BOSTON – An individual's microbiome is now recognized as a major factor in health, disease, and the effectiveness of medical treatment. At her Tuesday morning plenary talk at the 2018 Conference for Retroviruses and Opportunistic Infections (CROI), Nichole Klatt, assistant professor in the University of Washington's Department of Pharmaceutics, gave an overview of the female vaginal microbiome's role in both HIV infection and treatment.

Specifically, the vaginal microbiome affects the risk of contracting HIV in at least two distinct ways.

Through effects on both inflammation level and the physical integrity, the microbiome affects the vaginal microenvironment in ways that can make it more or less susceptible to HIV infection.

And, it turns out, certain members of the vaginal microbiome change the effectiveness of pre-exposure prophylaxis (PrEP).

PrEP, or taking low doses of antiretroviral drugs as a preventive measure, can reduce the risk of contracting HIV. Or not.

"In men, PrEP works very well," Klatt said, with an effectiveness of 80 to 90 percent.

In women, effectiveness is both lower overall, and varied more between individuals.

The outcome of the trials showing PrEP is less effective in women "has been attributed mainly to adherence," Klatt said.

But in a study Klatt and her colleagues published last year, they showed that the effectiveness of the microbicide gel tenofovir depended strongly on the user's vaginal microbiome.

As far as their vaginal microbiome was concerned, women in the CAPRISA study could be divided into two groups – those whose vaginal microbiome was dominated by Lactobacillus, and those who had microbiomes that were more diverse and dominated by Gardnerella vaginalis or other anaerobic bacteria.

Tenofovir gel was about 60 percent effective in preventing HIV acquisition in women with a Lactobacillus vaginalis-dominated microbiome, but less than 20 percent effective in those with a large fraction of G. vaginalis.

The reason was the G. vaginalis bacteria metabolized tenofovir, which led to a much more rapid decline in drug levels after administration.

In other studies, Klatt and her team showed that anaerobic microbiome bacteria also metabolized dapivirine, another antiretroviral drug that is delivered directly into the vagina for PrEP.

However, the bugs did not affect levels of oral PrEP agents in the vagina, and did not metabolize tenofovir's prodrug, tenofovir alafenamide, which is also used clinically.

There are a number of possible ways to manipulate the vaginal microbiome, including a vaginal fluid transplant, treatment with bacteriophages, gene editing, and the probiotic Lactin V (Osel Inc.), which is in clinical trials for the treatment of bacterial vaginosis, as well as urinary tract infections.

Klatt's plenary talk also illustrated the importance of having a good way to measure drug adherence. In the case of PrEP, adherence may play less of a role than researchers have suspected. But in other cases, adherence may offer a clue as to why certain treatments don't work, or work only sometimes.

Adherence monitoring, though, "has historically been by subjective reporting," Monica Gandhi, professor of medicine and associate division chief of the University of California at San Francisco's division of HIV, infectious diseases and global medicine, told the audience in an oral abstract session on Monday.

In her presentation, Gandhi showed data suggesting that one way to measure adherence more rigorously is by looking at hair samples.

The team took quarterly hair samples in the A5257 trial, which compared the efficacy of three different ART regimens.

The team took hair samples from trial participants quarterly, measured drug levels in the hair using mass spectrometry, and compared it both to self-reported adherence levels and to the likelihood that drug therapy was effective at suppressing viral levels.

They showed that adherence varied widely, and patients with low hair drug levels, suggesting poor adherence, were much more likely to have ineffective viral control than those with higher drug levels in their hair.

Hair, Gandhi pointed out, is easy to collect, can be stored and shipped at room temperature, and shows drug levels over long periods of time. The latter feature prevents what is known as "white coat adherence," or taking drugs shortly before a doctor visit only.

The study also reinforced the need for objective adherence measures. The team also asked patients to self-report their adherence. By that measure, the media drug adherence was 100 percent.