Abigail Norris Turner, BS¹, Kelly Blanchard, MSc², Jay S. Kaufman, PhD³, Charlotte Ellertson, MPA, PhD², and Barbara Friedland, MA⁴. (1) Department of Epidemiology, University of North Carolina at Chapel Hill, School of Public Health, CB #7435, McGavran-Greenberg Hall, Chapel Hill, NC 27599-7435, 919-929-3456, anorristurner@hotmail.com, (2) Ibis Reproductive Health, P.O. Box 1985, Parklands 2121, Johannesburg, 10017, South Africa, (3) Department of Epidemiology, UNC School of Public Health, 2104C McGavran-Greenberg Hall, Pittsboro Road, CB#7400, Chapel Hill, NC 275997400, (4) Population Council, One Dag Hammarskjold Plaza, New York, NY 10017

Background: Microbicides are being developed to reduce risk of sexual transmission of HIV. Products that decrease HIV risk, however, may simultaneously increase women's vulnerability to other infections, ultimately raising overall HIV susceptibility. Since bacterial vaginosis (BV) is an established risk factor for HIV, we assessed BV risk in a randomized, triple-blinded, placebo-controlled, Phase II safety trial of CarraguardŽ, a candidate vaginal microbicide. Methods: The study was conducted between October 1999 and January 2002 in two urban townships in South Africa. 400 women participated for up to 12 months, with monthly questionnaires and clinical exams. BV risk among CarraguardŽ users was compared with risk among users of the placebo product (methyl cellulose) using Generalized Estimating Equations (GEE) binomial regression modeling. BV was diagnosed using Nugent criteria. Results: Crude BV risk was similar among CarraguardŽ and placebo users (RR: 1.08, 95% CI: 0.86, 1.34). We evaluated several covariates: total number of sex acts, number of partners, condom use, vaginal product use (including desiccants and douching), presence of other vaginal infections, site, age, and parity. Using a change-in-estimate criterion of 10% between crude and completely adjusted estimates, no confounders were identified. We also assessed each covariate for effect modification and similarly found no evidence of modification. Conclusions: CarraguardŽ use did not increase BV risk, confirming previous safety conclusions. BV risk should be carefully addressed in all trials of candidate microbicides' safety and effectiveness, since first-generation microbicides may be only modestly effective against HIV, and increased BV risk could offset gains in HIV prevention.

Learning Objectives: At the conclusion of the session, the participant (learner) in this session will be able to

• Articulate the importance of appropriate consideration of bacterial vaginosis (BV) outcomes in trials of vaginal microbicides (due to interactions between current vaginal infections and increase risk of HIV acquisition).
• Recognize the myriad factors associated with increased risk of BV, and appreciate the benefits of evaluating those factors within the context of a randomized, controlled clinical trial.
• Weigh the benefits of a novel HIV-preventive product against the potential drawbacks associated with its use, particularly among vulnerable populations with few alternatives to protect themselves against sexually-transmitted infections.

Keywords: Reproductive Health, Women and HIV/AIDS

Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: The research involves a clinical trial of the candidate vaginal microbicide Carraguard, developed and tested by the Population Council (New York, NY).
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.