The effectiveness of Carraguard, a vaginal microbicide, in protecting women against high-risk human papillomavirus infection.

Background: A randomised, double-blind, placebo-controlled trial found the vaginal microbicide Carraguard unable to prevent HIV infection. A substudy assessed the association of genital high-risk human papillomavirus (HR-HPV) in women at study end with Carraguard use. Methods: Participants received Carraguard gel or placebo plus condoms, and were instructed to use gel plus condoms during each act of vaginal intercourse. HR-HPV detection on cervical samples from 1,723 women was by Digene Hybrid Capture 2 analysis. Poisson regression analysis assessed the prevalence of genital HR-HPV for individuals receiving Carraguard relative to individuals receiving placebo. Results: In the Carraguard arm (n=875) the end trial unadjusted HR-HPV prevalence was 23.5% (95% CI 20.8-26.3) and 23.0% (95% CI 20.2-25.8) in placebo arm (n=843). Significant risk factors for HR-HPV infection were younger age, being single, an abnormal pap smear, multiple sexual partners and promiscuous behaviour without the use of a condom. There were 348 compliant women (174 Carraguard, 174 placebo users), with relatively high adherence to gel use, who inserted 80% of their opened, returned applicators of test product with the proportion of applicator insertions to sex acts >30%. After adjusting for risk factors, these compliant Carraguard users were 0.62 as likely to be classified HR-HPV positive (95% Cl 0.41-0.94) as compliant placebo users. Conclusions: The prevalence of HR-HPV infection was lower in compliant Carraguard users than compliant placebo users. To our knowledge, this is the first report showing a negative association of HPV infection with a vaginal microbicide.