Improper stratification obscures true extent of risk for specific subtypes of endometriosis:

Nearly 120 years ago, pioneering endometriosis specialists, like Dr. Thomas Cullen, were already thinking about endometriosis subtype risk stratification concepts after noticing a potential association between histologically atypical endometriosis* and an elevated incidence of certain cancers. By 1925, Sampson expanded on these early investigations by identifying three of the four criteria we recognize today for the classification of endometriosis-associated malignancies.

Despite this century plus headstart in at least recognizing the oncogenic potential of endometriosis, progress in understanding these connections has generally been woefully behind other fields, due largely to decades of discriminatory healthcare policies that have left the field of endometriosis research scandalously under-funded.

To no one’s surprise, these historically iniquitous funding policies have translated to real world deleterious consequences for millions of patients worldwide. One such negative impact involves the systematic underestimation of cancer risk in endometriosis patients. The subject of underestimations made the news recently, when Barnard et al (JAMA, 2024) reported that endometriosis conferred an overall estimated 4.2-fold increased risk for developing any form of ovarian cancer, compared to those without endometriosis, whose relative lifetime odds have been cited by numerous studies as 1.31.

However, when risk stratifying by both macrophenotypic subtypes of endometriosis and ovarian cancer histotypes, even stronger associations were found, including a 19-fold (18.96) elevated association with type I ovarian cancers in cases of deep infiltrating endometriosis and/or ovarian endometriomas subtypes.

Put starkly, this is a striking reversal of much of the peer-reviewed literature, even from just a few years ago, which suggested odds ratios for developing ovarian cancer as low as 1.8 for those with endometriosis. Although follow-up studies are, of course, needed, naturally such concerning new reports should compel a full reevaluation of current patient guidelines, screening protocols, and public health policies, to ensure both patients and healthcare providers are provided with the most up-to-date information to help guide treatment decisions and identify potential preventive measures.

FLAWED RISK STRATIFICATION
The historical under-reporting of cancer risk in people with endometriosis involves many variables, but for this blog we’ll narrow our focus to addressing flawed risk stratification methods that have contributed to problematic calculations in the past.

Setting aside a small minority of exceptions, it has only been very recently that endometriosis researchers, like Barnard et al, have begun to utilize proper risk stratification techniques to assess cancer risk by endometriosis subtype. Instead, many studies often combined all subtypes of endometriosis together, which can obscure true risk, resulting in significant underestimations.

When other investigators over the years suggested that the incidence may be higher after all, many leading authorities offered dulcet reassurances to help allay any growing concerns that the established literature may not be accurate.

For example, a 2017 article in The Lancet concluded that, “…the lifetime risk of ovarian cancer among women with endometriosis is 1·80%—fewer than two women in 100… This small increase in absolute risk should thus reassure women with endometriosis that their lifetime ovarian cancer risk is quite low and, in absolute risk scales, is only negligibly different from women without endometriosis.”

As many may already know, subtype/phenotype risk stratification is the gold standard in oncology, as it helps researchers and physicians determine prognoses and identify the most effective treatments based on the type of biopsied cancer tumor. Even though endometriosis is not cancer, the same gold standard tissue analyses utilized in oncology must be applied to endometriosis research if we ever hope to achieve clinically meaningful and translational research to help improve the health outcomes for millions of people around the world.

This is especially true, considering that an estimated 10%-19% of epithelial ovarian cancer subtypes are associated with endometriosis (Kumar et al). As well, with ovarian cancer still the deadliest gynecologist cancer in the U.S., with an estimated 5-year survival rate for late-stage cases of only 30.8%*, according to the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) Program, these new findings underscore the urgent need to recognize this overlooked public health issue as a national and global priority.

*(Early stage ovarian cancer has a high 5-year rate of survivability of 93.1%, but unfortunately, most cases are not diagnosed until later stages with poor prognosis).

HISTORICAL VIEWS ON ENDOMETRIOSIS AND OVARIAN CANCER
With cancer being one of the most potentially distressing subjects in medicine, it wouldn’t be right to simply state that a heightened susceptibility exists, without providing more context. Therefore, before diving further into the subject, we thought a brief review of the literature on endometriosis-associated malignancies would help put any new information into perspective.

Though the following is not a comprehensive analysis by any means, we hope the summary of key findings from approximately the past century will help provide an insightful historical framework of our still-evolving understanding of endometriosis and its oncogenic potential, as well as equip patients with updated knowledge about potential risk factors to watch out for, as well as guidance on treatment options, potential preventive measures, and suggested screening protocols.

As noted, an association between endometriosis and certain cancers has been reported for well over a century. In fact, early researchers often misidentified some severe stages of endometriosis as cancer. Even Carl von Rokitansky, the renowned pathologist credited as the first to microscopically identify endometriosis (and adenomyosis)* in 1860, initially misclassified some of the biopsied samples as neoplasms. As a result, Rokitansky originally named these disease entities following the scientific nomenclature to indicate malignancy:  cystosarcoma adenoids uterinum polyposum and ovarian cystosarcoma.

Many are surprised to learn of the renowned pathologist’s initial misclassifications. Yet, as patients of today will attest, misdiagnoses like this still occur routinely now, sometimes leading to overly aggressive and unnecessary surgeries that can irreparably compromise fertility.

By the time Dr. John Sampson officially coined the term ‘endometriosis‘ in 1925, the medical consensus had shifted, largely rejecting Rokitansky’s initial sarcoma classification. Endometriosis was instead recognized as a benign condition with invasive potential and the ability to undergo malignant transformation, although its incidence remained unclear.

The field of study advanced considerably when Sampson later proposed the following three of the four criteria we recognize today for the classification of an endometriosis-associated malignancy: 1) evidence of endometriosis in close proximity to the tumor; 2) exclusion of metastatic invasion from another tumor source; and, 3) presence of tissue resembling endometrial stroma surrounding epithelial-like glands. Thirty years later, in 1953, Dr. RB Scott proposed a fourth criterion: “histological proof of transition from benign changes in endometriosis to malignant changes in cancer.” Currently,  all four criteria are still used to define an endometriosis-associated malignancy (EAM).

*(NOTE: Today, we would view some of Rokitansky’s subtypes as adenomyosis. However, during his era and up until fairly recently, these two distinct conditions were often referred to as endometriosis interna and endometriosis externa, after Sampson coined these terms in an effort to unify the theories and consolidate the nomenclature). 

MODERN-DAY KNOWLEDGE GAPS
Now that we are two decades into the 21st Century, the observed association between endometriosis and certain cancers has been well-established in the peer-reviewed literature, with leading endometriosis researchers widely recognizing that endometriosis increases the risk of developing certain cancers, although the exact magnitude of susceptibility and many other critical details continue to be the subject of intense academic debate.

While it’s now generally accepted that any form of endometriosis has the potential for malignant transformation, over the years certain subtypes have been recognized as conferring an increased chance of malignancy.  As researchers from a century ago already suspected, histologically atypical endometriotic lesions are indeed associated with an increased risk of malignant transformation. However, modern researchers now generally classify this type of endometriosis as precursor (premalignant) lesions as well. Endometriosis-associated malignancy phenotypes have also now been recognized, including endometriosis-associated ovarian cancer (EAOC), which includes ovarian clear cell carcinoma and ovarian endometrioid adenocarcinoma, the two histotypes most closely associated with endometriosis.

Given these generally accepted understandings, that different types of lesions confer variable risk, it’s disheartening to see that much of the peer-reviewed literature on endometriosis has, unfortunately, hasn’t always been done in past research during time periods when subtypes had been recognized.

EARLY SIGNS THAT CANCER RISK WAS HIGHER THAN BEING REPORTED
Early investigators throughout the 1980s-1990s were coming very close to uncovering the possibility of higher-than-reported incidences of ovarian cancer in endometriosis patients. In 1980, Mostoufizadeh & Scully were among the earliest in the modern era to suggest a higher incidence of ovarian cancer in endometriosis than was reported at the time.  Dr. Prof. Ron Batt was also an early proponent, who also contributed substantially to the history of Rokitansky’s work and other important historical findings.

Later, in 1993, Vercellini reported a high frequency of concomitant endometriosis in patients undergoing surgery for ovarian cancer, which indicated that the association was likely much stronger than was being reported at the time. In histotypes we now refer to as endometriosis-associated ovarian cancers, endometriosis was found in “26.3%, 21.1%, and 22.2%, respectively, in endometrioid, clear cell, and mixed subtypes.” In 1997, Fukunaga et al reported similar findings, observing that 24.1% of ovarian cancers in the study were associated with ovarian endometriosis. These are just a sampling of early investigations. Unfortunately, though, these and subsequent studies suggestive of stronger assocations seemed to have been perceived as anomolous outliers by the majority of academic medicine at the time.

Undettered, a new generation of researchers thankfully pursued the early leads. In 2018, Saavalainen and colleagues published Risk of Gynecologic Cancer According to the Type of Endometriosis, which was among the earliest studies to stratify ovarian cancer susceptibility by endometriosis subtype. When all forms of endometriosis were aggregated together to determine the incidence of all subtypes of ovarian cancer, the standardized incidence ratio was 1.76 (compared to those in the study without endometriosis). However, when Saavalainen et al stratified by the subtype of ovarian endometriomas, the probability of developing a particular histotype of ovarian cancer (clear cell) increased to 10.1.

In 2020, Murakami et al were also among the few to properly risk stratify to determine cancer risks in patients with endometriosis, Based on a large study cohort of 6,398 and follow-up of 12.8 years, Murakami and associates reported similar to Saavalainen, suggesting patients with endometriomas were 9 times more likely to develop ovarian cancer than healthy controls.

LANDMARK STUDY ON ENDOMETRIOSIS SUBTYPES AND CANCER RISK
It would take the landmark study by Mollie Barnard, Leslie Farland, and Bin Yan, (Endometriosis Typology and Ovarian Cancer Risk. Barnard, Farland, Yan, 2024, JAMA) to reawaken the world to potential missed connections by conducting a well-designed, risk-stratified analysis by both endometriosis subtype and ovarian cancer histotype, utilizing a large, single-state population cohort of 450,906 patients (including 78,476 with endometriosis and 372,430 matched controls).

As with the Saavalainen et al & Murakami et al studies, Barnard and colleagues also uncovered markedly elevated risk profiles for deep infiltrating (DIE) and/or ovarian endometriomas, which were 9.66-fold more likely to develop certain histotypes of ovarian cancer, compared to those without endometriosis. However, as noted earlier, the magnitude of these associations for DIE and/or endometriomas increased to an 18.96-fold elevated probability of developing specifically type 1 ovarian cancers. (See post-production notes below for important caveats).

BARNARD ET AL STUDY LIMITATIONS
As the authors noted, their study does have limitations, most notably relating to issues of potential misclassification of endometriosis subtypes. Without a reliable blood biomarker for endometriosis, let alone its many subtypes, this meant that the study relied on diagnostic methods of varying reliability throughout the duration of the study period (ie, imaging alone to diagnose vs the gold standard of video laparoscopy with biopsy).

These diagnostic constraints may help explain why only 1.7% of the study participants with endometriosis were diagnosed with DIE and/or ovarian endometrioma/s, when published prevalence rates are generally much higher. Just as one example, ovarian involvement, including endometrioma/s, is believed to occur in as many as 50-70% of all cases of endometriosis. As such, this would mean that the majority of endometriosis patients would fall into the study’s category linked to among the highest observed risks for developing certain histotypes of ovarian cancer.

Many epidemiologists and other public health thought leaders would also be quick to point out that the excess risks identified in the Barnard et al study are considered small, representing about 4-20 additional cases per 10,000 people in the affected population.  However, with approximately 200 million endometriosis sufferers worldwide, this means that the absolute number of individuals potentially impacted is likely not insubstantial in the least.

In fact, as one analyst noted, the strength of the observed associations “compares to the connection between smoking and lung cancer.”  The study authors even noted that further statistical analyses to test for bias actually indicated that the true associations between endometriosis and ovarian cancer may be stronger than what was observed in this cohort.

No magical age cutoffs, please
Another significant limitation includes the study’s truncated age range of 18-55. While it’s common for epidemiologists to limit age ranges to help manage large datasets, in the field of endometriosis research, this custom unfortunately sets us down a continued path of producing misleading data as a result of leaving out substantial populations of affected individuals. After all, endometriosis does not suddenly begin at precisely the age of 18, but in fact occurs in a large percentage (if not the majority) much earlier, including in pre-teen years. A small subset also experience premenarchal onset.

Similarly, endometriosis does not magically retire at age 55, but can continue to cause severe disability, serious, life-threatening comorbidities, and other serious health consequences throughtout the entire lifespan of those affected. Even in those who’ve had years of remission may be susceptible to postmenopausal reactivation of their endometriosis, particularly after exposure to some forms of hormone replacement therapy, as just one example found in the literature. Other cases of postmenopausal recurrence seem to have no clear triggering factor, but seem to arise out of nowhere.

Endometriosis arising in menopause is also associated with an elevated potential for malignant transformation, and so it’s frustrating to see this particularly vulnerable population routinely excluded from studies. Though data are limited, it’s possible that postmenopausal endometriosis may also be linked to more severe sequelae, including bowel obstructions that require surgical intervention.

The etiology of menopausal status matters 
Another factor to bear in mind is that the menopausal status of patients itself must also be carefully risk stratified, as the different etiologies involved confer different risks. For example, studies have shown that people with endometriosis who have undergone surgical menopause at an early age have more severe health consequences, including cancer risk, than those without endometriosis who experience natural menopause.

Even in those without endometriosis, the risk of ovarian cancer increases as populations become older, including after the age of 55.  As such, the age cutoff of 55 in this study introduces an additional bias that would result, generally, in a potential under-estimate of ovarian cancer risk.

Longer follow-up needed with endometriosis patients
As well, the study’s follow-up time of 12 years also introduces potential blindspots in the analysis. For example, some forms of iatrogenic scar endometriosis have been known to undergo malignant transformation more than 20 years after the index surgery. As such, it’s important that both researchers and patients are aware of the unique risk patterns that have been observed with some forms of endometriosis that often are excluded from the three main risk stratified subtypes.

As the author’s noted, another notable drawback was that they were unable to include in their analyses the exposure status of two of the most commonly prescribed medications for endometriosis patients: oral contraceptive pills and gonadotropic-releasing hormone analogues (GnRH). As a result, this particular study was not able to offer any new insights about how these pharmaceutical interventions may affect health outcomes or ovarian cancer risk.

The study was also constrained by the use of a database involving only the medical records of Utah residents. While single-state studies can be excellent for elucidating some associations, in the case of endometriosis research, it become problematic, considering that many patients travel outside of their resident states to receive treatment.  As a result, there may be important medical information that is not fully captured.  (This may help explain why, again, the incidence of DIE and/or endometriomas was lower than expected).

Lesion heterogeneity, plus anatomic location matters
For complex chronic diseases like endometriosis, risk stratification techniques also need to extend beyond just the three main macrophenotypic subtypes under investigation. With endometriosis known for its tremendous lesion heterogeneity, this speaks to potentially diverse molecular and genomic phenotypes that need to be investigated.

Even from a macro standpoint, other variables, such as anatomic location, lymphovascular involvement, fibrotic and adhesive status, recurrence patterns, and dozens of other characteristics, have been linked to potentially very different etiologies and clinical outcomes, including oncogenic potential.

For example, although the Barnard et al study did risk stratify DIE, the anatomical location/s of the DIE was not included in the analysis. As a result, this insufficient stratification may obscure the true nature of risks associated with specific forms of DIE. In deep infiltrating endometriosis of the rectosigmoid, for example, some studies have shown that lymphovascular invasion, a known red flag for heightened malignancy risk, actually occurs frequently, or at least more frequently than previously reported. As such, future studies should stratify for not only DIE anatomic locations, but also lymphovascular involvement, to help further elucidate the strength of these observed associations.

Extra-pelvic deep infiltrating endometriosis was also not tracked in this study, which means we remain largely in the dark about any associations it may have with ovarian or other malignancies as well.

Ovarian endometrioma sub-classifications
Ovarian endometriomas and DIE should also be risk stratified separately in future studies for the same reason, especially considering that there are at least four additional ovarian endometrioma histologic sub-classifications, which were first reported by Nezhat et al in 1992: Type 1, Type II, Type IIIa, and Type IIIb. When future studies expand to include modern molecular and genomic analyses, it’s likely that many phenotypically unique sub-classifications of ovarian endometriomas will eventually be identified. As such, modern endometriosis researchers must keep up with the fast-moving field of omics and be prepared to adjust their risk stratification methodologies accordingly.

Atypical Endometriosis linked to elevated malignant transformation
This is already a daunting list of challenges, yet we haven’t even scratched the surface concerning histolic differences that should be included in risk stratified analyses, as modern immunohistochemical tests have the potential to detect early forms of endometriotic tissue that may be at risk for undergoing malignant transformation.

In the case of atypical endometriosis (AE), as mentioned earlier, pathologists from more than a century ago recognized histologic features that suggested higher rates of potential malignant transformation.  Though debates are ongoing (see this 1979 article, for example), a majority of specialists today now generally classify atypical endometriosis as a precursor lesion (precursor to malignancy).

There’s a great deal of complexity on the topic of atypical endometriosis, but generally pathologists recognize two distinct histologic patterns: “cellular atypia, also known as cytologic atypia, and architectural atypia, commonly referred to as hyperplasia.”  The abnormal cytologic features AE may potentially exhibit includes cell nucleus atypia*, architectural atypia (hyperplasia), metaplasia, dysplasia, loss of heterozygosity, abnormal cell morphology, and many other aberrant characteristics that pathologists look for to determine whether a biopsied tissue sample may suggest potential or actual malignancy.

*(Note: Not all pathologists agree that atypical endometriosis should be classified as a precursor lesion, with some citing the fact that, while architectural atypia is common, the classical nucleus atypia found in most malignant tissue is not always discernible). 

Why is atypical endometriosis often excluded from ovarian cancer studies?
Despite this known variant of endometriosis, which has been linked to significantly elevated risk for malignant transformation for more than a century, it’s surprising to see just how many studies continue to exclude atypical endometriosis from subtype risk stratification. The root cause, though, is often beyond the researchers control (at this time), as, unfortunately, the previously limited ICD codes made it difficult to document different forms of endometriosis until approximately January 2022, when updated codes were finally introduced. (Not to mention the issue of having just one CPT code for endometriosis). This means that the electronic medical records large databases, which so many researchers rely on, often do not include these crucial distinctions.

Even with new coding, the other depressing reality is that not all biopsied endometriotic tissue samples are actually thoroughly checked by pathologists, partly due to many pathologists being unfamiliar with endometriosis protocols & partly due to economic pressures that result in severely under-staffed pathology departments.

As you can see then, activist-minded individuals are needed more than ever, as it will take continued pressure by the global community of endometriosis advocates to help drive forward wholesale restructuring in these deeply entrenched health systems that have been so resistant to change. (You can click here to become an EndoMarch Member & stay in the loop on upcoming campaigns to address these issues).

Not a comprehensive list of study limitations
This is not a comprehensive list of the study’s limitations, but hopefully this brief analysis will give just a glimpse of how concerning knowledge gaps may be potentially addressed in future studies. Although the many pathogenic mechanisms involved in both endometriosis and ovarian cancer remain poorly understood at this time, it’s at least encouraging to see so many new scientific teams, like Barnard et al, prioritizing endometriosis research and applying more robust risk stratification techniques.

CURRENT PATIENT GUIDELINES ARE BASED ON INACCURATE DATA
While this is just one study, with notable limitations, the Barnard et al 2024 findings (along with similar earlier reports) call into question peer-reviewed literature that stood nearly unchallenged for decades. By extension, this means that some of today’s leading patient guidelines are potentially incredibly misguided, as they have been based on the same flawed assumption, that endometriosis patients have only a slightly higher risk of ovarian cancer than the general population. 

As you can see, we are witnessing, in real time, what decades of discriminatory public health under-funding has done to the field of endometriosis research. The harmful impact includes potentially or actual catastrophic outcomes for millions of patients, who may not have been given critical information about necessary screening, potential preventive measures, and follow-up care, which would have been advised if the true extent of potential risk had been better understood all these years.

The overall lifetime chance of developing ovarian cancer is still much lower than other gynecologic cancers, like breast cancer. Nevertheless, we believe endometriosis patients deserve to be prioritized for a change, especially after centuries of historic neglect and entrenched barriers to quality care have left millions around the world in severe poor health, with increased risks of other serious comorbidities (including ovarian cancer), which perhaps could have been prevented had patients been taken seriously and adequate public health investments been made decades ago.

SYSTEM-WIDE BARRIERS TO BETTER RESEARCH NEED TO BE ADDRESSED
Of course, it’s important to point out that researchers may not have been able to conduct proper subtype stratification until recently. Many factors have contributed to this conundrum, ranging from a lack of consensus on endometriosis subtypes themselves, to issues with surgical biopsy tissue handling, variances in diagnostic modalities, and inconsistent use of more advanced immunohistochemical staining techniques.

Efforts to categorize different subtypes found at surgery have also been severely compromised by the extremely limited ICD codes, which, until recently, only allowed for a handful of coding options. While recent updates reflected in ICD-11 now offer over 100 different options, we are still stuck with just one CPT code for endometriosis: 58662.

As is evident from this endless morass of issues, it’s clear these issues will only compound into further disarray, with patients continuing to pay the ultimate price, unless there are urgent federal-level policy interventions.  As well, these entrenched systemic deficiencies will not be automatically solved with just increases in research funding. Rather, advocacy groups and thought leaders in the field of endometriosis research must work together to help build greater consensus and develop a plan to address these sorts of long-standing inequities that are deeply embedded within every health system worldwide.

CIRCLING BACK: A BRIEF REVIEW OF ENDOMETRIOSIS SUBTYPES
Following the gold standard of risk stratification modeling is crucial if we are to obtain clinically meaningful and translational research. This is especially true in cases of complex chronic diseases like endometriosis, which present with substantial heterogeneity involving multiple phenotypes/subtypes and multifactorial potential for pathogenesis.

Moreover, significant lesion heterogeneity exists not only between patients, but even within the same patient and even with the same tissue samples (ie, neoplasms with mixed genomic, molecular, and histologic characteristics).

Despite these tremendous variations, researchers have been able to identify three main macrophenotypic subtypes of endometriosis:

1) Superficial peritoneal lesions;
2) Ovarian endometriomas; and
3) Deep infiltrating endometriosis (DIE).

SUBTYPE LIMITATIONS, CONTROVERSIES AND CLINICAL IMPORTANCE 
For a more detailed overview of the currently recognized endometriosis subtypes, you can check out this 2023 article by Imperiele et al, which is paywall-free and provides an excellent summary.
It terms of key developments in the field of endometriosis research, it has certainly been a critical milestone to have come to somewhat of a consensus and recognize these three subtypes, especially in terms of guiding optimal treatment protocols.

For example, it’s been recognized for decades that symptomatic deep infiltrating endometriosis tends to be resistant to hormonal medications, making surgical intervention often the better option for alleviating symptoms.  The association of ovarian endometriomas with more severe disease, as postulated by Sampson nearly a hundred years ago, has also been consistently observed by modern specialists.

However, a few important caveats are in order. First, there is not actually full universal consensus that these three main subtypes actually represent different molecular or genetic phenotypes of endometriosis. Instead, some suggest they may merely represent different developmental stages of endometriosis. Studies demonstrating clonal relatedness of anatomically diverse endometriotic lesions provide additional evidence suggesting the deficiency of the current anatomically-based classification system.

There is also extremely limited or inconclusive research on the molecular profiles of these (mostly anatomic) subtypes, making it difficult to draw conclusions concerning etiologic origin and clinical outcomes from a modern, molecular medicine-based point of view. As well, the list leaves out other clinical variants which some researchers believe should be recognized as subtypes, including polypoid endometriosis, histologically atypical endometriosis, iatrogenic endometriosis, and abdominal wall endometriosis.

Intra-subtype variants have been recognized as well, including several subtypes of ovarian endometriosis which were identified in the 1980s by Nezhat et al. As such, the potential for substantial intra-subtype variation is an area of research that warrants further evaluation.

Given so many potential unknowns and competing theories, it’s crucial to keep in mind that we are truly only in our infancy in terms of understanding endometriosis, which may be better understood as potentially many disease phenotypes under one umbrella term. This is especially apparent when compared to the field of oncology, where advanced genetic and molecular mapping has led to the discovery of dozens of different tumor phenotypes with distinct etiologic origins and clinical behaviors, which has allowed researchers to identify more personalized treatment options with improved clinical outcomes.

DESPITE LIMITS, SUBTYPES KEY TO LANDMARK FINDINGS ON CANCER RISK
While there is ongoing debate about whether the three currently recognized endometriosis subtypes truly represent variant phenotypes, the clinical usefulness of these different disease manifestations has been made abundantly clear by Barnard et al’s seminal study, demonstrating the substantially heightened risk of ovarian cancer associated with specific forms of endometriosis.

OTHER RISK FACTORS
As mentioned earlier, other important clinical differences associated with various subtypes have also been observed for decades, including in symptomology, rate of disease progression, prognoses, and differential reactions to commonly prescribed medications for endometriosis.

Although the majority of cases of endometriosis generally fall into the three main subtypes listed above, this doesn’t mean that the risk profiles of the lesser known subtypes of endometriosis should be ignored. For example, several studies suggest that the risk for malignant transformation of certain forms of putatively iatrogenic endometriosis affecting old surgical scars appears to be higher than other types. Moreover, surgical scar endometriosis may be at risk of malignant transformation for nearly two decades after the original surgery. As such, patients should be made aware of these potential attenuated risks and encouraged to report any concerning changes to their surgical scars, such as pain, swelling, bleeding, or oozing, even if these symptoms arise many years later.

In cases involving ovarian endometriomas that repeatedly recur, the risk of malignant transformation is also elevated, compared to those with fewer recurrences. Other factors that raise the risk of malignancy in ovarian endometriomas include cyst complexity, such as solid parts mixed with fluid, multilocularity, fast growing cysts, and/or larger size masses, specifically those which are 9 centimeters or larger. (See below for postmenopausal onset of ovarian endometriomas as another potential red flag). Histologically atypical endometriotic lesions also confer a higher risk of malignant transformation, with some studies suggesting that up to 10% turn into cancer.

ENDOMETRIOMAS AFTER MENOPAUSE AT HIGHER RISK
As for patient demographics associated with elevated susceptibility to ovarian malignancies, studies also indicate that ovarian cancer risk is significantly increased in patients with postmenopausal onset or recurrence of endometriosis in general, but especially if the subtype is ovarian endometrioma/s. As well, an increased incidence of ovarian cancer has been observed in patients with both endometriosis and infertility – particularly with null parity – and this risk extends into menopause.

Since many researchers actually sometimes exclude postmenopausal patients from their studies (age cutoffs often end at age 55 when working with large databases), this continues to be a glaring knowledge gap that urgently needs attention and proper risk stratification for future studies. Though this particularly neglected cohort of endometriosis sufferers was not covered in the Barnard et al 2024 landmark study, a cursory review of recent literature suggests that substantial under-estimates of ovarian cancer risk may be also occurring in this historically marginalized cohort of older postmenopausal endometriosis patients.

Endometriosis sufferers have already been through so much; it’s outrageous to think that, as many of us approach our menopausal era, we again find ourselves subjected to still more instances of medical neglect.

And so, we’d like to spell it out plainly here: People with a current or prior history of endometriosis, and who are in or approaching menopause, deserve to have their potential health risks taken seriously, especially in light of emerging research on the increased ovarian cancer risk associated with endometriosis.  This means that, at the very least, screening advisories may need to be reevaluated and patients and their healthcare providers should meet to discuss these issues at length, so that all aspects of ongoing care can be thoroughly reassessed.

ESTROGEN AND PROGESTERONE RECEPTOR MUTATIONS 
Endometriotic lesions also display substantial heterogeneity in estrogen and progesterone receptor alterations, which may influence not only cancer risk, but also play a role in the vast differences we see in response to medications, disease severity, and prognosis. In oncology, risk stratification that includes steroid hormone receptor status is standard practice and has led to improved patient outcomes. It is vital for future endometriosis research to incorporate these similar receptor variances into standard stratification protocols, so that we can better predict high risk subtypes and identify safer and more efficacious treatment protocols.

GENETIC COUNSELING FOR ENDOMETRIOSIS PATIENTS NEEDED
Endometriosis research is also substantially lagging in counseling patients on gene variants associated with increased susceptibility to cancer. For example, many don’t realize that the well-known BRCA1 and BRCA2 genes associated with more aggressive forms of breast cancer are also associated with an elevated risk of endometriosis-associated ovarian cancers.

PREVENTIVE OPTIONS
While there is still no method to prevent endometriosis, the best we have is gold standard care, including early detection and early intervention. Unfortunately, millions around the world cannot even obtain a diagnosis for up to 7 years on average (and up to 10-11 years in the U.S.). While patient advocates around the world are fighting valiantly to change these appalling statistics, for those who suspect they may have endometriosis or adenomyosis, a good first step toward assessing your risk would be to take the free online Nezhat Endometriosis Advisor test, an AI-assisted screening technology developed by Dr. Camran Nezhat, one of the world’s most respected endometriosis specialists & Founder of Worldwide EndoMarch and The American and Global College of Endometriosis Subspecialists. To date, the Nezhat Endometriosis Advisor App test has demonstrated up to 90% accuracy in predicting the presence of endometriosis/adenomyosis.

Preventive therapies for ovarian cancer require very careful assessment by your medical provider, preferably including access to genetic counseling and second opinions. This subject is outside the scope of this blog, but we would do our best to provide referrals to specialists in this area, if you would like.

GOLD STANDARD TISSUE HANDLING, CODING, AND RESEARCH MATTERS     
Given the breadth and depth of potential risk variables, this is why it’s so crucial that surgeons, pathologists, and all researchers carefully distinguish between different lesions, not just in terms of where they were located in the body, but also with respect to lesion color, lesion depth, adhesiveness. fibrotic status, vascularity, aberrant decidualization of endometriotic lesions, and lymph node involvement, just to name a few characteristics. As well, careful detailing of clinical data is needed, such as the reported symptoms, patient age, age of menarche, age of symptom onset, menopausal status of the patient, surgical history, stage of disease, menstrual cycle phase, and whether the tissue was exposed to hormonal or other medications, just to name a few important variables that need to be considered in risk-adjusted models. All of this may seem obvious to many with a medical background. However, as endometriosis patients from around the world already know, it’s actually very common for many of these incredibly crucial details to be completely omitted in their medical records, even in cases of surgery.

For experienced endometriosis specialists and researchers, these standards are of course well-known. Unfortunately, the vast majority of patients are not able to seek care from experienced specialists. As a result, the patient data that end up being input into medical records databases continues to be incomplete and/or inaccurate. Without policy changes, this means that many serious knowledge gaps will persist, as researchers rely on large cohort databases derived from medical records to identify important epidemiological insights especially.

CALL TO ACTION – FOR MULTIPLE INSTITUTES !
The time has come to end the trivialization of endometriosis, including in terms of research standards. As you can see, the complexity of endometriosis naturally requires a more fastidious approach to research, beginning with tissue handling, but continuing with applying the highest standards to peer-reviewed publications, which exert such significant influence on clinical decision-making that directly impact patients.

These issues of tissue handling and proper risk stratification are so critical, that there’s even a global initiative of the Endometriosis Phenome and Biobanking Harmonisation Project). However, it’s clear renewed efforts are needed to help drive forward more consistent standardization in this crucial domain.

This is why our team is collaborating with endometriosis researchers from around the world, to help develop new endometriosis research protocols for all future endometriosis studies. Without proper stratification, we will miss important clues that affect treatment efficacy, prognosis, and the assessment of risk factors for cancer. The adoption of oncological principles, from surgical debulking with wide margins, to risk stratification, to tissue handling harmonization, are just a few of the many needed changes that would make a tremendous difference. In other words, gold standard care cannot be achieved if gold standard research methodologies are not followed. As such, we urge journal editors-in-chief and review boards to consider updating their standards for all future endometriosis studies. Otherwise, we’ll continue to see severely methodologically flawed endometriosis research published year after year, which, in turn, contributes to flawed clinical decision-making that can adversely impact patient outcomes.

We also urge the U.S. Preventive Services Task Force (USPSTF), as well as medical societies which publish endometriosis patient guidelines, to update their guidelines in light of these new data.

It’s imperative that predictive algorithms currently used in clinical settings, such as the ROMA test (the risk of ovarian malignancy algorithm) also be updated in view of the emerging research on endometriosis and ovarian cancer. At the present time, the ROMA test includes cancer antigen 125 (CA125), human epididymal protein 4 (HE4), and menopausal status to assign patients with adnexal masses into a high-risk or low-risk group, to help clinicians differentiate between benign cases versus epithelial ovarian masses. While the predictive accuracy of the ROMA test is actually quite limited, nevertheless many clinicians around the world utilize these tests to quickly screen their patients. Given the recent findings confirming an even greater risk than previously reported, it would be an opportune time to update the ROMA test (and other similar screening tools) to finally and officially include endometriosis, what is now one of the most well-recognized factors associated with up to 10-19% of all ovarian cancers.

Speaking of screening options, we are still without an effective ovarian cancer blood test, despite decades of valiant efforts by researchers around the world. In this era of AI and billion-dollar missions to Mars, surely we can join together to help expedite technological breakthroughs in these critical areas where progress has been stalled for years.

As for certain patient guidelines from a few years ago, which hit the endometriosis community like a wrecking ball when they claimed endometriosis has no association with increased cancer risks, we hope those involved in promulgating such potentially harmful policies will consider revising their guidelines in view of the preponderant, statistically significant, well-powered peer-reviewed evidence identifying endometriosis as a risk factor for developing one of the world’s most deadly gynecologic cancers.

POST-SCRIPT IMPORTANT NOTES
After posting this article on social media, we wanted to just follow-up with a few post-production additional notes:

We’ll be presenting ways to join the call to action regarding these concerning findings at next year’s ENDOMARCH-AGCES Conference in Atlanta (Buckhead), Georgia, USA, on March 28-30, 2025. You can click here to learn more or register to attend. (Virtual options will be added soon as well, if you are not able to travel to Atlanta this time around).

As for putting these concerning findings in context, here are additional notes regarding the Banard et al JAMA article under particular scrutiny: 

Due to many factors, it’s difficult to extrapolate the Banard et al estimates in a precise way to determine how many additional cases of ovarian cancer would occur in the global population of 200 million endometriosis  sufferers based on their reported increased risk profiles. As mentioned, epidemiological unknowns are legion due to decades of neglect & severe research under-funding. For example, we only have very rough estimates of what percentage of endometriosis patients suffer from severe forms of endometriosis (now or in future) involving either deep infiltrating endometriosis, ovarian endometriomas, or both. 

These endometriosis macrophenotypic subtypes are also still poorly understood from molecular/genomic/clinical standpoints & exhibit tremendous heterogeneity. In other words, not all nodules classified as DIE exhibit similar clinical/molecular/genomic patterns, so the risk stratification methodology used in the Barnard et al research are not necessarily applicable to all putative cases of DIE or even all ovarian endometriomas, which also vary widely.

Current discourse concerning endometriosis and the increased risk of certain cancers also often centers predominantly on just ovarian cancer, while minimizing discussions about other forms of gynecologic cancer, as well as non-gynecological cancers. For example, although debates are ongoing and there’s no consensus, there have been several preliminary studies pointing to a potential association of endometriosis with thyroid and colorectal cancer.

Considering that endometriosis is a whole-body (systemic), chronic inflammatory condition, the fact that researchers have reported associations with inflammation-linked cancers is not at all surprising. However, these studies seem to be largely ignored by many medical societies & researchers involved in establishing clinical guidelines for endometriosis healthcare around the world. In order to truly save lives and improve clinical outcomes, risk factors for all forms of cancer must be included in patient guidelines, not just those classified as gynecologic, such as breast cancer, cervical cancer, ovarian cancer, uterine/endometrial cancer.

The estimate of ~ 200 million with endometriosis worldwide is also now an outdated figure.  (We know this, as we are the research team that originally reported & published on the 200 million figure back in approximately 2012; other research teams from different institutes have reported 176 million and 190 million, respectively).  Not only has the population increased since 2012, but the estimated number of those with endometriosis has also increased, from 1 in 10, to approximately 1 in 5-7 women, girls, and individuals assigned female at birth. (Sources vary & references are forthcoming, but the 1 in 5 estimate is from a publication-in-progress by Nezhat et al team, while the 1 in 7 is from the Australian government). These global population estimates have also historically left out male-born, intersex, non-binary, and transgender populations who suffer from endometriosis and, as a consequence, may also be exposed to a significantly increased risk for certain cancers.

There’s also the complicated matter of the four criteria researchers use to classify what counts as an endometriosis-associated ovarian cancer. Although this is only preliminary research-in-progress from our team, but it’s our belief that many cases of endometriosis-related cancers are being left out of the endometriosis-associated cancers classification, due to outdated criteria (three of which were originally proposed by Dr. John Sampson in approximately 1925, with the fourth proposed in 1953 by Dr. RB Scott). Our call to action, therefore, will include evidence which suggests that the four criteria are in urgent need of updating with modern molecular & genomic methods.

WHAT DO THESE INCREASED RISK ESTIMATES MEAN IRL?
With these caveats in mind, a rough estimate of a 4-fold increase would translate to about 80,000 additional cases* of ovarian cancer in a population of 200 million with endometriosis worldwide.

In the cases of a 9.66-fold or 19-fold (18.96) increased risk, additional adjustments and assumptions would need to be made, as we do not have a truly accurate understanding of what percentage of endometriosis sufferers have the specific subtypes.

However, if we apply to the global population the observed 24.1% of the 10-fold group that was reported in the study, this would translate to about 48,200 additional cases of ovarian cancer worldwide.  For the group exposed to the 19-fold risk (DIE, endometriomas, or both), there were only 1.7% in the Banard et al study in this category, and so this would work out to 6,800 additional cases of type 1 ovarian cancers worldwide. (Note: Other studies have reported much higher incidences of DIE or endometriomas than the Banard et al study, but we are applying their observations for the sake of consistency).  

As mentioned, these are just very rough ballpark figures. Additional adjustments (like age-standardized incidence rate calculations) would be needed before hazard ratios/relative risk ratios from one study can be generalized to a large global population with diverse demographics, genetics, environmental factors, comorbidities, hormonal exposure, and other variables that can contribute independently to health outcomes in endometriosis patients. In other words, a more thorough evaluation of all of these factors is needed to truly understand the real-world implications of the Banard et al findings in particular. We will be applying these and other adjustments in the forthcoming published version of this website blog.

FINAL THOUGHTS BEFORE SIGNING OFF 
As is evident from these significantly increased risk estimates, it’s truly a moral imperative for the endometriosis thought leaders of the world, to come together as a matter of urgency, so that patient guidelines can be updated in light of these new data. 

This emerging research also adds more urgency to the question advocates have been asking for literally decades now: how can millions of patients, with a known higher risk for one of the deadliest cancers worldwide, continue to be denied access to surgical excision of the actual cancer-causing pathological growths that are conferring the significant excess risk in the first place?

Prevention of cancer is always talked about reverentially as a desired goal in public health circles. Yet, the one known, potentially preventive measure – surgical removal of the pathology – with one of the highest efficacy rates in the world for potentially reducing cancer risk in endometriosis patients, remains largely inaccessible to millions worldwide, who are told instead to lose weight and think positive thoughts as suggested preventive measures.

And, how can we continue to tolerate such inaction, when the preponderant canon of modern medical literature demonstrates that male-born patients are referred to surgery immediately in cases when pathological growths with a known heightened risk for malignant transformation have been found? (Sources forthcoming from a Nezhat et al publication-in-progress).

In these and so many countless case studies, it’s no longer possible to deny that unlawful gender-based healthcare discrimination is at play. Therefore, governmental agencies and policymakers must be reminded that they are duty-bound to protect endometriosis patients from these sorts of unlawful / unconstitutional denials of medically-necessary care, which includes potentially life-saving excision surgery by a qualified expert. The road is long, but we will pursue these urgently-needed policy reforms as if our ‘life, liberty, and freedom’ depend on it.

SHARE YOUR THOUGHTS OR INDEPENDENT RESEARCH HERE
We would love to hear what you think about the emerging research on elevated ovarian cancer risks in endometriosis patients. You can join the discussion in the comments section here on the social media version of this blog. As well, if you would like to share your own research about these topics, you could submit your abstract to AGCEScongress@gmail.com, for consideration at the next ENDOMARCH-AGCES conference. (If the abstract submission deadline has already passed, we encourage you to still submit what you have, and the conference organizers may be able to add you, if other qualifications are met, of course).

SOURCES
Please note that this blog post, originally published on May 15, 2024, is based on a publication in progress by our Nezhat et al team. We’ll be adding full references as soon as the article is published.

FOOTNOTES
1) *Prior to Sampson, who coined the term “endometriosis” in a landmark 1925 paper, endometriosis was referred to by many other names, including catamenial hemotocyles, adenomyomas externa, and even strangulation of the womb in ancient times).
2) ** Adjusted hazard ratios were used in the Barnard et al study, so some somewhat minor adjustments are needed in order to directly compare results from other studies which used relative risk or odds ratio calculations.