Background: Bacterial vaginosis (BV) recurrence rates may be attributed to factors including poor adherence to antibiotic therapy, inadequate dosing to eradicate BV-associated bacteria (BVAB) or antibiotic resistance. No published studies have linked vaginal concentrations of metronidazole to concentrations of indicator BVAB and BV outcomes. Methods: Participants with Amsel-BV were prescribed with intravaginal/oral metronidazole and returned after 6-15 days to determine if treatment was effective (test-of-cure) and 16-35 days to evaluate short-term BV recurrence. An antimicrobial assay was developed to assess vaginal metronidazole activity. Anaerobic vaginal bacteria were streaked on Brucella plates and bactericidal activity (clearance zone) of vaginal swab supernatants was used to indicate metronidazole use. Clearance noted on any prescribed days of intravaginal or oral metronidazole was considered any-documented-use while no clearance on any prescribed days was classified as no-documented-use. BVAB2 and Megasphaera spp. concentrations were measured using quantitative PCR (qPCR).
Results: Of 150 BV episodes in 88 participants, 98 (65.3%) had a test-of-cure visit; any-documented-use of metronidazole was noted in 78 (79.6%) and no-documented-use in 20 (20.4%) episodes. At test-of-cure visits, Amsel-BV was observed in 4/78 episodes (5.1%) among those with any-documented-use of metronidazole and 3/20 episodes (15%) among those with no-documented-use (p=0.15). Of 150 episodes, there was a follow-up visit (16-35 days) for 124 episodes (82.7%); any-documented-use was observed in 82 (66.1%) and no-documented-use in 42 (33.9%) episodes. At this follow-up, Amsel-BV was noted in 21/82 episodes (25.6%) among those with any-documented-use and 18/42 episodes (42.9%) among those with no-documented-use (p=0.041). We compared changes in BVAB concentrations in 150 episodes with daily qPCR data for 10 days post-initiation of antibiotics. Among those with any-documented-use, there was a decrease of 0.26 log 16S rRNA copies/swab of BVAB2 each day versus 0.13 log copies/swab in the no-documented-use group (difference in slope=0.13, p< 0.001). Likewise, there was a decrease in 0.34 log copies/swab of Megasphaera spp. in the group with any-documented-use compared to 0.13 log copies in the no-documented-use group (difference in slope=0.21, p< 0.001). Conclusion: A substantial fraction of participants lack evidence of use of metronidazole as indicated by our bioassay, leading to ineffective clearance of BVABs which may result in short-term BV recurrence.
