Medical management of endometriosis: emerging evidence linking inflammation to disease pathophysiology.

Endometriosis, clinically defined as the growth of endometrial glands and stroma outside the uterus, is a common gynecologic condition affecting millions of women worldwide. Despite having been recognized within the medical literature for more than 100 years, the disease remains a poorly understood condition affecting approximately 5–10% of all reproductive-age women 1 . Endometriosis can be debilitating, with many patients experiencing severe pelvic pain in addition to reduced fertility. Presently, clinical identification of endometriosis requires laparoscopic surgery and microscopic analysis of excised tissues, likely contributing to the near decade-long delay between onset of symptoms and accurate diagnosis 2 . Recognizing that the development of endometriosis is limited to menstruating species, Dr. John Sampson proposed that peritoneal deposition of displaced menstrual tissue was the primary cause of this disease 3 ; however, retrograde tissue flow is only a mechanical process occurring in most reproductive-age women, the majority of whom do not develop endometriosis 4 . Thus, alternative hypotheses regarding the etiology of this disease have been proposed, including the coelomic metaplasia theory and the development of disease following activation of embryonic cell rests 5 . More recently, it has been suggested that adult stem cells, which play a role in the recurrent process of endometrial self-renewal, may also contribute to the pathogenesis of ectopic sites of endometrial growth 6 . Finally, given the frequent occurrence of endometriosis in first-degree relatives, it is likely that a genetic predisposition and/or environmental factors may also influence the development of this disease 7 . Specifically, environmental toxicants capable of inducing epigenetic modifications during development have been suggested to negatively impact the relationship of the endocrine and immune systems within the adult reproductive tract 8 – 10 . Although it is likely that numerous converging factors ultimately determine a woman's individual risk for developing endometriosis, physicians and scientists generally accept Sampson's theory of retrograde menstruation as a common contributing mechanism to the development of ectopic endometrial growth. Nevertheless, since most cycling women exhibit retrograde menstruation, alterations in key biological processes must additionally be present that allow displaced endometrial tissue to successfully attach and survive ectopically in only a subset of women. In this regard, recent research suggests that the reduced response to progesterone noted in the eutopic endometrium of endometriosis patients 5 , 11 – 13 , combined with the altered nature of immune cells and their proinflammatory products within the peritoneal fluid 14 – 16 , may collectively promote the successful establishment of ectopic disease. Therefore, in this review we discuss the potential cooperative relationship of the progesterone resistant endometrial phenotype and a pro-inflammatory peritoneal microenvironment in the establishment and progression of this disease. Additionally, we describe how viewing endometriosis as a complex inflammatory disease may provide insight into the design of better strategies for targeted medical management. The human endometrium is distinct among adult organs, undergoing steroid-driven cycles of development-like tissue growth and differentiation in preparation for pregnancy, followed by tissue breakdown and bleeding in the absence of nidation. Clearly, in order to successfully initiate ectopic survival and growth, viable cells within fragments of retrograde menstrual tissue must evade peritoneal immune surveillance, avoid post-differentiation apoptotic processes, attach and invade the peritoneal mesothelium and subsequently establish a vascular supply. To this end, the maturational state of various endometrial cell types residing at the interface between the retained basalis region versus the post-mature, shedding functionalis region can be quite different; thus shedding of tissues from the regenerative basalis would likely contribute to a menstrual effluent with the capacity to survive ectopically. Equally relevant, more than a decade ago, scientific reports began to suggest that certain phenotypic properties of the functionalis region of the eutopic endometrium differ between women with and without endometriosis (for example, 5 , 8 , 17 , potentially providing the former tissue a survival advantage within the peritoneum. Therefore, to design better medical and surgical therapies for the treatment of endometriosis we must clarify the phenotypic characteristics of refluxed endometrial cells and how these characteristics affect the interface of endometrial cells with both immune and somatic cells within the peritoneal micro-environment. At this juncture, it is not known whether the phenotype of individual cells within the menses of certain patients contributes to the development of endometriosis via differential expression of specific bioactive agents. Prior to menstruation, the behavior of each individual cell type within the eutopic endometrium is influenced directly or indirectly by their sequential exposure to the ovarian steroids estrogen and progesterone 18 – 20 . Specifically, at the end of each nongravid menstrual cycle, the declining anti-inflammatory effects of progesterone leads to activation of resident immune cells and affects the recruitment of non-resident immune cells; together these cells create a heightened state of inflammation involving the release of multiple cytokines and chemokines that set the biological stage for endometrial breakdown 20 – 21 . Thus, the cyclic loss of endometrial tissue occurs at menstruation as a consequence of inflammation-driven expression and activation of proteolytic enzymes, including members of the matrix metalloproteinase (MMP) family 22 . Multiple members of the MMP family are expressed in a cell-specific pattern and these enzymes are critical for normal endometrial tissue remodeling across the cycle, with the highest levels of MMP expression associated with tissue breakdown at menstruation 22 . Importantly, menstruation represents a controlled inflammatory event and members of the MMP family are intimately involved in mediating various aspects of tissue inflammation in a manner that is independent of extracellular matrix (ECM) degradation within somatic tissues. Indeed, our emerging understanding of the MMP system in regulating active tissue inflammation has led some investigators to consider this family of enzymes to be key components of the overall innate immune system 23 – 24 . The ability of progesterone to largely suppress the MMP system within the endometrium is critical to controlling proinflammatory cytokine activation of these enzymes as immune cells migrate to the human uterus during the secretory phase of the menstrual cycle, in preparation for pregnancy 10 , 22 , 25 . Reflecting the importance of progesterone action, in the absence of nidation, the highest levels of expression and activation of MMPs occurs as the anti-inflammatory action of this steroid is lost, resulting in menstruation 22 . Following each episode of endometrial breakdown, inflammation-related MMP expression persists under the influence of estrogen and focal expression of these enzymes mediates ECM remodeling during the proliferative phase, as endometrial repair and re-growth of the functionalis occurs. After a variable period of focal MMP expression related to estrogen-mediated reconstruction of the glandular architecture of the functionalis region, progesterone rapidly acts to stabilize the endometrium by limiting MMP expression during the invasive establishment of pregnancy. Although menstruation and implantation are each inflammatory processes, MMP expression during pregnancy establishment must be tightly regulated in order to preserve endometrial integrity 22 . In contrast to healthy women, the elevated endometrial MMP expression noted during the time of secretory maturation in tissues acquired from women with endometriosis 12 , 26 strongly suggests a disease-related failure of progesterone to appropriately regulate the cross-talk between the endometrial endocrine and immune systems. The ability of progesterone to balance the endocrine/immune physiology of the eutopic endometrium is a critical component of the function of the entire female reproductive tract. Among endometriosis patients, the specific failure of progesterone to act appropriately during endometrial differentiation ultimately affects the phenotype of tissue shed at menstruation, a key risk factor that affects not only the likelihood of successful ectopic growth but progression of disease and development of its associated symptoms. For example, initial studies comparing endometrial tissues from women with and without endometriosis examined circulating progesterone levels relative to expected histological responses across the secretory phase of the menstrual cycle 27 – 29 . These investigations revealed that while women with endometriosis exhibit normal circulating ovarian progesterone levels, the endometrium's ability to respond appropriately to this steroid appeared to be reduced 27 – 29 . Subsequent studies confirmed that endometrial tissues from women with endometriosis did not exhibit the changes in specific gene and protein expression normally expected during the progesterone-dominated secretory phase 12 , 30 – 31 . Perhaps not surprisingly, altered expression of genes and proteins in endometriosis patients was reported to be associated with changes in the expression pattern of progesterone receptor (PR) isotypes (PR-A and PR-B), at both eutopic and ectopic sites of endometrial growth 11 , 32 – 33 . At present, the biological origin of reduced endometrial progesterone responsiveness among women with endometriosis remains to be fully elucidated; however, a number of research groups have begun to examine whether chronic inflammatory processes may promote the development of endometrial resistance to this steroid. Within the reproductive tract, an important component of steroidal regulation of inflammation involves cellular signaling by members of the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kappaB) family. This signaling network has been suggested to play a critical role in triggering, or enhancing, the inflammatory processes associated with the development and progression of endometriosis 34 . The NF-kappaB family is best described as a protein complex that controls DNA transcription, resulting in the regulation of a wide array of genes that collectively mediate cellular processes such as cell proliferation, adhesion, apoptosis, angiogenesis and immune responses 35 . Although NF-kappaB activity varies throughout the normal menstrual cycle 36 , a mutual suppressive effect has been reported between NF-kappaB and progesterone, suggesting that reduced PR protein expression in endometriosis patients would lead to an increase in inflammation 34 . More specifically, at sites of ectopic endometrial growth in women with endometriosis, continuous expression of pro-inflammatory cytokines would promote NF-kappaB activation 34 ; thereby mediating a loss of PR expression, resulting in a failure to suppress NF kappaB action. Thus, constitutive activation of NF-kappaB could contribute significantly to the development of the progesterone resistant “endometriosis phenotype” and negatively influence multiple PR-dependent biological processes involved in the pathogenesis of this disease. Although the studies noted above support the concept that inflammatory processes may represent a potential trigger for the loss of progesterone sensitivity related to endometriosis, the precise cellular and molecular mechanisms leading to this disease phenotype remain elusive. In this regard, a number of recent observations suggest that epigenetic modification, mediated by chronic inflammation, could explain the progesterone resistant endometrial phenotype observed in women with endometriosis. Specifically, several studies have examined whether epigenetic modifications might affect the transcriptional regulation of the PR isoforms PR-A and PR-B; nuclear isotypes that exhibit distinctly different effects on uterine gene expression across the menstrual cycle. The truncated isoform, PR-A has been associated with transcriptional inhibition of progesterone action while the ligand-bound PR-B isoform promotes many of the unique anti-inflammatory effects of this steroid that supports endometrial differentiation 37 . Therefore, epigenetic modifications leading to alterations in PR isotype expression would likely negatively impact normal progesterone-responsive gene expression and promote the development and progression of endometriosis 8 , 18 , 30 . In vivo studies have clearly demonstrated a shift in the PR-A/PR-B ratio at eutopic and ectopic sites of endometrial growth in tissues acquired from women with endometriosis 11 , 32 – 33 . Recent in vitro studies further suggest that reductions in PR-B expression are linked to hypermethylation of the PR promoter and that partial methylation of this gene could be induced by prolonged stimulation of cells with tumor necrosis factor alpha (TNF-α) 38 – 39 . Alternatively, recent studies from the Mendelson laboratory suggest a relationship between increased expression of the inhibitory PR-C isoform and the heightened inflammatory response observed in women with endometriosis 40 . Clearly, disease-related epigenetic modifications, specifically affecting expression levels of PR-A, PR-B or PR-C would likely alter the anti-inflammatory actions of progesterone at both eutopic and ectopic sites of endometrial growth. As discussed above, emerging evidence suggests that the pathogenesis of endometriosis is related in part to a loss of progesterone's anti-inflammatory actions, a defect that not only impacts the progression of ectopic disease but also affects the function of the eutopic endometrium. Once inflammation-related patterns of cell-cell communication are established the local proinflammatory microenvironment may further contribute to the progesterone-resistant phenotype, creating a negative feedback loop that promotes disease progression. In women with active endometriosis, reduced progesterone sensitivity within the reproductive tract likely affects the progression of ectopic disease by impacting critical elements of immune function within the peritoneal cavity, including: apoptosis, immune surveillance, attachment/invasion and establishment of vasculature through angiogenesis ( Figure 1 ). Apoptosis is an important mechanism by which autologous cells are eliminated after death without eliciting an inflammatory response. Since apoptosis plays a critical role in maintaining tissue homeostasis during periods of endometrial growth, differentiation and at menstruation, the cycling human endometrium normally exhibits variations in apoptotic activity. For example, endometrial expression of the B cell lymphoma/leukemia-2 (BCL-2) gene, which suppresses apoptosis, is highest during the maximum period of growth that occurs during the proliferative phase. In contrast, late secretory phase endometrial tissue, under the influence of declining progesterone, exhibits an increased expression of pro-apoptotic proteins 41 , a biological response that normally promotes cell death and thus increases phagocytosis of sloughed menstrual tissue within the peritoneal cavity. Therefore, apoptotic activity is normally lowest during periods of endometrial proliferation and highest immediately prior to menstruation 42 , suggesting that cell death acts to limit the likelihood of ectopic survival and growth. Importantly, the activity of BCL-2 is opposed by BCL-2 associated X protein (BAX) and the resultant apoptotic index within different regions of the endometrium can also vary during the menstrual cycle related to the BCL-2/BAX ratio 41 . For example, in response to estradiol, the human endometrium regenerates from the basalis region during each menstrual cycle; therefore, as opposed to the cyclic regulation of BCL-2 and BAX within the endometrial functionalis , the basalis exhibits continuous expression of BCL-2 and minimal apoptosis 43 . Since cells arising from the basalis region, as opposed to the functionalis , avoid cell death associated with terminal differentiation, it is logical to postulate that ectopic endometrial growth could principally arise from tissue fragments shed from this compartment during menstruation. Alternatively, the apoptotic indices have been found to be reduced within the endometrial functionalis region of women with endometriosis compared to disease-free women, primarily due to a decrease in apoptosis during the late secretory/menstrual and early proliferative phases. For this reason, the endometrial functionalis of women with endometriosis may exhibit basalis -like characteristics and thus be better able to survive ectopically compared to tissues from disease-free women 44 – 45 . These reported differences in the cyclic patterns of apoptosis-related proteins in the eutopic endometrium of women with and without endometriosis suggest that alterations in progesterone-regulated apoptotic gene expression may be part of the pathophysiology of this disease. Specifically, the progesterone resistant endometrial phenotype of women with endometriosis would be expected to result in a loss of progesterone mediated regulation of apoptotic proteins during endometrial differentiation. Consistent with this theory, the number of non-apoptotic cells arising from either the basalis or functionalis region and flowing into the peritoneal cavity during retrograde menstruation appears to be greater in women with endometriosis 46 . Following menstruation and the deposition of this tissue into the peritoneal cavity, these non-apoptotic cells may further contribute to an altered innate immune response that allows displaced menstrual tissues to avoid immunosurveillance. Following menstruation, immune surveillance within the peritoneal cavity by resident and migrating immune cells provides an important line of defense against the development of endometriosis by removing displaced endometrial tissues. Although alterations within the immune system of endometriosis patients has been broadly noted by many investigators 47 , a lack of appropriate experimental models limits our current understanding of the role of the innate immune system within the peritoneal microenvironment related to a woman's risk for developing this disease. While prospective human studies are not possible, we recently demonstrated in a humanized model of experimental endometriosis that normal immune cell function within the peritoneal cavity acts to limit the development of ectopic endometrial growth 48 . This finding suggests that endometrial cells arising from the endometrium of women with endometriosis exhibit an altered capacity to interact with peritoneal immune cells which promotes the establishment of persistent sites of ectopic growth. In this regard, the inflammatory mediator, prostaglandin E2 (PGE2), has been proposed as a master-regulator of endometriosis 49 , due in part to the impact of this potent cytokine on macrophage behavior. PGE2 is abundant in the peritoneal fluid of women with endometriosis and serves to inhibit MMP-9 activity as well as the production of this enzyme by peritoneal macrophages 50 – 51 . Macrophages are the major resident immune cell population within the peritoneal cavity which act to eliminate apoptotic cells and debris, including endometrial tissues deposited via retrograde menstruation. While the concentration of peritoneal macrophages has been shown to be increased in patients with endometriosis compared to disease-free women 4 , their phagocytic capacity and uptake of debris has been shown to be decreased due to a reduced expression and activity of MMP-9 52 and a downregulation of the scavenger receptor, CD36 50 . Thus, while defects in macrophage phagocytic behavior could be a critical factor in the initial establishment of endometriosis, these cells continue to produce cytokines, growth factors and potent angiogenic factors which may promote the ectopic growth of endometrial tissue fragments that survive phagocytosis 53 . Supporting this possibility, investigators using an experimental mouse model of endometriosis observed that the peritoneal environment can dramatically influence the differentiation of macrophage precursors towards alternatively activated mature macrophages, cells that can impact the vascularization and growth of ectopic lesions 54 . Defects in multiple cell types within the innate immune system, including natural killer (NK) cells, may also affect the clearance of endometrial tissue within the peritoneal cavity of endometriosis patients. Although there are conflicting reports on whether or not the number of peritoneal NK cells is altered in women with endometriosis, most studies indicate that NK cells from these patients display reduced cytotoxicity 55 – 56 resulting from an increased expression of killer inhibitory receptor (KIR) and altered antigenicity due to over-expression of HLA class I 57 – 58 . Finally, the cytotoxicity of T cells is reduced in women with endometriosis 59 and the peritoneal fluid of women with this disease may contribute to the survival of displaced endometrium by inducing apoptosis in cytotoxic lymphocytes via the Fas-FasL pathway 60 . Collectively, macrophages, NK cells and cytotoxic T-lymphocytes in women with endometriosis may provide a more immunotolerant peritoneal environment than would normally exist, thus facilitating rather than inhibiting the disease process 61 . Although endometriosis-related changes in peritoneal macrophage function leads to a reduction in phagocytic activity, the continued ability of these cells to produce proinflammatory cytokines, as noted above, may mediate the recruitment of additional inflammatory immune cells to sites of peritoneal disease which may further exacerbate an excessive inflammatory microenvironment 62 . For example, studies have indicated a higher percentage of peritoneal neutrophils in women with endometriosis compared with disease-free women 63 , although the activation status and function of these cells remains unclear. In addition to neutrophils, type 17 T-helper (Th-17) cells and regulatory T-cells (Tregs) have a suspected, though not well-studied, role in the pathogenesis of endometriosis. Recently, Hirata et al. 64 demonstrated recruitment of type 17 T-helper (Th-17) cells to endometriotic tissues, cells that produce inflammatory signals affecting the recruitment, activation and migration of neutrophils. At this juncture, it remains to be determined whether Th17 cell-mediated signaling affects neutrophil migration and inflammatory behavior at sites of ectopic endometrial growth among endometriosis patients. As studies begin to focus on whether altered endometrial cell interactions with peritoneal immune cells determines the risk for establishment of endometriosis, understanding the potential role(s) of each specific immune cell populations will be necessary. In this regard, emerging information suggests that Tregs may represent one of the most important immune cells in the pathogenesis of endometriosis due to the potential role of these cells in the regulation of disease-related inflammatory responses. In disease-free women, Tregs are most prominent during the estrogen dominant proliferative phase, while their number is significantly reduced during the progesterone-dominant secretory phase of the menstrual cycle. However, Tregs remain abundant within the endometrium during the secretory-phase in women with endometriosis, perhaps reflecting the reduced progesterone responsive endometrial phenotype associated with this disease. It has been proposed that preservation of Tregs in women with endometriosis decreases the ability of newly recruited immune cell populations to effectively recognize and target endometrial antigens during menstruation, potentially contributing to the survival and implantation of shed endometrial cells 62 . Similarly, suppression of local immune responses by a Treg cell dependent mechanism could underlie deficient clearing of ectopic tissues within the peritoneal microenvironment as discussed above. At this juncture, it remains to be determined how the trafficking and function of various interactive immune cell populations is impacted by the progesterone resistant endometrial and peritoneal microenvironments associated with endometriosis. However, progesterone provides key immunosuppressive actions within the reproductive tract, therefore some of the immunopathologies that have been noted among women with endometriosis likely reflect the biological consequence of endometrial tissue microenvironments that are resistant to this steroid. For instance, progesterone has been shown to antagonize estrogen-regulated neutrophil recruitment and function within the uterus of mice 65 – 66 while blocking progesterone action promotes neutrophil recruitment and activation 67 . Peripheral NK cells contain both PR isoforms, while peritoneal and endometrial macrophages also express PR 68 – 69 . These and other related studies strongly suggest that the anti-inflammatory action of progesterone not only regulates the behavior of somatic cell types but also affects their interaction with multiple immune cells across the menstrual cycle. The studies noted above demonstrate that alterations in the ability of progesterone to appropriately regulate the trafficking and function of immune cells within the eutopic endometrium as well as within the peritoneal cavity likely contributes to the pathophysiology of endometriosis. However, it must also be considered that genetic or epigenetic defects within various somatic and/or immune cell populations may equally impact an individual's risk of developing endometriosis by disrupting endocrine-immune cell communication 8 , 18 , 30 . To accept Sampson's theory of retrograde menstruation as a primary mechanism for establishment of endometriosis, it is necessary to also accept that cells within fragments of endometrial tissue are not only able to avoid apoptosis and immune defenses but also successfully accomplish a significant invasive event in a short period of time. Specifically, for continued growth and disease progression, endometrial cells which survive the peritoneal defenses must attach and invade a mesothelial surface site and rapidly acquire a vascular supply. Since it is not practical to examine each of these specific cell-cell interactions directly in women diagnosed with endometriosis, investigators have turned to various in vitro and in vivo model systems. More than a decade ago it was hypothesized that an injury of the mesothelial cell surface would be necessary for successful invasion by menstrual endometrial cells 70 . Nevertheless, endometrial cells obtained from either the proliferative or secretory phases of the menstrual cycle were shown to have the in vitro ability to invade an intact mesothelium 71 – 73 . Using a similar in vitro approach, Schenken and colleagues further demonstrated that viable menstrual endometrial cells obtained from women with endometriosis exhibit a greater capability for attachment to peritoneal mesothelial cells compared to cells obtained from the tissue of disease-free women 74 . In addition to the potential that endometriosis patients exhibit a unique endometrial cell phenotype, at least two studies have suggested that TNF-α, a cytokine more abundant in the peritoneal fluid of these patients 75 , can increase the adherence of endometrial stromal cells to mesothelial cells in vitro 76 . In contrast to these studies, another research group demonstrated a dose-dependent inhibition of endometrial stromal cell adhesion to a mesothelial cell monolayer by TNF-α and other pro-inflammatory cytokines 77 . Currently, it is difficult to reconcile the conflicting findings of these various in vitro studies; however, cell culture models of endometrial cell attachment to mesothelial cells may not represent the same in vivo challenges that endometrial fragments face within the peritoneal cavity at the time of retrograde menstruation. The in vitro studies noted above represent important steps in our understanding of how individual endometrial cell behavior in response to various biomolecules contributes to peritoneal invasion during initiation of endometriosis. However, in vivo studies are equally necessary to reveal the physiological role(s) of complex interactions that occur between multiple cells types that characterize this invasive disease. As discussed below, some of the most relevant experimental observations in regard to the capability of human endometrial fragments to successfully establish ectopic sites of growth have been made using chimeric models in which human endometrial tissue is injected into the peritoneal cavity of immunocompromised mice. Chimeric models have allowed investigators to explore whether an altered peritoneal cytokine/chemokine microenvironment may represent a key contributory factor to the initial establishment of this disease by endometrial cells entering the peritoneal cavity as well as modulate the action of steroids on disease progression. To address the potential interactive role of various endocrine and immune factors in the establishment of endometriosis our laboratory established an in vivo model using immunocompromised nude mice, animals which accept human tissue xenographs. Using this system, we initially demonstrated that under the influence of estrogen, cells within normal human endometrial tissue fragments readily attach and invade an intact mesothelial surface resulting in establishment of viable ectopic sites of growth 78 . Significantly, compared to exposure of endometrial tissue fragments to estrogen alone, additional exposure to IL-1α, an abundant proinflammatory cytokine within the peritoneal fluid of women with endometriosis 55 , led to the development of larger and more numerous ectopic lesions in our experimental model 79 . Taken together, numerous in vitro and in vivo studies from multiple laboratories now suggest that invasion of an intact peritoneal mesothelial lining is unlikely to represent a significant barrier to the establishment of endometriosis 80 . Additionally, supporting in vitro observations, endometrial tissue obtained from endometriosis patients has been shown to exhibit a greater capacity for the establishment of experimental endometriosis in vivo when compared to endometrial tissue acquired from disease-free tissue donors 60 , 66 . Therefore, it is likely that the endometrial phenotype observed in endometriosis patients together with a permissive inflammatory-like peritoneal microenvironment ultimately determines the nature of the disease process leading to medically significant disease. A less understood, yet equally critical component in the successful establishment of endometriosis, is the nature of the biological host site response of the mesothelial and other cells at the site of ectopic invasion. In this regard, we recently reported that one of the earliest events associated with ectopic endometrial lesion establishment in mice was an apparent encapsulation of the human tissue fragment by the murine mesothelium within 16 hours of tissue injection 81 . Certainly, as discussed above, in vitro studies indicate that eutopic endometrial tissue acquired from patients with endometriosis exhibits significantly more invasion-related adhesive 80 and proteolytic activity compared to similar tissue acquired from control individuals 82 . Additionally, we have previously shown that inflammation-related production of MMPs is a critical component of peritoneal invasion and suppressing the expression of these enzymes by endometrial tissue fragments with progesterone or blocking their activity with TIMP-1, a natural inhibitor of MMP action, effectively inhibits the establishment of experimental endometriosis 78 . However, our histological observations suggest that the “wound-like” mesothelial cell reaction at the initial attachment site of endometrial tissue fragments within the peritoneal cavity may actively participate in successful establishment and survival of the ectopic lesion. Given that patients with endometriosis exhibit a generalized increase in peritoneal fluid proteolytic activity compared to matched control samples 83 , proteolytic enzymes arising from peritoneal sources likely increase the capacity of retrograde menstrual tissue from these patients to successfully penetrate the peritoneal mesothelium. While fully understanding the role played by the peritoneal mesothelial lining and other cells at the host site to successful establishment of endometriosis will require further study, this information should contribute significantly to the development of better therapeutics for the prevention or treatment of endometriosis. As discussed above, although multiple biological triggers are involved in the initiation and progression of endometriosis, the cellular and molecular mechanisms specifically responsible for the progesterone resistant endometrial phenotype appear to involve inflammation-like patterns of cell-to-cell and tissue-to-tissue signaling. Thus, an appealing therapeutic approach for the medical management of this disease would be to utilize anti-inflammatory agents capable of normalizing the progesterone resistant endometriosis phenotype or which block key elements of the inflammatory processes associated with successful ectopic growth. For example, among anti-inflammatory agents in current use for other diseases, statins are known to decrease levels of various mediators and markers of inflammation including c-reactive protein, TNF-α, several interleukins and monocyte chemotactic protein-1 (MCP-1) 84 – 86 . Thus, we and others have explored the therapeutic potential of different statins to reduce disease burden using in vivo models of experimental endometriosis 81 , 87 – 89 . In our study, we found that simvastatin treatment of mice bearing experimental endometriosis led to a significant reduction in the number and volume of ectopic lesions, partly through the ability of this statin to protect against inflammation 81 . Although the precise mechanism(s) of simvastatin action in our experimental endometriosis model has yet to be determined, statins are known to regulate multiple processes associated with initiation of endometriosis, including cell proliferation, apoptosis, cell morphology and motility/invasiveness. Following successful attachment, ectopic survival of endometrial tissues within the peritoneal cavity requires acquisition of a vascular supply. Thus, another critical area of endometriosis research related to inflammation is to unravel the biological crosstalk between the host invasion sites and displaced endometrial tissue fragments which promote new blood vessel growth. Angiogenesis is a mandatory process in the pathogenesis of endometriosis in women 90 as well as in the development of experimental endometriosis in mice 91 . Therefore, it is not surprising that vascular endothelial growth factor (VEGF) expression is increased in the peritoneal fluid from patients with endometriosis compared with control women 92 – 93 , that VEGF levels correlate with the stage of disease 92 – 93 and that this growth factor appears to play a prominent role in vascularization of endometriotic tissues 94 . Therapeutically, anti-inflammatory agents would be expected to inhibit vascular development at ectopic sites of endometrial growth since a number of pro-inflammatory cytokines modulate the expression of VEGF by activated peritoneal macrophages, migrating neutrophils and by somatic cells within the endometriotic lesions. In a collaborative study using an experimental endometriosis model, we found that soluble flt-1, a VEGF receptor antagonist, was quite effective in blocking the formation of ectopic human lesions 95 . This is consistent with findings using a similar experimental endometriosis model in which murine vessels from the peritoneum were found to invade endometrial implants 5–8 days after human tissue introduction into mice, coinciding with an increase in VEGF production 96 . More recently, we examined the influence of the anti-inflammatory drug pioglitazone on vascularization in experimental endometriosis in nude mice 97 . Mice receiving pioglitazone treatment not only exhibited markedly fewer lesions, but lesions that were present had a significant reduction in microvessel density compared to lesions from control mice. Taken together, these studies suggest that an additional benefit of anti-inflammatory agents for the treatment of endometriosis may be to limit vascularization at ectopic sites of growth. As noted above, inflammatory processes play a key role in the early establishment of experimental endometriosis and numerous studies now provide evidence that the endometriosis-related endometrial phenotype and peritoneal inflammation may work in concert to promote ectopic endometrial growth. Nevertheless, the key role that the peritoneal microenvironment plays in the establishment of endometriosis is often underappreciated in the clinical management of this disease. For example, in two recent studies, we examined the peritoneal microenvironment associated with a recent surgical injury on the development of experimental endometriosis using a nude mouse model 97 – 98 . In the initial study, mice were injected with human endometrial tissues at various time points following ovariectomy or sham surgery. Mice injected with human endometrial fragments within 16 hours of ovariectomy exhibited more extensive ectopic disease compared to animals receiving tissues after 36 hours from the time of surgery or following sham surgery. Importantly, surgical injury had a profound effect on microvessel density (MVD), which was greatest in lesions established closest to the time of peritoneal surgery. In a second study we demonstrated that the anti-inflammatory action of dietary fish oil supplementation can reduce peritoneal inflammation and thus limit the both the development of experimental endometriosis and related adhesions 98 . Although surgical treatment of endometriosis is common, the independent influence of the surgical procedure itself on the inflammatory state of the peritoneal microenvironment is generally not considered as a potential trigger for the reestablishment or progression of disease. Despite the apparent lack of appreciation of this relationship, it is likely that the success of our fish oil therapy was also associated with a reduction in peritoneal inflammation. As noted previously in this review, by utilizing a severely immunocompromised murine model which allows the adoptive transfer of both human immune cells and endometrial tissues, we demonstrated an important role of normal immune cell function in controlling peritoneal inflammation and preventing development of endometriosis 48 . Certainly, the occurrence of endometriosis in only certain women suggest that the behavior of immune cells that participate in peritoneal inflammation may be altered in women at risk for developing this disease and numerous groups have found that anti-inflammatory agents are effective in reducing disease burden in experimental models of endometriosis 81 , 99 . In future studies, it will be important to examine the influence of immune cells acquired from women with endometriosis relative to the establishment and progression of experimental disease. Specifically, would immune cells acquired from women with active endometriosis fail to impede the survival of ectopic disease? Answering this question would no doubt contribute significantly to our understanding of the role that immune system disruption plays in the initiation and progression of this disease and provide a basis for the development of new therapeutic approaches.

Given the potential for many years of individual suffering as well as significant costs to the healthcare system, there is an urgent need to better understand the basic cellular processes leading to the development of endometriosis in order to improve medical management of this disease. The experimental studies presented herein illustrate the evolving view that the development of endometriosis is both an endocrine and immune disease. As such, the proinflammatory nature of the endometriosis phenotype appears to contribute not only to survival and propagation of ectopic endometrial tissues, but also to the infertility, adhesive disease and pelvic pain which are frequently associated with this disease. Proinflammatory cytokines are known to reduce endometrial progesterone responsiveness, further exacerbating the disease and its symptoms. Importantly, it has been suggested that the proinflammatory nature of the peritoneal fluid of women with endometriosis may promote endometrial cell escape from immune surveillance, allowing survival and establishment of disease 100 . Thus, inadequate removal of refluxed menstrual debris, coupled with an enhanced ability of sloughed endometrial tissue to evade immune surveillance and invade a peritoneal site, likely represent the most critical factors in determining an individual woman's risk for developing endometriosis. Since predicting and treating inflammatory processes related to the development of an endometriosis phenotype prior to recognition of disease symptoms is not presently feasible; current medical strategies focus on treatment of existing disease. However, treatment options for women with endometriosis remain limited and frequently involve either hormonal manipulation and/or surgery. For many women, the side effects of medical therapy are poorly tolerated, while surgical treatment is generally non-curative with a high rate of recurrence. Developing a greater insight into the cellular and molecular mechanisms by which a hyper-inflammatory state is initiated in both the reproductive tract and peritoneal cavity of women with endometriosis is likely a prerequisite to designing better and more effective medical management strategies in the future. As we gain insight into the initial processes associated with the early establishment of ectopic growth, it may be possible to design specific therapeutic targets which can prevent this debilitating disease.