Angiogenesis signaling in endometriosis: Molecules, diagnosis and treatment (Review)

At present, reliable laboratory biomarkers for EM pathology are difficult to obtain and there are no uniform international standards ( 1 ). The gold standard for diagnosis of EM remains the pathological findings following surgical treatment ( 61 , 62 ). There is often a 6–12-year hiatus between onset and diagnosis of EM, with the delay being caused by multiple factors, including economic deprivation and uneven distribution of health services ( 63 , 64 ). Reliable biomarkers found in the biological fluids of affected patients may be an expedient diagnostic tool for EM, which would also facilitate an objective assessment of treatment efficacy. Although lacking both specificity and sensitivity for this pathology, the most representative glycoproteins used as biomarkers for EM include cancer antigen CA-125 and CA-199, which have similar specificity and reflect the severity of disease ( 65 , 66 ). However, in total laparoscopic ectopic lesion removal, high CA-199 expression is markedly associated with a high postoperative recurrence rate; monitoring the expression of CA-199 may help predict the progression of EM ( 67 ). Vascular cell adhesion molecule-1 (VCAM-1) and intercellular adhesion molecule-1 (ICAM-1) are members of the integrin adhesion protein family, expressed by endothelial and other cells, and are involved in inflammatory responses, immunity, tumor immune escape and other processes ( 68 , 69 ), potentially due to involvement of the NF-κB pathway in TNF-α-induced upregulation of ICAM-1 and VCAM-1 in vascular endothelial cells ( 70 , 71 ). Kuessel et al ( 72 ) showed that the mRNA levels of both VCAM-1 and ICAM-1 are high in normal peritoneal samples of patients with EM and that serum soluble VCAM-1 (sVCAM-1) levels are not influenced by the entity of the lesion, the stage of menstrual cycle or the severity of the disease. The aforementioned study further indicated that the specificity and sensitivity of serum VCAM-1 for diagnosis of EM is 80 and 84%, respectively. It is noteworthy that the diagnostic specificity and sensitivity of sVCAM-1/sICAM-1 ratio, which further improved this predictive performance, were 86.7 and 90.3%, respectively. In addition, elevated levels of sICAM-1 may impair natural killer cell activity, which in turn accelerates the progression of EM ( 70 ). VEGF, one of the most potent pro-angiogenic factors, is expressed at elevated levels in patients with EM and may be a potential biological molecule for the diagnosis of EM ( 10 , 50 ). A previous study demonstrated that degree of VEGF elevation is associated with the stage of EM, with additional marked elevation in advanced stages ( 24 ). Serum microRNA (miRNA/miR) is a non-coding RNA consisting of 21–23 nucleotides widely found in eukaryotes, which bind exosomes or to specific protein complexes to protect against endogenous RNase degradation, playing an active role in the regulation of gene expression and cell cycle ( 73 , 74 ). Several studies ( 75 , 76 ) have shown that multiple miRNAs regulate angiogenesis and inflammatory responses in the ectopic endometrium by regulating expression of molecules such as VEGF, MMPs, IL and TGF-β1. Consequently, dysregulation of their expression serves an essential role in the pathogenesis of EM ( 75 , 77 ). Compared with conventional ultrasound, miRNAs have the advantages of improved stability and specificity, simpler detection methods and lower detection costs, they are becoming potential molecular indicators for non-invasive identification of EM ( Table II ). An increasing number of studies have been conducted to detect expression of one or more specific miRNAs in plasma by reverse transcription-quantitative PCR to investigate their diagnostic value in EM and it has been hypothesized that the miR-125b, −451 and let-7 families may serve as potential circulating biomarkers for EM ( 78 – 80 ). A previous study showed that serum miR-125b-5p has a sensitivity and specificity of 100 and 96%, respectively, for the diagnosis of EM ( 74 ). In patients with EM, upregulation of miR-125b-5p and downregulation of let-7b-5p expression may be associated with an increase in cytokines such as TNF-α and IL-1β and −6 ( 81 , 82 ). These cytokines are hypothesized to be important components involved in angiogenesis ( 31 , 39 ). Migration inhibitory factor (MIF) is a pleiotropic inflammatory cytokine with upstream regulatory roles in immunity and has been shown to be involved in the development of EM. miRNA-451, which is hypothesized to target MIF, is a mitogenic cytokine that creates a proliferative and angiogenic phenotype for growth of ectopic endothelium ( 76 ). Compared with the healthy population, miRNA-451 expression is decreased in patients with endometrial co-infertility, suggesting that miRNA-451 expression levels are negatively associated with co-infertility in patients with EM and may be used as a marker for disease prognosis ( 83 , 84 ). In addition, elevated miR-145-5p expression levels in EM are associated with upregulation of VEGF-A and decreased levels of epidermal growth factor 2, phosphatase and tensin homolog and C-X-C chemokine receptor type 4 ( 75 ). MiR-199a is another possible suitable biomarker for EM. miR-199a downregulates the expression of VEGF-A in ESCs under hypoxic conditions and partially attenuates angiogenesis in hematopoietic stem cells under hypoxic conditions by inhibiting the HIF-1α/VEGF-A pathway ( 85 ). Maged et al ( 86 ) showed that serum miR-199a and −122 have high sensitivity and specificity and miR-199a has 100% sensitivity and specificity for the diagnosis of EM. Another study showed that VEGF-A is a direct and functional target of miR-199a-5p and elevated miR-199a-5p expression inhibits the proliferation and angiogenesis of MSCs, as confirmed in vitro experiments and EM female C57BL/6j mouse models ( 87 ). Notably, HIF-1 upregulates miR-20a; miR-20a not only increases the expression of the VEGF-A angiogenic gene, but also downregulates anti-angiogenic thrombospondin 1 to promote angiogenesis ( 76 , 88 – 90 ). However, miR-20 as a biomarker for the diagnosis of EM remains to be investigated. In addition, dysregulation of miRNA-202-3p ( 91 ), miRNA-126 ( 92 ), miRNA-200 ( 75 , 93 – 95 ) and other genes mediate VEGF expression to promote angiogenesis and invasion of ectopic endothelium by promoting cell proliferation and angiogenesis. The aforementioned studies suggest that epigenetic abnormalities and dysregulation of miRNAs may play a key role in the formation of EM by affecting different physiological processes, which may also be beneficial for further studies on mechanisms of EM. With the progress of miRNA research, diagnostic value of miRNAs has been further confirmed in the clinical setting. Moustafa et al ( 79 ) first demonstrated in a prospective study that plasma miRNAs accurately distinguish EM from other gynecological diseases, and miR-125b-5p, −150-5p, −342-3p and 451a are notably higher in patients with EM, while miR-3613-5p and let-7b are expressed at lower levels, miRNA expression levels are not markedly associated with menstrual cycle or hormonal drug use. To improve the diagnostic value of miRNA, co-diagnostic methods have been applied ( 96 – 99 ). Results of several studies remain controversial and even contradictory due to the irreproducible nature of the studies ( 12 , 100 ). At present, no single or groups of miRNA biomarkers has sufficient specificity and sensitivity in the diagnosis of EM ( 101 ). Future studies need to adopt uniform and standardized methods to clarify the value of miRNA as a biological marker for early diagnosis and prognosis prediction of EM and for clinical translation and application. The Internet of Medical Thing (IoMT) is the most emerging era of the Internet of Things ( 102 , 103 ). IoMT can generate large amounts of data without human intervention through connected smart medical devices, allowing healthcare professionals to facilitate disease detection by learning complex models and extracting meaningful information from large amounts of data ( 103 ). Based on this, researchers have trained on pulmonary embolism detection datasets and obtained satisfactory results in terms of sensitivity and specificity ( 102 ). To the best of our knowledge, however, its application in EM has not been reported yet. EM is a chronic inflammatory disease and using IoMT to establish a safe and effective system may contribute to the examination and treatment of the disease. Numerous studies reveal ( 69 , 104 , 105 ) the importance and effectiveness of extracting various putative biomarkers from the biological fluids of affected patients. Nevertheless, due to the large heterogeneity between different studies, there is no uniform conclusion and further studies with larger sample sizes are needed to identify sensitive and specific indicators for transition from the laboratory to the clinic. As vascular mechanisms of ectopic endometrium are refined, newer and more accurate diagnostic markers for EM are expected to be discovered.

Endometriosis (EM) is an estrogen-dependent disorder characterized by the presence of endometrioid epithelium and glands outside the endometrium and myometrium ( 1 ). It affects 10–15% of women of reproductive age worldwide and is one of the most frequently occurring diseases among this group ( 1 ). The main clinical manifestation of EM is pain, including typical but non-specific pain, such as dysmenorrhea and chronic pelvic pain, and ~1/3 of patients have infertility, with advanced EM potentially leading to gynecological malignancy, such as ovarian cancer ( 2 , 3 ). The lack of specific clinical manifestations of EM makes it difficult to differentiate from conditions characterized by chronic pelvic pain in clinical practice, which makes the diagnosis and treatment of EM challenging ( 3 ). This results in psychological burden to patients and affects physical and mental health, as well as their quality of life ( 4 , 5 ). The pathophysiology of EM disease is not thoroughly understood. The potential pathogenesis of EM is hypothesized to involve retrograde menstruation, uterine stem cells and somatic epithelium ( 1 ). Among them, the theory of retrograde menstruation is the leading theory of the pathogenesis of EM ( 6 ). However, development of EM is associated with immune alterations and a proinflammatory peritoneal milieu, which determines the progression to EM ( 7 ). It is noteworthy that the formation and growth of ectopic lesions depends on nutrient support, with neovascularization serving an important role in the development of endothelial ectopic lesions ( 2 , 8 ). Vascular permeability is notably increased in ectopic lesions of EM and a large number of new blood vessels appear ( 2 ). This indicates that angiogenesis is a key link in the pathogenesis of EM, which represents a potential target for development of future diagnostic and therapeutic strategies. Recent study of angiogenic mechanisms in EM has revealed the importance and potential role of angiogenesis-related pathways and markers of EM for diagnosis and treatment ( 9 – 12 ). That has provided a new direction for the discovery of novel targets of anti-angiogenic action and therapy and novel lines of thought and strategies for clinical treatment of EM. The present review aimed to summarize advances in the investigation of EM angiogenesis, including availability of non-invasive molecular markers for the diagnosis of EM, for the potential clinical value of anti-angiogenic therapy as an alternative to non-hormonal therapies for treatment of EM.

Literature searches were performed in PubMed ( pubmed.ncbi.nlm.nih.gov/ ) and ISI Web of Knowledge ( webofscience.com ) for English articles with the key words ‘endometriosis’, ‘endometriotic lesions’, ‘angiogenesis’, ‘vascularization’, ‘anti-angiogenic’, ‘anti-angiogenesis’, ‘anti-angiogenic therapy’ and ‘diagnosis’. The search included animal and human studies focusing on EM angiogenesis, diagnosis and treatment, and no restriction was set regarding the publication date.

Although EM is a benign lesion, the mechanism of its angiogenesis has similarities with pathologic angiogenesis mediating tumor and metastasis, which is complicated by multiple factors and mechanisms. The improvement of understanding of the EM angiogenic network regulation system has brought new hope for its diagnosis and treatment. However, the majority of studies are still at the laboratory stage, and to the best of our knowledge, there are no satisfactory clinical trials due to differences in research data and non-reproducibility, making it challenging to apply non-invasive biomarkers and anti-angiogenic drugs in the clinic. In addition, clinical evidence for the efficacy of anti-angiogenic treatment strategies in EM is lacking. Anti-angiogenic therapy may adversely affect normal physiological angiogenesis such as ovulation and wound healing ( 134 , 135 ). Research is required to elucidate the interactions between factors in the angiogenic microenvironment to discover more effective targets for drug therapy. In addition, clinical trials are needed to evaluate the therapeutic value of these serum markers and anti-vascular drugs and to strengthen multi-targeted combined diagnosis and treatment to inhibit angiogenesis.

Angiogenesis is defined as generating new blood vessels from the germination of pre-existing vessels. This is initiated by vascular endothelial growth factor (VEGF), which activates resting endothelial cells in microvasculature to release matrix metalloproteinase (MMP), thereby mediating degradation of the vascular basal membrane so that endothelial cells migrate into the surrounding tissue; additionally, tubular branches of endothelial cells form vascular rings and novel basement membranes, ultimately leading to new vessel formation ( 13 ). Angiogenesis is key for normal physiological processes by spontaneous endometrial thickening and development, serving an important role in the menstrual cycle ( 14 ). In patients with EM, the endometrium forces menstrual blood to the peritoneal cavity leading to induction of oxidative stress and immune responses, releasing related inflammatory factors and cytokines. This results in disrupted homeostasis of the peritoneal microenvironment and a relative increase in pro-angiogenic factors, which leads to neovascularization of ectopic endometrial vessels and establishment of the microvascular network ( 2 , 15 ). EM lesions develop in a complex and dynamic environment that is regulated by a number of molecules, cytokines and associated signaling pathways ( Table I ) ( 16 ). At present, although the etiology of EM is unclear, immune dysfunction has been cited as a key contributing factor in the growth of ectopic lesions of endometrial debris following retrograde menstruation ( 17 , 18 ). Density of immunocompetent cells such as T lymphocytes, natural killer cells, dendritic cells (DCs), macrophages and neutrophils is markedly increased in EM lesions and the peritoneal cavity ( Fig. 1 ) ( 17 , 19 ). Abnormal alterations in inflammatory and immune cells cause alterations in cytokines in the peritoneal fluid of patients with EM. Compared with healthy patients, the expression of interleukin (IL) −1β, −6, −8, −10 and −17, tumor necrosis factor-α (TNF-α) and macrophage migration inhibitory factor is notably elevated in the peritoneal cavity or serum of patients with EM ( 20 – 27 ). These cytokines are not only involved in the chronic inflammatory response to EM, but also promote the implantation and growth of ectopic lesions by upregulating expression of cyclooxygenase 2 (COX-2) and/or VEGF. Previous studies have suggested that enhanced concentration of neutrophils in the peritoneal fluid of patients with EM may be the result of elevated concentrations of potent neutrophil chemokines, such as IL-8, in plasma and peritoneal fluid ( 28 , 29 ). Moreover, neutrophils secrete cytokines such as VEGF and IL-8 and −17 to promote the proliferation and invasion of endometriotic cells and angiogenesis ( 30 ). In this process, estrogen is involved in neutrophil activation, and it was found that in vitro culture in the presence of IL-6, TNF-α, lipopolysaccharide (LPS) and estrogen-enhanced VEGF release from isolated peritoneal macrophages and neutrophils ( 31 ). Conversely, inhibition of estrogen signaling activity decreases the number of neutrophils and expression of proinflammatory cytokines, thereby inhibiting the progression of EM ( 32 ). Immune dysfunction characterized by hyperactive peritoneal macrophages with altered phagocytic ability represents a key point in EM angiogenesis ( 33 , 34 ). Macrophages are hypothesized to be one of the sources of VEGF and fibroblast growth factor (FGF), which promote the growth of diseased blood vessels in EM and accelerate the process of EM development ( 24 ). Previous studies have found that accumulated macrophages are recruited to the ectopic environment, primarily using the alternatively activated macrophage (M2) phenotype, which promotes and enhances proliferation and clonogenic capacity of endometrial stromal cells (ESCs) ( 33 – 36 ). Additionally, M2 cells produce a variety of stimulating factors to induce peritoneal inflammation, thus forming a complex loop of regulatory mechanisms to maintain the altered peritoneal microenvironment of EM to promote immune escape, adhesion and angiogenesis of the ectopic endothelium ( 31 , 37 ). By contrast, nanovesicles derived from M1 macrophages directly or indirectly inhibit migration and invasion of endometrial mesenchymal stromal cells (E-MSCs) to decrease formation of blood vessels ( 35 ). Contrary to macrophages, the role of DCs in pathogenesis of EM is not yet clear. Suen et al ( 38 ) showed that IL-10 secreted by plasma cell-like DCs enhances ESC migration and promotes angiogenesis via the secretion of pro-angiogenic factors, leading to the growth of endometriotic lesions. Furthermore, DCs may promote EM angiogenesis by secreting cytokines such as IL-6 and −12 and transforming growth factor-β (TGF-β) ( 39 ). Regulatory T cells and helper T 17 (Th17) cells are essential for immune defense and immune homeostasis ( 18 ). Imbalance in expression of Th2 and Th1 cells is one of the essential promoters of ectopic endothelial growth ( 40 ). Th17 cells produce IL-17, a strong proinflammatory mediator that stimulates expression of angiogenic and proinflammatory cytokines, and promotes angiogenesis in the ectopic endothelium ( 30 , 41 ). In vitro results suggest that IL-17 may be a stimulus that induces pathogenic polarization of macrophages towards the M2 phenotype ( 42 ). This suggests a synergistic role of inflammation and the Th17 immune response. Previous studies have found numerous similarities between EM angiogenesis and tumor pathologic angiogenesis ( 8 , 43 ). Ectopic endometrial tissue initially has a degree of hypoxia comparable with that of cells in growing tumor centers ( 16 , 44 ). Hypoxia is one of the most potent stimuli for upregulation of angiogenic growth factors; it prevents proteasomal degradation of hypoxia-inducible factor-1α (HIF-1α) ( 44 , 45 ). HIF-1α is required for oxygen-regulated transcriptional activation of genes encoding VEGF to enhance hypoxia-induced angiogenesis. HIF-1α enters the nucleus to form hypoxia response elements and binds the promoter region of the VEGF gene, which in turn upregulates VEGF expression and promotes angiogenesis ( 9 ). In the hypoxic microenvironment of the peritoneal cavity, HIF-1α downregulates expression of chicken ovalbumin upstream of the promoter transcription factor II, which leads to elevated angiopoietin expression and promotes neovascularization of ectopic lesions in EM ( 46 ). Additionally, HIF-1α is involved in immune regulation and inflammatory responses and it can upregulate expression of IL-8 and COX-2, thus initiating IL-8- and COX-2-mediated angiogenic mechanisms and indirectly promoting ectopic endothelial neovascularization ( 47 ). The involvement of HIF-1α in regulating disease is complex and HIF-1α is also involved in cellular autophagy. HIF-1α promotes mesenchymal cell migration and invasion in EM via upregulation of autophagy ( 45 ). Reactive oxygen species (ROS) are a key element in stabilizing HIF-1α. The release of ROS and the expression of VEGF are enhanced under hypoxic conditions, which forms a positive feedback mechanism between ROS and VEGF in enhancing angiogenesis ( 48 ). As a consequence, HIF-1α is an important regulator that promotes angiogenesis and is involved in angiogenesis in EM via the complex regulation of multiple pathways. VEGF is an important pro-angiogenic factor that induces vascular endothelial cell neogenesis and cell proliferation by binding to vascular endothelial cells, increases vascular permeability and promotes blood vessel formation ( 43 , 49 ). VEGF protein family includes VEGF-A, -B and -C, virus-encoded VEGF-D and placental growth factor, with VEGF-A playing pivotal roles in regulation of angiogenesis ( 43 ). The VEGF signaling pathway regulates activity of multiple kinases by binding to cell surface tyrosine kinase receptors, VEGF receptor (R)-1, VEGFR-2, and VEGFR-3, resulting in different biological effects that ultimately result in angiogenesis ( 43 ). In recent years, expression of VEGF has been studied in different experimental EM models and tissue samples from patients with EM ( 10 , 50 – 52 ). Numerous researchers have found that increased expression of VEGF in the serum and peritoneal fluid of patients with EM can serve as an auxiliary diagnostic indicator of EM ( 9 , 10 , 50 ). Li et al ( 10 ) found that angiogenic factors released from ESCs may be influenced by expression of fibrinogen α chain (FGA) and that the expression of pro-angiogenic factors, such as VEGF, platelet-derived growth factor (PDGF), FGF and MMP-2 and −9, is reduced in FGA-knockdown hEM15A cells. FGA may activate the VEGFA/VEGFR-2/FAK signaling pathway in EM and promote angiogenesis by regulating expression of VEGF-A and MMPs in EM ESCs. In addition, VEGF-C expression is upregulated in EM cells, and it has been demonstrated that VEGF-C enhances lymphangiogenic capacity of lymphatic endothelial cells via extracellular vesicle transport in an autograft mouse model of EM, whereas blocking the VEGF-C signaling pathway attenuates development of local chronic inflammation and EM ( 50 ). The expression of VEGF in EM is regulated by expression of multiple factors and related pathways and plays a vital regulatory role in the process of neovascularization in EM. The stimulation of endometrial and endometriotic cells leads to activation of different intracellular pathways and associated signaling molecules. Nuclear factor-κB (NF-κB) is a transcription factor that mediates inflammatory signaling pathways and is a driver of inflammation and plays a crucial role in the development and regulation of inflammation ( 2 ). NF-κB is expressed in trace amounts in normal endometrium, but it is expressed at high levels in EM, suggesting that activated NF-κB serves an important role in regulating the development of ectopic endometrium ( 26 , 53 ). Gou et al ( 33 ) showed that estrogen receptor β (ERβ) regulates production of C-C motif chemokine ligand 2 through activation of the NF-κB signaling pathway in B cells, thereby recruiting macrophages to ectopic lesions in ESCs to promote pathogenesis. Additionally, ERβ stimulates expression of genes associated with the unfolded protein response in normal endometrium, inhibits the IL-6/JAK/STAT3 signaling pathway and suppresses the TNFα/NF-κB signaling pathway, leading to endometrial dysfunction associated with EM ( 54 ). In addition, oxidative stress is associated with development of EM and is a key inducer of the NF-κB signaling pathway in EM cells, extracellular high mobility group box-1 (HMGB-1), a prototypical molecule of damage-associated molecular patterns, inducing an inflammatory response via HMGB-1/TLR4/NF-κB axis ( 2 , 53 ). Excessive production of ROS in the pelvis of patients with EM is an important inducer of NF-κB-mediated chronic inflammatory responses, and NF-κB activation is primarily regulated by ROS, regulating the expression of cytokines, such as IL-10, in EM by activating the NF-κB pathway ( 26 ). Oxidative damage to DNA in EM increases DNA fragmentation and activates NF-κB signaling to promote transcription and expression of inflammatory factors such as TNF-α, and IL-1, −8 and −6 ( 26 , 55 ), while NF-κB is activated by TNF-α, IL-1β and LPS ( 56 ), thus creating positive feedback to promote alteration of the abdominal inflammatory microenvironment and the development of ectopic endothelial lesions. Additionally, the NF-κB signaling pathway decreases expression of antioxidant enzymes by activating the nitric oxide synthase (NOS)/NO signaling pathway, which is involved in oxidative stress ( 53 ). In addition to VEGF, other factors have been shown to be involved in angiogenesis in EM. COX-2 serves as a rate-limiting enzyme for prostaglandins (PGs), induces the production of PGE2 and E2, which increase VEGF expression ( 57 ). COX-2 is rapidly upregulated in response to proinflammatory and pathogenic stimuli, and it is able to induce VEGF synthesis upon stimulation by cytokines ( 26 , 57 ) Furthermore, COX-2 has been shown to serve a vital role in the pathological process of diseases such as endometrial carcinoma and endometrial carcinoma ( 53 , 58 ). COX-2, which is either not expressed or expressed at low levels in normal endometrium, is aberrantly expressed in patients with EM and may regulate VEGF to serve a role in the process of disease development ( 54 , 57 ). MMPs play a key role in various pathophysiological processes, such as adhesion and invasion of ectopic endothelium ( 13 , 59 ). Furthermore, COX-2 can regulate angiogenesis via regulation of MMP-2 activity in EM by the COX-2/PGE2/phosphorylated AKT axis ( 60 ). Blocking the expression of COX-2 and/or AKT inhibits MMP-2 activity and endothelial tube formation and inhibition of MMP-2 and COX-2 notably decreases the number of lesions in a mouse model of EM ( 60 ). In conclusion, the pathogenesis of EM is complex and its angiogenesis is driven by a variety of cytokines mediating pro- and anti-angiogenic effects, such as hormonal, immune and inflammatory oxidative stress. These cytokines interact with each other to form a complex network of regulatory mechanisms and are involved in development of neovascularization in EM via multiple cellular pathways ( Fig. 2 ). Further studies on pathogenesis of EM will contribute to identification of biomarkers or biomarker groups to improve and the diagnostic process in a non-invasive manner.

The current medical treatment for EM focuses on either hormonal modulation to induce a low estrogen state or surgical treatment to remove the ectopic lesions. However, even if the ectopic lesions are removed by surgery, certain patients will still have recurrence and even face the possibility of reoperation ( 106 ). Nevertheless, targeted therapeutic strategies based on anti-angiogenic therapy have been demonstrated in oncology and ophthalmology and have been widely used in clinical practice ( 8 , 43 ). As aforementioned, angiogenesis serves a key role in the development of EM and is essential for survival of ectopic endothelial implants. This shows that inhibition of ectopic endothelial angiogenesis is an important method to treat EM and if the mechanism can be regulated or inhibited or its pathway expression blocked, ectopic endothelial angiogenesis can be inhibited, improving the symptoms of patients and having value for clinical treatment. Currently, the primary anti-angiogenic drugs used in targeted therapy include anti-VEGF antibodies, VEGFR tyrosine kinase and COX-2 inhibitors and dopamine agonists (D2-ags; Table III ). Since VEGF is the most potent angiogenic factor and has an important role in EM ( 24 ), VEGF and its signaling pathway are considered to be the most effective targets for anti-angiogenic therapy, and in treatment of kidney, lung, rectal and cervical cancer ( 13 , 107 ). Bevacizumab and ranibizumab are monoclonal antibodies that bind and selectively neutralize VEGF activity by inhibiting binding of VEGF to VEGFR, primarily VEGFR2 or kinase insertion domain receptor, thereby reducing angiogenesis; additionally, both antibodies are effective against EM ( 51 , 52 ). Blocking VEGF signaling and inhibiting tyrosine kinase activity are potential targets for anti-angiogenic therapy ( 107 – 109 ). Research on VEGFR tyrosine kinase inhibitors has been ongoing and tyrosine kinase inhibitors such as sorafenib and sunitinib have been shown to be effective against EM ( 108 , 109 ). In comparing the effects of pazopanib, sorafenib and sunitinib on the VEGF/VEGFR protein kinase pathway and their role in EM, Yildiz et al ( 109 ) noted that pazopanib is more effective than control and other treatments, reducing EM lesions by ≥45%, but sorafenib exhibited improved modulation of VEGF. Anlotinib is a novel oral multitarget tyrosine kinase inhibitor developed independently in China that can effectively inhibit VEGFR, platelet-derived growth factor receptor, FGFR and c-Kit and exert antitumor angiogenesis effects and has been applied in gynecological tumors such as ovarian cancer; to the best of our knowledge however, no studies have reported on its application in EM ( 107 ). E-MSCs express D2, with notably higher levels of ectopic endometrial expression relative to normal endometrial tissue ( 110 ). D2-ags decrease tumor size by targeting abnormal angiogenesis in pathological tissue, showing that D2-ags could be used to treat EM ( 111 , 112 ). Cabergoline, a dopamine agonist, exerts anti-angiogenic effects by inducing endocytosis of VEGFR-2 in endothelial cells, leading to decreased and inactivated VEGF-VEGFR-2 binding ( 112 ). D2-ags are as powerful as standard anti-angiogenic compounds in interfering with angiogenesis and lesion size ( 111 , 113 ). In addition, in a prospective clinical trial, D2-ags were shown to be superior to luteinizing hormone-releasing hormone agonists in decreasing size of endometrial tumors, with fewer drug side effects, easier administration and lower price ( 112 ). Quinagolide is a non-ergot-derived D2-ag that has been reported to reduce or eliminate peritoneal EM lesions in patients with EM ( 110 ). Another study found notable inhibition of E-MSCs by quinagolide, which decreases invasion and endothelial differentiation via the AKT signaling pathway, further supporting the theoretical basis for the use of this drug in treatment of EM ( 110 ). The role of quinagolide in EM is currently undergoing phase 2 clinical trials (trial nos. NCT03749109 and NCT03692403 ), which are expected to provide a novel treatment strategy for the future treatment of EM ( 110 ). Celecoxib, a potent COX-2 inhibitor, has been shown to inhibit EM lesions by markedly decreasing the size of ectopic lesions and vascular density in a study exploring the effects of celecoxib and rosiglitazone on implantation and growth of EM-like lesions in mice with EM ( 114 ). Additionally, treatment with peroxisome proliferator-activated receptor γ (PPARγ) activator rosiglitazone has also been shown to inhibit the growth of EM implants, and the combination of celecoxib and rosiglitazone is more potent than single application of celecoxib in the inhibition of ectopic EM lesions ( 114 , 115 ). Telmisartan, a partial agonist of PPARγ, also blocks angiotensin II type 1 receptor (AT1R); combined blockade of AT1R and activation of PPARγ markedly inhibits angiogenesis and growth in EM-like lesions in mice ( 116 ). In addition, Nenicu et al ( 117 ) found that combination of telmisartan and parecoxib in the treatment of endometrioid lesions increases the rate of lesion regression and notably improves therapeutic effects compared with treatment with telmisartan alone. This suggests that the combination of ≥2 drugs may achieve more desirable therapeutic effects than a single drug against angiogenesis. NF-κB, a key factor in signaling, can regulate growth of ectopic endothelium through multiple pathways ( 53 ). Therefore, NF-κB is considered an important target for anti-angiogenic therapy. BAY11-7085, an NF-κB inhibitor, inhibits the viability of EM cells and induces apoptosis in endometriotic cyst stromal cells by inhibiting anti-apoptotic proteins in an in vitro experimental study; its effect on normal endometrial cells is not notable, but, to the best of our knowledge, clinical studies for its treatment have not been reported ( 118 ). Pyrrolidine dithiocarbamate is another potent NF-κB inhibitor that inhibits NF-κB signaling in EM cells, suppresses COX-2 expression, decreases PGE2 production and suppresses EM cell proliferation, angiogenesis and inflammatory responses ( 119 ). In addition, natural extracts such as ginsenoside Rg3 and curcumin decrease EM by regulating apoptosis and inhibiting VEGF expression and angiogenesis via the NF-κB signaling pathway ( 120 – 122 ). The low risk of adverse effects of these natural products indicates their potential clinical value in the treatment of EM. It has been demonstrated that the inhibitory effect of nobiletin on EM is achieved by inhibiting activation of the NF-κB pathway, which markedly decreases the regulation of the expression of angiogenic factors, including VEGF and E-cadherin, in ectopic endometrium ( 123 ). Dienogest is an artificially developed, fourth-generation progestin with strong anti-proliferative effects on EM implants, as well as anti-angiogenic and anti-inflammatory properties ( 124 ). Its mechanism of action may be associated with inhibition of the NF-κB pathway, which is effective in decreasing the size of endometriomas and EM-related pain and has a good safety and tolerability profile, with improved clinical efficacy in young patients with endometriomas who have not yet had children ( 124 ). Moreover, 2-cyano-3,12-dioxooleana1,9-dien-28-oic acid effectively inhibit EM by increasing ectopic cell apoptosis and decreasing angiogenesis by modulating Nrf2 and NF-κB pathways and has anti-fibrotic, anti-inflammatory and antioxidant effects in EM ( 55 ). Numerous researchers have noted involvement of immune inflammatory factors in ectopic endothelial vascularization ( 18 , 27 , 104 ). Therefore, modulation of the immune system may also serve a role in EM treatment. Etanercept, a fusion protein consisting of the human recombinant soluble TNF receptor 2 bound to human Fc antibody subunits, is effective in decreasing the volume of ectopic lesions in rats by decreasing serum VEGF and IL-6 levels, and TNF-α activity ( 125 ). An experiment investigating the effect of rapamycin (RAPA) on EM lesions in severe combined immunodeficient mice found that RAPA inhibits the growth of EM lesions, potentially by suppressing expression of VEGF in the lesions and thus angiogenesis ( 126 ). In addition, similar pharmacological efficacy of interferons (IFNs) and pentoxifylline for treatment of EM has been observed, IFN-β 1a has a notably stronger in vitro inhibitory effect on ESC proliferation and migration than IFN-α 2b, while the efficacy and safety of pentoxifylline in treatment of EM has been reported, but there is insufficient evidence to support its effectiveness in the treatment of infertility and pain in EM ( 127 , 128 ). The data from the aforementioned studies provide a theoretical basis for future clinical trials. With in-depth research on development and targets of EM, the important role of oxidative stress in EM has become increasingly recognized ( 2 , 53 ). In parallel, studies have demonstrated that antioxidants inhibit the development of ectopic endothelium and achieve efficacy in the treatment of EM ( 25 , 129 – 131 ). The effects of antioxidant melatonin and vitamins E and C on EM are all demonstrated in clinical randomized placebo-controlled trials; they notably improve the reduction of chronic pelvic pain and decrease expression of inflammatory markers in peritoneal fluid in patients with EM ( 129 – 132 ). Resveratrol is a naturally occurring synthetic polyphenolic compound with antitumor, anti-inflammatory, antioxidant and anti-angiogenic properties ( 25 ). In a randomized controlled clinical trial study, resveratrol was found to decrease angiogenesis and inflammation in endometrial tissue of patients with EM by downregulating expression of VEGF and TNF-α ( 25 ). Although melatonin, resveratrol and combined vitamin C and E supplements have shown good results in treatment of EM, their use requires further investigation. Recently, Li et al ( 133 ) demonstrated that the iron death inducer erastin could induce ectopic endometrial stromal cell death in C57BL/6 female mouse model of EM, and the ectopic lesions are reduced after treatment with erastin, suggesting that erastin may be a potential drug for treatment of EM, which provides a new idea for the targeted treatment of EM. In conclusion, anti-angiogenic therapy has become a focal point of medical research and has an improved effect in the treatment of EM, which may provide a potential novel treatment for EM ( Fig. 3 ). However, most of the drug application research results are at the stage of in vitro and animal experiments; although some drugs are at the stage of clinical experiments, there is still lack of drug safety and efficacy assessment. Therefore, more clinical trials are needed to assess the value of these drugs clinically.