536    

Background:  There is a clear need for vaginal microbicides that will prevent HIV transmission. An important attribute will be the effect of such products on the vaginal flora. We describe the effects of 2 microbicide candidates, cellulose sulfate gel (CS) and tenofovir gel (TFV), on the pH and vaginal flora of sexually abstinent and sexually active HIV^– and HIV⁺ women in 2 completed phase I studies, HIV Prevention Trials Network (HPTN) 049 and HPTN 050.

Methods:  Eligibility criteria included negative pregnancy test, normal Pap smear, no sexually transmitted infections in the preceding 6 months, and no vaginal symptoms or discharge. A pelvic examination, including pH measurement, wet mount, and Gramm stain of vaginal fluid, were conducted prior to and after 14 days of product use. Both studies included cohorts of sexually abstinent and sexually active women.

Results:  We randomized 59 HIV⁺ women enrolled in HPTN 049 to 6% CS or placebo (KY gel) once or twice a day. Eleven were sexually active. Assignment to CS or placebo was double-blind. The mean vaginal pH was 5.0 at baseline and 5.2 by day 14. Of the 53 women with an available Nugent score, 23 (43%) had bacterial vaginosis at enrollment. By day 14, 16 (70%) of these 23 no longer had bacterial vaginosis. Bacterial vaginosis clearance was similar for CS (89%) and placebo (57%), and did not differ by frequency of gel use or by sexual activity. We assigned 84 women enrolled in HPTN 050 (60 HIV^– and 24 HIV⁺) to 0.3% or 1% TFV once or twice a day:  60 were sexually abstinent and 24 were sexually active. Mean vaginal pH was 5.0 at enrollment, and 4.9 by day 14. Of the 76 women with an available Nugent score, 30 (39%) had bacterial vaginosis at enrollment. By day 14, 15 (50%) of these 30 women no longer had bacterial vaginosis. Bacterial vaginosis clearance did not differ by gel dose or frequency, or by HIV status or sexual activity level.

Conclusions:  Asymptomatic women enrolled in these microbicide studies had a high prevalence of elevated vaginal pH and incidental bacterial vaginosis at enrollment. In both studies, clearance of bacterial vaginosis was observed in at least half of those with bacterial vaginosis at baseline, regardless of type of product used, including placebo. There was no substantial change in vaginal pH, suggesting either a dilution effect or a pH-independent effect of the gels. As bacterial vaginosis may act as a co-factor in the heterosexual transmission of HIV, the impact of vaginal microbicides on bacterial vaginosis warrants further study.

Keywords: vaginal flora; tenofovir; cellulose sulfate