Effectiveness of COL-1492, a Nonoxynol-9 Vaginal Gel, on HIV-1 Transmission in Female Sex Workers: A Randomised Controlled Trial

Although the male condom, when used consistently and correctly, provides high levels of protection against HIV-1 and other STDs, negotiating its use is not always feasible for many women. Therefore, there is a need for a female-controlled method for prevention of HIV-1. Research on microbicides is part of this global effort.

The product that has been tested most, nonoxynol-9, is an over-the-counter spermicide that has in-vitro anti-HIV-1 activity. Results of studies of its effectiveness have been inconclusive. The current study aimed to assess effectiveness of this vaginal gel.

Researchers did a randomized, placebo-controlled, triple- blinded, phase 2/3 trial with COL-1492, a vaginal gel containing 52.5 mg nonoxynol-9, in 892 female sex workers in four countries: Benin, Côte d'Ivoire, South Africa and Thailand. The gel covers the cervix and vaginal walls via an applicator and gives immediate availability of the drug. By random allocation, 449 women received nonoxynol-9 gel and 443 women received a placebo gel. Analysis was by intention to treat.

The primary analysis included 765 women. During follow-up, there were 104 seroconversions, of which 59 were in the nonoxynol-9 group. A higher incidence of HIV-1 in the nonoxynol-9 group than in the placebo group was noted in three of the four centers in the phase three study. Overall incidence of gonococcal infection was almost a quarter higher in the nonoxynol-9 than in the placebo group, and overall incidence of chlamydial infection was slightly increased in the nonoxynol-9 group. Researchers analyzed site data and found that condom use and anal sex did not confound the results.

Overall incidence of lesions, combining those with or without epithelial breach, did not differ significantly between the two treatment groups: 34.3 percent versus 29.4 percent in the nonoxynol-9 and placebo groups, respectively.

To test the hypothesis of dose-dependent toxic effects of nonoxynol-9, researchers divided the mean gel use per working day into three categories based on tertiles, assessing HIV-1 incidence per treatment group and per category of gel use. HIV-1 incidence increased most rapidly with increasing gel use in the nonoxynol-9 group versus the placebo group. In the nonoxynol-9 group, HIV-1 incidence rose from 8.8 per 100 woman-years in women reporting mean use of 3.5 or fewer applicators per day to 30.6 in women reporting a higher mean daily use. In the placebo group, HIV-1 incidence in those categories was 8.1 and 14.5 per 100 woman-years, respectively. Two hundred eighty-nine women (32 percent) reported use of a mean of more than 3.5 applicators per working day, and in these women, risk of HIV-1 infection in nonoxynol-9 users was almost twice that in placebo users. HIV-1 risk did not differ between the two treatments for the 516 women (68 percent) who used the gel less frequently than 3.5 times a day. In addition, the researchers reported no significant effect of nonoxynol-9 on N. gonorrhoeae or C. trachomatis infections.

Researchers divided lesions into those with or without an epithelial breach and assessed whether incidence of lesions with a breach also increased with increasing gel use. Incidence of lesions rose with increasing gel use. The increase in incidence of lesions was seen in both groups, but it happened most rapidly in the nonoxynol-9 group. Hazard ratio of HIV-1 infection in women who had at least one episode of a lesion with an epithelial breach (irrespective of treatment group) was twice that in those who never had such a lesion. Lesions without an epithelial disruption did not increase a woman's risk of HIV-1 infection.

Researchers conclude that nonoxynol-9 increased risk of HIV- 1 infection compared with placebo. Risk was especially high in women who used the study drug gel more than 3.5 times per day and who had a high incidence of lesions with epithelial disruption. This finding, researchers said, suggests that nonoxynol-9 has an adverse effect on vaginal integrity when used frequently, thus increasing women's susceptibility to HIV-1 infection. At low frequency use, nonoxynol-9 had no effect, either positive or negative, on HIV-1 infection. Their conclusion did not change after adjustment for sexual behavior (frequency of vaginal sex not protected by condoms or unprotected anal sex). Assessment of other microbicides should continue.

Back to other CDC news for October 17, 2002

Previous Updates