New Clues to an Old Mystery: Recent Gains in Endometriosis

In 1927, American gynecologist John Sampson published his theory of the etiology of endometriosis, postulating that retrograde flow of endometrial debris flows backward through the fallopian tubes during menses into the peritoneal cavity. Sampson's notion remains the main paradigm today, mentioned still in recent articles on the topic, but it has a flaw: Although the theory may account for how endometrial tissue escapes the uterus, a 1984 study revealed that this phenomenon occurs in 90% of women. Why, then, do only 10% of women suffer from endometriosis?

Endometriosis describes a condition in which endometrial tissue lining the uterus is found outside the uterus. The disease can be painful, even crippling. As many as 30% of women in their reproductive years who have endometriosis are infertile as a consequence. The hallmarks of the condition are superficial peritoneal lesions of varying color, cysts in the ovaries, deeper nodules accompanied by scarring and adhesion, primarily in the pelvis but sometimes appearing outside the pelvis. The syndrome can be challenging to identify, requiring laparoscopy for definitive diagnosis.

John Sampson aside, scientists have struggled for the past century to identify the cause, or causes, of endometriosis. Hormones clearly play a role in its development, and women with endometriosis have an elevated risk of clear-cell and endometrioid ovarian cancer and autoimmune diseases. Immunodeficiency also could be to blame, if a faulty immune system fails to find and remove endometrial tissue outside of the uterus. A class of chemicals known as endocrine disruptors have been linked to endometriosis, but not definitively. Twin studies have demonstrated that as many as 50% of cases have a genetic basis, while mice with surgically induced endometriosis have been found to have a higher ratio of harmful to beneficial bacteria in their gut.

Several studies published this year point to new insights into the old mystery — with possible implications for ways to treat the disorder.

Perhaps the most surprising came out earlier this month in Science Translational Medicine, as a team of researchers in Japan reported that invasive infection by bacteria of the genus Fusobacterium may cause at least some cases of endometriosis.

**Is Fusobacterium the New Helicobacter pylori?**

The researchers, from Nagoya University, are the first to suggest that not only might a single bacterial genus cause endometriosis, but that antibiotic treatment could prevent progression of the disease too. Using endometrial tissue obtained from 79 women undergoing hysterectomy for endometriosis and 76 women undergoing hysterectomy for other reasons (such as cervical cancer), the team started with gene expression profiling to explore differences between the two sets of samples.

They uncovered an interesting chain of cellular events: macrophages found in endometriotic lesions were secreting transforming growth factor-beta (TGF-beta). TGF-beta in turn stimulated high levels of expression of a gene called TAGLN in fibroblast cells from women with endometriosis but not in fibroblasts from women without endometriosis.

Turning on TAGLN transformed these previously inactive cells into active myofibroblasts, leading to increased proliferation, mobility, and attachment to mesothelial cells, the layer of cells that line body cavities and internal organs. In short, they identified some key players in an environment that seemed very favorable to the development of endometriosis.

Dr Yutaka Kondo

"So, the question is: Why are macrophages activated?" said Yutaka Kondo, MD, PhD, the senior author of the study and a professor in the division of cancer biology at the Nagoya University Graduate School of Medicine. "We think that there are always bacteria in the endometrium."

After reviewing data from a previously published study, they used quantitative polymerase chain reaction to rule out one candidate, Erysipelothrix, but scored on their next attempt, identifying Fusobacterium species in endometrial tissue from 64% of the women with endometriosis, compared with fewer than 10% of the controls.

To confirm that the bacteria could cause disease and were not simply bystanders, Kondo's team turned to a mouse model for endometriosis, in which endometrial cells are surgically removed from the uteri of mice and injected into the peritoneum of recipient mice, leading to the formation of endometriotic lesions. When mice received further injections of uterine tissue from mice that were infected with F nucleatum, their lesions were more numerous when compared with mice that received injections of uninfected uterine tissue. Furthermore, antibiotic treatment with metronidazole or chloramphenicol immediately after surgery largely prevented progression to endometriosis, Kondo and his colleagues reported.

Kondo likened this relationship between Fusobacterium and endometriosis to that of the link between Helicobacter pylori and peptic ulcers but acknowledged that he doesn't have all the answers.

"We need more clinical trials, and also we have to know what kind of treatment might be the most effective for the treatment of endometriosis," Kondo said, pointing out that other therapies should still be pursued in addition to antibiotics, as not all the samples from women with endometriosis harbored Fusobacterium. "It might be possible that other mechanisms are also involved."

Don't Write off Gut Microbiota

Ramakrishna Kommagani, PhD, associate professor of pathology and immunology at Baylor College of Medicine in Houston, agreed. "Endometriosis is a complex disease, which appears to be impacted by many factors, including genetic, epigenetic and environmental factors," Kommagani said.

In earlier work, Kommagani colleagues showed that broad-spectrum antibiotics could reduce disease progression in mice with a mouse model similar to that Kondo's group used.

Dr Ramakrishna Kommagani

A key difference between his work and Kondo's is his focus on gut microbiota, whereas the Japanese team looked at bacteria in the vagina and endometrium. But Kommagani said he thinks both could play a role. "Maybe the vaginal microbiome might have a direct impact on disease similar to what we showed on the gut," he said.

But he said at least part of the answer to why some women develop endometriosis may have to do more with the balance of beneficial and harmful bacteria in the gut rather than because of a single family of microbes like Fusobacterium.

Most recently, by dovetailing a mouse model for inducing endometriosis in mice treated with antibiotics to deplete their gut microbiome, Kommagani's lab expanded on its previous work: They showed that the animals developed fewer of the typical lesions seen in endometriosis than those that did not receive antibiotics before all of the mice underwent the surgical procedure used by researchers to induce endometriosis — possibly because they had no bacteria in their gut triggering the inflammatory response required for the development of endometriosis.

But after oral feedings with fecal matter from mice without endometriosis, the microbiota-depleted rodents began developing lesions typical of endometriosis, suggesting that altered gut flora from mice with endometriosis appeared to promote the disorder. Meanwhile, their microbiota-depleted counterparts who were fed fecal matter from mice without endometriosis did not develop the typical lesions.

Kommagani's team then compared metabolites from bacteria in stool from mice with and without endometriosis and investigated the in vitro effect of these metabolites on cells from human endometriotic lesions. One of them, quinic acid, increased the proliferation of human endometriotic epithelial cells.

"Some metabolites such as fiber-derived short-chain fatty acids have beneficial effects; they inhibit the disease," Kommagani said. "But maybe an amino acid derivative such as quinic acid, [may] promote disease, and these are generated because there is a gut dysbiosis."

This statement hints at some of the possible therapeutic approaches for endometriosis, such as a high-fiber diet to promote healthy gut flora, or perhaps antibiotics to eradicate unhealthy bacteria. But as with other conditions that have been linked to dysbiosis, like inflammatory bowel disease, use of antibiotics is a bit like balancing on a tightrope; although antibiotics may remove harmful bacteria, their use may negatively affect the beneficial bacteria.

Clues in Genetic Variants

Krina Zondervan, DPhil, professor and head of the Department of Reproductive and Genomic Epidemiology at the University of Oxford, England, focuses on genomic, molecular, and epidemiologic approaches to understanding endometriosis.

Dr Krina Zondervan

In her most recent paper, published earlier this year in Nature Genetics, Zondervan and colleagues started with a meta-analysis of 29 genome-wide association studies involving over 200,00 cases of endometriosis from Europe, Australia, the United States, and Japan. They identified 42 genetic regions linked to endometriosis, nearly four times the number previously identified.

"It's one thing identifying risk variants and the next question is, okay, well, what do those variants actually do in terms of biology?" Zondervan said. The Oxford team next explored how the identified genetic variants affect gene expression and the proteins generated, drawing on previously collected data on gene expression from samples of human blood and endometrial and uterine tissue.

They found many of the genes implicated in the risk for endometriosis code for proteins that affect sex hormones, uterine development, transformation of healthy cells into cancerous tissue, inflammatory adhesion molecules, and factors promoting development of new blood vessels. All of that, she said, explains how a few endometrial cells making their way into the pelvis can attach to ovaries, ligaments, and peritoneal surfaces; proliferate; and acquire a blood supply to ensure their survival.

"We were able to identify a whole host of things that were likely causal to the disease," Zondervan said. And that finding lead to her next question: "Are there particular genes or areas around them that can be targeted with certain medications?"

The surprising answer was that several of the genes linked to endometriosis share pathways with clinical syndromes that often occur in women with endometriosis. Many of these are chronic pain conditions — such as migraines, headaches, and back pain — but also include inflammatory illnesses such as asthma and osteoarthritis.

As Zondervan explained, "A lot of the variance that we see for endometriosis is also experienced for low back pain and migraine, and that clearly has something to do with pain perception and pain mechanisms."

A connection between the development of neural pathways and endometriosis has been proposed before, as researchers have found that endometriotic lesions can develop their own nerve supply, creating a direct interaction between the lesions and the central nervous system. And some clinicians have been employing treatment strategies that employ multimodal therapies — employing physical therapists, mental health practitioners, nutritionists, and pain specialists prior to and following surgical removal of lesions — to improve overall success rates of treatment.

But Zondervan's team is the first to uncover an important clue about how this happens.

The study findings also provide solid clues to researchers about which genes and proteins to focus on for drug target discovery. In particular, the gene pathways shared by endometriosis and various pain conditions could allow for repurposing of drugs developed for other conditions for treating endometriosis, reported Zondervan.

Zondervan's other important conclusion, echoed by Kondo and Kommagani, is that endometriosis is not one disease. Rather, it appears to be akin to cancer in terms of the heterogeneity of how it presents and the different subtypes of diseases. The Oxford study corroborated this belief, identifying certain genes that were closely associated with cystic lesions in ovaries, but failing to turn up a genetic link to other types of lesions in the pelvis long considered to be part of the spectrum of endometriosis disease.

Zondervan agreed that the potential link between Fusobacterium is a fascinating area given the critical role of inflammation in the pathogenesis of endometriosis, although she'd like to see the work replicated with larger sample sizes. "From a personal point of view, I'd be really fascinated to see how genetics interplays with this," she added.

What's Next?

The chief limitation of human studies looking at mechanisms of endometriosis is that they are correlational: Tissue samples are collected from women with and without endometriosis, often through an invasive procedure such as laparoscopy or biopsy, at one point in time. If a researcher identifies a factor that is more common in women with endometriosis — a particular bacterium or environmental exposure — proving causality is difficult. Currently, the best tools for proving causation are animal models of endometriosis, such as the those used by Kondo's and Kommagani's teams.

Better diagnostic tools would solve that problem. The ultimate goal is a noninvasive test for endometriosis that would allow clinicians to follow women over time and permit the monitoring of disease progression, or regression, without the need for painful procedures. Such a diagnostic tool would facilitate rigorous longitudinal studies evaluating mechanisms of disease, as well as monitoring outcomes of clinical trials of new treatments.

Could stool samples be the answer?

The Japanese team found that women harboring Fusobacterium in endometrial tissue also had Fusobacterium in vaginal samples taken at the time of their hysterectomy — and stool samples can pick changes in the gut microbiome.

"Vaginal swab or stool tests are probably the best and easiest for noninvasive early detection," Kommagani said.

Spit tests for DNA would be even easier to obtain. Polygenic risk scores could be developed to estimate an individual's risk of disease based on the number of variants, but Zondervan cautioned that not all the genes that account for endometriosis are known.

"The things that we found altogether explain about 5% of disease variability, basically — which is still not an awful lot," she said.

Kondo's work was supported by the Grant-in-Aid for Scientific Research, the Japan Society for the Promotion of Science, and the Research Grant of the Princess Takamatsu Cancer Research Fund. A patent method for detecting bacteria of genus Fusobacterium in order to diagnose endometriosis (WO2023/ 042714), was submitted (international publication date, March 23, 2023).

Kommagani's work was funded, in part, by National Institutes of Health/National Institute of Child Health and Human Development grants R01HD102680, R01HD065435, and R00HD080742. He has no other conflicts of interest. Zondervan received funding from The Wellcome Trust (216767; 104036; 084766; 212904; 076113 and 085475) and also reported grants from Bayer AG, AbbVie Inc, Volition Rx, MDNA Life Sciences, and Roche Diagnostics Inc. outside the submitted work.

Ann Thomas, MD, MPH, is a pediatrician and epidemiologist living in Portland, Oregon.

For more news, follow Medscape on Facebook, Twitter, Instagram, YouTube, and LinkedIn

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.