The systematic review was designed based on the Preferred Reporting Ideas for Systematic Review and Meta-analyses (PRISMA) (Page et al., 2021) systematic review checklist and was registered on PROSPERO (Date of registration: July 22nd 2025, ID: CRD420251042858, review protocol link: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251042858).

Original research articles written in English and published between 1st January 2000 and 31st December 2025 were eligible for inclusion. We included studies reporting any relation between hormonal therapy to treat endometriosis and the endometrial, cervical and vaginal microbial profile. Studies regarding women diagnosed with endometriosis based on clinical evaluation, imaging or surgical confirmation (e.g., laparoscopy) and under hormonal therapy were considered. Intervention included the administration of: GnRH analogs, estrogen-progestin pills, only progesterone pills. Whereas, studies based on the use of intra-uterine devices were excluded. Control population included patients with endometriosis who did not receive hormonal treatments. Randomized and nonrandomized studies conducted in hospitals, clinics and universities worldwide were taken into consideration. Articles not published in English, whose full text was not available, letters to the editor, narrative and systematic reviews, preprint, guidelines, book chapters, authors’ replies, case reports, and poster presentations were excluded. Systematic reviews concerning the topic of the study have been examined to identify eligible articles which were not found by database searching.

To ensure a comprehensive retrieval of all the studies relative to endometriosis and the microbial profile of female reproductive tract, we chose to exploit three relevant and reliable databases: PubMed (MEDLINE), Scopus and Google Scholar. The combination of MeSH terms searched in the databases included “endometriosis,” “vaginal microbiome,” “endometrial flora,” “bacterial vaginosis,” “vaginal dysbiosis,” “GnRH agonist,” “progestin,” “combined oral contraceptive”, “oral hormonal therapy”. The detailed search terms for all databases are provided in Supplementary Table 1.

Duplicate articles were removed from the results of the literature search. Two independent authors screened the titles and abstracts of the remaining articles to ensure that the eligibility criteria were met. Any discrepancies between the authors were identified and discussed (with inputs from a third author if required). The remaining included articles were assessed by full-text screening by two independent authors, using the same eligibility criteria.

Risk of bias assessment was carried out by using Version 2 of the Cochrane risk-of-bias tool for Non-randomized Studies of Interventions (ROBINS-I-V2) (Risk of Bias Tools, n.d.). Three reviewers (EA, SL and GC) independently assessed the studies for potential biases, including selection and reporting biases. The overall risk of bias is a synthesis of the three independent evaluations. Any discrepancies were resolved through discussion.

The extracted data were synthesized according to the design and outcome measures of the included studies. Given the paucity of studies identified after selection, and the limited number of patients analyzed in each study, a meta-analysis could not be performed. The main outcomes evaluated for each study were: microbiota composition and relative abundance of bacterial communities in cervical, vaginal and endometrial samples; α-diversity and β-diversity; imbalances or disruptions in the normal microbiota, including a decrease in beneficial bacteria or an increase in harmful microbes (e.g., Lactobacillus dominance and specific taxa colonization); Nugent score for bacterial vaginosis. The findings were synthesized in tables by using a qualitative narrative approach.

The systematic search identified a total of 577 articles: 137 articles were available in PubMed, 290 studies in Scopus, and 150 in Google Scholar. Among those, 34 were duplicates and 272 articles were excluded since the publication type did not meet the eligibility criteria (non-English articles, reply to the editor, narrative and systematic reviews, preprint, guidelines, conference abstracts, case reports, book chapters). Two independent authors screened a total of 271 articles for their relevance in the topic by assessing the title, abstract or full-text. Among the 271 articles, 267 studies were excluded since unrelated to the effect of hormonal therapy on the microbial profile of endometriosis patients. During the screening process, upon the evaluation of systematic reviews concerning the object of the study, 2 articles not identified by the database searches but relevant for the present review were found and added manually to the list. As a result, 6 articles were included and analyzed in this systematic review (Figure 2).

The literature search was conducted over the period from 1st January 2000 to 31st December 2025. The 6 identified studies included in the analysis were all published after 2014. This is probably due to the fact that the importance of the microbiome in reproductive biology gained more attention in the last decade. Furthermore, the technological advances of recent years, particularly concerning next generation sequencing (NGS) methods, have enabled the characterization of the complete microbiota of the female reproductive tract with greater precision and cost-effectiveness.

Among the selected papers, three investigations were conducted in Japan, two in the USA and one in Australia (Table 1). Despite the inclusion criteria for literature searching not specified a range of age, all the studies included women spanning from 18 to 51 years old. In the papers from Khan et al. (2014), Khan et al. (2016), Wee et al. (2018), and Khan et al. (2021), all patients analyzed had a regular menstrual cycle (28–32 days), whereas Do et al. (2024) and Le et al. (2021) did not provide this information.

In four papers out of six, endometriosis was diagnosed by laparoscopy and the staging of the disease based on the revised classification of the American Society of Reproductive Medicine (r-ASRM) (American Society for Reproductive Medicine, 1997). In one paper, endometriosis was assessed through laparoscopy, however the classification used for disease staging was not specified. The paper from Wee et al., on the contrary, based on the clinical history of the patients collected through a self-completed questionnaire, without considering the stage of the disease.

Since the six papers selected were based on non-randomized studies, we used the ROBINS-I-V2 tool for risk of bias assessment (Risk of Bias Tools, n.d.; Figure 3). Before evaluation, we preliminarily determined the confounding factors to be relevant for the intervention-outcome relationship under study. In detail, we considered ethnicity, body mass index (BMI), range of age, menstrual cycle phase (Proliferative/Secretive/Menstrual/Amenorrhea), endometriosis stage, use of therapies for chronic diseases, previous or concomitant use of immunosuppressive or antimicrobial agents as substantial confounding factors. Whereas age, menstrual cycle regularity (28–32 days), history of chemo-radio therapy and surgical gynecological interventions (not related to endometriosis) were assigned a comparatively lower degree of importance. Additionally, we took into consideration whether all treated patients followed the same intervention protocol.

In relation to Domain 1, the two studies from Do et al. (2024) and Le et al. (2021) showed a low level of risk of bias due to confounding factors. On the contrary, the risk of bias of the remaining four papers analyzed was assessed as serious, mainly due to the lack of information about the use of antimicrobial agents, BMI and ethnicity of patients. Consequently, the aforementioned papers also exhibited a critical risk of bias attributable to missing data (Domain 5).

Due to the lack of information about the type of oral contraceptive used, Wee et al. (2018) was assigned a moderate level of risk of bias in classification of interventions (Domain 2). All other studies showed a low risk of bias in the same domain.

The risk of bias for Domain 3 (Risk of bias in selection of participants into the study or into the analysis), Domain 4 (Risk of bias due to deviations from intended interventions) and Domain 6 (Risk of bias arising from measurement of the outcome) was found low in all the selected papers. For Domain 7 (Risk of bias in selection of the reported result), all papers were attributed a low or moderate risk of bias. In particular, the study from Khan et al. (2014), in which the outcome was measured through the assessment of Nugent score, was assigned a low level of risk since there is only one possible way in which the outcome domain can be measured. The other five papers, based on NGS techniques, have been attributes a moderate level of risk, due the existence of multiple strategies to analyze the outcome.

In consideration of these results, the overall risk of bias for Do et al. (2024) and Le et al. (2021) was categorized as moderate. Conversely, the overall risk for Khan et al. (2014), Khan et al. (2016), Wee et al. (2018), and Khan et al. (2021) was deemed critical.

Research studies investigating the impact of hormonal treatments on the composition of the reproductive tract microbiome in women with endometriosis reported heterogeneous findings, largely depending on therapeutic approach and methodological differences. Interventions include GnRHa treatment and contraceptive therapy (COCs and progestins), whereas methodologies include standard culture methods, Gram staining and NGS techniques, as illustrated in the following sections and in Table 2.

Endometriosis is a common gynecological disorder affecting around 10% of pre-menopausal women worldwide (Li R. et al., 2025). This condition is the leading cause of chronic pelvic pain, from adolescence to adulthood, and is often associated with infertility. In both adult and young patients, the recommended therapy includes oral hormonal administration of COCs or progestins only pills, and, as second option, GnRH agonists. Therapy must be personalized based on the peculiar characteristics of the patients, the presence of comorbidities and future reproductive aspirations.

Increasing evidence pinpoint the microbiome of the reproductive tract as a crucial determinant of the development and progression of endometriosis. In line with this view, lower richness and α-diversity of cervical microbiome was found in patients with more severe symptoms and infertility, suggesting that a more diverse microbial environment, with prevalence of beneficial species, might favor better clinical outcomes (Chang et al., 2022).

In partial accordance with these findings, the studies included in this systematic review evidenced a different microbial profile in patients with endometriosis compared with not affected women. Indeed, the presence of endometriotic lesions results accompanied by sub-clinical uterine infections, increased amount of Streptococcaceae and Moraxellaceae and augmented α-diversity in the endometrium (Khan et al., 2014, 2016, 2021). In women with endometriosis, Khan et al. registered an increased vaginal pH, suggestive of a dysbiotic state (Khan et al., 2014), and a decreased abundance of Lactobacillaceae in cystic fluids (Khan et al., 2016). Conversely, Le et al. observed a reduced α-diversity along with a higher abundance of Lactobacillus, indicative of a healthy balanced vaginal microbiome in endometriosis patients. These discrepancies might be partially explained by the different types of samples analyzed, vaginal swab vs. cystic fluids, and by the ethnicity of the study populations, known to have an impact on microbial profiles (Wei et al., 2024). Another crucial aspect which should be taken into consideration relies on the fact that both studies did not specify the dominant Lactobacillus specie, and this information is noteworthy since it is known that different species characterize different types of vaginal ecosystems (Zhu et al., 2022). Precisely, Lactobacillus crispatus provides protection against pathogens, whereas Lactobacillus iners and Lactobacillus gasseri often co-occurs with dysbiosis-associated microbes (De Seta et al., 2019).

In support of the hypothesis that endometriosis is associated with a microbial disequilibrium in the reproductive tract, the recently published review from Wang et al. highlighted the correlation between microbial communities and the onset of the disease, discussing the multiple mechanisms behind, which include the modulation of inflammatory/immune responses and the regulation of estrogen metabolism (Wang et al., 2025). For instance, in preclinical studies it has been observed that the inoculation of the pathogenic specie Fusobacterium nucleatum increased the numbers and weights of endometriotic lesions in mice, which were effectively reduced by antibiotic treatment (Muraoka et al., 2023). In addition, it has been assessed that BV-associated bacteria (Gardnerella vaginalis, Prevotella bivia and Mobiluncus spp.) and Lactobacillus spp. depletion in the cervico-vaginal microbiome associate with endometriosis (Uzuner et al., 2023). In this context, Khan et al. unveiled that the use of GnRHa in women with endometriosis further alters vaginal flora and promotes the establishment of an alkaline milieu, ideal substrate for the proliferation of potentially pathogenic bacteria (Khan et al., 2014; Figure 4). The treatment with GnRHa, by establishing a hypo-estrogenic state along with a reduced anti-microbial peptides release, may promote the growth of potentially pathogenic microbes in vaginal and endometrial districts.

A similar effect is ascribable to the administration of OCs, which reduces vaginal Lactobacillus and favors the colonization of microorganisms as Prevotella, Streptococcus and Ureaplasma (Figure 4). Such dysbiotic environment has the potential to contribute to the discomfort experienced by affected women. Importantly, endometriosis patients under OCs therapy, had a higher Bacillota/Bacteroidota ratio. β-glucuronidase-producing Bacillota are known to deconjugate estrogen metabolites elevating bioactive estrogen levels in the serum (Elkafas et al., 2022). The increase of vaginal Bacillota in endometriosis women treated with OCs may lead to the establishment of a therapeutic paradox cycle, in which the administration of hormonal therapy leads to an hyper-estrogenic condition, overall potentially interfering with the efficacy of the treatment (Li Z. et al., 2025; Wang and Wang, 2025).

It is noteworthy that the reproductive microbial profile can exert a significant influence on fertility, as well as the potential outcomes of future reproductive plans (Venneri et al., 2022). This association is supported by a recent meta-analysis which took into consideration 15 studies and evidenced a correlation between low abundance of vaginal Lactobacillus and female infertility (Hong et al., 2020). Accordingly, other authors posed the alteration of the vaginal microbial homeostasis as a risk factor for pregnancy related complications, like preterm delivery (<37 weeks) and late miscarriage, and a negative predictor for in vitro fertilization (IVF) success in terms of pregnancy rates (Bernabeu et al., 2019; Venneri et al., 2022). Increasing evidence are emerging indicating the beneficial effect of oral probiotics and specific dietary habits on the vaginal microbiome (Noormohammadi et al., 2022; Djusse et al., 2025; Udjianto et al., 2025). In particular, a study is exploring whether the administration of oral probiotics could restore the vaginal environment increasing the abundance of beneficial Lactobacillus species in IVF patients, thereby improving their outcomes (van Haren et al., 2025).

Similarly to the vaginal environment, an endometrial microbiota not dominated by Lactobacilli but colonized by Bifidobacterium, Gardnerella, Staphylococcus and Streptococcus is associated with adverse outcomes in IVF procedures (Moreno et al., 2022; Xiao et al., 2024). In accordance with these data and with the observations raised from the present review, a recent study evidenced that the use of a GnRHa protocol for endometrial preparation prior to embryo transfer in infertile endometriosis patients associated with Staphylococcus and Escherichia-Shigella colonization, and with worse obstetrics outcomes as miscarriage, cesarean delivery, and hypertensive disorders, compared to other interventions (Zhang et al., 2025). Given these observations, the choice of hormonal therapy for endometriosis needs to be discussed with the patient, who should be adequately counseled on the protection of future fertility.

The strengths of this systematic review include its rigorous methodology and comprehensiveness in cataloging the existing evidence. Additionally, it allowed to define and highlight important confounding factors which are often overlooked in studies regarding the reproductive microbiome. The included papers regarding the effect of OCs on vaginal microbiome adopted the same analytical method to measure the outcomes (NGS), giving us the opportunity to draw considerations based on homogeneous data. Nonetheless, the depth of microbial characterization varied among the selected investigations, with authors reporting changes at different taxonomic levels (e.g., family, genus or species). Even species that belong to the same family or genus can differ in terms of their pathogenicity or beneficial role. For this reason, it is recommended that future investigations use NGS methods to precisely characterize the microbiome and determine the composition at species level.

A limitation of the present work relies on the lack of consideration of several confounding factors in the selected studies, as evidenced through risk of bias assessment performed with ROBINS-I V2 tool. For instance, four papers did not register any previous treatment with either immunosuppressing or antimicrobial agents for enrolled patients. This issue has been considered a bias of great importance since the recent use of antibiotics may clearly impact the outcomes. In addition, probably due to the limited sample size, none of the investigations reported a stratification of the patients according to patients age, ethnicity, menstrual cycle phase or BMI. Conversely, these features are known factors influencing the microbiome of the reproductive tract (Garg et al., 2023; Kumar et al., 2023; Dubé-Zinatelli et al., 2024; King et al., 2024; Wei et al., 2024; Landolt et al., 2025). Moreover, another factor which has been registered in all the studies but not taken into account for patient stratification is the stage of endometriosis. In this regard, recent findings have identified the vaginal microbiome as a potential tool for disease staging. For instance, Perrotta et al. (2020) reported that patients with ASRM stages III-IV exhibited a significantly different microbial composition and an enrichment for Anaerococcus during the menstrual phase when compared with stages I-II.

Considering the studies based on the use of OCs, the conclusions may be affected by the heterogeneous types of adopted formulations. Most of the results based on the use of COCs therapy, however, Do et al. also used GnRHa and progestin only (norethindrone acetate) medications, even if in only two participants, and Wee et al. did not specify the therapeutic regimen. There are minimal and controversial published data indicating whether COCs or progestin only therapy could exert different effects on the vaginal microbial environment (Vodstrcil et al., 2013; Ratten et al., 2021; Bakus et al., 2023), nonetheless this aspect might not be underestimated and further studies are mandatory to better understand their impact on the microbiome of endometriosis patients.

Importantly, when interpreting the results of the four studies analyzing endometrial and cervical samples, it is crucial to consider the risk of cross-contamination with the vaginal environment. Although sample collection was performed in accordance with protocols aimed at minimizing this risk, no contamination controls were applied to strengthen the robustness of the conclusions.

Ultimately, due to the unfeasibility of conducting a meta-analysis, the implementation of specific analytical tools for the evaluation of the certainty of evidence was not possible. Nonetheless, some considerations should be raised on the limited strength of the evidence emerged from the present systematic review. Indeed, the interpretation of the overall results require caution, given critical risk of bias, inconsistency across studies in the characteristics of the populations, interventions, methods, or outcomes measurement, and imprecision deriving from the low number of participants.

To the best of our knowledge, this is the first systematic review which highlights the up-to-date evidence on the effect of hormonal therapy on the taxonomic composition of microbial environment of vaginal and uterine cavities in women affected by endometriosis. Despite the aforementioned limitations, the overall results suggest that both the administration of GnRHa and OCs might negatively affect the reproductive microbiome, establishing a milieu dominated by potentially harmful bacteria (Figure 4).

Since the microbial profile is tightly linked to the efficacy of the therapy and to the global health status of women, these considerations may be relevant for a better clinical management of patients. Despite our observations suffer of the criticisms due to heterogeneity of the analyzed studies, they lay the foundations for a future debate among clinicians on the routine screening of microbial dynamics alongside treatment. Similarly, in light of the findings raised from this review, it would be advisable to better investigate on the beneficial effect of probiotics supplementation or an appropriate dietary regimen to further improve patients’ outcomes. Moreover, as endometriosis onset often occurs at young age, the implications of hormonal therapy-associated dysbiosis might be considered in regard to fertility and reproductive health.

Importantly, our work unveils significant factors that need to be considered when designing investigations on reproductive microbiome in endometriosis. For instance, it will be recommended to define inclusion and exclusion criteria taking into account the recent use of antimicrobial or immunosuppressive agents and the type of intervention. Moreover, it will be mandatory to stratify the analyses by age, ethnicity, BMI, stage of endometriosis and menstrual cycle phase. To grant the appropriate homogeneity of the results, standardization of the microbiome assessment methodology is strongly suggested, in particular the use of NGS technology should be preferred to achieve species-level classification.