Effect of metronidazole on concentrations of vaginal bacteria associated with risk of HIV acquisition

Abstract

Several bacterial vaginosis (BV)-associated bacteria have been associated with elevated risk of human immunodeficiency virus (HIV) acquisition; however, susceptibility of these bacteria to antibiotics is poorly understood. Vaginal samples were collected from 22 persons daily for 2 weeks following BV diagnosis. Metronidazole treatment was prescribed for 5-7 days. Changes in bacterial concentrations were measured with taxon-specific 16S rRNA gene quantitative PCR (qPCR) assays. A culture-based antimicrobial assay confirmed presence of antibiotics in vaginal swab samples. Bacterial DNA concentrations decreased during antibiotic administration for all 13 bacterial taxa tested. Comparison of bacterial DNA concentrations in samples before administration of antibiotics to samples taken on the last day of antimicrobial assay-confirmed antibiotic presence showed a 2.25-4.78 log10-fold decrease across all taxa. Concentrations were frequently reduced to the qPCR assay's limit of detection, suggesting eradication of bacteria. Mean clearance time varied across taxa (1.2-7.9 days), with several bacteria (e.g., Sneathia spp., Vaginal TM7, and Eggerthella-like sp.) taking >7 days to suppress. Metronidazole reduces quantities of bacterial taxa associated with increased HIV acquisition risk.IMPORTANCEHuman immunodeficiency virus (HIV) transmission through sex remains a major public health challenge despite efforts at risk reduction and use of anti-retroviral pre-exposure prophylaxis. Many bacterial vaginosis (BV)-associated vaginal bacteria have been associated with increased HIV infection risk among women. If these bacteria help mediate HIV infection risk, then eradication of these bacteria is one potential strategy to reduce this risk. However, the best approach to eradicate HIV-high risk bacteria from the vagina is not known. We analyzed vaginal swabs collected daily from women with BV to determine the impact of metronidazole treatment on 13 vaginal bacterial taxa linked to elevated risk of HIV infection through use of taxon-directed quantitative PCR assays. We conclude that eradication of high-risk vaginal bacteria using metronidazole is one promising avenue for reducing HIV acquisition risk, and we provide evidence that a 5-7-day treatment course may not be sufficient to suppress all bacteria.

Keywords: HIV risk; bacterial vaginosis; metronidazole; quantitative PCR; vaginal microbiota.

Fig 1

Concentrations of vaginal bacteria over time during metronidazole treatment for a representative treatment course in one person with BV.

Fig 2

Two visualizations of the same data set, illustrating changes in concentrations of 16S rRNA gene copies per swab, comparing “pre-metronidazole” samples (baseline measurements collected 1 day prior to confirmed antibiotic presence) to “post-metronidazole” samples (collected on the final day of antibiotic activity detection). (A) Graph showing decreases in concentrations of bacterial DNA for each taxon assayed, from “pre-metronidazole” to “post-metronidazole” samples. Statistical significance assessed by Wilcoxon matched pairs test (Table S1). (B) Box plot displaying log-fold change in bacterial concentrations, from “pre-metronidazole” to “post-metronidazole” samples. Total number of treatments in which a bacterial taxon appeared is indicated to the right of each bacterial taxon.

Fig 3

Box plot (A) showing time to suppression for each bacterial taxon (total number of treatments in which a bacterial taxon appeared is indicated along base of x-axis); scatter plot (B) showing relationship between baseline bacterial concentration and time to suppression.

Fig 4

Box plot showing time to suppression between different bacterial taxa, split by method of antibiotic administration (total number of treatments in which a bacterial taxon appeared is indicated along base of x-axis). Statistical significance assessed by Mann-Whitney U-test (Table S2).