Tenofovir-only and tenofovir/levonorgestrel intravaginal rings are unlikely to impact the genital microbiota of sub-Saharan women

BACKGROUND: Multipurpose HIV/HSV-2 prevention intravaginal rings (IVRs) with and without contraceptive co-formulation are under development. We investigated the impact of 90-day tenofovir (TFV) IVRs with/without levonorgestrel (LNG) relative to placebo on the genital microbiota of Kenyan women.
METHODS: In this phase 2a, double-blind, placebo-controlled, randomized trial, 27 women ages 18-34 years (non-pregnant, negative for HIV, Treponema pallidum, Chlamydia trachomatis, Neisseria gonorrhoeae andTrichomonas vaginalis, Nugent score <7, and without any contraindications to LNG or TFV) were randomized 2:2:1 to continuous use of IVRs containing: TFV (n=11); TFV/LNG (n=11); or placebo (n=5). Using lateral vaginal and IVR swab samples from each participant at the time of ring insertion and removal, the absolute abundance of bacteria per swab was determined by real-time PCR of the 16S region, and 16S rRNA gene sequencing was used to determine the microbial composition of the female genital tract.
RESULTS: Women used the IVR for a median of 67 days (range: 15-91 days). Vaginal total bacterial burden increased between insertion and removal of the placebo ring (0.32 log increase) and decreased with both TFV and TFV/LNG IVR use (0.57 and 0.27 log decrease respectively; all p>0.05). Compared to the genital lateral wall, the TFV/LNG IVR tended to have a higher biomass, the placebo IVR a lower biomass, and no difference with the TFV IVR. Median Shannon a-diversity decreased among women randomized to use the TFV/LNG IVR (1.79 vs. 1.29; p=0.26), increased in women using the TFV IVR (0.85 vs. 1.56; p=0.07) and remained stable in women with placebo IVR (1.80 vs. 1.88; p=0.77). Overall, the TFV/LNG IVR had a 'stabilizing' effect on the FGT microbiota whereby 50% of the participants' microbiota composition did not change and 50% shifted toward more lactobacillus-dominant states. More specifically, TFV/LNG IVR use was accompanied by increased abundances of L. gasseri and the genital probiotic species L. fermentum, and a decrease in Streptococcus spp (all false discovery rate-adjusted p<0.01).
CONCLUSIONS: We found that the TFV and TFV/LNG IVRs had no sign of detrimental impact on genital microbiota, with the latter being associated with increased lactobacillus abundance. These findings warrant further investigation in next-phase studies of these IVR.