A multi-strain probiotic kept bacterial vaginosis from returning in two-thirds of women

For Caroline Mitchell, the frustration was familiar. As director of the Vulvovaginal Disorders Program at Massachusetts General Hospital, she had spent years watching the same cycle repeat. A patient would come in with bacterial vaginosis - the discharge, the odor, the irritation. Antibiotics would clear it. And within six months, the infection would be back. Up to 60% of the time, it came back.

The problem was not that antibiotics failed to kill the harmful bacteria. They did that just fine. The problem was what they left behind: a vaginal environment stripped of its protective microbes, vulnerable to the same disruption all over again. Antibiotics cleared the infection but did nothing to rebuild the ecosystem.

Mitchell wanted to change that. Working with a global team spanning Mass General Brigham, the Centre for the AIDS Programme of Research in South Africa (CAPRISA), and collaborators from the Vaginal Microbiome Research Consortium, she helped design a trial to test whether reseeding the vagina with protective bacteria after antibiotics could break the cycle. The results, published in Cell Host & Microbe, suggest the approach works - and works better than previous single-strain attempts.

Why Lactobacillus crispatus matters

Bacterial vaginosis (BV) is not a simple infection caused by a single pathogen. It is a disruption of the vaginal microbiome - the community of microorganisms that inhabit the vagina. In a healthy vaginal environment, one species typically dominates: Lactobacillus crispatus. This bacterium produces lactic acid that keeps the vaginal pH low, creating conditions hostile to the harmful anaerobic bacteria that cause BV symptoms.

When BV develops, L. crispatus is displaced by a diverse consortium of anaerobes. Antibiotics can knock back those anaerobes, but they do not restore L. crispatus dominance. After treatment, very few women have an optimal vaginal bacterial community. The door is open for the harmful bacteria to return.

BV affects approximately 30% of women globally. Beyond the bothersome symptoms, it carries serious health consequences: increased risk of preterm birth, higher susceptibility to HIV acquisition, and association with abnormal cervical cell growth. In sub-Saharan Africa, where HIV remains a major challenge among young women, reducing BV could be a meaningful piece of an HIV prevention strategy.

From single-strain failure to multi-strain design

The idea of using probiotics to prevent BV recurrence is not new. Earlier phase 2 clinical trials tested a live biotherapeutic product containing a single strain of L. crispatus. Those trials showed some success in reducing BV recurrence after 12 weeks of dosing, but the beneficial bacteria failed to remain in more than half of participants. Colonization - getting the protective bacteria to actually take up residence and persist - proved to be the stubborn bottleneck.

Mitchell's team hypothesized that using multiple strains of L. crispatus might improve colonization success. Different strains bring different metabolic capabilities and different competitive advantages. A multi-strain product could cover more ecological niches and be more likely to find a foothold in the diverse conditions of individual vaginal environments.

The trial, called VIBRANT (Vaginal lIve Biotherapeutic RANdomized Trial), enrolled 90 participants from two sites: a rural research clinic in Vulindlela, South Africa, run by CAPRISA, and Boston. Each participant first received standard antibiotic treatment for BV. Then they were randomly assigned to one of three groups: placebo tablets, 6-strain L. crispatus tablets, or 15-strain tablets, taken over seven days.

Two-thirds colonized within five weeks

The researchers used genetic sequencing to test for the presence of the probiotic bacteria weekly for five weeks after treatment. The results exceeded their expectations for a phase 1 trial.

Two-thirds of participants (66%) had detectable protective L. crispatus bacteria in their vaginal microbiome within the first five weeks. Nearly half of that group still carried the bacteria at the 12-week follow-up, even though some participants had received only three days of active treatment out of the seven-day protocol (the remaining days were filled with placebo tablets to maintain blinding).

The clinical payoff was clear: participants who had protective bacteria at five weeks were significantly less likely to experience BV recurrence during the study period. The bacteria were not just present - they were doing the job they were designed to do, maintaining a protective environment that kept BV-causing organisms at bay.

A window into vaginal microbiome biology

Beyond the clinical implications, the VIBRANT trial produced something rare in vaginal health research: detailed, longitudinal data on how introduced bacteria interact with existing vaginal ecosystems. The weekly genetic sequencing created a time-lapse view of colonization, revealing which strains took hold, how quickly they established dominance, and how the broader microbial community shifted in response.

Mitchell considers this scientific aspect at least as important as the clinical results. Research into the basic biology of the vaginal environment has been chronically underfunded and underexplored. The VIBRANT trial is one of the few methods available to actually study beneficial vaginal bacteria in a controlled setting - by introducing known strains and watching what happens.

From South Africa to FDA approval: what comes next

The trial's South African arm was led by Disebo Potloane, MB, ChB, at CAPRISA's rural clinic. The site was chosen deliberately. HIV remains a disproportionate burden for young women in sub-Saharan Africa, and BV is an established risk factor for HIV acquisition. If a vaginal probiotic could reduce BV recurrence, it could potentially reduce HIV vulnerability as well - a goal that has been difficult to achieve through other means.

The research team is now planning a follow-up trial to optimize the treatment regimen before initiating larger-scale clinical trials aimed at FDA approval. Key questions remain: What is the optimal number of strains? How many days of treatment produce the best colonization rates? Does the probiotic need to be combined with a specific antibiotic regimen? And critically, does the protection last beyond 12 weeks?

Conflicts and limitations worth knowing

This was a phase 1 trial with 90 participants - small by the standards needed for regulatory approval. Phase 1 trials are designed primarily to assess safety and feasibility, not to provide definitive evidence of efficacy. The colonization and recurrence results are encouraging but must be replicated in larger, longer studies before any clinical claims can be made.

The trial was funded by the Gates Foundation. Mitchell has financial ties to several companies in the vaginal therapeutics space, including consulting for Freya Biosciences and DIVA Inc, and a financial interest in Ancilia Biosciences, which develops live biotherapeutic products. Several co-authors are named inventors on a patent for the biotherapeutic product tested in the trial. These conflicts do not invalidate the results, but they are relevant context for evaluating the findings.

The study also did not include a head-to-head comparison with the earlier single-strain products, making it difficult to directly quantify how much the multi-strain approach improved colonization over previous attempts.

Still, for a condition that affects nearly one in three women worldwide and has resisted durable treatment for decades, a 66% colonization rate from a brief course of treatment represents a genuine step forward in a field that has had precious few of them.