How Bacterial vaginosis treatment tackles individual species

Science Spotlight

How Bacterial vaginosis treatment tackles individual species

Vaccine and Infectious Disease Division

June 19, 2017

Bacterial vaginosis (BV) is vaginal infection made up of multiple species of bacteria. Prevalence of BV is high, affecting millions of reproductive-aged women worldwide often reoccurring after treatment. BV can cause adverse health outcomes, including premature birth in pregnant women, cervicitis, and HIV-1 acquisition. BV is associated with various species of bacteria; however the exact bacterial combination is unknown and probably various among women. With the advent of new molecular techniques, such as 16S ribosomal RNA detection our knowledge of BV has increased. Nevertheless, it is still largely unknown how treatment affects individual bacterial species. In a paper published in The Journal of Infectious Disease researchers from Fred Hutchinson Cancer Research Center (Vaccine and Infectious Disease Division) and collaborators looked at the effects of BV intervention on vaginal bacterial species. The study specifically focused on the impacts of periodic presumptive treatment (PPT) on the vaginal microbiome looking at the species level. Lead author Dr. Balkus said, “Evidence suggests that specific vaginal bacteria associated with bacterial vaginosis (BV) may increase the risk of adverse health outcomes in women, including HIV acquisition and preterm birth. Interventions that modify the vaginal mircobiome by reducing such species may decrease the occurrence of these adverse health outcomes. This analysis provides data on the impact of a vaginal health intervention on detection and the quantity of key bacterial species.“

Women from either the US or Kenya made up two groups of the study; 111 in the PPT arm and 110 in the placebo arm. The PPT arm consisted of treatment using topical metronidazole 750mg and intravaginal miconazole 200mg for five consecutive nights, each month, lasting one year. Patients were evaluated each month and samples were taken to assess vaginal bacterial microbiome using bacterial quantitative PCR and Gram-stained slides. Bacterial load was scored using both detection value, called the lower limit of detection (LLD) and by receiver operating characteristic (ROC). ROC curves were the quantity of species versus the presence of BV. This allowed for a predictive measure to identify a quantity threshold, predictive of BV and could be used to with the LOD to predict BV. When comparing the arms of the study, patients from each group had similar baseline characteristics (i.e. bacterial species and prevalence), however there was a difference of characteristics between the US and Kenyan participants. 

Graphs showing the bacterial species having values over the LLD for each arm of the study.

Graphs showing the bacterial species having values over the LLD for each arm of the study.

Graph provided by Dr. Balkus

Using an unadjusted bacterium-specific model the team found that more BV-associated species were below the LLD in the PPT arm compared to placebo specifically BVAB1/2, A. vaginae, Leptotrichai/Sneathia and Megasphaera species. When looking by study site the results were similar however the breakdown by country deferred. Kenyan participants had a more pronounced reduction in BV-associated bacteria while in US participants, the effect size was small leading to no significant differences between arms. The researchers suggest that the difference observed between US and Kenyan participants could be due to adherence, which was reported as 99% compared to 88% in Kenyan and US participants, respectively. When looking at ROC cutoffs the team noted a decrease in the proportion of visits with bacterial quantities over the ROC cutoff in women in the PPT arm. Similar to LLD the ROC assay produced a bigger difference in the Kenyan cohort with no significant difference in the US participants.

Overall, PPT treatment significantly reduced bacterial colonization with BV-associated bacteria. Since many samples were below the LLD the group used the novel predictive method to determine thresholds by ROC curves. This furthered the understating of changes within the vaginal microbiome after treatment, better linking changes in bacteria over baseline with clinical outcome. Further work needs to be done to assess this change over time after treatment completion. The importance of this work as stated by Dr. Balkus is, “In light of recent findings demonstrating associations between detection of specific vaginal bacteria and increased risk of adverse outcomes, the role of interventions to improve vaginal health and reduce these adverse health outcomes in women should be explored.“

This work was funded by the National Institutes of Allergy and Infectious Disease, NIH, and the University of Washington Center for AIDS Research.

Balkus JE,Srinivasan S,Anzala O,Kimani J,Andac C,Schwebke J,Fredricks DN,McClelland RS. 2017. Impact of Periodic Presumptive Treatment for Bacterial Vaginosis on the Vaginal Microbiome among Women Participating in the Preventing Vaginal Infections Trial. Journal of Infectious Diseases, 215(5), 723-731.