Vaginal microbial shifts are unaffected by oral pre-exposure prophylaxis in South African women

Abstract

Vaginal microbiota have been shown to be a modifier of protection offered by topical tenofovir in preventing HIV infection in women, an effect not observed with oral tenofovir-based pre-exposure prophylaxis (PrEP). It remains unclear whether PrEP can influence the vaginal microbiota composition. This study investigated the impact of daily oral tenofovir disoproxil fumarate in combination with emtricitabine for PrEP on the vaginal microbiota in South African women. At baseline, Lactobacillus iners or Gardnerella vaginalis dominant vaginal communities were observed in the majority of participants. In cross sectional analysis, vaginal microbiota were not affected by the initiation and use of PrEP. Longitudinal analysis revealed that Lactobacillus crispatus-dominant "cervicotypes 1 (CT1)" communities had high probability of remaining stable in PrEP group, but had a higher probability of transitioning to L. iners-dominant CT2 communities in non-PrEP group. L. iners-dominant communities were more likely to transition to communities associated with bacterial vaginosis (BV), irrespective of PrEP or antibiotic use. As expected, BV-linked CTs had a higher probability of transitioning to L. iners than L. crispatus dominant CTs and this shift was not associated with PrEP use.

Figure 1

Heat map of the most abundant vaginal bacteria identified by 16S rRNA sequencing in 100 women at baseline. Relative abundance of key taxa across all of the samples was used to determine each of the four CTs: L. crispatus (CT1), L. iners (CT2), G. vaginalis (CT3), and mixed communities not dominated by other taxa (CT4). BV and STI statuses were also depicted.

Figure 2

Composition and transitions between cervicotype in PrEP (n = 64) and non-PrEP groups (n = 36). (A) Box plots showing CTs assigned to individual participants over 3 visits at baseline, 3 months and 6 months. Blue colour depicts CT1, green CT2, pink CT3, and orange CT4. PID number that starts with 082 refers to women enrolled in CAPRISA 082 and 084 refers to CAPRISA 084 cohort. (B) Transition probabilities between CT’s from one visit to another in PrEP and non-PrEP groups who did not use or used antibiotics during the study.

Figure 3

Non-metric Multidimensional scaling (NMDS) of Bray–Curtis showing no clustering dissimilarity of species level relative abundances in PrEP and non-PrEP groups at baseline, 3 months and 6 months. The pink colour represents non-PrEP group and green colour represents PrEP group. Open shape represents participants on antibiotics.