard use was not associated with increased genital findings or with abnormal vaginal flora. The study showed lower HIV viral load, leukocyte and hemoglobin levels, and epithelial cell counts in CVLs 15 min after Carraguard and placebo application compared with no-product use. CVL HIV viral load reductions persisted at day 7 but disappeared by day 14. Although these reductions may be clinically significant, they may also be due to gels covering the cervicovaginal mucosa during CVL collection, thereby reducing the amount of viral RNA captured in the CVL specimen. Both gels had similar effects on CVL viral loads despite having different anti-HIV effects in preclinical experiments [2–4]. Additionally, all CVL analytes showed the same reduction pattern, with the greatest reduction soon after gel application. In-vitro testing prior to study initiation showed that the HIV RNA extraction and testing procedure used in this study was not inhibited by the presence of gels . Carraguard does not appear to increase genital shedding of HIV RNA, which is consistent with Carraguard safety data from Durban, South Africa . Additional studies are required to determine the cause for reductions in HIV viral load and whether these reductions can be correlated with reduced HIV infectiousness. With frequent follow-up and questioning, genital symptoms were commonly reported in all treatment arms, including the no-product arm, underscoring the importance of careful assessment of microbicide acceptability.
The advantage of the crossover design is that each woman serves as her own control, which removes variability among participants who receive different treatments, and allows for smaller sample sizes . Our study had a limited sample size, and results should, therefore, be interpreted as exploratory. Participants were carefully selected in a single setting and may not be representative of HIV-positive women in general. Several key outcomes (such as adherence and sexual activity) were assessed by self-report, which is prone to social desirability bias. Finally, the potential for bias in ascertainment of clinical outcomes cannot be excluded, particularly for the no-product month during which treatment assignment could not be blinded.
Carraguard gel appears to be well tolerated for once-daily vaginal use by HIV-infected women. The observed reduction in CVL viral load in the gel months may not be clinically relevant but may have resulted from study gel interference with specimen collection; this needs further investigation.
The authors would like to thank the study participants for their contribution of time, energy, and commitment to this effort. The authors would also like to acknowledge colleagues who contributed to study implementation: the clinical study team at the Chiang Rai Health Club, Tepnaruemit Medtanavyn, Barbara Friedland, Robin Maguire, Clyde Hart, Tammy Evans-Strickfaden, Liesbeth Bollen, Philip Mock, Chalinthorn Sinthuwattanawibool, Nancy Young, Wanna Leelawiwat, Punneporn Wasinrapee, Beth Bell, Elizabeth Unger, and Catherine Ley. The authors thank Maya Sternberg for assistance with study design and statistical support. The financial support for this study was provided by the US CDC and the Bill and Melinda Gates Foundation.
C.M., J.H.H.MvdW., H.E.J., S.J.W., S.B., J.W.T., L.E.M., and P.H.K. conceived and designed the study and were involved in study implementation and analysis. J.M.K. was the study statistician. J.M.M. was responsible for laboratory testing. S.J.W., J.A., and S.C. were the on-site investigators. C.M. and J.H.H.MvdW. wrote the manuscript. All authors contributed to the interpretation of the data and reviewed and approved the manuscript.
None of the authors have a commercial association with Carraguard or other conflicts of interest. Carraguard is being developed by the Population Council, which is a not-for-profit nongovernmental organization, under a US Food and Drug Administration Investigational New Drug Application (IND 58441).
This trial was registered in the US National Institutes of Health clinical trials registry under registration number NCT00213044.
The findings and conclusions in this manuscript are those of the authors and do not necessarily represent the views of the US CDC or of the other collaborating institutions.
1. Harrisons P, Rosenberg S, Bowcut J. Topical microbicides for disease prevention: status and challenges. Clin Infect Dis 2003; 36:1290–1294.
2. Maguire RA, Bergman N, Phillips DM. Comparison of microbicides for efficacy in protecting mice against vaginal challenge with herpes simplex virus type 2, cytotoxicity, antibacterial properties, and sperm immobilization. Sex Transm Dis 2001; 28:259–265.
3. Pearce-Pratt R, Phillips DM. Sulphated polysaccharides inhibit lymphocyte-to-epithelial transmission of human immunodeficiency virus-1. Biol Reprod 1996; 54:173–182.
4. Zacharapoulos V, Phillips D. Vaginal formulations of carrageenan protect mice from herpes simplex virus infection. Clin Diag Lab Immunol 1997; 4:465–468.
5. Kilmarx PH, van de Wijgert J, Chaikummao S, Jones H, Limpakarnjanarat K, Friedland B, et al. Safety and acceptability of the candidate microbicide Carraguard in Thai women: findings from a phase II clinical trial. J Acquir Immune Defic Syndr 2006; 43:327–334.
6. van de Wijgert J, Braunstein S, Morar N, Jones H, Madurai L, Evans Strickfaden E, et al. Carraguard vaginal gel safety and effect on HIV RNA shedding among HIV-positive women and men in South Africa. J Acquir Immune Defic Syndr 2007; 46:538–546.
7. Kilmarx PH, Blanchard K, Chaikummao S, Friedland BA, Srivirojana N, Connolly C, et al. A randomized, placebo-controlled trial to assess the safety and acceptability of use of Carraguard vaginal gel by heterosexual couples in Thailand. Sex Transm Dis 2008; 35:226–232.
8. Skoler-Karpoff S, Ramjee G, Ahmed K, Altini L, Plagianos MG, Friedland B, et al. Efficacy of Carraguard for prevention of HIV infection in women in South Africa: a randomised, double-blind, placebo-controlled trial. Lancet 2008; 372:1977–1987.
9. Fernández-Romero JA, Thorn M, Turville SG, Titchen K, Sudol K, Li J, et al. Carrageenan/MIV-150 (PC-815), a combination microbicide. Sex Transm Dis 2007; 34:9–14.
10. Mauck C, Rosenberg Z, Van Damme L, International Working Group on Microbicides. Recommendations for the clinical development of topical microbicides: an update. AIDS 2001; 15:857–868.
11. Senn S. Crossover trials in clinical research. 2nd ed. West Sussex, England: John Wiley & Sons; 2002.
12. Coetzee N, Blanchard K, Ellertson C, Hoosen A, Friedland B. Acceptability and feasibility of Micralax applicators and of methylcellulose placebo for large-scale clinical trials of vaginal microbicides. AIDS 2001; 15:1837–1842.
13. CONRAD and World Health Organization. Manual for the standardization of colposcopy for the evaluation of vaginal products. Update 2004. Geneva, Switzerland: WHO; 2004.
14. Bollen L, Kilmarx PH. Photo atlas for microbicide evaluation. Bangkok, Thailand: Thailand MOPH – U.S. CDC Collaboration; 2002.
15. Leelawiwat W, McLean CM, Tasaneeyapan T, Sinthwattanawibool C, Evans-Strickfaden T, Kilmarx PH, et al. In vitro effect of Carraguard, a candidate microbicide, on standard HIV RNA Roche Amplicor PCR testing and on as modified assay (NasMo) in simulated HIV cervicovagina lavage samples. In: 13th IUSTI Asia Pacific Conference; 6–9 July 2004; Chiang Mai, Thailand.
clinical trials; HIV; prevention of sexual transmission; Thailand; vaginal microbicides; virus shedding
© 2010 Lippincott Williams & Wilkins, Inc.
Highlight selected keywords in the article text.