Does the Road to Treating Endometriosis Start in the Gut?

Researchers may be on track to develop a much-needed tool for studying endometriosis: A noninvasive stool test that could replace the current gold standard of laparoscopy.

Their approach, which focuses on the link between the gut microbiome and endometriosis, also identified a bacterial metabolite they said might be developed as an oral medication for the condition, which affects at least 11% of women.

In previous research, Rama Kommagani, PhD, an associate professor in the Department of Pathology & Immunology at Baylor College of Medicine in Houston, worked with a mouse model in which endometrial tissue from donor rodents was injected into the peritoneal space of healthy rodents to induce the disorder.

Transfer of fecal microbiota from mice with endometriosis to those without the condition induced the trademark lesions, suggesting the microbiome influences the development of endometriosis. Treating the animals with the antibiotic metronidazole inhibited the progression of endometrial lesions.

Kommagani speculated the microbes release metabolites that stimulate the growth of the endometrial lesions. “Bad bacteria release metabolites, you know, which actually promote the disease,” Kommagani said. “But the good bacteria might release some protective metabolites such as short-chain fatty acids.”

In a new study, Kommagani and his colleagues sought to identify a unique profile of bacteria-derived metabolites that could reliably diagnose endometriosis. Using stool specimens from 18 women with the condition and 31 without the disease, his team conducted whole metabolic profiling of the gut microbiota.

After identifying hundreds of metabolites in the samples, further analysis revealed a subset of 12 metabolites that consistently differentiated women with and without endometriosis.

These findings led to more questions. “If a metabolite is lower in women with endometriosis, does it have any functional relevance?” Kommagani said.

One candidate was 4-hydroxyindole (4HI), which was found in lower levels in the stool of patients. This substance is a little-understood derivative of its parent compound, indole, which occurs naturally in plants and has a wide range of therapeutic uses.

Using a mouse model, Kommagani’s lab demonstrated that feeding mice 4HI before receiving an endometrial transplant prevented the development of lesions typical of endometriosis. Mice given 4HI after they had developed endometriosis showed regression of lesions and decreased response to painful stimuli.

“In a nutshell, we found a specific set of bacterial metabolites in stool, which could be used towards a noninvasive diagnostic test,” Kommagani said. “But we also found this distinct, specific metabolite that could be used as a therapeutic molecule.” 

Tatnai Burnett, MD, an assistant professor of obstetrics and gynecology at Mayo Clinic in Rochester, Minnesota, who is not associated with the study, said a noninvasive test would have several advantages over the current methods of diagnosing endometriosis. Clinicians could detect and treat women earlier in the course of their disease. Secondly, a test with sufficient negative predictive value would be helpful in deciding whether to initiate treatment with hormones or other oral medications or go straight to surgery. “I would choose not to do a surgery if I knew with enough certainty that I wasn’t going to find anything,” said Burnett.

Lastly, a test that was quantitative and showed a response to treatment could be used as a disease activity marker to monitor the course of someone’s treatment.

But Burnett said more data on the approach are necessary. “This is a fairly small study, as it goes, for developing a screening test,” he said. “We need to see what its positive predictive value, negative predictive value, sensitivity, and specificity are in a bigger group.”

The road to a cure is even longer than the path to developing a screening test. Kommagani’s lab is now conducting more studies in mice to elucidate the pharmacokinetics and toxicology of 4HI before human trials can be attempted.

And as Burnett pointed out, although mouse models are great for experimentation and generating hypotheses, “We’ve seen way too many times in the past where something’s really exciting in a mouse model or a rat model or a monkey model, and it just doesn’t pan out in humans.”

Kommagani received funding from National Institutes of Health/Eunice Kennedy Shriver National Institute of Child Health and Human Development grants (R01HD102680, R01HD104813) and a Research Scholar Grant from the American Cancer Society. Burnett reported no financial conflicts of interest.

A former pediatrician and disease detective, Ann Thomas is a freelance science writer living in Portland, Oregon.