Harnessing Vaginal Microbiota to Protect Women From HIV: What We Know and Don't Know

If you're a provider, you want to offer patients the latest, evidence-based ways to reduce their risk for HIV -- including women who may have increased risk not because of their behavior or even the prevalence of unsuppressed HIV in their community, but because of the microbes in their genital tracts.

But, if you ask R. Scott McClelland, M.D., M.P.H., associate director of the University of Washington Center for AIDS Research, the role of the microbiome in women's HIV risk is still not totally clear.

"Moving forward, investigating the effect of the [vaginal] microbiome together with the effects on the immune system is critically important," he said. "I think that to some extent the microbiota shapes the mucosal immune response, and the mucosal immune response may shape the microbiota. Fully disaggregating those relationships may be difficult, but I do think that's a critical way forward."

Related: The Company You Keep: Do Social Networks Influence HIV Status?

At the Conference on Retroviruses and Opportunistic Infections (CROI 2017) last month, however, we got closer to understanding some of those relationships -- as well as the limits of current research.

A Milieu Whose Time Has Come?

The vaginal microbiome became a central topic of discussion at 2016's International AIDS Conference, when Salim Abdool Karim, M.D., Ph.D., director of the Center for AIDS Research Program in South Africa (CAPRISA), released an analysis showing that the bacteria Prevotella bivia was associated with a 13-fold increase in HIV prevalence. At the same conference, McClelland presented a poster showing that other bacteria -- Eggerthella species type 1, Gemella asaccharolytica, Leptotrichia/Sneathia, Megasphaera and Mycoplasma hominis -- were associated with statistically significant increases in HIV prevalence, but no where near a 13-fold increased risk.

Then, in January, researchers at the Harvard Medical School lab of Douglas Kwon, M.D., Ph.D., found that diverse microbiomes -- that is, microbiomes not dominated by Lactobacillus crispatus -- were associated with a four-fold increase in HIV prevalence. What's more, the report, published in the journal Immunity, found that such microbiomes were also associated with increased activation of mucosal CD4 cells. In other words, those bacteria were associated with the availability of more HIV target cells on the surface of a woman's genital tract.

When Kwon's team looked more closely, they saw that some of the same bacteria described by Karim and McClelland -- Prevotella, Megasphaera and Sneathia, for instance -- were associated with increased vaginal inflammation and HIV prevalence.

The findings, the team wrote in the paper, could have bearing for wide swaths of women in sub-Saharan Africa.

"Our findings further indicate that individuals within a particular [bacterial community] group are at similar risk of acquiring HIV," they wrote. "This suggests that more than half the population of young black South African women is at increased HIV infection risk based on the composition of their cervicovaginal microbiome."

Kwon's lab's study was one of several McClelland included in his survey on the vaginal microbiome and HIV risk. Here's what we do know, based on numerous presentations and posters released at CROI last month.

Vaginal Microbiota May Thin the Defenses Against HIV

It's long been known that domination of the vaginal microbiota by Lactobacillus crispatus (L. crispatus) is associated with lower pH and lower HIV acquisition. But in a poster presented at CROI on Wednesday, Feb. 15, Julie K. Shade and colleagues at Johns Hopkins University presented data showing that it's not just the low pH that inhibit HIV acquisition; something else, something that's not totally clear yet, is at work.

The study asked women aged 18-45 to collect a sample of their vaginal mucus. Then, the researchers challenged the samples with HIV to see how mobile the virus is in L. crispatus microbiomes versus more diverse microbiomes, sometimes referred to as vaginal dysbiosis or simply bacterial vaginosis (BV).

They found that HIV slipped through the mucus of women with BV easily but was "trapped" in the lower-pH mucus of women with Lactobacillus-dominant microbiota. The more the virus is trapped in the mucus, the less likely it is to reach whatever target cells are present on the surface of the genital tract.

Then, to test whether the trap was set by the pH or the bacteria itself, Shade and colleagues altered the pH of both the samples containing high amounts of Lactobacillus and those samples without. Again, they challenged the samples with HIV.

They found that even the Lactobacillus-laden mucus thinned at a higher pH. But when the pH again dropped? The trap was again set.

"In contrast," the researchers wrote, "acidification of cervicovaginal mucus from women with BV did not lead to the trapping of HIV."

In other words, it's not just the pH; something specific about L. crispatus inhibits HIV from reaching the genital surface.

BV Is Hard to Dislodge

In his presentation, McClelland described the other challenge of diagnosing a woman with BV. Namely, even with the approved treatments for BV, there's only an 80% initial cure rate.

"And then, in women initially cured, [there is] about a 50% failure rate by six months recurrence and even higher over longer periods," he said. "So, simply treating BV once isn't going to be the answer."

He added that there's not much in the pipeline in terms of new BV regimens, though some groups are working on probiotics and other approaches to manage the microbiome to become a more advantageous ecosystem.

Circumcision's Role in Altering the Microbiome

The vaginal microbiota doesn't exist in isolation, after all. McClelland has told TheBodyPRO.com in the past that BV may simply be a marker of condomless sex with uncircumcised men. So, the penile microbiota is important to a woman's HIV risk, as well.

At CROI, Lance Price, Ph.D., a professor of science and engineering at George Washington University, presented data led by researcher Cindy M. Lui testing the hypothesis that male circumcision alters the penile microbiota, thereby decreasing the production of cytokines and target cells and resulting in lower HIV risk for men overall.

What Liu's team found was that Prevotella shows up again -- this time in the penile microbiota of men living with HIV at baseline, but not in their HIV-negative counterparts.

A year after circumcision, the swabs returned a totally different result.

"The microbial communities changed dramatically a year after circumcision," Price said. Specifically, he said, they saw significant decreases in anaerobic bacteria -- that is, the bacteria associated with BV in women.

"You'll recognize some of these names: Prevotella, Porphyromonas, Finegoldia, etc.," he said.

Not only that, circumcised men showed evidence of new bacteria taking their place, including Lactobacillus.

This also points to a challenge in dislodging high-risk microbes from women's microbiota, Price told TheBodyPRO.com: They had to literally remove skin to dislodge high-risk microbes from men's penises. There's no such option available to women who wish to change their microbiome.

Oral Tenofovir Works Regardless of Vaginal Microbiome, but Topical Applications May Not

So, if the microbes that live in the vagina can help or hinder HIV risk, do those microbes matter when it comes to HIV prevention? CAPRISA's Salim raised this question at the International AIDS Conference last year when, in addition to showing that P. bivia is associated with higher HIV prevalence in women, he showed that another microbe, Gardnerella, ["was just gobbling up" http://www.pbs.org/newshour/updates/bacteria-stoking-hiv-cases-blocking-hiv-treatment/] tenofovir (Viread) gel, leaving less to protect women from HIV.

This raised the question of whether the microbiome could blunt the effectiveness of other HIV prevention methods containing tenofovir, such as oral Truvada (tenofovir/emtricitabine). Plus, some researchers were just plain skeptical of Gardnerella's tenofovir-gobbling abilities.

So, at CROI, researchers presented the first evidence to either support or disprove Karim's findings.

First, Renee Heffron, Ph.D., M.P.H., an assistant professor of global health and epidemiology at the University of Washington and an investigator on the PARTNERS pre-exposure prophylaxis (PrEP) trial, took the samples they'd collected from women over the years and tested them for Gardnerella, the absence of Lactobacillus and other signs of BV. Then, they looked at the HIV acquisition rates among the women by vaginal pH and BV status.

What they found was that even though 24% of participants met the criteria for BV, there was no significant difference in oral tenofovir efficacy between groups. Essentially, what's happening in the microbiota has no influence on tenofovir when it's taken systemically.

But what about when it's applied topically? The CAPRISA tenofovir gel study ended several years ago, and a few tenofovir gel-based vaginal rings are in early development, so the absorption rate of tenofovir by the microbes in the vagina might still matter.

So, Sharon Hillier, Ph.D., at the University of Pittsburgh Magee Women's Research Institute, took the samples she had from her organization's participation in the CAPRISA study, tested them for presence of Gardnerella and Atopobium vaginae and assessed trough tenofovir concentrations after six days of gel use.

"I was surprised," Hillier said. "I had some skepticism about the [IAC] data, it's safe to say."

She didn't expect to see a difference by BV status, but there was. In fact, Gardnerella did, indeed, seem to reduce the amount of tenofovir gel available for protection, both in the cervicovaginal fluid and in blood plasma. The same was true of Atopobium vaginae.

This doesn't mean topical tenofovir couldn't be effective in women with BV, Hillier told TheBodyPRO.com. It could just mean that dosing of the drug would have to be more frequent if women have BV.

"But it's obviously not good," she said. "Anything that decreases the available drug isn't good."

What This Means for Women and Their Providers

At the beginning of his presentation, McClelland showed the audience a Venn diagram of the bacterial diversity that can make up BV. In the upper left hand corner was a circle with the bacteria M. hominis in it. In the lower left, were BVAB2 and Megasphaera species; in the lower right was Prevotella timonensis; in the upper right were Prevotella buccalis, Prevotella disiens and BVAB1. In the Venn diagram these species didn't overlap. M. hominis is associated with higher vaginal pH, but not with clue cells. BVAB2 is associated with clue cells but not with a positive whiff test -- in other words, the odor associated with BV. And P. disiens is associated with odor but not with vaginal discharge, which is the purview of P. timonensis.

Only Leptotrichia amnionii and Eggerthella on his slide hit all the marks.

"Different bacteria have different clinical effects on the cardinal signs and symptoms of BV," he said.

This is important for two reasons: One, McClelland said that presence of asymptomatic BV is associated with increased acquisition of other sexually transmitted infections (STIs), such as herpes, which in turn has been found to increase the likelihood that a woman will acquire BV over time. Both further increase women's risk for acquiring HIV.

Two, HIV risk doesn't decrease when a woman is asymptomatic. In a poster presented at the conference, researchers from ICAP at Columbia University in New York City asked whether waiting for symptoms to test for reproductive tract infections might be "antiquated in the HIV era."

What they found was that of the 314 Tanzanian women living with HIV in the study, 37.3% reported symptoms of reproductive tract infections, including BV. But when they tested all the women for STIs, they found that 78.7% of women actually had an STI. Syndromic diagnosis, they concluded, underestimates infections significantly.

This led one provider to ask McClelland why clinical guidelines don't reflect this evidence.

"Current guidelines state that if the symptoms resolve, no follow up is needed for bacterial vaginosis," the provider said. "But given [these facts], do you think that's an adequate strategy? Or do you think that should be reconsidered and maybe we need to call [women] back in and rescreen these folks?"

McClelland's answer might surprise you.

"We're not at a stage where we can tell women that they need to come back in to make sure that the BV is gone," he said. "If they don't have symptoms from a clinical perspective, I think we're done."

But, he added, if funding and research continues on the interplay between the microbiome with the immune system, that could change.

"I would love at some point in my career," he said, "to be able to say that we need to control asymptomatic BV."