Endometrial cancer (EC) is the most common gynecological malignancy in developed countries, second only to cervical cancer in the developing part of the world. Unlike many other cancers, the incidence of EC and associated mortality have been increasing in recent decades (1). Most EC cases (around 80%) are histologically endometrioid, present early, and have a more favorable prognosis. These cases are known under the group name Type 1 EC tumors, first described by Bokhman (2). The rest - serous, clear cell, mucinous, and several others - were grouped under the name Type 2 EC tumors. Type 1 tumors usually contain sex steroid receptors and arise from hyperplastic preconditions, such as endometrial hyperplasia (EH). Their development is traditionally related to unopposed exposure to estrogens. The latter may be the consequence of the action of risk factors, including, but are not limited to, obesity, nulliparity, estrogen therapy, etc.

However, in light of new publications and the discoveries of the new genomic atlas, endometrial cancers seem to be more heterogeneous (3). The four distinct subtypes were described: DNA polymerase epsilon ultra-mutated, microsatellite instability hypermutated, copy-number low, and copy-number high. The recent studies suggest that microbiota can also play role in EC tumorogenesis (4, 5). While the vaginal microbiota is well-known to be active and dominated by the Lactobacilli genus, the uterine cavity was previously considered sterile (6, 7). The advent of next-generation sequencing (NGS) technologies revealed that the uterine microbiota is more scarce than the vaginal (10³-10⁵ times lower), but simultaneously more diverse (8). The origin of the intrauterine microbial community and its role in health and disease is an emerging area of research. For instance, recent studies suggested that the endometrial microbiota can modulate uterine receptivity and therefore influence fertility (9, 10). Furthermore, the composition of the uterine microbiota has been associated with various gynecological conditions, including endometriosis, dysfunctional menstrual bleeding, and cancer (8, 11). The current study was conducted to provide a comparative analysis of the reproductive tract microbiota between patients with EH and EC.

In the current study we investigated differences in microbiota composition both in the vagina and the uterus between EH and EC patients. The results indicate the presence of a scarce but diverse microbial community within the uterus, as shown by several intragroup diversity parameters, the Chao1, Shannon, and Simpson index. Furthermore, both the vaginal and endometrial microbial composition per se differs between the groups with possible explanations and implications to be described below.

The existence of bacteria in the uterus has long been a controversial topic, with early research relying on culture-based methods suggesting the uterus was sterile. This may be due to the imperfections of the method, as well as the fact that some bacteria are known to grow poorly on artificial media. However, recent culture-independent studies using next-generation sequencing have revealed the presence of diverse microbial communities in the uterus, referred to as the uterine microbiota (21). In healthy women, the uterine microbiota appears to be low in biomass, but highly diverse, containing bacteria belonging mainly to the Lactobacillus (Bacillota), Prevotella (Bacteroidota), Gardnerella, and Atopobium (both belong to Actinomycetota) (22). The specific composition of the microbiota varies according to factors such as the menstrual cycle, parity and menopausal status (23). In postmenopausal women, the uterine mileue was enriched in Flavobacterium (Bacteroidota) compared to the cervicovaginal and anorectal microbiota (24). In contrast, in nulliparous women with abnormal uterine bleeding Prevotella, Fusobacterium and Jonquetella were the most abundant taxa (25). Infertility, endometriosis, and oncogynecological diseases are among the most studied areas in the context of microbial influence (26–28).

The uterus is an immunologically priveleged organ: it can accommodate tissue invasion by immunologically semiforeign placental cells, yet maintain mucosal immune defenses against ascending foreign organisms, and provide a system to efficiently clear the endometrial detritus that results from menstruation. The relations between the human body and the resident microbiota are an emerging scientific topic, which has not yet been described. Sometimes, the microbe-host interactions are simplified by saying that these two entities apply oppositely directed forces: The microbes strive to thwart host defense and multiply infinitely, while the host puts all efforts into eradicating the microorganisms. Although this might be acceptable or at least pragmatic for using this model for acute infections, the pattern of how the host interacts with the resident microbiota is not linear and is best described by the Nash dynamic equilibrium (29). In such a model, a host sequesters a microbe community within defined locus (should that be the enteric lumen or endometrial cavity) and does not favor their spread beyond these borders.

In our study members of both Porphyromonadaceae and Prevotellaceae families were among the most prevalent bacteria in the vagina and the uterus. This is in accordance with the previous studies. For example, the study by Caselli et al. showed that both Atopobium vaginae and Porphyromonas somerae induced the expression of pro-inflammatory cytokines in endometrial cells, which could have implications for the development of endometrial cancer (30). Based on sequencing of the 16S rDNA V3-V5 region from the reproductive tract, Walther-António et al. suggest that the detection of Atopobium vaginae and Porphyromonas sp. in the gynecologic tract combined with a vaginal pH greater than 5, had a sensitivity of 100% and specificity of 60% to detect EC (31). In a recent study by the same group (32) Porphyromonas was described as a robust biomarker of EC. Porphyromonas uenonis was more common in vaginal specimens from patients with both high-grade and low-grade endometrial cancers than from those with benign conditions (33). Wang et al. studied the microbiota in EC and adjacent tissue samples and found that several genera were enriched in malignant loci—Porphyromonas, Prevotella, Atopobium, Anaerococcus, Dialister, and Peptoniphilus (5). Li et al. provided evidence that tumor burden is associated with the presence of a specific endometrial microbiota, which is distinguished by the enrichment of Pelomonas and Prevotella in patients with EC (34). Transcriptomic data identified the upregulation of eight genes involved in pathways associated with fibrin degradation, including those with proteolysis and transcription factor activities. Hence, the authors concluded that Prevotella may promote fibrin degradation by inducing the expression of specific genes. It should be also noted, that Porphyromonas was previously described as carcinogenic within the digestive tract, due to its ability of inducing immune responses in the host (35).

In the current study, potential biomarkers for endometrial cancer have only been identified in the vaginal microbiota. Typically absent from the normal vaginal microbiota, including Porphyromonas, Prevotella corporis, Streptococcus, Moryella, Peptococcus niger, Criibacterium and Clostridiales bacterium S5 A14a. Notably, vaginal Porphyromonas and Prevotella, along with Bacteroides, Mycoplasma, Bacillus, Dialister, and Anaerococcus, have recently been identified as indicators of cervical cancer (36). Prevotella corporis, an established human pathogen, along with Prevotella bivia and Prevotella disiens, originates from the urogenital tract (37), but can also be an odontogenic pathogen (38). Prevotella corporis, an anaerobic bacterium rarely mentioned in the literature, was previously found to be associated with bacterial vaginosis, including in pregnant women (39), and has never been reported to be associated with EC until the current study. However, other Prevotella species and the Prevotella genus in general have previously been found to be associated with various gynecological cancers. Prevotella bivia was one of two species, along with Fusobacterium ulcerans, significantly more abundant in high grade compared to low grade endometrial carcinoma (33), and even earlier was found associated with cervical cancer (40). As previously shown, the abundance of vaginal Prevotella and Streptococcus was increased in the group of gynecological cancers (41). Although Prevotella spp. generally cause minimal proinflammatory epithelial activation, they may induce changes to the cellular physiology and integrity of the endometrial epithelium in a species-specific fashion. Prevotella spp. can form biofilms, alter the epithelial barrier, and impact the colonization of secondary colonizers (42).

Moryella, currently represented by a single species Moryella indoligenes, is a minor component of the vaginal microbiome (43), closely correlated with bacterial vaginosis (44). This bacterium, along with Prevotella, was previously described as being abundant in the vaginal samples from the females with EC (45). The association with gastrointestinal cancer was also reported (46).

Peptococcus niger, is an established human pathogen, previously found in a healthy vaginal ecosystem (47). This species has not previously been reported to be associated with gynecological cancers, however, it was previously identified in oral swabs in patients with potentially malignant disorders (48).

Criibacterium a genus with a single specie Criibacterium bergeronii isolated from human clinical samples, was shown to be enriched in the EC tissues, as mentioned in the work by Wang et al. (5). Criibacterium bergeronii, besides Streptococcus pyogenes, was the only other bacterial species enriched in idiopathic cutaneous ulcers in children after community-based mass treatment with azithromycin (49).

Clostridiales bacterium S5 A14a and Anaerovoracaceae family to which this bacterium probably belongs were found to be abundant in vaginal swabs in patients who had previous malignancies (50). In addition, Clostridiales bacterium S5-A14a was one of the taxa differentiating healthy control and post-treatment cervical swab samples in premenopausal women with squamous cell carcinoma of the cervix (51).

Intriguingly, the genes of Porphyromonas, Prevotella, Streptococcus, and Peptococcus niger can be classified as part of the estrobolome. The first three taxa can encode β-galactosidases, and Peptococcus niger can encode sulfatases and convert estrone sulfate to estrone, influencing active estrogen levels - the major risk-factor for EC (52, 53). In a recent study, for example, changes in gut Porphyromonas and Prevotella corporis correlated with postmenopausal hot flashes, which may also be related to their effect on estrogen levels (54).

Limitations of the study include the limited number of samples and differences in age and menopausal status between the groups. On the one hand, it reflects the natural distribution observed clinically, as patients with endometrial cancer tend to be older than those with hyperplasia (55, 56). In particular, Walsh et al. (32) also reported that postmenopausal status and age were significant modifiers of the microbiome composition. At the same time, we believe that, since the results of our research are consistent with previous studies, especially regarding Porphyromonas and Prevotella, menopause should be considered as only one of several factors that influence the outcome. The hypoestrogenism typically observed in postmenopause might potentially modulate the microbiota through interactions betwen estrogen and intraepithelial glycogen. The latter serve as a substrate for Lactobacilli, which act as antagonists to other bacteria through producing hydrogen peroxide and antimicrobial compounds. This interaction becomes even more complex when considering that multiple bacterial species can affect estrogen reabsorption and enterohepatic circulation by their β-glucuronidase and β-glucosidase enzymes that deconjugate estrogens (57, 58). Therefore, it can be hypothesized that increased activity of the deconjugating bacteria could lead to increased estrogen levels, a known risk factor for EC.

In conclusion, both vaginal and endometrial microbiota are different in terms of composition between patients with EH and EC. Several taxa were enriched in vaginal samples in patients with EC: Prevotella corporis, Porphyromonas, Streptococcus, Moryella, Criibacterium, Peptococcus niger and Clostridiales bacterium S5-A14a. We believe that these differentiating taxa found may be a hallmark of EC-associated vaginal dysbiosis. Our results therefore support previous findings suggesting differences in the composition of the genital tract microbiome, which could potentially be used for the diagnosis of EC, including the differential diagnosis of EH.