NK cells are large granular lymphocytes that are characterized by CD56⁺, CD16^+/− and CD57^+/− expression, and positivity for natural cytotoxicity trigger receptor 1 (NCR1), otherwise known as CD335, but not CD3 or surface T cell receptor (68,69).

NK cells can be divided by their expression level of CD56 into cd56bright and cd56dim, which have increased and reduced expression levels of CD56, respectively. Cd56bright NK cells produce more abundant cytokines, while cd56dim NK cells have increased cytotoxicity and increased expression levels of FC γ receptor III (fcgr3), also known as CD16 (70,71). NK cells spontaneously recognize and eliminate infected, ectopic, tumorigenic and stress responsive cells so as to automatically monitor for viral infections, ectopic tissue and malignant cells (72–74). Repeated exposure to the same target results in increased accumulation of NK cells and the production of a specific recall response by NK cells that is characterized by the enhancement of the functional activity of NK cells against the target (75–77). The binding of inhibitory killer cell immunoglobulin-like receptor (KIR) on NK cells with MHC class 1 or human leukocyte antigen (HLA)-1 (KIR/HLA-1) inhibits the activation of NK cells and allow for further intrinsic interactions. KIR therefore has an important role in distinguishing autologous cells from diseased and foreign cells to avoid non-selective killing of autologous healthy cells (78,79). The maturation and cytotoxic function of NK cells is based on the interactions between KIR and autologous MHC molecules, which is called licensing (80). Once licensed and functionally mature, NK cells are inhibited by inhibitory receptors that bind to the autologous MHC. Cells without MHC class 1 expression will be eliminated by activated NK cells (81). If NK cells are not stimulated by interaction with autologous MHC they may lose their normal function (82). However, the inhibition of KIRs and MHC is not absolute in mature NK cells and can also be eliminated or offset by a much stronger active stimulator. Other MHC receptors are also involved in NK cell cytotoxicity regulation against target cells, such as leukocyte immunoglobulin-like receptor subfamily B (LILRB) and natural killer group protein 2 (NKG2) (83–85).

Among the receptors that activate NK cells, FCGR3, which is expressed in almost all NK cells, is key for antibody dependent cytotoxicity. FCGR3 expression levels alone are sufficient to induce interferon γ and TNF, making it is one of the most effective activating receptors of NK cells. Other receptors that can activate NK cell cytotoxicity include NCRs, which are divided into NCR1, NCR2 and NCR3 (or NKp46, NKp44 and NKp30, respectively) (86).

In addition to licensing through the interaction between receptors on the surface of NK cells and the autoantibodies of the body, exposure to cytokines is also essential to activate the cytotoxicity of immature NK cells and promote cytokine secretion (87). IL-2 and IL-15 secretion by macrophages can activate and trigger the maturation of NK cells and promote their proliferation (88–90). Simultaneous exposure to IL-12 and IL-18 is not only able to activate NK cells, but can promote IFN-γ secretion (91,92). Conversely, increased interferon secretion can also enhance the cytotoxicity of NK cells, such as the anti-tumor response of NK cells that is mediated by the cyclic GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) pathway activation (93). Activated NK cells can induce apoptosis by releasing cytolytic particles against target cells. They can also by cytotoxic to target cells through a Fas L-mediated mechanism with the help of the CD95 receptors on the surface of target cells (94,95).

Several reports have shown that NK cells in patients with EM have reduced abilities both in clearing out of ectopic endometrium and in participating in local and systemic immunity, which creates a favorable environment for the survival and growth of ectopic endometrial tissue (96–98). NK cytotoxicity decreases not only in ectopic endometrial tissue, but also in the peripheral blood and peritoneal fluid (99).

Decreased NK cytotoxicity is currently debated, as to date, studies on this topic have not adequately shown this cytotoxicity to occur (48,100,101). It has been speculated that the upregulation of NK cell inhibitory receptors and the downregulation of stimulatory receptors in ectopic endometrium may be caused by cytokines such as IL-2, IFNs and TGF-β. This indicates that cytokine therapy targeting NK cell inhibitory or stimulatory receptors is feasible (102,103). Most studies show that the proportion of NK cells decreases in patients with EM. However, there are also studies showing that the proportion of NK cells in patients with EM increases when compared with that in healthy controls (104,105). These results suggest that the decreased NK cytotoxicity in patients with EM is not the result of decreased NK cell infiltration, but rather the abnormal expression level of NK cell activation receptor and/or inhibition receptor. However, only the upregulation of NK cell inhibitory receptors is supported by current research, while the regulation of stimulatory receptors remains unclear due to the lack of studies and considerable results.

Overexpression of inhibitory receptors is considered vital for modulating immune evasion and maintaining immune tolerance in EM. Increased levels of HLA-1 in the glandular and stromal cells of endometrial tissue was observed in the study by Vernet-Tomas Mdel et al (106), which may result in increased resistance to NK cytolysis in patients with EM (106). The case-control study by Wu et al (107) finds that the levels of KIR in the peritoneal NK cells of patients with EM also increased (including NKB1 and EB6), thus further reducing the cytotoxicity of NK cells (107). Overexpression of other inhibitory receptors and their ligands were observed in patients with EM, including the inhibitory receptors KIR2DL1, CD94/NKG2a and LILRB1 on peritoneal NK cells, and their separate endogenous ligands HLA-C, HLA-E and HLA-G (108–110). The altered presence and distinct combined presence of different KIR genes contributes to a unique genetic background of patients with EM. It should be noted that not all KIR/HLA binding promotes the development of EM. For example, receptor KIR2DS5 in combination with its ligand HLA-C C2 has a protective effect against EM (111).

Decreased expression levels of NKG2D, a stimulatory receptor of NK cells, was also reported in patients with EM (112). It is plausible that the downregulation of NKG2D is the result of increased levels of TGF-β and ectopic endometrial tissue-derived IL-15 (113,114). However, the ligands for NKG2D, including MHC class-I chain-associated proteins (MIC)A and B that were upregulated in the peritoneal fluid of patients with EM (115), paradoxically exert an inhibitory effect on NK cells (116,117). The precise mechanism behind how the regulation of NK cell stimulatory receptors influences the pathogenesis of EM requires further verification and examination.

IL-12 can regulate the immune recognition of NK cells in the endometrium. IL-12 is composed of two heterologous polypeptide chains, p40 and p35. A study found that the concentration of IL-12 in patients with EM is similar compared with that of healthy controls. However, increased levels of free p40 in the peritoneal fluid of these patients indicated that overexpression of the p40 subunit alone could reduce the cytotoxicity induced by IL-12 (118). The ratio of FCGR3-negative NK cells to FCGR3-positive NK cells in the peritoneal cavity of patients with EM was increased (119). Based on these findings, the increased expression levels of inhibitory receptors and ligands has a key role in the decline of NK cytotoxicity.

Platelets also regulate the function of NK cells (111). Platelets release TGF-β during retrograde menstruation in patients with EM, thereby suppressing the expression level of a stimulatory receptor of NK cells, NKG2D and reducing their cytotoxicity (Fig. 2) (120–123).

Studies on the changes and mechanisms of NK cells in patients with EM provide a solid experimental basis for the targeted treatment of EM. The present focus of immunotherapy targeting NK cells is to restore their cytotoxicity. To reach this goal, three possible aspects should be considered: i) Blocking inhibitory receptors; ii) anti-inhibitory or stimulatory cytokine therapy; and iii) immune checkpoint therapy.

KIR2DL1, LILRB1/2 and CD94/NKG2a are inhibitory receptors that are overexpressed in the NK cells of women with EM. The ability to interfere with the binding of the inhibitory receptors and their ligands to improve the cytotoxicity of NK cells has been tested by multiple studies (124,125). Disruption of the inhibitory receptors on NK cells contributing to enhanced cytotoxicity was found in a human cell model treated with 5-aza-2′-deoxycytidine reported by Binyamin et al (124) in 2008. The study observed increased NK cell cytotoxicity when the inhibitor KIR2DL1 was applied to the NK cells of a healthy woman. It was also observed that blocking KIR2DL1 can enhance the effects of rituximab (an anti-CD20 monoclonal antibody known to recruit the immune system to attack and kill B cells) by increasing the cytotoxicity of NK cells (124,126). The study by Andre et al (125) targeted NKG2A on NK cells with monalizumab combined with cetuximab (an EGFR inhibitor) in patients with head and neck carcinoma, resulting in increased NK cell cytotoxicity (125). Whilst neither of these studies were based on EM, this blocking of inhibitory receptors therapy may be a promising treatment for EM and has been shown to be effective against some malignancies (125,127); however, further studies on their role in EM are required to make conclusive statements.

Previous studies tried to identify cytokines that affect the regulation of NK cell activity as new targets for immunotherapy. Some ILs (such as IL-2 and IL-12) and IFNs are NK cell stimulative cytokines. Intraperitoneal injection of IL-2 in surgically implanted EM rat models was found to be capable of recruiting leukocytes into EM lesions and reducing lesion size (128,129). However, existing studies on the therapeutic effects of IL-2 on EM are based in animal models, therefore, further studies and analysis are needed on the applicability to human patients. IL-12 is another important NK cell stimulative cytokine. Researchers pretreated NK cells with an IL-12 heterodimer to reduce the ratio of free p40 to IL-12 and enhance the cytotoxicity of NK cells in ectopic endometrial tissue, resulting in suppressed development of ectopic endometrial tissue. IL-12 is therefore considered a potential specific target for correcting the increase in free p40 levels in patients with EM (118). Type I IFNs, which include IFN-α2b, IFN-β1a and type II IFN (IFN-γ), can activate NK cells and enhance their cytotoxicity. However, to the best of our knowledge, no work has studied the feasibility of using the NK cell activating effects of IFNs to treat EM. The study by Dicitore et al (130) showed that IFN-β1a is superior to IFN-α2b at inhibiting the proliferation and migratory activities of endometrial stromal cells.

A case-control study by Wu et al (131) found that treatment with GnRH-a could restore the damaged immune function of the peritoneal fluid in patients with EM, and proposed their hypothesis according to their findings (131). In this study, women with EM who used GnRH-a long-term were found to have increased levels of the CD3⁻CD69⁺ subpopulation of peripheral blood mononuclear cells and the CD3⁺CD69⁺/CD3⁺CD24⁺ subpopulation of activated T cells. It was suggested that the increased level of activated T cells induced by GnRH-a secreted increased levels of IL-2 and IFN-γ, which led to restoration of NK cell activity in the peritoneal fluid.

TGF-β suppresses the cytotoxicity of NK cells and the function of other immune cells (132,133). TGF-β secretion is upregulated in women with EM and is considered important to EM pathogenesis (134,135). Anti-TGF-β therapies are currently being evaluated clinically as treatments for malignancies and other diseases, such as diabetes. However, the results are generally unsatisfactory, reporting non-responsiveness and potential systematic side effects (136,137). There are also concerns that anti-TGF-β therapies would cause systematic suppression and result in severe systematic side effects due to the important role that TGF-β has in multiple vital signaling pathways, such as cell proliferation and differentiation. In vitro and in vivo studies are needed to further study anti-TGF-β therapy for EM.

Other immunotherapies may restore NK cell function in EM, most of which are based on immunotherapy models of other diseases. In 2004, the study by Clayton et al (138) first proposed the possibility of using Mycobacterium to restore NK cell activity in EM (138). However, the hypothesis was only supported by in vitro studies and further verification is required.

Immune checkpoint blocks (ICBs) are also new immunotherapies that researchers are currently evaluating. PD-1/PD-L1 pathway-associated inhibitors are a type of checkpoint blocking therapy. Previous studies have reported increased PD-1 expression levels in the peripheral blood cells of patients with EM, and increased PD-L1 expression levels in both the ectopic and non-ectopic endometrial tissues of patients with EM (139,140). This indicates that the peripheral tolerance caused by PD-1/PD-L1-induced T cell suppression may contribute to the immune abnormalities noted in EM. ICB therapy using PD-1/PD-L1 inhibitors is a promising treatment for preventing immune tolerance to EM (141–143). However, studies have reported that PD-1/PD-L1 inhibitor treatment can lead to adverse reactions in a variety of tissues and organs throughout the body. The use of ICBs in the treatment of EM should therefore be carefully selected, and inhibitors with the strongest specificity for EM should be utilized (144,145). In addition to ICB therapy, the use of genetically modified NK cells, such as chimeric antigen receptor (CAR)-NK cells, in tumor immunotherapy has also attracted increased attention (146–148). However, engineering a CAR-NK structure requires a biomarker specifically expressed on the surface of ectopic endometrial cells, which, to the best of our knowledge, has not yet been discovered.

Although the aforementioned targets for the treatment of EM have been supported in theory by in vitro and animal experiments, limited clinical trials have been reported.

The exact mechanism behind how immunosuppression in ectopic endometrial tissue and its environment damages the cytotoxicity of NK cells is unclear, which makes it difficult to identify an appropriate immunotherapy target. Three main specific inhibitory NK cell receptor families have been identified: KIR, LILRB and NKG2. To the best of our knowledge, there are no reports on the inhibition of these receptor/ligand interactions. Cytokine therapy and the upregulation of associated activated receptors also requires further research. The possibility of utilizing immunotherapy in the treatment of EM needs further analysis due to the lack of tissue/cell specificity, which results in systemic side effects. It requires investigation on the epigenic differences between the ectopic and eutopic cells to develop treatments with increased specificity.

It should be noted that whether enhanced NK cell cytotoxicity is associated with abortion is debatable (149,150). Further examinations and analyses are needed before NK cell treatment can enter clinical research.

Macrophages are a late differentiation cell type of the mononuclear-phagocyte system, which have an important role in both the non-specific and specific immune response. Macrophages were previously considered to be solely derived from blood monocytes, which are widely distributed and participate in the innate immunity of the body (151). This notion has changed due to the discovery of macrophages derived from and residing in specific tissues without the participation of circulating monocytes (152). Macrophages can be polarized into different directions based on the effects of different microenvironments and stimulating factors. Based on the surface markers of polarized macrophages and their functions, polarized macrophages can be categorized into two types: Classically activated macrophages (M1) and alternatively activated macrophages (M2) (153,154). M1 has a pro-inflammatory effect on the early stages of inflammation, phagocytizes and digests foreign pathogens, secretes pro-inflammatory factors, activates the T cell-dependent immune response and promotes the Th1 immune response. M2 can promote tissue repair and wound healing, regulates the Th2 immune response and contributes to disease recovery during the later stage of inflammation, which results in an anti-inflammatory effect (153,154).

There are two types of macrophages in the female pelvis: Endometrial and peritoneal. Endometrial macrophages (eMs) are involved in triggering and regulating the process of endometrial breakdown, and the subsequent repair of the endometrial functional layer by facilitating cell proliferation and angiogenesis (155,156). eMs function in the following three ways: i) Production and release of VEGF to promote angiogenesis; ii) participation in triggering and controlling the shedding process; and iii) facilitating and rebuilding the functional layer (157). Peritoneal macrophages (pMs) are distributed in ectopic endometrial tissues outside of the reproductive tract. The increased macrophages in patients with EM are mainly pMs. pMs have an immune monitoring role on the peritoneal surface. pMs can be classified into resident pMs and monocyte-derived pMs of bone marrow origin (158,159).

Based on the differences in MHCII and F4/80 expression levels, pMs can be divided into two phenotypes: Big, tissue-resident pMs and small, monocyte-derived pMs. Both types of macrophages can be either polarized into M1, which is pro-inflammatory, or M2, which is anti-inflammatory, depending on the stimulation of pathogen-associated molecular patterns (51). The pro-inflammatory M1 phenotype of pMs, similar to the classification of helper T cells, is activated mainly through the activation of IFN-γ, LPS, TNF-α or a combination of the three. The anti-inflammatory M2 phenotype of pM is mainly activated by IL-1, IL-10 and IL-13. The polarization of pMs produces corresponding molecular markers, which allows researchers to detect the regulation of macrophages (160–162). Another type of macrophage, tumor-associated macrophage, has a role in the nutrition and angiogenesis of patients with EM and endometrial cancer (Fig. 3) (64).

Although the number and activation of pMs in patients with EM are increased, the phagocytic capacity of these pMs is still unable to remove the ectopic endometrial tissue debris. pMs obtained from women with EM show a reduced capacity for phagocytosis due to the decreased expression level and activity of matrix metalloproteinase-9, which is regulated by prostaglandin E2 (PGE2) and is the enzyme that is necessary in the degradation of the extracellular matrix (163–165).

Ectopic endometrial tissue is in a hormonal environment that contains abnormal concentrations of estrogen and androgens. The secretion of C-C motif chemokine ligand 2 (CCL2) by endometrial stromal cells is upregulated in the ectopic milieu, which has been confirmed to be mediated by estrogen (166). CCL2 mediates the polarization of macrophages to M2 instead of M1 (167). The abnormal EM environment also promotes the elevation of distinct anti-inflammatory phenotypes of macrophages, forming an immunosuppressive microenvironment by stimulating the proliferation of epithelial and stromal cells in endometriotic foci, and promoting angiogenesis (168). However, chronic inflammation is still observed in the lesion microenvironment. It could be possible that the upregulation of M2 is compensatory, induced by persistent inflammation and tissue repair. According to the macrophage depletion study by Bacci et al (169), anti-inflammatory M2 induced by macrophage colony-stimulating factor and IL-10 is considered to be of importance to the growth and development of ectopic EM tissue, while pro-inflammatory M1 induced by IFN-γ is capable of eliminating the ectopic tissue (169). The phenotypic plasticity of pMs makes it possible to investigate potential therapeutic targets for EM based on the suppression of the M2 phenotype in pMs or the activation of the M1 phenotype. The suppression of M2 polarization has already been proposed as chemical therapy for colon tumors, such as by using ovatodiolide to prevent the polarization of M2 tumor-associated macrophages (170).

Estrogen receptors on macrophages can be classified into surface receptors and nuclear receptors. Estrogen nuclear receptors can be divided into ER-α (ER1) and ER-β (ER2). ER2 promotes inflammation and disease progression by increasing the production of inflammatory cytokines, including IL-1β and IL-6 (171,172). IL-6 mediates the recruitment of monocytes and their differentiation into macrophages, which contributes to the increased macrophage infiltration into the EM lesions. ER2 also inhibits apoptosis by interacting with the NLRP3 sensor, caspase 1 and apoptosis signal regulated kinase-1 (173). However, chloroindazole, an ER2 ligand developed in 2015, can suppress inflammation and angiogenesis within the EM lesion, thereby suppressing EM progression (174). These data indicate that the activation of ER2 can also be anti-inflammatory and can serve as a possible target for EM treatment.

ER1 has two main roles. Firstly, ER1 promotes the secretion of pro-inflammatory cytokines, such as IFN-1, contributing to the inflammatory response (175,176). Secondly, ER1 activation also inhibits the NF-κB pathway, which limits the extent of the inflammation (177,178).

Upregulation of ER2 expression levels and downregulation of ER1 expression levels in macrophages and endometrial stromal cells will result in an extremely low ratio of ER1:ER2. There are controversies on whether the influence the ER1 deficiency and ER2 overexpression have an inflammatory or anti-inflammatory effect on the EM environment due to the opposing findings of previous studies (172,179). Despite these controversies, the consensus is that the dysregulation of estrogen and ERs contributes to inflammation in EM. The regulation of inflammatory pathways and immune cells by ERs and estrogen in endometriotic stromal cells will be discussed in further detail below.

Estrogen receptors on the cell surface are G protein-coupled ERs (GPERs), which are expressed on the surface of macrophages in this hormonal environment (180). GPERs are seven transmembrane-spanning receptors that bind to estrogen and mediate rapid non-genomic signaling pathways, such as the mitogen-activated protein kinases (MAPK) pathway and the phosphatidylinositide-3-kinases/Akt (PI3K/Akt) pathways, which can be initiated within seconds and rapidly induce a physiological response in target cells (181). GPER expression levels in macrophages within ectopic endometrium are increased, suggesting that this abnormality may be important to the regulation of the macrophage immune response (182). The roles of GPERs in EM have been gradually discovered and reported, making them a promising target for EM therapy. Activation of GPERs by their agonist G-1 can inhibit the secretion of TNF-α and IL-6 that was induced by LPS, resulting in an anti-inflammatory effect on GPER-expressing macrophages (183).

Continuously elevated estrogen levels also lead to the synthesis and secretion of inflammatory cytokines by macrophages, such as IL-1, IL-6 and TNF-α, which trigger a series of pro-inflammatory responses (173,184). Macrophages have a two-way response to estrogen: Upregulation of pro-inflammatory cytokines is induced by comparatively low concentrations of estrogen and inhibited by increased concentrations of estrogen (185). It has been hypothesized that the function of macrophages in ectopic endometrial tissue may be estrogen-dependent, and that estrogen may regulate the immune response through the GPERs and ERs of macrophages in ectopic endometrial tissue (182,186). These observations indicate that whether these estrogen-dependent events have a pro- or anti-inflammatory role on macrophages in EM depends on the types of ERs and the local concentration of estrogen within the lesion.

The chronic inflammatory environment in ectopic endometrial tissue triggers the secretion of inflammatory cytokines, such as IL-1β, IL-17A and TNF-α (187). Increased inflammatory cytokines can lead to the abnormal activation of the mTOR/PI3K signaling pathway, and the abnormal activation and proliferation of keratinocytes (188). Whether the same inference can be made on macrophages in EM requires further study. In addition, inflammatory cytokine IL-6 in inflamed tissue acts as a superior coordinator of protein synthesis capacity and cell growth rate by stimulating the translation of c-Myc mRNA, an oncogenic transcription factor that activates transcription via all three nuclear RNA polymerases. RNA polymerase I-associated transcription factors are recruited to rDNA by IL-6 when quiescent cells are stimulated to re-enter the cell cycle. Stimulated c-Myc mRNA will therefore eventually lead to an upregulation of rRNA transcription and enhanced proliferation of macrophages (189).

Macrophages are also involved in the induction of the immunosuppressive peritoneal environment in EM. IL-8 secreted by macrophages increases the expression level of Fas ligand (FasL) in endometrial cells, and the binding of FasL to Fas on T cells triggers apoptosis (190). It has been reported that the mRNA level of IL-8 in the peripheral blood and peritoneal fluid of patients with EM is considerably increased. The expression level of FasL on endometrial cells is also increased, which contributes to the formation of the immunosuppressed and immune tolerant microenvironment that is conducive to the adhesion of ectopic cells (191). Serum IL-8 levels have the potential to be an early indicator of EM (192). However, there seems to be little research assessing relevant immunotherapeutic targeting of Fas/FasL for EM. Other immunosuppressive cytokines are also secreted by macrophages in ectopic endometrial tissues, including IL-10 and TGF-β, leading to the inhibition of NK cells in the peritoneal cavity (Fig. 4) (59).

Similar to malignant tumors, the expression level of oncogenes and tumor suppressor genes in EM is considerably altered. Studies have shown that c-Myc, a recognized oncogene, is overexpressed in most patients with EM. It has been proposed that c-Myc also participates in the pathogenesis of EM (189,193).

In addition to the aforementioned inflammatory cytokines that have important roles in the pathogenesis of EM, Tie2-expressing macrophages in ectopic endometrial tissue inhibit endothelial cell apoptosis by preventing the caspase-3 activation of neovascular endothelial cells. This may also be used as a potential therapeutic target for EM (194).

EM is characterized by considerably increased levels of macrophage migration inhibitory factor (MIF), a multipotent protein that has a range of immune regulatory functions and is a key upstream regulator of both the non-specific and specific immune responses. During the onset of premenstrual syndrome, patients with EM have elevated MIF levels in the normal endometrium, early ectopic endometrium, peritoneal fluid and systemic circulation (195–198). Studies have shown that MIF and its specific inhibitors can be used not only to improve the accuracy of EM diagnosis, but also to develop new therapeutic strategies against EM. The study by Seeber et al (199) found that the combined use of MIF and factors such as CA-125, monocyte chemoattractant protein 1 and leptin, can improve the accuracy of EM diagnosis to 93% (199). The concentration of macrophages in the normal endometrium of patients with EM is considerably reduced when compared with that of healthy controls, which predisposes patients to a poor prognosis (200).

It has been suggested that abnormal macrophage regulation may also be associated with the various clinical features of EM. For example, the increased concentration of pMs may disturb normal fertilization and lead to infertility in women with EM (201). Decreased insulin-like growth factor-1 (IGF-1) production by macrophages may also be associated with pelvic pain associated with EM (202).

In conclusion, patients with EM have increased macrophage levels and activation mediated by the mTOR/PI3K signaling pathway, c-Myc oncogene expression levels, and its resulting ribosome biogenesis in EM lesions and peritoneal fluid. Macrophages that co-exist with ectopic EM cells in vitro are immunosuppressive and macrophages in the peritoneal fluid of women with EM exist as a mix of both pro-inflammatory and anti-inflammatory cells. Anti-inflammatory macrophages in peritoneal fluid, which are M2, promote the development of EM lesions, while pro-inflammatory macrophages, which are M1, are antagonistic. Increased concentrations of estrogen in the ectopic EM microenvironment promote re-polarization from M1 to M2, which further contributes to the growth of the lesion.

Several theories for immune-associated therapies targeting the upregulation of pMs in EM have been proposed. The main hypotheses regarding the upregulation of pMs in the peritoneal fluid of patients with EM include the estrogen dependency theory, the mTOR/PI3K signaling pathway theory, overexpression of the oncogene c-Myc, ribosome biogenesis and the overexpression of MIF. Several potential medication therapies for managing these etiologies have been proposed.

Regarding the estrogen dependency theory, estrogen replacement and other associated therapies are being investigated and introduced into clinical management. The estrogen replacement therapy is currently the most commonly used method for treating EM that can achieve complete ovarian suppression, and has been in use since it was first reported in 1948 (203). Low-doses of combined estrogen and progestin or progestin alone can effectively relieve the clinical symptoms of pelvic pain caused by EM, and can also reduce the adverse effects of low estrogen that are induced by GnRH agonists (204). However, GnRH therapy can lead to several clinical side effects, including increased follicular development (205).

In addition to the estrogen-dependent theory, the mTOR/PI3K signaling pathway theory is also important, where mTOR/PI3K functions as the upstream regulator of ribosome biogenesis and plays a key role in protein synthesis (206). The mTOR/PI3K inhibitor GSK2126458 and the RNA polymerase 1 inhibitors CX5461 and BMH21 have been developed, all of which have shown very good therapeutic effects in a mouse model of human EM (206).

In terms of the MIF theory, the study by Khoufache et al (207) showed that the specific antagonist of MIF (ISO-1) can effectively reduce the growth and progression of ectopic endometrium, indicating that this agent has good clinical potential (207). The research and development of other associated drugs is still in the experimental stage and requires additional attention.

The complement system is an indispensable part of the innate immune response and is involved in the identification and elimination of pathogens and abnormal cells, such as apoptotic and necrotic cells (208–210). The complement system recognizes and tags pathogens and altered or transformed self-cells, thereby activating the inflammatory response and modulating the adaptive immune response, and ultimately leading to the lysis of target cells or pathogens (211). The complement system is a functionally complex system that can trigger a severe immune response or inflammatory process (211). This system may be harmful to the body when it is excessively or abnormally activated in conditions such as inflammation or tissue damage (212–215). The complement system has a considerable role in peritoneal inflammation, which is associated with the early stages of EM (216).

The complement system is formed from a group of small proteins that demonstrate enzymatic activity after activation and exist in the serum and tissue fluid of healthy individuals and animals. The components of the complement system are extremely complex and variable. The study by Aslan et al (217) reported that 23 out of 84 immune response genes were upregulated in patients with EM, two of them considerably so. Some of these differentiating molecules were later confirmed to be members of the complement system (217).

In a previous study, most components of the complement system that are associated with EM were found to be upregulated in patients with EM (218). To the best of our knowledge, a limited number of complement components were found to be decreased. Such decreased components included mannose/mannan binding lectin-associated serine protease-1 (MASP-1), and several remain controversial (219).

In a previous study, the quantities of C1q and C1INH in the peritoneal fluid of patients with EM at various stages were considerably increased compared with those in normal controls (220). Moreover, increased levels of C1q and C1INH were found in the peritoneal fluid of early stage EM (220). These results suggest that immunoglobulins participate in the initiation of the classical pathway in ectopic endometrial tissue, especially during early EM (220). Furthermore, C1-associated genes (including C1QA, C1QB, C1R and C1S) and C2 genes were increased in ectopic endometrial tissues compared with those in healthy controls (221). C3 is usually expressed in the ectopic tissue of patients with EM despite its regular expression levels in the glandular epithelium of normal endometrium (222,223). The overall C3 levels in the peritoneal fluid and peripheral blood of patients with EM were increased compared with those of normal controls, especially C3c and C3b (219,224–227).

A growing body of evidence has reported that the C3 levels in the serum of patients with EM were considerably upregulated, in particular among patients with mild EM, when compared with that in healthy subjects and patients with severe EM (228–231). Other studies found that the C3 levels in the eutopic endometrium of patients with EM were also considerably increased under the influence of ectopic endometrial tissue (228,232). Elevated iC3b in the peritoneal fluid of patients with EM may negatively regulate NK cell activity via the iC3b/CR3 signaling pathway, thereby downregulating NK cell cytotoxicity (219,233).

The levels of other members of the complement system, such as C5, C6, C7 and C8A, were also upregulated in the ectopic endometrial tissue of patients with EM (217,221). C6 levels were considerably increased in patients with early stage EM compared with those in their healthy counterparts (234). C7 was also upregulated in ectopic endometrial tissue (221).

The expression levels of complement factor (CF)B, CFD, CFH and CFI in the complement system are upregulated in ectopic EM tissue, while the expression level of MASP-1 is downregulated (Fig. 5).

There are controversial theories about the relationship between C4 and EM. Studies have shown that the concentration of C4 in the peripheral blood and peritoneal secretions of women with EM are increased compared with those of normal controls (219,224,225). C4a was considerably decreased in the peritoneal fluid of patients with peritoneal and ovarian EM (218,235). The C4A/B gene expression level was upregulated in ectopic endometrial tissues, while complement component 4 binding protein α (C4BPA) expression level was reduced (221).

Two studies have examined the membrane attack complex (MAC; also known as SC5b-9) in patients with EM. One study found that the MAC levels in the peritoneal fluid and peripheral blood of patients with EM were increased. The concentrations of terminal complex were also increased in patients with advanced EM (219). However, another study found no considerable difference in the MAC level in the peritoneal fluid in patients with EM compared with that in normal controls (236,237).

The relationship between mannose-binding lectin (MBL) and EM is also controversial. Some studies have shown that equivalent MBL levels exist between normal controls and patients with EM (238,239), while the study by Sikora et al (220) observed an increased level of MBL in the peritoneal fluid of patients with EM. Furthermore, the concentration of MBL in patients with early EM was increased compared with that in patients with late EM (218,220,236,237,240).

Complement C3 inhibitors can interrupt the inflammatory cascade at its earliest stage and reduce the production of iC3b, which weakens NK cytotoxicity. The blockade of C5a and C3a to induce macrophage activation, C1q inducing the transformation of macrophages to the M2 phenotype and angiogenesis in EM lesions via complement immune therapy could all be promising targets for EM treatment (241,220).

Estrogen and progesterone are two key sex hormones that are closely associated with the occurrence and progression of EM. As previously discussed, estrogen mainly exerts its functions by interacting with ER and inducing an inflammatory environment. ER2 is associated with the inhibition of the inflammatory response. Increased ER2 activity can promote cell survival by inhibiting TNF1-mediated apoptosis, participating in growth factor signaling and promoting epithelial mesenchymal transition (242).

The close relationship between sex hormones such as estrogen and progesterone and the immune system has been frequently demonstrated. Estrogen can induce the activation of the immune response and immune cells through nuclear receptors. The dysregulation of estrogen and progesterone signaling in EM are termed estrogen dominance and progesterone resistance (243,244).

The binding of progesterone to progesterone receptor (PR) in epithelial and stromal cells inhibits epithelial cell proliferation and promotes decidualization (245). These effects of progesterone are achieved by the integration of the response through two functionally different subtypes of PR: PR-A and PR-B. These two subtypes share the same gene but have separate promoters, which makes their structure and function distinct from one another (246). PR-A is recognized as the initial driver of uterine PR function, while PR-B is key to progesterone-induced morphogenesis during pregnancy, and mainly improves progesterone reactivity by maintaining an appropriate ratio to PR-A (247).

Estrogen and its two nuclear receptor subtypes have already been briefly introduced in the preceding sections. Estrogen promotion of epithelial cell proliferation and endometrial stromal decidualization is also mediated by the binding of estrogen and its receptors. The two receptor subtypes, ER1 and ER2, are transcribed by different genes (248).

ER1 is expressed in most cells of the immune system, while ER2 is limited to certain cell types of some immune organs, such as lymphocytes in human lymph nodes, bone marrow and thymus. Therefore, ER1 has a stronger impact on the immune system than ER2 (220). Both ER subtypes are expressed in the endometrium, with the expression levels of ER1 outnumbering those of ER2 (249). ER1 also has a more important role in promoting the proliferation of endometrial epithelial cells, implantation and fertilization than ER2 (221). The response induced by estrogen binding to ER1 may be mediated by slower genomic signaling pathways such as the IGF-1-PI3K/AKT pathway (250). As aforementioned, ER2 is associated with the inhibition of the inflammatory response. Increased ER2 activity can promote cell survival by inhibiting TNF1-mediated apoptosis, participating in growth factor signaling and promoting epithelial mesenchymal transition (242).

In addition to the two aforementioned types of sex steroid hormones, the abnormal elevation of prostaglandins (another hormone that induces an inflammatory response) is also involved in the pathophysiological changes of EM. Increased PGE2 is detected in both the eutopic and ectopic endometrial tissues of patients with EM. PGE2 participates in the direct and indirect induction of pain through positive feedback with estradiol (E2). In this positive feedback loop, E2 activates cyclooxygenase II (COX2) to promote the production of PGE2, and the upregulation of PGE2 level in turn promotes the expression of steroidogenesis-associated genes and aromatase, thereby increasing E2 production (251). Sensitivity to IL-1β is important in the regulation of COX2, which contributes to the maintenance of sex hormone-associated inflammation in EM lesions (165).

Abnormal regulation of sex hormone nuclear receptors in patients with EM has been reported (252). The main dysregulations of sex steroid hormones in EM can be classified into two types: Estrogen dominance and progesterone resistance.

Estrogen dominance refers to estrogen-induced cell proliferation and inflammation. The estrogen response is primarily triggered by ER1 and ER2. These two receptors have different behaviors, and thus the expression level and ratio of these two receptors are important to determine effects of estrogen on EM. The ratio of ER1:ER2 is decreased in ectopic endometrial tissue of the ovary. This decreased ratio is caused by the upregulation of ER2 and the downregulation of ER1 due to changes in the methylation level of their promoters (249). Decreased methylation of the ER2 promoter leads to increased expression levels of ER2, while increased methylation of the ER1 promoter leads to decreased ER1 expression levels (251,253). In addition to epigenetic changes, crosstalk between ER1, ER2 and PRs has also been found to be important. ER2 directly downregulates the expression of ER1 by binding to the promotor region of ER1 (254). The downregulation of ER1 contributes to the reduction of PR and further promotes the development of EM and infertility (255).

ER2 upregulation in EM can activate a variety of proliferation- and inflammation-associated signaling pathways, such as the COX2-PGE2 feedback loop, which may be the main reason for increased lesion survival, cell proliferation and inflammation (249). Other research has found that ER2 can interact with inflammatory factors to regulate apoptosis and the inflammatory response, which is also associated with the pathogenesis of EM (172).

The other sex steroid hormone dysregulation in EM is progesterone resistance, in which normal and ectopic endometrial tissues in patients with EM do not respond to progesterone (256). Little is known about the mechanism behind progesterone resistance. Several studies have suggested that the downregulation of progesterone receptors may be a potential contributor to progesterone resistance, but this remains controversial (257,258).

Studies have shown that the hormonal treatment of EM is feasible (249,259). Current hormone therapies include GnRH agonists, aromatase inhibitors, COCs containing progesterone and E2, progestin-based therapies and androgen therapy (25,260,261). Hormone therapy aims to inhibit lesion growth or control pelvic pain by reducing the estrogen response and promoting the progesterone response (25,262). However, hormone therapy interferes with ovarian function, homeostasis and individual immunity, resulting in numerous side effects that include weight gain, androgen effects, reduced bone density, infertility and other adverse effects (262). Clinically, hormonal therapies are currently the most effective drugs for the treatment of EM (1,260). COCs are generally composed of a specific proportion of estrogens and progestogens that can inhibit steroid production in the ovary to treat chronic pelvic pain caused by ectopic endometrium (263). Although COCs are effective against the post-operative recurrence of EM, the estrogen contained in COCs still carries a risk of aggravating progestin resistance. Progestogen-based hormone therapy is another ideal treatment for EM. Medroxyprogesterone acetate (MPA) has been effective at reducing the pain caused by EM and reversing the decreased bone mineral density caused by low estrogen levels (264). A decrease in ER and an increase in PR in endometrial tissue can be found in patients with EM using MPA and other progestin-based drugs (265,266). GnRH agonists are second-line drugs after hormone therapy, which can reduce the production of estrogen, weaken E2 dominance and downregulate pituitary function through negative feedback (261). Although GnRH agonists are effective at alleviating the pain caused by EM, their side effects, such as bone loss, greatly limit their clinical use. As aforementioned, there are still several obstacles that block the treatment of EM with hormone therapy, including unresponsiveness to progesterone caused by progesterone resistance and the adverse effects of hypoestrogenemia.