Endometrial microbiome in reproductive failure: The possibility of metagenomic analysis

Abstract

With the advent of metagenomics and 16S ribosomal RNA sequencing, growing attention has been dedicated to the endometrial microbiome. Research involving a relatively large cohort of healthy female participants has reported Lactobacillus dominance (LD) in the endometrial microbiome. Multiple studies have also shown that the loss of LD and/or increased microbial diversity within the endometrium are associated with reproductive failure. This phenomenon may stem from the loss of the immunomodulatory effects of Lactobacillus and the rise of proinflammatory responses triggered by pathogenic proliferation. Recent research has employed personalized antibiotic therapy followed by probiotic supplementation, tailored to the endometrial microbial composition of women with repeated implantation failure. The findings suggest that metagenomic analysis of the endometrial microbiome could be a valuable tool in addressing reproductive failure.

Introduction

Implantation is a process that requires close interactions between the embryo and the endometrium. Failure of implantation can be attributed to embryonic or maternal factors. Preimplantation genetic testing for aneuploidy (PGT-A) has enabled in vitro fertilization and embryo transfer (IVF-ET) to achieve clinical pregnancy rates (CPRs) of over 60% following an ET cycle [1]. However, even after addressing various maternal factors, including anatomical abnormalities of the uterus and systemic immunologic and thrombogenic issues, the rate of repeated implantation failure (RIF) with the transfer of euploid embryos remains approximately 5% [2].

It is widely acknowledged that certain bacterial infections can establish an inflammatory environment in the female genital tract, potentially leading to infertility and adverse obstetric outcomes [3,4]. However, until recently, our understanding of the microbiomes in the female genital tract was primarily limited to a few pathogens detectable with conventional culture techniques or polymerase chain reaction. The Human Genome Project revealed that approximately 9% of the total body microbiome is derived from the female genital tract, with many species remaining undetectable through conventional culture methods [5]. Recent advances in metagenomic studies, particularly those employing 16S ribosomal RNA (rRNA) sequencing, have facilitated a deeper exploration of the female genital tract microbiota and its association with various reproductive diseases.

In this review, we aim to summarize the association between the female genital tract microbiome and reproductive failure (RF), focusing on the endometrial microbiome.

Vaginal microbiome and reproductive health

Over a century ago, Dr. Albert Doderlein reported the predominance of Lactobacillus in the vaginal microbial community. Since then, extensive research has shown that Lactobacillus dominance (LD) in the vagina promotes reproductive health [6]. Conversely, dysbiosis of the vaginal microbiota, characterized by the loss of LD and an increase in α-diversity (that is, a greater variety and evenness of species), has been associated with a number of gynecological conditions, including bacterial vaginosis, sexually transmitted infections, cervical cancer, and endometriosis [7-9]. Additionally, this state has been linked to multiple adverse pregnancy outcomes, such as preterm birth, preterm premature rupture of membranes, and premature cervical dilatation [10-12].

Recently, research has focused on the relationship between the vaginal microbiome and miscarriage. A prospective study that analyzed the vaginal microbiome starting at 5 weeks of gestation found that first-trimester miscarriage was associated with a decreased abundance of Lactobacilli and increased α-diversity [13]. Additional research has indicated that a reduction in LD within the vaginal microbial community is linked to euploid miscarriage and elevated levels of proinflammatory cytokines [14,15].

During IVF, vaginal samples can be readily collected and analyzed using metagenomic methods. One study revealed a positive correlation between LD in the vaginal microbiome and the CPR in women undergoing IVF-ET [16]. Furthermore, a recent study observed that women with RIF exhibited higher vaginal microbial diversity compared to women who achieved clinical pregnancy in their first frozen ET cycle [17].

The initial discovery of the microbiome in the uterine endometrium, which was based on 58 samples from hysterectomy procedures performed for benign disease [18], dispelled the misconception that the uterus was sterile. This finding paved the way for subsequent studies focused on the endometrial microbiome. With the advent of next-generation sequencing (NGS) of bacterial 16S rRNA, research has shifted from the vaginal microbiome to the endometrial microbiome, particularly regarding its implications for reproductive health and implantation [19].

Endometrial microbiome and reproductive failure

The initial characterization of the uterine endometrial microbiome from the hysterectomy sample provided irrefutable evidence of its independent existence, as it eliminated the possibility of contamination from the vagina [18]. Subsequent research using hysterectomy and simultaneous vaginal samples, along with a large-scale study involving samples from 110 women at six sites within the female reproductive tract, identified a similar microbial composition (including LD) between the vagina and endometrium. However, this research also noted a gradient of microbial diversity from the lower to the upper reproductive tract [20,21].

A study published in 2000 found that a higher live birth rate was associated with a positive culture for H₂O₂-producing Lactobacillus in 91 women undergoing IVF [22]. Subsequent research using 16S rRNA sequencing revealed a relationship between the abundance of Lactobacilli in the endometrial microbiome and more favorable pregnancy rates in women undergoing ET [23-25], although some findings did not reach statistical significance [26]. In contrast, another study observed no significant differences in implantation and miscarriage rates between groups with non-LD and LD endometrial microbiota compositions, even though a higher prevalence of LD was noted in fertile controls [27]. Furthermore, the study reported seven pregnancies in the non-LD group without any intervention.

The endometrium is regulated by sex hormones. A notable study from 2016 assessed the differences in endometrial microbiome composition between the pre-receptive phase (2 days after the luteinizing hormone [LH] surge) and the receptive phase (7 days after the LH surge) using 44 paired endometrial samples. The study found no differences in microbial composition within the same woman between the pre-receptive and receptive phases [24]. Another study also reported a stable endometrium, with regard to LD, throughout the menstrual cycle [26]. However, a different study employing meta-transcriptomics, as opposed to metagenomics with 16S rRNA, identified a significant difference in the microbiome between the early follicular phase (days 6–8 of the menstrual cycle) and the mid-luteal phase (7–9 days post-LH surge) [28]. Furthermore, other researchers have observed similar cyclic changes in the composition of the vaginal microbiome [29,30]. Interestingly, all studies that showed a significantly positive correlation between LD of the endometrial microbiome and favorable CPR in IVF cycles involved endometrial sampling at the time of ET or during the preceding mid-luteal phase—essentially, within the window of implantation [22-25]. In contrast, studies that did not suggest this correlation did not acquire samples during these times [26,27]. These findings suggest potential variations in the microbial composition of the endometrium throughout the menstrual cycle.

Similar to a study on the vaginal microbiome [31], the endometrial microbiome was analyzed using 16S rRNA sequencing to compare the microbial composition between individuals with missed abortion and those with induced abortion at the time of abortion [32,33]. The endometrial microbial composition of the missed abortion group was distinct from that of the induced abortion group. Additional research has examined the endometrial microbiome in women with a history of recurrent pregnancy loss (RPL), contrasting it with that of fertile controls. These studies were all conducted during the mid-luteal phase and revealed differences in the endometrial microbial composition between women with RPL and fertile control participants, although the details were not consistent across studies. Some research identified specific pathological microbial elevation in women with RPL [34], while others reported variations in the abundance of Lactobacillus spp. [35,36] or a loss of LD in this population [37].

RIF, RPL, and unexplained infertility are collectively classified as a single disease entity known as RF. Numerous studies have suggested the existence of certain common etiologies across these conditions [38,39]. Among investigations of the endometrial microbiome, several studies have explored RPL while also incorporating RIF or infertility [23,37]. These studies are listed in Table 1. In contrast, some research has focused exclusively on RIF. One such study compared the endometrial microbiome of 28 women with RIF to that of 18 infertile women undergoing their first IVF-ET cycle. The findings indicated higher β-diversity and the presence of Burkholderia spp. solely in the RIF group [40]. Another study observed increased α-diversity in the endometrial microbiome of 117 women with RIF compared to 17 healthy controls [41]. Subsequent research has explored the potential for treating RIF with antibiotics. One study demonstrated that antibiotic therapy improved the CPR and ongoing pregnancy rate in 125 women with RIF who had pathogenic bacteria in their endometrium [42]. Another study involving 131 patients with RIF, who were analyzed using EMMA—a commercial service for endometrial microbiota analysis employing NGS of microbial 16S rRNA—reported a relatively high cumulative CPR of 64.5% within two additional ETs after a personalized regimen of antibiotics followed by probiotics [43].

Table 1.

Endometrial microbiome composition in women with reproductive failure

Study	No. and type of patients	Sampling time	Findings
Shu et al. (2022) [32]	17 Missed abortions vs. 12 induced abortions	Just before D&C	Higher β-diversity in the missed abortion group
Wang et al. (2023) [33]	38 Early missed abortions vs. 18 induced abortions	Just before D&C	Significant difference in EM and vaginal microbiome between two groups
Shi et al. (2022) [34]	67 RPL	Mid-luteal phase prior to pregnancy	More Ureaplasma spp. in the abortion or preterm birth group
Masucci et al. (2023) [35]	40 RPL vs. 7 fertile controls	Mid-luteal phase	Less Lactobacillus acidophilus and the presence of L. inners only in the RPL group
Peuranpaa et al. (2022) [36]	47 RPL vs. 39 fertile controls	Mid-luteal phase	Less L. crispatus and more Gardnerella vaginalis in the RPL group
Vomstein et al. (2022) [37]	20 RPL and 20 RIF vs. 10 fertile controls	Follicular, periovulatory, and luteal phase	Decreased α-diversity during periovulatory and luteal phase only in controls
Kitaya et al. (2019) [40]	28 RIF vs. 18 infertile women in 1st attempt of ET	Mid-luteal phase	Presence of Burkholderia spp. only in a portion of RIF group (25%)
Ichiyama et al. (2021) [41]	145 RIF vs. 21 fertile controls	Early & mid-luteal phase	Higher α-diversity in the RIF group but no difference in the proportion of Lactobacillus between the two groups
Zou et al. (2023) [42]	141 RIF women	Mid-luteal phase	Most women with RIF (88.7%) had pathogenic bacteria in EM
Iwami et al. (2023) [43]	Among 195 RIF, 131 tested with ‘EMMA’a) and treated vs. 64 not tested	Mid-luteal phase	In the following ET, higher CPR and ongoing pregnancy rates in the tested and treated group
Moore et al. (2000) [22]	91 Infertile women undergoing IVF, culture methodb)	Just before ET	Higher LBR associated with isolated H2O2-producing Lactobacillus
Moreno et al. (2022) [23]	342 Infertile women undergoing IVF	Mid-luteal phase	Higher LBR associated with Lactobacillus
Moreno et al. (2016) [24]	35 Infertile women undergoing IVF	Periovulatory vs. mid-luteal phase	Different EM microbiomes in periovulatory and mid-luteal phases, lower implantation rate and higher LBR in the LD group
Diaz-Martinez et al. (2021) [25]	48 Infertile women undergoing IVF	Luteal phase	Lactobacillus spp. more abundant in pregnant women
Kyono et al. (2018) [26]	102 Infertile women	Various phases, mainly follicular	LD in 38% of IVF patients, 78.9% in non-IVF patients, and 85.9% in controls, similar LBR in LD and non-LD groups
Hashimoto et al. (2019) [27]	99 Infertile women undergoing IVF	Mock transfer (no other description)	Similar CPR, IR, and MR in dysbiotic and eubiotic endometrium group

D&C, dilatation and curettage; EM, endometrial; RPL, women with a history of recurrent pregnancy loss; RIF, women with a history of repeated implantation failure; ET, embryo transfer; CPR, clinical pregnancy rate; IVF, in vitro fertilization; LBR, live birth rate; LD, Lactobacillus dominant; IR, implantation rate; MR, miscarriage rate.

^a)‘EMMA’, a commercial service for endometrial microbiota analysis; ^b)All studies employed next-generation sequencing with 16S rRNA sequencing except one study marked as ‘Culture method.’

Potential mechanism

Despite a general consensus regarding the association between the endometrial microbiome and RF, the underlying mechanism remains elusive. Research has uncovered two phenomena contributing to this association. The first involves LD in the endometrial microbiome. The protective role of Lactobacillus spp. in preventing pathogen colonization is well-documented [44,45], with lactic acid production by Lactobacilli being the primary mechanism. While the healthy vaginal cavity is acidic due in part to LD, the endometrial cavity typically is not, even when Lactobacilli are plentiful [21]. Recent serial studies of systemic lupus erythematosus (SLE) have uncovered various immunomodulatory effects of Lactobacilli. Administration of probiotics containing certain Lactobacillus spp. was associated with increased regulatory T cells and forkhead box P3 (Foxp3) expression and decreased interleukin 6 levels in a mouse model of SLE [46]. Additionally, treatment with Lactobacilli in macrophages and dendritic cells from patients with SLE resulted in anti-inflammatory shifts in cytokine secretion [47,48].

The second potential phenomenon linking the endometrial microbiome to RF is microbial diversity, including pathogens. Research indicates that Lactobacilli do not trigger inflammation or cause cell death in endometrial cells [49]. Furthermore, an increase in pathogens within the endometrium has been associated with RIF and RPL [34,50]. Pathogenic infections can induce proinflammatory mediators in the endometrium, leading to a disruption of the immune milieu that is crucial for successful implantation and the maintenance of pregnancy. However, cases of chronic endometritis may occur without evidence of specific pathogenic infection, presenting instead as increased α-diversity, the loss of LD, and the heightened expression of proinflammatory and apoptotic genes [50-52].

Limitations of current research

To date, numerous studies have identified associations between the endometrial microbiome and RF, including infertility, RPL, and RIF. Notably, some research has highlighted differences in the endometrial microbial composition between fertile and infertile women, as well as among women with varying pregnancy outcomes, such as missed abortion versus induced abortion and RIF versus successful pregnancy following a first attempt at IVF-ET. These findings suggest that deviations in the endometrial microbiome from that of healthy fertile controls may influence RF. However, without definitive evidence of causation, it remains difficult to ascertain whether changes in the endometrial microbiome directly contribute to RF. Moreover, inconsistencies in the timing of endometrial microbial sampling across studies hinder the ability to draw consistent conclusions. Only a few studies have compared pregnancy outcomes by transferring exclusively euploid embryos via PGT-A to minimize bias from embryonic factors. Consequently, to establish a causal link between the endometrial microbiome and fertility outcomes, and to elucidate the mechanisms involved, it is essential to standardize the timing and methodology of endometrial sampling, ensure the quality of transferred embryos to mitigate the influence of embryonic factors, and conduct longitudinal follow-ups of reproductive outcomes to clarify the cause-and-effect relationship.

Recently, several studies explored the potential effectiveness of treatments targeting abnormal endometrial microbiomes in cases of RIF and RPL. However, the value of these treatments remains unclear, and the evidence supporting their efficacy is limited. Moreover, the optimal treatment approach has yet to be established. While most research has focused on eradicating pathogens using antibiotics, some studies have concentrated on restoring a healthy LD environment by administering Lactobacillus spp. following antibiotic treatment. The effectiveness of oral Lactobacillus supplementation for re-establishing LD in the female genital tract microbiome is still a topic of debate [53]. Recent research has pivoted towards the vaginal administration of Lactobacillus [43,54]. Additionally, vaginal microbiome transplantation has been attempted [55]. Further research is necessary to ascertain whether interventions targeting the endometrial microbiome can improve reproductive outcomes in populations affected by RF.

Sampling the endometrial microbiome is more invasive and susceptible to contamination than obtaining samples from the vaginal microbiome. Additionally, the data derived from the endometrial microbiome cannot yet be deemed superior to those from the vaginal microbiome. These factors present challenges to researchers and clinicians in accurately characterizing the endometrial microbiome and clarifying its role in reproductive health.

Conclusion

The advent of metagenomics and 16S rRNA sequencing has provided growing evidence supporting the role of the endometrial microbiome in RF. However, further research is needed to fully comprehend its implications and to develop potential therapeutic interventions. Addressing challenges associated with study design, sample collection, and data interpretation is critical to advance our knowledge in this area.