Uterine Endometrium Microbiome in Women with Repeated Implantation Failure Complicated by Endometriosis

This prospective study was conducted between March 2021 and December 2023 according to the guidelines of the Declaration of Helsinki, and was approved by the Institutional Review Board of Teine Keijinkai Hospital (No. 2-020387-00, date of approval: 14 March 2021), and we obtained informed consent from all participants. We defined RIF as the failure to achieve pregnancy after ≥2 embryo transfers, and 43 women with RIF were enrolled in this study. Of these 43 women, 12 had endometriosis (EM group), whereas 31 had no endometriosis (non-EM group). None of the women with RIF had received prebiotics/probiotics within 3 months of study participation. Clinical information was obtained from electronic medical records. Of the 12 women with endometriosis, 6 had ovarian endometrioma, 2 had adenomyosis, 2 had pelvic endometriosis, 1 had ovarian endometrioma and adenomyosis, and 1 had adenomyosis and pelvic endometriosis ( Table 1 ). Endometriosis was diagnosed based on magnetic resonance imaging, intraoperative findings, and/or postoperative pathology. Uterine endometrium specimens were obtained via aspiration during the mid-luteal phase, confirmed by transvaginal ultrasonography and the last menstrual cycle, as previously reported [ 17 ]. The samples were placed in the middle part of the aspiration tube and soaked in an OMNIgene ® -VAGINAL microbiome kit (DNA Genotek Inc., Ottawa, ON, Canada). They were promptly sent to Varinos Inc. (Tokyo, Japan), where the uterine endometrium microbiome was analyzed using 16S rRNA sequencing. The 16S rRNA gene is a component of the small subunit of the ribosome, which is common to bacteria and archaea. The gene has a conserved region, which is common to all bacteria, and a highly mutated region; 16S rRNA sequencing reveals the structure and diversity of microbial communities by analyzing specific regions of the gene. Conserved regions and variable region 4 (V4) of 16S rRNA, were amplified by polymerase chain reaction (PCR) using DNA extracted from tissue specimens [ 21 ]. The amplified PCR samples were identified following the Illumina 16S Metagenomic Sequencing Library Preparation protocol [ 22 ]. For histopathological findings of CE, after defining CD138-positive cells in the endometrium as plasma cells by immunohistochemical staining and counting the number of CD138-positive cells/10 mm 2 , CE was diagnosed when the plasma cell count was >5.15/10 mm 2 , according to the criteria set by Liu [ 13 ], who was the first to demonstrate an association between CE and uterine endometrium microbiome in women with infertility. Histopathological analyses for CE were performed at Sapporo Clinical Laboratory Inc. (Sapporo, Japan). Endometriosis was diagnosed by magnetic resonance imaging examination when the lesion appeared as a single or multifocal cystic lesion with a high signal on both T1- and T2-weighted images or as a uniform high signal on a T1-weighted image and a low signal, called shading, on a T2-weighted image, and was not suppressed on fat-suppression imaging. The biodiversity of a bacterial community in a certain niche can be characterized and quantified by its richness and evenness [ 23 ]. Species richness denotes the number of bacterial species present, while evenness pertains to the relative abundance of these organisms. The relative abundance of bacterial species reflects their relative dominance, and mathematical parameters such as species richness and evenness serve as comparative measures. The Shannon diversity index, a widely employed estimator of species richness and evenness [ 24 ], was utilized to compare the diversity of the uterine endometrium microbiome. Data were analyzed using JMP version 16 (SAS Institute Inc., Cary, NC, USA). Categorical variables were compared by Fisher’s exact test and the Mann–Whitney U test. Patient backgrounds, the number of plasma cells/10 mm 2 with CD138 staining, the frequency of CE (Liu’s method), and the uterine endometrium microbiome were compared between EM and non-EM groups. Multivariable logistic regression analysis was performed to determine the bacterial species associated with endometriosis. All p -values were two-sided, and p -values < 0.05 were considered significant.

Endometriosis, which affects 10–15% of women of reproductive age, is a lifelong condition that reduces a woman’s quality of life and causes dysmenorrhea, pelvic pain, infertility, and ovarian cancer, with an increased risk of future cardiovascular events [ 1 , 2 , 3 , 4 ]. Endometriosis is an estrogen-dependent pelvic inflammatory disease thought to affect gynecological organs and is a cause of infertility; however, the complex mechanism and variety of factors involved limit our understanding of the condition. Endometriosis is considered a chronic inflammatory disease, and immune abnormalities in its pathogenesis have been pointed out [ 5 ]. Although the gut microbiome in women with endometriosis is associated with pathogenesis, immune dysfunction, and inflammatory autoimmune diseases [ 6 , 7 ], the intrauterine microbiome has not been evaluated extensively because of the limited detection capability of bacterial culture methods. Recently, microbiome analyses with 16S ribosomal RNA (rRNA) sequencing using a next-generation sequencer have been performed in the reproduction field and have made progress in the study of uterine endometrium microbiome [ 8 , 9 ]. Chen et al. examined the microbiome of the uterine cervix, uterine lumen, fallopian tubes, and peritoneal fluid along the female reproductive tract using 16S rRNA sequencing analysis and demonstrated that the microbial community is present in each area [ 10 ]. Franasiak JM et al. first examined the microbiome of post-embryo transfer catheter tips in ART patients by 16SrRNA sequencing analysis and found that the most common bacteria were Lactobacillus species [ 11 ]. Moreno et al. first showed that there was a correlation between the endometrium microbiome and subsequent reproductive outcome. They reported that rates of implantation, pregnancy, and live births after in vitro fertilization embryo transfer were higher in women with Lactobacillus -dominant microbiome (>90%) in the endometrium than in those with non- Lactobacillus -dominant microbiome, using 16S rRNA sequencing [ 12 ]. Liu Y et compared the uterine endometrial microbiome in women with chronic endometritis (CE) associated with implantation failure to uterine endometrial microbiome in women without CE and demonstrated that the percentage of Lactobacillus species was significantly decreased in women with CE than that in non-CE (CE vs. non-CE, 1.89% vs. 80.7%) [ 13 ]. Recent studies also suggest that abnormalities in the uterine endometrium microbiome are associated with implantation failure; therefore, evaluation of the endometrium microbiome is considered to be important for improving the reproductive outcomes in infertility [ 14 , 15 , 16 ]. We also examined the uterine endometrium microbiome and CE in women with repeated implantation failure (RIF), recurrent pregnancy loss (RPL) and fertile controls, and revealed increases in the prevalence of CE in women with RPL and in percentages of Lactobacillus iners and Ureaplasma species in women with CE [ 17 ]. The generally accepted theory of endometriosis is that the endometrium refluxed to the pelvic cavity via the fallopian tubes during menstruation grows, proliferates, and develops, and the intraperitoneal cavity of women with endometriosis has traditionally been thought to be sterile. However, a technique using 16S rRNA sequencing identified bacteria in the intraperitoneal fluid and menstrual blood of women with endometriosis [ 18 ]. This suggests that the bacteria originate in the uterine cavity and that uterine endometrium microbiome may be involved in the development of endometriosis. Since an increased number of bacterial species at various sites is associated with endometriosis, the endometrial microbiome may be related to endometriosis development [ 19 ]. However, studies on the uterine endometrium microbiome in women with endometriosis are still limited, particularly for women with infertility [ 10 ]. RIF is an intractable form of infertility believed to be caused by various factors, such as a fertilized egg, uterus, immunity, hormones, and other unknown factors [ 20 ]. The intrauterine environment is important during the implantation of a fertilized egg into the endometrium; therefore, the intrauterine microbiome may be involved in the pathophysiology of implantation. In this study, we hypothesized that endometriosis is involved in the pathogenesis of CE and affects the uterine endometrium microbiome in women with RIF. To evaluate this, we compared the uterine endometrium microbiome and the prevalence of CE between women with RIF complicated by endometriosis and those without endometriosis.

The clinical characteristics and backgrounds of EM and non-EM groups were compared ( Table 1 ). No differences in age, body mass index, gravidity, parity, number of previous miscarriages, and number of implantation failures were found between the two groups. The frequency of CE and the uterine endometrium microbiome were compared between EM and non-EM groups ( Table 2 ). The prevalence of CE determined by Liu’s method (plasma cell count > 5.15/10 mm 2 ) or the number of CD138-positive cells per 10 mm 2 was not different between the two groups. The EM group had a higher number of bacterial species than the non-EM group (median [range] 6.0 vs. 3.0 [3–11, 1–16], p = 0.009); however, there was no significant difference in the Shannon diversity index. There were no significant differences in the relative dominance rate of Lactobacillus species or in the frequency of the Lactobacillus -dominant microbiome defined as >90% of the relative dominance rate of Lactobacillus species. The relative dominance rate of Lactobacillus crispatus was lower in the EM group (median 0 [range 0–98.2]) than in non-EM group (0.15 [0–99.9]); however, the difference was not significant ( p = 0.053). Among the infertility-associated bacteria, the rate of presence of Dialister species (41.7% [5/12] vs. 3.3% [1/31], p = 0.004) and Streptococcus species (58.3% [7/12] vs. 16.1% [5/31], p = 0.017) in the EM group was higher than in the non-EM group. The rates of presence of Ureaplasma , Mycoplasma , Gardnerella , Prevotella , Atopobium , Bifidobacterium , Anaerococcus , Escherichia , Enterococcus , Bacteroides , Corynebacterium , Finegoldia , Pseudomonas , Peptoniphilus , or Sphingomonas species were not statistically different between the two groups. Multivariable logistic regression analysis was performed to determine whether the presence of Dialister or Streptococcus species in uterine endometrium microbiome was associated with endometriosis in RIF ( Table 3 ). The results revealed that only Dialister species were found to be associated with endometriosis in RIF (odds ratio 10.97 [95% confidence interval 1.17–249.37], p = 0.036). In the EM group, the number of bacterial species (median [range], 10 vs. 5 [6–11, 3–7], p = 0.021) and Shannon diversity index (0.50 vs. 0.20 [0.19–1.39, 0.03–0.46], p = 0.026) were higher in the five women with Dialister species than in the seven without Dialister species. The relative dominance rate of Lactobacillus species or prevalence of each bacterial species was not different between the two groups.

This cohort study, for the first time, demonstrated that the number of bacterial species and rates of presence of Dialister and Streptococcus species increased in the uterine endometrium microbiome of women with RIF complicated by endometriosis as compared with those without endometriosis. The presence of Dialister species in women with RIF complicated by endometriosis was related to increases in the number of bacterial species and in the Shannon diversity index in the uterine endometrium microbiome. These results suggest that the presence of Dialister and Streptococcus species, together with increases in the number of bacterial species and Shannon diversity index in the uterine endometrium microbiome, are pathologically associated with RIF in women with endometriosis. Compared with women without endometriosis, those with endometriosis had dysbiosis in the gut microbiome with increases in the abundance of harmful bacteria and decreases in beneficial bacteria, resulting in a loss of bacterial diversity [ 25 ]. A study using 16S rRNA sequencing reported that the diversity of the uterine endometrium microbiome was higher in women with endometriosis who had surgically confirmed endometriosis lesions than in women with endometriosis but no confirmed lesions (symptomatic control) [ 26 ]. In the female reproductive tract, bacteria are classified into five community groups (community state types [CSTs] I–V), distinguished by the abundance of Lactobacillus species and the diversity of anaerobes [ 27 , 28 ]. Chen et al. reported that women with endometriosis had more CST IV bacteria than women without endometriosis [ 10 ]. Similarly, in the present study, the number of bacterial species and rates of presence of CST IV, Dialister, and Streptococcus species increased in the uterine endometrium microbiome of women with RIF complicated by endometriosis. The microbiome is known to play an integral role in the immune system, creating the permissive environment necessary for successful implantation. A complex microenvironment, influenced by infection and inflammation, is established by cytokines involved in both endometrial receptivity and embryo development [ 29 ]. Dialister species are anaerobic Gram-negative bacteria belonging to CST IV, found primarily in the human oral cavity and intestine, and are associated with bacterial vaginosis and preterm delivery [ 30 ]. Dialister species are also identified in the placentas of women with periodontal disease, spreading hematogenously from the oral cavity, and increase the risk of preterm delivery [ 31 , 32 ]. The presence of Dialister species in the intestinal microbiome reflects the activity and degree of inflammation in autoimmune diseases such as spondyloarthritis [ 33 ]. Dialister species may act on various immune cells to promote the production of proinflammatory cytokines and chemokines. On the other hand, in inflammatory skin diseases such as atopic dermatitis and rosacea, Dialister is reported to act in a protective manner, because it produces propionic acid, induces regulatory T-cells, and controls inflammation by suppressing the production of inflammatory cytokines such as IL-6 and TNF [ 34 ], suggesting that the presence of Dialister species indicates the intensity of inflammation, but whether it is a cause or a consequence remains unclear at this stage. The presence of Dialister species reflects the inflammatory state and signifies an intrauterine environment that is difficult to implant; therefore, it is considered important to make the intrauterine environment Lactobacilus species dominant. Streptococcus species is a Gram-positive coccus, also classified as CST IV, and is involved in bacterial vaginosis, preterm delivery, and neonatal infections [ 35 ]. Streptococcus species have been detected in the uterine endometrium microbiome of women with endometriosis [ 36 ] and in ovarian endometrioma [ 37 ]. Certain Streptococcus species are known to cause endometritis [ 38 , 39 ]. Since intrauterine inflammation prevents implantation by interrupting the endometrial switching of proliferation to differentiation, which is essential for embryo acceptance [ 40 , 41 ], treatment for dysbiosis with inflammation in the endometrium is required prior to embryo transfer. Pathological endometriotic lesions include several sites, such as ovarian endometriomas, adenomyosis, and pelvic endometriosis. The endometrial tissue in adenomyosis is inflamed [ 42 ]. The prevalence of Lactobacillus species decreases, and bacteria belonging to CST IV increase in the uterine endometrium microbiome of women with adenomyosis [ 10 ]. In the present study, all four women with adenomyosis had Dialister and/or Streptococcus species. Women with endometriosis and adenomyosis have a high risk of preterm delivery, which may be caused by these bacteria in the intrauterine environment [ 43 ]. In RIF women with endometriosis, treatments with antibiotics, probiotics and prebiotics, as well as surgery and hormonal therapy for endometriosis may more effectively restore uterine endometrium to the Lactobaccilus -dominant microbiome. This prospective study has several limitations. Since the number of RIF women with endometriosis is clearly smaller than the number of non-endometriosis cases, it is difficult to conclude that the results obtained from this study alone are correct. Further studies are needed and should enroll a larger sample of RIF women with endometriosis. The relationship between Dialister and/or Streptococcus species in uterine endometrium microbiome and endometriosis in women with RIF needs to be clarified to determine whether it is a cause or a consequence. Basic researches including several biomedical and pharmaceutical approaches will elucidate these pathologic mechanisms [ 44 , 45 , 46 ]. The investigation of implantation and pregnancy rates after treatments for abnormal uterine endometrium microbiome is needed in future studies.

This study was the first to reveal that the number of bacterial species and rates of presence of Dialister and Streptococcus species increased in the uterine endometrium microbiome of women with RIF complicated by endometriosis. These changes seemed to worsen the inflammatory state of the intrauterine environment, causing implantation failure. A major limitation of this study is the small number of women with endometriosis; therefore, future studies with a larger sample size are needed in order to prove the results of this study. As a treatment strategy for RIF-complicated endometriosis, analyses of the uterine endometrium microbiome using 16S rRNA sequencing and treatments for endometriosis with surgery and/or hormonal therapy and for abnormal endometrium microbiome with antibiotics, prebiotics, and probiotics can be performed.