Presence of BV-related bacterial species in the vaginal microbiota may contribute to the vulnerability of African women to HIV

The association between disruptions of vaginal microbiota and vulnerability to STIs has been widely discussed. Tamarelle & Astagneau (STI) report a study based in a French STI clinic showing some evidence of greater vulnerability to Chlamydia in the case of microbiota not dominated by Lactobacillus crispatus. The more important question of an association of BV-related bacteria with vulnerability to HIV has been in the air for a long time (Schmid & Koumans (STI))Gosmann & Kwon, in a recent study of South African women showed a four-fold decreased vulnerability among women with L. crispatus-dominated microbiota compared with women with microbiota dominated by non-Lactobacillus communities other than Gardnerella.

Now McClelland & Fredericks (M&F) have reported a study of vulnerable South and East African women based on five cohorts assembled for the purposes of some recent well known HIV studies, including Partners in Prevention and Partners PrEP. These included sex-workers, women in sero-discordant relationships and pregnant or post-partum women. The two-stage design involved: first, using abundance data to show the association of bacterial diversity with HIV and to guide selection of bacteria for further investigation; second, testing the hypothesis of an association of concentrations of each of twenty selected bacteria with HIV risk by subjecting vaginal samples from 55 HIV infected and 55 HIV non-infected women to taxon-directed rtPCR assays. In the studies mentioned earlier, associations are sought not with individual taxa, but with five or so overall microbiota types.

C&F found bacterial diversity to be higher in HIV infected women: median 1.3 as against 0.7. There were significant associations with HIV acquisition for seven of the 20 bacterial taxa; for the third tertile of concentration, adjusted odd ratios for HIV association were: Parvimonas species type 1, 4.64; P. species type 2, 2.18; Gemella asaccharolytica, 3.03; Mycoplasma hominis, 2.76; Leptotrichia/Sneathia, 2.59; Eggerthella species type 1, 1.53; vaginal Megasphaera, 1.32. As might be expected from the earlier studies mentioned above, there were strongly correlations in the concentrations of many of the 20 taxa, excluding the lactobacilli (crispatus; iners; jensenii).

There has been some debate in the literature about whether the greater vulnerability associated with certain microbiota types has more to do with the presence of disruptive bacteria or a deficiency of protective lactobacilli. M&F note that their primary analysis using rtPCR assays in the full dataset showed no significant association between concentration of lactobacilli and HIV acquisition. So, in the light of the negative correlation between concentrations of lactobacilli and BV related bacterial species, they are inclined to favour the view that increased susceptibility to HIV is the result of the presence of disruptive bacteria rather than the deficiency of protective ones.

On a more general level, M&F see their study as suggesting an explanation of the apparently much greater vulnerability of African women to HIV infection than women of European and Asian origin: higher diversity vaginal bacterial communities not dominated by lactobacilli are more common in this racial group. M&F suggest that vaginal dysbiosis may account for as much as 20-30% of the population attributable risk of HIV acquisition in African women.