Background: Bacterial vaginosis (BV) is a common vaginal condition that is characterised by a non-optimal, lactobacillus-deficient vaginal microbiota. BV recurrence following first-line antibiotics is common. Studies exploring the contribution of the vaginal microbiota to recurrence are limited and conflicting. We investigated if the vaginal microbiota composition pre-treatment or immediately post-treatment was associated with BV recurrence within one-month of treatment.
Methods: This was a secondary analysis of vaginal samples collected from women who participated in three clinical trials that evaluated new interventions for improving BV cure, including concurrent antibiotic treatment of regular sexual partners (RSP). Women attending Melbourne Sexual Health Centre who were diagnosed with BV (≥3 Amsel criteria and Nugent Score=4-10) received first-line BV-treatment, and self-collected vaginal swabs pre-treatment and immediately post-treatment. 16S-rRNA gene sequencing was used to characterise the vaginal microbiota. Differences in the pre- and post-treatment microbiota composition between women who experienced recurrence within one-month of enrolment and women who remained cured were examined using ANOSIM. Logistic regression, adjusted for confounders, was performed to explore the association between BV recurrence and 1) bacterial diversity and 2) centre-log-ratio transformed abundance of individual bacterial taxa before and after treatment.
Results: Data from 121 women were analysed; 16 women (13%; 95%CI:8-21%) recurred within one-month of enrolment. Women with an untreated RSP were more likely to recur than women with no RSP (9/28 vs 3/44; P=0.008) or women with a RSP who received partner treatment through two of the parent trials (4/49; P=0.011). Neither the diversity nor composition of the microbiota pre-treatment (adjusted odds ratio [AOR]=1.99, 95%CI:0.77,6.71, p=0.203 and ANOSIM-R=0.070, p=0.166, respectively) or immediately post-treatment (AOR=1.37, 95%CI:0.60,2.97, p=0.435 and ANOSIM-R=0.037, p=0.318, respectively) differed between recurrence and cure cases. However, a higher abundance of Prevotella spp. pre-treatment (AOR=1.35, 95%CI:1.05,1.91, p=0.041) and Gardnerella spp. immediately post-treatment (AOR=1.23, 95%CI:1.03,1.49, p=0.026) were both associated with increased odds of BV recurrence after adjusting for partner type and treatment status.
Conclusion: Our findings suggest that specific taxa may play a role in the high rates of treatment failure following recommended BV therapy. Further work should prioritise understanding the contribution of Prevotella and Gardnerella species to BV recurrence.
