Pharmaceuticals targeting signaling pathways of endometriosis as potential new medical treatment: A review

1EM is a disease caused by functional endometrial tissues growing in other areas outside the uterine cavity. It is a chronic disease that affects productivity and quality of life in women. 1 The typical presenting symptoms in women with EM include chronic pelvic pain, abnormal menstruation, and dyspareunia. EM occurs frequently in women of reproductive age, and the incidence is approximately 10%. 2 Approximately 40%–60% of women with EM experience dysmenorrhea, and 20%–30% are complicated with infertility. 3

2Although EM presents as benign clinical and pathological manifestations, it has similar characteristics to cancers, including dissemination, invasion, and hyperplasia. It is generally accepted that EM is a hormone‐dependent disease. 4 Estrogen (E2) augmentation and progesterone resistance feature EM pathology, but the mechanism of how this occurs is unclear. Nevertheless, EM has been observed even in the absence of increased E2 production in postmenopausal women. 5 The pathogenesis of EM is dominated by the theory of ectopic implantation of the endometrium, along with multiple factors, such as endocrine, immunity, invasion, and angiogenesis. Retrograde menstruation theory suggests reflux of endometrial tissue through the fallopian tubes during menstruation and implantation into the peritoneal cavity.6, 7 Lymphatic and vascular dissemination theories suggest that endometrial cells disseminate via lymphatic or blood circulation. 8 Stem cell origin theory suggests that undifferentiated peritoneal tissue, ovarian surface epithelial tissue, and endometrium mesenchymal stem cells transform into endometrial‐like tissue in response to retrograde menstrual blood flow and stimulation from chronic inflammatory factors. 9

3EM development is also associated with a combination of genetic variation and environmental factors. First‐degree relatives of women with EM have a seven fold greater risk of developing EM than those without a family history, and the risk of developing the disease in identical twins of women with EM is as high as 75%.10, 11 In recent years, the increased incidence of EM is also thought to be associated with exposure to environmental pollutants. Tetrachlorodibenzo‐p‐dioxin (TCDD) is the most prevalent air pollutant worldwide, and it promotes cytokine secretion. Endogenous E2 exacerbates the effects of TCDD and the interaction of the two chemicals provokes inflammatory responses, induces toxicity, and thus increases the severity of EM.12, 13, 14 Therefore, the pathophysiology of EM is complex, interrelated, and specific, thereby requiring multiple targeted therapies.

4Regardless of EM theories, endometrial cells must complete a serial process of immune escape, survival, adhesion, invasion, and angiogenesis to develop and grow in the ectopic sites. 15 Signaling pathway refers to a series of enzymatic reaction pathways that pass molecular signals into cells through the cell membrane to exert corresponding effects. EM‐related signaling pathways, together with their upstream and downstream regulatory factors, constitute a large and complex transduction system and play an important role in the occurrence and development of EM. Abnormalities in these pathways and their interactions can lead to abnormal proliferation, apoptosis, autophagy, adhesion, invasion, fibrosis, angiogenesis, reactive oxidative stress (ROS), immune system, and inflammatory responses of the ectopic endometrial tissues, thereby promoting its growth and development. Hormonal‐related enzymes, growth factors, inflammatory cytokines and chemokines, such as tumor necrosis factor (TNF)‐α, transforming growth factors (TGF)‐β, prostaglandin E2 (PGE2), prostaglandin‐endoperoxide synthase (COX)2 play important roles in these processes.16, 17 They induce local immune imbalance in the microenvironment to tolerate immune clearance and promote the survival of ectopic lesions. Downstream molecules, such as hypoxia‐inducible factors (HIF)‐1α, matrix metallopeptidase (MMPs), and vascular endothelial growth factors (VEGFs), are dysregulated and play roles in the angiogenesis and growth of EM lesions.2, 15, 16, 17

5Current treatments for EM include surgical and medical therapies. Conservative surgery removes the EM deposits but increases the risk of impairing ovarian reserve, harming other organs, and imposing postoperative recurrence. 18 Therefore, medical therapy (Table 1) always comes first into consideration, and the choices depend on multiple factors, such as symptom severity, conceive desire, and comorbidities. Generic classes of medical therapies for EM include hormonal therapy, including oral contraceptives (COC), progesterone and gonadotropin‐releasing hormone (GnRH) agonist and antagonist, and nonhormonal therapies such as nonsteroidal anti‐inflammatory drugs (NSAIDs).

6The available reports on the effectiveness of NSAIDs on pain relief in EM are very limited, and there is no strong evidence to support a conclusion. 1 Among all medical treatments, combined COC and progestin monotherapy represent the first‐line therapy, which can be applied to most women clinically diagnosed with EM with or without a surgical diagnosis. 26 Continuous COC effectively reduces the recurrence of dysmenorrhea, 27 and progestin suppresses ovulation by maintaining a hypoestrogenic state. Women with risk factors such as thrombosis and myocardial infarction may tolerate the side effects of progestin better than those of COC. 28 To date, few derivatives of progesterone, namely, depot medroxyprogesterone acetate and norethindrone acetate, have been approved by the US Food and Drug Administration (FDA) as the sole therapy for EM.29, 30

7Although GnRH is an effective hormonal treatment for EM, severe hypoestrogenic symptoms limit long‐term compliance.31, 32 GnRH agonists are second‐line hormonal therapies that exert strong action on the GnRH receptor, leading to an initial short stimulation and subsequent suppression of gonadotropin secretion. Decreased hormone levels result in the dormancy of endometriotic lesions. Owing to its long‐term adverse effects, especially osteoporosis, an add‐back therapy is recommended. 33 Recently, the FDA approved elagolix, a nonpeptide small molecule GnRH receptor antagonist that suppresses luteinizing hormone and follicle‐stimulating hormone and correspondingly reduces E2 and progesterone, as a treatment for moderate to severe EM‐associated pain. Its efficacy was shown after a 6‐month treatment, but it also caused a significant decrease in bone mineral density as the main side effect. 34 To overcome EM refractory to current hormonal treatments and NSAIDs, there have been extensive research of new medicines in recent years. Other than therapeutic efficacy, the potential use of a drug as a preventive treatment after surgery is also desirable. The recurrence of EM and the associated symptoms within 5 years after laparoscopy is approximately 19% in patients with endometrioma, 35 and up to 10% of women require secondary surgery after 1 year, 36 emphasizing the need for new medical treatments to prevent a recurrence.

8In summary, to identify and develop new pharmaceuticals for EM treatment, understanding the dysregulated molecular and signaling pathways in EM development is essential (Figure 1). Numerous studies have focused on the antiproliferation mechanism and related targeted therapies in EM models and/or endometrial cells.37, 38 Owing to the interaction of different signaling pathways, the efficacy of potential pharmaceuticals in promoting or inhibiting a single signaling pathway is often very limited. Therefore, pharmaceutical targeting multisignaling pathways in EM has become important in the medical treatment of EM. An overview of the molecular pathways involved in the pathophysiology of EM has been reported by various publications,39, 40, 41 which provides a high quality evidence of the underlying pathophysiology of EM. However, previous publications only focused on currently available pharmaceuticals. In this review, we aimed to present an updated summary of studies focusing on new potential pharmaceuticals, including preclinical studies, clinical trials, as well as studies on marketed pharmaceuticals. In‐depth studies of signaling pathways targeted by pharmaceuticals are currently an emerging research direction, which will open up broad prospects for the new generation of EM treatment.

9The choice of investigation models considerably influences the translational potential of preclinical research. Endometriotic and endometrial tissue cells with specific cell characteristics, defined by their morphology and phenotypes, confirmed by immunocytochemistry allow in vitro investigations of the mechanism of hormonal expression, cytokine secretion, cell proliferation, and differentiation. 42 Romano et al. 43 critically analyzed different EM culture models of samples from peritoneal, ovarian, and deep infiltration EM and recommended a guideline for assessing the quality of both primary endometriotic cells and immortalized endometriotic cell lines. Culture conditions can imitate EM in situ; for example, endometrium undergoing menstruation, 44 macrophage activation, 45 epithelium mesothelium transformation, 46 and cell–cell interactions.47, 48 In addition, in vivo animal experiments provide a biological system with an integrative environment and complete cellular and molecular network for lesion development and growth in vivo. It mimics the conditions in humans in the hopes that the results can be translated from bench to clinic. The application and limitations of various EM animal models, including autotransplantation of uterine tissues and xenotransplantation of human endometrial tissues into the peritoneal cavity or subcutaneous pocket in ectopic sites of rodent models, as well as in the primate model have been assessed, and the choice of the appropriate model for studies depends on the research questions. 49

10Apart from the appropriate model, positive control of current pharmaceuticals should be included for comparison, which will serve as experimental evidence of the efficacy of new drugs. When choosing a positive control, pharmaceuticals with relevant actions to the examined molecular and signaling pathways should be considered. For example, dienogest can be used as a positive control to compare the inhibition of NF‐κB activation, enhancement of apoptosis, or inhibition of MMP‐2/‐9,50, 51 leuprolide acetate to compare the inhibition of promitogenic cytokines, 52 and celecoxib to compare the proliferation‐inhibitory and apoptosis‐enhancing effects. 53

11In addition to the efficacy, the pharmacokinetic profile of a drug with respect to absorption, distribution, metabolism, and excretion should be available to support its clinical use. 54 The bioavailability of a drug and its active metabolites in systematic circulation and local tissues should be quantified to justify the therapeutic dosage for clinical application. 55 The relationship between drug potency and pharmacological effects on the body and action site should be evaluated to prevent off‐target toxicities. 56 The possible adverse effects on other tissues also need to be determined. Medications with specific efficacy on the ectopic endometrium and minimal side effects on the eutopic endometrium are preferable for EM treatment, as these medications will affect reproductive cycles the least. In animal experiments, adverse effects on reproductive tissues and functions should be carefully monitored. As a short‐term measure, no significant change in body weight should occur in the test animals, and as a long‐term measure, the animals should be able to conceive and deliver. For women with EM who prefer symptomatic medical therapy, such side effects should be limited and well‐tolerated. Medicines that regulate E2 levels usually result in hypoestrogenism and are associated with side effects such as hot flushes and vaginal dryness, which are acceptable, but not preferable. 57 Other common adverse effects, such as osteoporosis and venous thromboembolism, should be avoided. The effect of the medications on fertility should also be monitored; however, the current data are very limited.

12Several studies have systematically recorded the direct and indirect costs of EM treatment and highlighted its long‐term economic burden on the society, healthcare system, and affected women. 58 This has raised awareness of the disease and increased the demand for cost‐effective EM drugs. However, the choice of treatment depends not only on patients’ desired outcomes but also on treatment affordability. A cost‐effective medication is that with equivalent monetary value and efficiency. Therefore, as the most cost‐effective treatment for EM is considered for use as a standard hormonal treatment, 59 a potential new pharmaceutical should be affordable and easily accessible to the market, in addition to showing good efficacy with fewer side effects.

13In summary, a potential new pharmaceutical should be well‐studied in terms of not only action mechanism and efficacy in vitro and in vivo, but also safety, efficiency, and cost‐effectiveness. Progress in this area is expected to provide clear and effective insights for policy‐making and for decision‐making in the individualized treatment of EM.

14Medications investigated in ongoing or completed clinical trials on EM are summarized in Table 2. Most drugs are mainly symptomatic. Outcome measures used in these studies are pain score, levels of dysmenorrhea and dyspareunia, and quality of life, except for epigallocatechin gallate (EGCG) and quinagolide, whose efficacy in reducing endometriotic lesions will be determined. To the best of our knowledge, there is limited clinical trial to examine the pathophysiology or signaling pathways targeted by the drugs. Moreover, heterogeneous pathophysiology among patients affects their responsiveness to drug treatment; therefore, the development of personalized medicines to specific patients based on EM pathophysiology is desirable. 39 These further emphasizes the demand for new pharmaceutical that is for symptomatic management, as well as targets specific pathophysiology and signaling pathways to eliminate the endometriotic lesions.

15Here, we discuss the pathophysiology and molecular targets that are directly or indirectly associated with the drugs, as well as their effects on the corresponding signal transduction pathways in the treatment of EM. In Table 3, we distinguished potential drugs as a repurposed or a de novo drug of EM. A new drug is defined as a chemical that has not been studied in clinical trials for other diseases before EM and a repurposed drug is defined as a chemical that has been studied in clinical trials for other diseases before EM. We provided sufficient scientific data of their efficacies in reducing endometriotic cell viability in vitro or lesions in vivo, as well as in regulating specific signaling pathways and molecules involved in the pathophysiology of EM. The advantages, side effects, and limitations of the drugs are also highlighted.

16Increased levels of E2 reduce progesterone and inhibit endoplasmic reticulum stress in endometrial cells. 157 Increased expression of estrogen receptor (ER) isoforms has been observed in endometriotic lesions,158, 159 suggesting their contribution in regulating proliferation of the lesions. E2 is mediated by ERα and ERβ as well as by G protein‐coupled receptor 30 (GRP30), which is a seven‐transmembrane receptor. It activates phosphoinositide 3‐kinases (PI3K) and mitogen‐activated protein kinase (MAPK) through the transactivation of the epidermal growth factor receptor (EGFR) in the plasma membrane.160, 161 Chloroindazole and oxabicycloheptene sulfonate are two new chemicals bound to ERα and ERβ, respectively, and both inhibited E2‐driven proliferative and inflammatory activities in a dual action manner in ectopic lesions. This experimental study demonstrated great potential owing to their high potency and efficacy as preventive and therapeutic treatments. In addition, they do not exert any undesirable effects on the reproductive system. Co‐treatment of either ligands with letrozole enhanced the regression of ectopic endometrium, but it did not affect eutopic uterine tissues as with only letrozole. 64 Increased COX‐2 and aromatase (P450AROM) levels stimulate E2 synthesis.162, 163, 164 P450AROM inhibitor maintained a low E2 level and reduced EM lesion size. 57 COX‐2‐targeted treatment with celecoxib and indomethacin, which are two available NSAIDs, showed multiple effects on EM.60, 165 The drugs inhibited COX‐2‐mediated prostaglandin E2 (PGE2), which regulates E2 formation, 53 but caused side effects including reproductive failures and cardiac adverse conditions. 166

17EGCG and resveratrol are both natural products that have been studied for EM treatment in clinical trials. They act as anti‐E2 agents, but with reduced side effects compared with hormonal drugs.65, 69 High doses of resveratrol reduced proliferation by interacting with ERα, and its expression in the endometrium epithelium was reduced to a profound level similar to that achieved with progesterone treatment. Nevertheless, progestogen did not reduce Ki‐67 expression in the endometrium stroma, whereas resveratrol reduced its expression in both the epithelium and stroma. 66 EGCG inhibited E2‐stimulated proliferation and VEGF expression in cultured endometriotic glandular cells as well as angiogenesis and lesion growth via VEGF in mouse models.70, 129, 167

18NF‐κB is a protein that promotes cell proliferation and inhibits apoptosis in endometrial and endometriotic cells.168, 169, 170, 171 It is activated by cytokines, including TNF‐α, interleukins (IL)‐1β, and lipopolysaccharide. These stimulate inhibitors of NF‐κB (IκB) to be phosphorylated by IκB kinase (IκK). 171 NF‐κB binds to DNA and transcripts the genes of angiogenic and adhesion factors, cytokines, growth factors, and inducible enzymes such as nitric oxide synthase and COX. 172 Dienogest is a pregestational steroid of NF‐κB inhibitor, and it inhibits IL‐8 production to attenuate NF‐κB activation in endometriotic stromal cells in vitro. 50

19In EM, MAPK is activated to mediate the intracellular transmission of extracellular signals and induce cellular processes, 127 as shown by a high phosphorylated extracellular signal‐regulated kinase (ERK) level.173, 174 RAS binds to RAF and activates ERK kinase (MEK1/2) to phosphorylate ERK, which is a major MARK signaling cascade. 127 ERK1/2 regulates c‐fos and c‐jun expression to regulate mitosis and cell viability in endometrial cells under EM. 174 E2, IL‐1β, and TNF‐α stimulated the phosphorylation of ERK1/2 in endometriotic stromal cells, but not in normal endometrial cells. 175 Protease‐activated receptor 2 also activated ERK1/2 in cultured ectopic endometrial stromal cells. 176

20Sorafenib has completed phase IV clinical trials in several types of carcinoma, and it significantly abrogated the phosphorylation of RAF kinase by 64% via the MAPK/ERK pathway in stromal cells of EM patients. However, weight loss was observed in the xenograft EM mouse models. 87 Vemurafenib is FDA‐approved for the treatment of metastatic melanoma and significantly inhibits ERK phosphorylation by over 60% in both endometriotic stromal cells and epithelial cells. 89 U0126 is an MEK1/2 inhibitor that increases progesterone receptor (PR)‐αβ levels in endometriotic stromal cells. 91 However, although the above treatments were shown to have significant efficacy, treatment of EM with MAPK inhibitors induced adverse effects on reproductive functions in animal models, including ovulation inhibition, embryotoxicity, and teratogenicity. 127 Puerarin is a natural product that strongly binds to ERs; its binding affinity to ERs is one‐third that of E2, and it suppresses E2‐induced endometriotic stromal cells by 30% via the ERK pathway in vitro, 93 and results in reduced adverse effects.

21PI3K phosphorylates phosphatidylinositol 4,5‐bisphosphate (PiP2) into phosphatidylinositol 3,4,5‐trisphosphate (PiP3) and activates protein kinase B (Akt). 177 Mammalian target of rapamycin (mTOR), a downstream protein kinase of Akt, is overexpressed in ectopic lesions.178, 179 Reduction of the phosphatase and tensin homolog deleted from chromosome 10 (PTEN) by mutation 180 enhanced the phosphorylation of Akt, thus promoting proliferation, inhibiting apoptosis, and reducing PR expression in EM. 91 MK2206, an Akt inhibitor, is a drug candidate for cancer treatment that acts by increasing PRβ and PRαβ levels and decreasing the viability of endometriotic stromal cells, without affecting normal cells. 91 WIN 55212–2 is a nonselective cannabinoid agonist that binds to cannabinoid receptor (CB)1 or CB2 to inhibit Akt levels and Akt phosphorylation, suggesting the inactivation of the Akt pathway. However, although it reduced the proliferation rate of endometriotic cells, it also reduced that of eutopic endometrial stromal cells. 98

22The Hippo/YAP pathway is important for balancing cell proliferation and apoptosis. Upregulation of this pathway increased the viability of endometriotic cells, whereas knockdown of YAP increased apoptosis and decreased B‐cell/B‐cell lymphoma 2 (Bcl‐2) expressions. 181 Verteporfin, a YAP1 inhibitor, inhibited the proliferation of endometriotic stromal cells, production of E2, and infiltration of immune cells. 75 It is an FDA‐approved drug for the treatment of subfoveal choroidal neovascularization. EM mouse models showed decreased vessel tube formation and cell migration, with no reported effects on reproductive organs, infertility, or transgenerational influence. 75 YAP1 is a potential target protein but has not been widely studied in EM.

23The metabolic pathways toward increased lactate and dysregulation of glycolysis were shown as contributing factors for cancer progression. 182 Lactate induces angiogenesis and supplies nutrients to proliferate tumor cells.182, 183 In EM peritoneal mesothelial cells, increased glycolysis, decreased mitochondrial respiration, decreased pyruvate dehydrogenase activity, and increased lactate was also observed. 46 Dichloroacetate, a nonhormonal treatment or recurrence prevention of EM, reversed the pathophysiology of EM by inhibiting pyruvate dehydrogenase kinase to activate pyruvate dehydrogenase. 46 Although dichloroacetate has completed a phase III clinical trial for lactic acidosis in 1998 and has widely studied in cancer, it has not been approved by the FDA for therapeutic use in cancer.

24Apoptosis is a programmed cell death process that maintains the balance between the growth and differentiation of cells for tissue renewal. It is regulated by selective chromatin internucleosomal cleavage to shrink the cells. 184 Apoptosis is important in the normal endometrium to remove dysfunctional cells and repair tissues during the menstrual cycle. 185 Apoptotic cells were found to be more predominant in the endometrial epithelium glands than in the stroma. 186 Cell apoptotic activity was found to be relatively low in EM. 187 This can be explained by the reduced expression of proapoptotic factors (e.g., Bcl‐2‐associated X [Bax] and Bcl‐2 associated agonist of cell death [Bad]), overexpression of antiapoptotic factors (e.g., Bcl‐2), and dysregulation of cell cycle.188, 189 Endometriotic cells could not express surface receptors to trigger proapoptotic proteins; neither apoptotic signals were appropriately transduced, leading to proliferation being triggered instead. 188 GnRH agonists, such as leuprolide acetate or the preclinical nonhormonal drug propofol, increased the levels of proapoptotic proteins, as well as decreased the levels of antiapoptotic proteins and promitogenic cytokines.52, 100 Melatonin is highly effective in amplification of apoptotic activity via regulating MMP‐3 signal, was able to regress EM at either early or late stage. Melatonin in high dose and long‐term treatment shows no adverse effects in EM rodents. 63

25The signaling pathways MAPK/ERK, PI3K/Akt, and NF‐κB also regulated apoptosis in endometriotic cells. ERK1/2 was activated as an antiapoptotic protein in eutopic and ectopic endometrial glands throughout the menstrual cycle. 174 Non‐E2 targeted treatments, such as selective PGE2 inhibitors, inhibited PGE2 receptor (EP) 2 and EP4 to induce apoptosis via multiple pathways including ERK1/2, AKT, and NF‐κB. They enhanced apoptosis by approximately 50% in both epithelial and stromal cells; thus, owing to their efficiency as selective or combination inhibitors, they are ideally used to treat stage I and II EM. 95 PGE2 inhibitors are preferred to COX‐2 inhibitors due to their fewer adverse effects and lack of hypoestrogenic effects. 95

26BAY11‐7085 is an NF‐κB inhibitor that inhibits cell viability and enhances apoptosis in endometriotic cells by seven fold but induced less profound results in normal endometrial cells. 62 The activity of Bay11‐7085 has been shown in in vitro studies, but there has been no clinical study of its therapeutic potential. Ginsenoside Rg3, genistein, and curcumin are found in natural products, and they enhanced apoptosis via regulation of the NF‐κB pathway.79, 85 In addition, ginsenoside Rg3 neutralized the effect of TNF‐α to regulate proliferation. 79 Genistein had a similar capability as a GnRH agonist, that is, inhibiting transforming growth factor (TGF)‐β to regulate NF‐κB. Expression of Bcl‐2 was suppressed, whereas that of Bax was enriched. It also reduced COX‐2 and PGE2 expression to levels comparable to those in the control group. 85 Curcumin inhibited MMP‐3 and increased the Bax/Bcl‐2 ratio by upregulating tumor protein p53 (p53). 81 These natural products are highly attractive owing to their potential efficacies and minimal side effects reported in in vitro and in vivo studies of EM and clinical studies of other diseases.

27Akt is a pleiotropic regulator of apoptosis that increases endometriotic cell survival and decreases apoptosis. 190 Akt/mTOR can be inhibited by endoplasmic reticulum stress. 51 Tunicamycin enhanced TNF‐related apoptosis‐inducing ligand‐induced apoptosis by inducing endoplasmic reticulum stress in endometriotic stromal cells. The effect was more potent in endometriotic cells than in eutopic endometrial cells 73; however, its action mechanism and pharmacokinetics profile have yet to be elucidated.

28Transcription of antiapoptotic Bcl‐2 protein was increased by E2 through the promotion of thymic stromal lymphopoietin. 191 Extranuclear kinases are activated by an elevated ER complex to trigger a rapid nongenomic signaling cascade and inhibit apoptosis in stromal and epithelial cells.192, 193 Genistein interfered with the E2/ER pathway to induce apoptosis and apoptotic proteins in endometrial hyperplasia. 194 Moreover, the activity of E2 strongly regulates TNF‐α‐induced effects. In a healthy endometrium, TNF‐α stimulated apoptosis. In eutopic and ectopic endometriotic cells, TNF‐α stimulated proliferation and inhibited apoptosis instead. 195 In endometriotic lesions, apoptosis signal‐regulating kinase (ASK‐1), a TNF‐α induced apoptosis complex I, interacted with ERβ as well as serine/threonine kinase receptor‐associated protein (STRAP) and 14‐3‐3 proteins. 62 Formation of this complex disrupted the association of TNF‐receptor‐associated factor 2 (TRAF2) and ASK‐1 for TNF‐α‐induced apoptosis.196, 197 TNF‐α‐induced apoptosis complex I, complex II, and apoptosome were all inhibited, which dysregulated apoptosis and activated the invasiveness of lesions for survival. In addition, in endometriotic tissue, TNF‐α from macrophages and natural killer (NK) cells induced the generation of the steroid receptor coactivator (SRC)‐1 isoform from cleaved MMP‐9. 198 ERβ interacted with the caspase (CASP)‐8 and SRC‐1 isoforms to prevent activation of TNF‐α‐induced apoptosis complex II in endometriotic lesions in ectopic sites. 62 Therefore, inhibition of ASK1/ERβ/STRAP‐14‐3‐3 and ERβ/CASP8/SRC‐1 protein complex are potential therapeutic targets to regulate apoptosis via the E2/ER/TNF‐α pathway.

29Autophagy is a process related to nonapoptotic cell death and is defined as self‐degradation. It balances the energy sources by removing misfolded proteins, damaging organelles, and eliminating intracellular pathogens. It promotes the proteolytic degradation of cytosolic components at the lysosome. 199 Recently, there have been more studies on the role of autophagy in both accelerating and decelerating the pathogenesis of EM. To examine the pathophysiology of autophagy in regulating EM, we divided the section into antiautophagy and proautophagy to discuss the controversies in the progression of EM.

30Downregulation of apoptosis favors the stimulation of autophagy, thus promotes EM growth. 200 A significant reduction in apoptosis inducer p53 mediated by Akt, an increased lapidated microtubule‐associated protein 1A/1B light chain 3 (LC3)‐II, and a significant decrease in sequestosome 1 (SQSTM1) were observed in ovarian endometrioma. 200 LC3‐II is a standard autophagy marker while SQSTM1 is an autophagy adaptor protein that transfers ubiquitinated proteins to the autophagic machinery and is degraded via autophagy to indicate the activation of autophagic flux.197, 198, 199, 200, 201 Hypoxia upregulated autophagy in endometriotic cells to induce HIF‐1α. 202 Overexpression of HIF‐1α under normoxic conditions also induced autophagy. 203 Decreased expression of homeobox A10 (HOXA10) induced autophagy in EM, which was attributed to excessive inflammation. 204 It contributed to mitochondrial damage, increased PGE2, and increased mitochondrial ROS,203, 204 all stimulated autophagic processes. An increased oxidant heme oxygenase (HO)‐1 was observed in ovarian endometrioma to activate an adaptive defense mechanism and negatively modulate inflammation and apoptosis, and positively stimulated autophagy. 200 A combination therapy with MK2206 and chloroquine was found to be more effective than with either MK2206 or chloroquine in reducing endometriotic cell viability and preventing regrowth by inhibiting autophagy. 102 SQSTM1 expression was significantly upregulated. 102

31Upregulation of autophagy promoted apoptosis and suppressed cell growth and invasion in EM.205, 206 Endometriotic stromal cells have an abnormal response to progesterone, which suppresses PTEN expression, suppresses autophagy, and reduces apoptosis in the menstrual cycle via the AKT/mTOR pathway. 207 Increased expression of YAP significantly decreased autophagy through the mTOR pathway in eutopic endometrium stromal cells. 208 Mullerian‐inhibiting substance (MIS) induced autophagy and apoptotic cell death and inhibited proliferation in vitro. 104 MIS, also known as anti‐Müllerian hormone, however, was found to be increased in EM lesions and in the serum of women with ovarian endometrioma and promoted inflammation. More preclinical data are required before the clinical application of this agent in EM treatment.209, 210 A ginsenoside metabolite, protopanaxadiol (PPD), reduced ERα expression and induced PRα expression in vitro and in vivo, which then induced autophagy and suppressed lesion growth, resulting in a significantly different expression of autophagy‐related genes, including downregulated estrogen receptor (ESR1), SQSTM1, and TGF‐β levels as well as upregulated CASP‐3, ATG ‐3/‐5/‐12 after treatment. 103

32Cell migration and invasion are critical processes for EM establishment according to the implantation theory. EM is believed to occur due to the shedding of endometrial cells and then migration to ectopic sites.15, 211 Migration of endothelial cells mediates angiogenesis and plays a role in the pathophysiology of EM. 212

33The wingless‐type mouse mammary tumor virus integration site family (Wnt) plays a role in developmental processes and homeostasis. β‐catenin is crucial in regulating the cell cycle, which includes proliferation, differentiation, and migration in ectopic lesions. 213 In the presence of Wnt ligands, an accumulation of β‐catenin translocates to the nucleus and interacts with T‐cell factor/lymphoid enhancer‐binding factor (Tcf/LEF) transcription factors to activate the Wnt/β‐catenin signaling pathway. 107 Wnt/β‐catenin was found to be abnormally activated in EM. 214 Multi‐drug resistance protein 4 (MRP4) regulates Wnt/β‐catenin signaling by stabilizing β‐catenin activity. It was involved in the pathogenic transformation of EM endometrium, confirmed in the ectopic lesion. In MRP4‐knockdown endometrial epithelial cells, reduced activity of β‐catenin was found to downregulate Wnt/β‐catenin signaling. 215

34Overexpression of T‐cadherin inhibited the invasion and migration of cells in EM, and the phosphorylation of heat shock protein (HSP)‐ 27 and c‐Jun N‐terminal kinase (JNK)‐1/2/3 was promoted. MMP‐2/‐9 and vimentin expression was lowered in endometriotic cells. 216 MMP‐2/‐9 and Cyclin D1 are targets of Tcf/β‐catenin genes and were found to be upregulated in endometrial epithelial or stromal cells of EM, 107 MMPs are responsible for regulating migration, invasion, and angiogenesis by balancing growth factors and cytokines, and high expression of MMPs in EM favors lesions. 217 PKF115‐584 and CGP049090 are fungal derivatives and were screened through high‐throughput assay to disrupt Tcf/β‐catenin complex, 218 significantly inhibiting MMP‐9 activity and cell invasiveness in epithelial and stromal cells to a level close to that of normal endometrium. 108 However, they have not been widely studied in EM. Genistein regulates invasion and migration through downregulation of MMP‐2/‐9 by targeting NF‐κB, as shown in an in silico study and in an EM mouse model. 112 In addition, Wnt/β‐catenin is important for stem cell maintenance and tissue homeostasis 107; thus, an antagonist of Wnt/β‐catenin may have potential side effects. Moreover, Wnt2 is secreted by ectopic stromal cells, which induces β‐catenin signaling activity in ectopic endometrial epithelial cells, as well as expressions of the growth‐associated proteins in endometrial epithelial cells. 48 Wnt2/β‐catenin is a pathway involved in the communication of stromal and epithelial cells in EM, which can be modified by metformin. 48

35Ras homolog family member A/Rho‐associated coiled‐coil kinase (RhoA/ROCK) plays a major role in cell migration via phosphorylation of cytoskeletal regulatory proteins and results in actin depolymerization, actomyosin contraction, endothelial cell adhesion, and migration. 219 ROCK in the endothelial cytoskeleton is activated by proangiogenic stimuli. 219 VEGF stimulates RhoA/ROCK and mediates endothelial cell migration. 220 ROCKII regulates cell body contraction during migration. 221 It is a positive regulator of p27 to further activate cell migration in endometrial stromal cells and regulate RhoA in the cells. 222 Fasudil is a ROCK inhibitor used clinically and has the potential for treating EM by reducing endometriotic cell viability and inducing apoptosis by targeting Rho/ROCK. 115

36Recently, EM was defined as a profibrotic condition, as a new concept by Vigano et. al.223, 224 The crucial role of fibrosis and differentiation of myofibroblasts in the progression of EM lesions have been reported. Fibrosis is the pathological activity when activated myofibroblasts accumulate leading to contraction of the collagenous extracellular matrix and anatomical structure disruption. The event of fibrosis justifies EM‐associated morbidity and adhesiveness and is considered a potential EM therapeutic target.223, 224

37TGF‐β1 is a stimulating factor that triggers the production of collagen and transition of epithelial to mesenchymal phenotype, leading to fibrosis. TGF‐β1 was found to be significantly increased in the peritoneal fluid of women with EM. 225 Wnt signaling is required for TGF‐β1‐mediated fibrosis. 226 Matsuzaki and Darcha 110 also showed that the Wnt/β‐catenin pathway was involved in fibrogenesis in EM. PKF1150‐584, CGP049090, and EGCG significantly inhibited TGF‐β1‐induced fibrotic markers, including α‐smooth muscle actin (α‐SMA), type I collagen, fibronectin, and connective tissue growth factor in stromal cells via Wnt/β‐catenin signaling, with EGCG showing the greatest effects.110, 111 ICG‐001 and C‐82 are CBP inhibitors. They are also metabolites of PRI‐724, which is under clinical trials for its antitumor activity, 227 they inhibited fibrosis via Wnt/β‐Catenin signaling, with ICG‐001 showing great efficacy in upregulating apoptosis and inhibiting migration as well as C‐82 showing potent inhibition of proliferation and cell viability. 106

38Activation of the Rho/ROCK signaling pathway was associated with fibrosis in EM. 228 Heparin has completed phase IV clinical trials for several medical indications, including anticoagulation and cancer. It inhibited RhoA, ROCKI, ROCKII, and α‐SMA expression and activated the Rho/ROCK pathway to attenuate endometriotic stromal cell contractility, differentiation, and fibrosis. 117

39Angiogenesis is highly regulated in the female reproductive system, and it is a process that results in the formation of new blood vessels from existing ones. Three mechanisms of angiogenesis have been described as sprouting, elongation, and intussusception. 229 It provides neovascularization to deliver essential nutrients and oxygen supply for the growth of endometriotic lesions. 230

40VEGF regulates angiogenic, endothelial cell‐specific mitogenic, and vascular permeability activities of endometrial and endometriotic cells through vascular endothelial growth factor receptor (VEGFR) 1–3 on the microvascular endothelial cell surface. The VEGF family helps in establishing and maintaining endometriotic foci.231, 232, 233 VEGFR2 is a highly active kinase that plays a major role in angiogenesis. VEGF binds to two proximal VEGFR2 receptors, promotes vascular permeability, increases the migration and proliferation of endothelial cells, and contributes to the formation of new blood vessels.234, 235 Sunitinib, SU6668, SU5416, sorafenib, and pazopanib were originally indicated as anticancer drugs but were further repurposed and tested for efficacy against EM.119, 120, 123, 126, 128 SU5416 selectively bound to VEGFR, and only reduced graft size by 5%.123, 126, 236 SU6668, a multi‐kinase inhibitor, reduced the endometrial graft by 25% by blocking VEFFR‐2, fibroblast growth factor receptors (FGFR)‐1, and platelet‐derived growth factor receptors (PDGFR)‐β.123, 236 Sunitinib regulates angiogenesis and apoptosis through multi‐kinase inhibition, regressing 50% cyst in an EM rat model. 120 The effects of pazopanib, sunitinib, and sorafenib on VEGF/VEGFR protein kinase pathways and their actions in EM were compared by Yildiz et. al., 128 which showed that pazopanib had better efficacy than the control and other treatments, reducing EM lesions by at least 45%, but Sorafenib was better in regulating VEGF. 128 However, tyrosine kinase inhibitors that regulate VEGFR are associated with a significant risk of treatment toxicities. 237 In sarcoma, pazopanib treatment led to a higher incidence of adverse effects, including fatigue and hypertension, compared with sunitinib or sorafenib. 237 An alternative VEGF regulator, EGCG, significantly inhibited lesion growth by suppressing VEGFC/VEGFR2 signaling. Overexpression of VEGFC induces migration of endothelial cells, increases vascular permeability, and induces angiogenesis and endometriotic lesions growth.70, 129 Prodrug of EGCG (ProEGCG), reduced lesion size, weight, and VEGF concentrations in plasma to a greater extent than the parent EGCG molecule. More importantly, there were no signs of side effects on reproductive tissues. 130 Quinagolide, a dopamine receptor 2 agonist, is under phase II clinical trial to examine its efficacy in reducing EM lesion size and related pain. It can completely reverse the size of lesions by downregulating the VEGF/VEGFR2 pathway in EM. 135 It has an acceptable safety profile and does not stimulate serotonin receptor subtype 2b to proliferate fibroblasts in cardiac valve tissues, and thus holds great potential as an alternative of tyrosine kinase inhibitor for EM. 135

41Under normoxic conditions, HIF‐1α is regulated via proteasome‐mediated degradation. 238 However, under hypoxic conditions, HIF‐1α escapes ubiquitination and binds to hypoxia‐responsive enhancer on VEGF genes to upregulate their expression. 239 HIF‐1α was upregulated in lesions, thus promoting VEGF expression in an EM mouse model. 240 Increased VEGF secretion was observed in hypoxia‐induced endometrial stromal and glandular cells compared with that under normoxic conditions. 241 Oxidative stress also increased VEGF secretion, as shown by the results obtained after incubating endometrial epithelial cells with oxidized low‐density lipoprotein. 242 In the EM peritoneal environment, PGE2 was upregulated to elicit cell signals through upregulation of VEGF and FGFR. It induced the expression of COX‐2 and synthesis of E2 in ectopic endometrial cells to increase the production of MMP, thus enhancing VEGF expression and inducing angiogenesis.229, 243 TNF‐α mediates the angiogenic activity of macrophages, which stimulates endothelial cell migration and induces the release of VEGF and the formation of bloodvessel. 244 Pyrrolidine dithiocarbamate inhibited NF‐κB activation and attenuated TNF‐α‐mediated VEGF and MMP‐9 expressions. 132 Pentoxifylline attenuated TNF‐α mediated effects in other diseases, requires further investigation of this in EM, but it suppressed angiogenesis by reducing VEGFC and VEGFR2 (Flk‐1) expression levels in glandular cells of endometriotic lesions. 133 Pentoxifylline is an immunomodulatory agent and has completed a phase III clinical trial of EM‐associated infertility. The clinical trial did not present any data on lesion progression or recurrence, only on pregnancy rate. 245

42The Rho/ROCK pathway regulates VEGF‐mediated endothelial cell activation and vessel stability.233, 246 RhoB mitigates VEGF‐induced vessel sprouting via the RhoA/ROCK signaling pathway. 247 RhoA/ROCK activity can be blocked by inhibiting protein prenylation in endothelial cells to reduce migration and adhesion. 248 Avian myelocytomatosis virus oncogene cellular homolog (C‐Myc) is a target of the PI3K/Rho/ROCK signaling pathway and regulates VEGF expression. Under hypoxic conditions in EM, guanosine triphosphate (GTP)‐bound Rho was regulated in a PI3K‐dependent manner to induce VEGF by binding to C‐Myc without suppressing the induction of HIF‐1α. 249 ROCK activation initiates E2‐induced angiogenesis. 219 Inhibitors that target the Rho/ROCK pathway should be further investigated for their ability to regulate angiogenesis, migration, invasion, and fibrosis of endometriotic cells.

43Oxidative stress is the imbalance between ROS production and antioxidant function, which plays a main role in EM progression. 250 ROS are molecules that have unpaired electrons and can damage lipids, nucleic acids, and proteins.251, 252 Apoptotic endometrial tissues and macrophages induced oxidative stress in EM through retrograde menstruation. 253 Oxidative stress causes DNA hypermethylation and histone modification, which are linked to aberrant endometrium development in EM. 254

44The production of ROS in the progression EM could be achieved through several pathways, including activation of inflammatory cytokines, MMPs, and transcriptional factors, such as NF‐κB. 255 Environmental factors, such as reproductive toxins, can also increase oxidative stress and decrease the expression of antioxidant enzymes. 256 Di‐2‐ethylexyl phthalate is used as a plasticizer and solvent in cosmetic and consumer products, and they altered the NF‐κB signaling pathway and expression levels of ER and PR in human endometrial stromal cells via activation of the MAPK/ERK and PI3K/Akt signaling pathways. 257

45Balancing ROS production and antioxidant function by inhibiting free radicals or increasing antioxidant levels is important to regulate oxidative stress. Antioxidants include the enzymes superoxide dismutase (SOD), glutathione peroxidase (GPx), HO, and catalase, as well as nonenzymatic molecules such as vitamins A, C, and E.250, 252, 255 Significantly lower levels of antioxidants were found in the peritoneal fluid of EM, which indicated that women with EM had a low free radical‐scavenging ability.255, 258, 259 N‐Acetyl cysteine (NAC) is an antioxidant that inhibits ROS as well as abrogates ERK activation and proliferation. 137 EGCG and resveratrol are both well‐known natural antioxidant supplements for the treatment of EM and other diseases, including cancer.130, 138, 260 Resveratrol acts as a radical scavenger, and it significantly regressed endometriotic implants, decreased lipid peroxidation, and increased at least 50% endogenous antioxidant capacity in tissues and serum in EM. 138 The antioxidant effects of ProEGCG against EM were significantly greater than those of EGCG and at least four fold that of the control. 130 Melatonin reduced the level of oxidative stress markers and increased level of antioxidants in EM. EM implants in rat were significantly regressed.142, 143 Caffeic acid found in plants exerted similar antioxidant effects in EM and an enhanced nuclear translocation of nuclear factor erythroid 2–related factor 2 (Nrf2) regulated antioxidant enzymes in EM.141, 261

46NO is a vasodilator that mediates endothelium‐dependent vasodilation and angiogenesis. 262 NO has an unpaired electron and is a highly reactive free radical. The formation of NO requires NO synthase (NOS), l‐arginine, oxygen, and a number of cofactors, including nicotinamide adenine dinucleotide phosphate, flavin mononucleotide (FMN), and flavin adenine dinucleotide. 263 Macrophages increased IL‐10 in EM and stimulated NO 264; NO and NOS levels were increased in endometrial tissues in EM. 265 Increased E2 level activated the formation of NO, 266 implying that macrophages and E2 regulate NO‐mediated oxidative stress in EM.

47Iron carries hemoglobin throughout the body, an overproduction of it not only enhances epithelial cell proliferation but also induces oxidative stress.267, 268 Ferritin is a cellular iron storage that generates hydroxyl radical via Fenton reaction to initiate a free radical chain reaction, namely lipid peroxidation.37, 269 McKinnon et al. 37 reviewed studies on the implication of mTOR on iron homeostasis and suggested the dysregulation of mTOR in EM could overload iron levels to stimulate oxidative stress. Currently, there is lack of pharmaceuticals targeting iron‐mediated oxidative stress in EM and regulating the mTOR signaling pathway might reduce oxidative stress.

48Immune system dysregulation and chronic inflammatory response are characterized in EM.270, 271 The adaptive immune system with increased quantity of regulatory T (Treg) cells and a shift towards type 2 immune response fail to recognize the endometriotic cells in the peritoneal cavity. 272 On the contrary, the innate immune system in EM is characterized by an enhanced activation state of macrophages, along with upregulated cytokines, but downregulated phagocytosis, 273 as well as a reduced cytotoxicity of natural killer (NK) cells, 274 and an altered population of dendritic cells. 275 These promote inflammation and contribute to the implantation process in EM and new drugs that modulate these specialized cells hold promise as a novel immunotherapy for EM.

49Macrophages exert their inflammatory effects against tumors via host defense mechanisms. 276 IL‐1 is a proinflammatory cytokine secreted by activated monocytes, macrophages, or NK cells, and is responsible for activating lymphocytes to reduce immune surveillance and stimulate PGE2 via COX‐2 in EM stromal cells.271, 277, 278 IL‐6 is responsible for stimulating B‐cell activity and T‐cell differentiation. The levels of IL6 in serum and peritoneal fluid are high, but its receptor is reduced in EM. However, endometriotic cells are resistant to its growth‐inhibitory effects.271, 279, 280 VEGF and TNF‐α are proinflammatory cytokines secreted by activated lymphocytes, neutrophils, NK cells, and macrophages to initiate the inflammatory cascade. 271

50Niclosamide is an FDA‐approved nonsteroidal therapy for antihelminth 281 and was found to inhibit the proliferation and growth of endometriotic lesions. It reduced MAPK, WNT, and inflammation signaling‐related genes, such as NF‐κB and signal transducer and activator of transcription 3, in an EM mouse model. No disruption to reproductive function was observed, indicating potential therapeutic efficacy and safety for EM treatment. 154 NAC regressed lesions by suppressing COX‐2 and MMP‐9 expression. Its side effect is mild and seemed to not interfere with fertility in vivo. 144 Crocin, curcumin, and metformin inhibit proinflammatory cytokines and chemokines, including TNF‐α, IL‐1β, IL‐6, VEGF, and so forth.148, 149, 155 These are responsible for recruiting and activating macrophages, neutrophils, and NK cells to the EM site and further enhancing angiogenesis and inflammation.155, 282 Acai, a natural product found in plants from the Amazon region, has completed phase III clinical trials as an antioxidant agent. It reduces EM lesions by targeting active macrophages, VEGF, and COX‐2. 156 Resveratrol inhibited inflammatory responses by reducing peritoneal and serum cytokines, 150 as well as activating sirtuin 1 (SIRT1) to significantly suppress IL‐8 in TNF‐α‐induced endometriotic stromal cells via NF‐κB. 151 SIRT1 has a dual function as a tumor suppressor or promoter, 283 and it is a potential target protein, considering that it is a strong regulator of the inflammatory responses, apoptosis, and oxidative stress in EM. 284 Macrophage migration inhibitory factor (MIF) is a proinflammatory cytokine that is upregulated in peritoneal fluid in women with EM. It activates the MAPK/ERK pathway, stimulates COX‐2, and produces PGE2 in ectopic endometrial cells. MIF also contributes to angiogenesis via its effect on endothelial cell proliferation. 285 ISO‐1 is a MIF antagonist and a leading molecule discovered to treat sepsis. 286 It inhibited angiogenic and proinflammatory pathways via VEGF/VEGFR in peritoneal EM in vivo, without interrupting the reproductive cycle. 287

51In EM, increased ESR2/E2 induces COX‐2 and PGE2, upregulates macrophages and NF‐κB,288, 289 and leads to oxidative stress. 255 ERβ modulates macrophage infiltration via NF‐κB in EM 290 and induces IL‐1β by interacting with the inflammasome complex to evade immune surveillance and promote the attachment of lesions at the endometriotic sites. 196 E2 activates thymic stromal lymphopoietin and induces the secretion of endometrial stromal cells‐associated growth‐promoting cytokines, including monocyte chemoattractant protein 1 and IL‐8, via the JNK and NF‐κB pathways. 291 IL‐6 reduces E2 production in human granulosa tumor cells via the MAPK signaling pathway, 292 implying its possible targeting of E2 biosynthesis in EM. Puerarin reduced the level of ERβ, but not ERα, by inhibiting P450AROM. In a rat model, simultaneous reduction of E2, COX‐2, and PGE2 expression levels, as well as enhancement of the metabolism of E2 into estrone, 152 led to the inhibition of lesion growth in the ectopic endometrium tissues. In another study, ginsenoside PPD inhibited the E2 signal, thus activating the cytotoxicity of NK cells against ectopic endometrial stromal cells to regulate cell death. This was also confirmed in peritoneal fluids of the EM mouse model. 103

52In the treatment of EM, targeting a specific pathway, or multiple pathways alleviate the lesions. Targeting a single molecule can lead to several anti‐EM effects, as downstream transduction elements are usually connected to a series of molecular events as secondary responses. However, owing to synergistic effects, a multiple target therapy may have a greater suppressive effect on lesions compared with a single targeted therapy. 196 Table 4 summarizes several single pharmaceuticals with multiple molecular targets, which affect multiple signaling pathways in a complex disease such as EM.

53Melatonin is a natural substance produced by plants. It is also a hormone produced in the pineal gland to regulate neuroendocrine functions and inhibits LH and FSH secretion from the anterior pituitary gland. 293 Melatonin acts as an antioxidant and anti‐inflammatory agent and is currently under phase 2 clinical trial for reducing EM‐related pain. Another randomized, double‐blind, and placebo‐controlled clinical trial of melatonin was completed in 2013. The results of the study showed that melatonin acts as an analgesic and can relieve EM‐related chronic pain. 294 Melatonin receptor (MR)1A and MR1B are significantly upregulated in peritoneal EM lesions compared with those in the eutopic tissue. Melatonin has been shown to reduce EM lesions in various studies. It inhibits cell proliferation and modulates endometrial epithelial cell function. 295 Melatonin also inhibits angiogenesis via VEGF and oxidative stress via regulating radical scavenging activity and amplifies apoptotic activity via CASP3 mediated pathway in vivo and in vitro in EM.63, 142, 143 Melatonin has no adverse effects on reproductive functions, instead, it can improve ovarian functions, and thus has potential to treat EM‐related infertility.296, 297 High‐dose intravenous treatment of pain and sepsis with melatonin showed no adverse effects. 298 Its bioavailability is 15%. 299 Long‐term therapeutic investigation of melatonin in EM should be conducted to elucidate its ability to regulate E2 functions in EM.

54Metformin was shown to target multiple pathways 300 by regulating stromal‐epithelial cell communication in EM via Wnt2‐mediated signaling 48 and exerted an anti‐inflammatory effect through regulating cytokines. 149 Although a mild side effect was implied, 149 metformin regulated reproductive functions, 301 and improved conception in EM patients by inhibiting serum cytokine production. 149 Metformin is available in the market as a treatment for type 2 diabetes and PCOS in women. Considering its low cost, metformin was advocated to be used as a long‐term treatment. 302 NAC, an acetylated form of cysteine, has been prescribed as an antidote since the 1960s. It replenishes intracellular glutathione levels and modulates the redox environment; therefore, NAC is a strong antioxidant. 303 In EM, it acts as an antioxidant, antiproliferative, anti‐inflammatory, and anti‐invasiveness agent via ROS‐scavenging mechanism or through regulating cytokines in vitro and in vivo.137, 144 NAC is highly efficacious at low doses, and with no adverse effects in EM. 144 Long‐term adverse effects are also limited, including no effect on fertility.137, 144 NAC is considered to have a good safety profile and has been evaluated in phase 4 clinical trials for treating gastrointestinal and metabolic diseases.303, 304 Its pharmacokinetics and toxicity profiles are available; its terminal half‐life is 6.25 h after oral administration and bioavailability is 9.1%. 305 NAC is commercially available and cost‐effective as a dietary supplement in the market; however, studies on its efficiency in EM are limited, requiring more preclinical evidence.

55Natural products have a long history of use in the management of medical conditions. Research advances in analytical and synthetic chemistry have improved the identification and isolation of active compounds from natural products. EGCG is a polyphenol catechin from green tea and a well‐known antioxidant. It exerts efficacy against diseases including cancer, diabetes, and inflammation. 306 In EM, it exerts ant antiangiogenetic effect via the VEGFC/VEGFR2 pathways,70, 129 antioxidant effects via ROS‐scavenging mechanism, 130 antiproliferative effect via reduction of E2 production, and anti‐migration and anti‐invasion effects via TGF‐β1‐induced phosphorylation of ERK1/2 and MAPK pathways, 167 thus inhibiting the development and growth of lesions. Promising evidence of its high potency and efficacy, and without major side effects in reproductive functions were reported.130, 167 EGCG is currently under phase 2 clinical trial for reducing lesion size and pain as well as an evaluation of its safety profile in EM. On the contrary, EGCG act as an adjuvant that brings synergistic effects, as well as reduces adverse effects in cancer treatment. 307 This suggests a potential role of EGCG in combination therapy with current EM treatment. However, the low bioavailability of EGCG has limited its attractiveness in the market. 308 ProEGCG, is a prodrug of EGCG, shows higher bioavailability and greater efficiency than EGCG to reduce lesions in vivo. 130 More studies should be conducted to confirm the underlying mechanism of ProEGCG in the treatment of EM. Resveratrol is a polyphenol found in grapes. In EM, it reduces proliferation via an anti‐E2 mechanism targeting ESR1, 66 inhibits inflammatory responses via radical scavenging, 138 and inhibits angiogenesis by reducing the cytokines COX‐2 and VEGF 150 and activating SIRT1. 151 Resveratrol has completed a phase 4 clinical trial in EM and is safe and effective in relieving EM‐related pain, as well as in reducing serum CA125 and prolactin levels. Resveratrol is well‐known for its chemopreventive property. It also has a promising clinical profile in cancer treatment, nevertheless, the rapid metabolism rate of resveratrol has limited its efficacy in vivo. 309

56Curcumin, genistein, ginsenoside, and puerarin are not under any EM clinical studies but have been clinically evaluated in breast cancers, endometrial carcinoma, endothelial functions, and so forth. They have been studied for their action mechanism against EM in primary cells, cell lines, and animal models. Curcumin, which is found in ginger and turmeric, enhances apoptosis by increasing the Bax/Bcl2 ratio through targeting of MMP‐3 via NF‐κB 81 and regulates angiogenesis and inflammation by targeting chemokines and cytokines 155 in EM. It has multiple biological effects in different diseases, including cancer and inflammatory diseases. It establishes a good safety profile with no acute toxicity. 310 Genistein is an isoflavone that acts as an E2 agonist or antagonist to manage postmenopausal symptoms. In EM animal models, genistein downregulated MMP‐2/‐9 and regulated cell invasion and migration by targeting NF‐κB. 311 It also regulated NF‐κB by inhibiting TGF‐β. 85 Long‐term treatment with genistein lower the incidence of endometrial hyperplasia and provided support for bone formation in postmenopausal women.83, 312 It acts as a chemopreventive and chemotherapeutic agent against cancers and has synergistic effects with other anticancer drugs. 313 Toxicity of high dose is minimal, so it needs more study to test the safe range. 314 Ginsenoside RG3, extracted from ginseng, restored TNF‐α‐induced effects by inhibiting NF‐κB, VEGF, and CASP‐3 in EM, which are responsible for cell proliferation, apoptosis, and angiogenesis. 79 Ginsenoside PPD regulated ERα and PRα expression to suppress autophagy and lesion growth in EM. 103 It also targeted E2‐induced NK cell cytotoxicity to regulate the immune system in EM. 103 Ginsenoside also possess synergistic effect with anticancer drugs, as well as prevents toxicity and morbidity from chemotherapy. 315 Puerarin is a phytoestrogen, binds to ERs via the ERK pathway to regulate proliferation in EM. 93 It also regulates inflammation in ectopic endometrium by inhibiting P450AROM and COX‐2 and promoting ERβ expression to facilitate E2 metabolism in EM. 152 Its therapeutic effects are studied extensively in diseases including cancer and cardiovascular disease. 316

57Most of these products have known toxicity or pharmacokinetic profiles and act via multiple targets, making them beneficial as anti‐EM agents. However, their poor aqueous solubility and low oral bioavailability in vivo is the major challenge to be potential EM treatment.82, 92, 315 There are several approaches available currently to progress the bioavailability of drugs, which include prodrug approach, 130 solid dispersions approach, 317 lipid‐based formulation approach. 318 On the contrary, the long‐term safety of natural products in reproductive function and EM recurrence profiles should be further elaborated in future studies. Nevertheless, minimal side effects and available as an over‐the‐counter dietary supplement and routine remedies make them preferable to hormonal medicines.

58EM is a complex clinical challenge, and recently, more signaling pathways have been identified to contribute to its pathophysiology. EM drugs that target only one receptor have inadequate therapeutic efficiency. However, although multitarget drugs present potent efficacy in suppressing the progression of EM lesions, they also pose a risk of side effects such as binding to undesirable drug targets and bringing off‐target toxicities. 56 Therefore, designing a drug that targets the appropriate pathways with high selectivity is highly desirable. For this purpose, it is essential to understand the compound‐target pathway‐disease relationships.

59In the theory of Chinese medicine, EM is defined as a blood stasis syndrome that leads to the formation of endometriotic lesion and other associated symptoms. Stagnation of Qi (energy) is believed to be one of the causes of EM. TCM aims to lessen the chronic pain experienced by women with EM. Therefore, studies on the action mechanism of TCM are focused mainly on the alleviation of inflammation and oxidation. TCM decoctions containing several herbs in different compositions, which are varied according to the condition of the patient, are a combinational approach that can target various pathophysiology. Fang et al. 319 and Tsai et al. 320 have identified the decoctions commonly used for treating EM in Taiwan, which included Gui‐Zhi‐Fu‐Ling‐Wan, Dang‐Gui‐Shao‐Yao‐San, Jia‐Wei‐Xiao‐Yao‐San, Shao‐Fu‐Zhu‐Yu‐Tang, and Wen‐Jing‐Tan. The therapeutic efficacy and pathophysiology of TCM in cancer and other diseases have been widely evaluated in vitro and in vivo; however, there are limited studies on the efficacy of TCM for EM.

60Most of the herbs exert anti‐inflammatory effects by inhibiting the production of proinflammatory cytokines. Poria has been confirmed to exert antitumor activities against various cancers. It binds to cytokines and effector immune cells to regulate immunity and upregulate apoptosis. 321 Angelicae Sinensis Radix exerts anti‐inflammatory effects by reducing TNF‐α inflammatory cells. 322 Ligusticum Rhizoma inhibits inflammation and reduces PGE2 production. 323 Moutan Cortex, Glycyrrhizae Radix, Paeoniae Alba Radix, and Bupleuri Radix suppress proinflammatory cytokines via the NF‐κB signaling pathways.324, 325, 326, 327 Paeoniae Alba Radix and Bupleuri Radix also exert such effect via MAPK signaling pathways.

61Atractylodis Ovatae Rhizoma exerts antioxidant effect by activating the MAPK cascades and inhibiting the production of radicals by 2,2‐diphenyl‐1‐picrylhydrazyl and catalases, thus inhibiting the activity of free radicals. 328 Glycyrrhizae Radix and Poria act as radical scavengers against superoxide and hydroxy radicals.328, 329, 330 Ligusticum Rhizoma acts as a reducing agent via the Nrf2 and NF‐κB pathways. 331

62Angelicae Sinensis Radix exerts antiproliferative and proapoptotic effects; it induces mitochondrial‐dependent apoptosis and inhibits the Akt/mTOR pathway. 332 Atractylodis Ovatae Rhizoma induces apoptosis by upregulating ROS. 333 Moutan Cortex exerts proapoptotic effects by increasing Bax/Bcl‐2 expression and decreasing MMP via the formation of apoptosome and cytochrome c, activation of CASP, and the adenosine monophosphate‐activated protein kinase pathway. 323 It also induces apoptosis via activation of CASP‐3/‐8. 334 Paeoniae Alba Radix induces apoptosis via activation of CASP‐3/‐9 293 and exerts antiproliferative activity via cell cycle arrest and Fas/Fas ligand‐mediated apoptotic pathway. 334 It also downregulates the antiapoptotic protein Bcl and upregulates the apoptotic proteins Bax and CASP‐3. 335

63TCMs have great potential as multitarget drugs. As TCMs consist of herbal formulas with various combinations of herbs, they have multiple mechanisms of action, which can be beneficial to reduce the concentration of each herb, thus, drug toxicity. 336 However, the costs and availability vary for different herbs, which limits its acceptability in Western countries at present. Furthermore, there is a lack of clinical management methods to evaluate their clinical effectiveness and standardized regulations of TCM practice.

64This is the first review article combining medicinal research based on EM pathophysiology and the related signaling pathways. Our review revealed the challenges in EM management and the need for various available medical treatment options. Most of the medications prescribed by the FDA to treat EM are hormonal, such as contraceptives, progesterone, and GnRH. However, current hormonal medicines raise a major concern in the case of long‐term treatment. Therefore, new nonhormonal pharmaceuticals with relatively safer and few side effects are urgently needed.

65Our aims in this review were to facilitate the research and development of novel treatments for EM based on an understanding of the pathological process. To compare new and old pharmaceuticals, an effective scale to evaluate parameters between different treatments as well as to align outcome measures from preclinical to clinical studies is needed. There is a lack of experimental and clinical evidence to support the effectiveness, pharmacokinetic, and pharmacodynamic profiles of potential drugs in alleviating the pathophysiology of EM, compared with that of drugs already available in the market. Good practices such as the Endometriosis Phenome and Biobanking Harmonization Project, derived by the World Endometriosis Society, can help facilitate a large‐scale collaboration project worldwide. 337 It is a platform to ensure that the protocol is sufficient and consistent enough to maintain high research quality, datasets are shared to ensure data reproducibility, and results can better support the development of translational medicine. Moreover, multicenter collaboration can increase research visibility and avoid data integrity issues.

66The nonhormonal treatments reviewed in this paper were only studied in vitro or in animal models or are still under clinical trials. The drugs mentioned in this review article showed significant efficacy in reducing ectopic endometrium cell viability and endometriotic lesion size; however, severe adverse effects were not elaborated in‐depth. High efficacy and innovative approach do not guarantee final success. Data from legal regulation and patients’ demand for available resources are as important as the pharmacological profile of medicines. In many countries, a new drug must be regulated and approved by the relevant authority before it is launched in the market. 338 Thus, apart from efficacy and safety data, the medical and financial burden of EM to women have raised the awareness on EM and accelerated the scientific research on this disease, which are key factors considered by R&D investors. To maximize a drug's value and cost‐effectiveness in the market while maintaining its affordability, fulfilling the society's demand, and making scientific advances, modification of lead compound or bioactive compound derived from natural products holds great potential because only the functional groups are modified, whereas the original core structure is conserved.

67Considering both the medicinal and commercial perspectives of drug development, there is a huge pressure in the development of a new drug, starting from the synthesis or discovery stage to clinical trial, to proceeding with legal regulations, and to launch in the market. A drug requires 10–17 years of development, with less than 10% success rate to pass clinical trial. 339

68Taking advantage of big data mining, drug repurposing is a strategy to identify new therapeutic use of a drug that is approved or under clinical trial, which comprises 30% of newly FDA‐approved drugs and vaccines.340, 341 These drugs can bind to the same target owing to the similar pathophysiology of different diseases, or these drugs can have multiple targets and are thus relevant to other diseases. 342 A repurposed drug offers sufficient preclinical pharmacology profile and safety reports, leading to a greater potential for phase III and IV clinical trials, which can reduce the time of drug development and cost of investment. Nevertheless, some advantages of de novo drug development outweigh the benefits of drug repurposing. A constant influx of chemicals via synthesis or extraction from natural products offers novel medical options for patients. A growing understanding of the pathophysiology of EM favors structure‐based or ligand‐based drug designs, in which by modifying lead compounds based on structure–activity relationships, the efficacy, potency, and selectivity can be compromised. However, a more in‐depth research is needed to study the underlying mechanisms and drug targets to support their potential as new EM treatments.

69In conclusion, this review provides an update on the pathophysiology of EM and shows the efficacy of various medicines in treating EM. Increasing attention has been focused on understanding the pathophysiology of EM and the action mechanisms of potential pharmaceuticals; however, many of these are still not completely understood. With this review, we hoped to raise awareness on the missing puzzle pieces and to promote related research that can further advance diagnosis and treatment for better management of EM and improve the quality of women's lives.

70Chi Chiu Wang is an active member of the World Endometriosis Society and an advisor of the Aptorum Group.

71Sze Wan Hung, Ruizhe Zhang, and Chi Chiu Wang participated in research design. Sze Wan Hung participated in data evaluation, extraction and interpretation. Sze Wan Hung, Ruizhe Zhang, and Zhouyurong Tan participated in data validation and in drafting the manuscript. Tao Zhang participated in designing figure 1. Sze Wan Hung, Tao Zhang, Jacqueline Pui Wah Chung and Chi Chiu Wang critically revised the manuscript. All authors approved the final version of the manuscript.