Chapter 49 Endometriosis
INTRODUCTION
The prevalence of endometriosis in the general population is unknown and will remain so until a reliable, noninvasive test can be applied to a large unselected population. Because the standard method of diagnosing endometriosis is an invasive surgical procedure, any studied population will be symptomatic and will represent a biased segment of the overall population. Within these symptomatic populations, the reported prevalence has been increasing over the past two decades. This is likely due to the recognition of more subtle forms of endometriosis and the more liberal use of diagnostic laparoscopy.1 This observed increased prevalence might also represent a real increase in the incidence of endometriosis due to sociological, environmental, and genetic factors.
TYPES OF ENDOMETRIOSIS
The microscopic definition of endometriosis implies the presence of endometrial glands and stroma outside the endometrial cavity and uterine musculature. After the introduction of laparoscopy in the late 1960s, black puckered peritoneal lesions were the first lesions to be recognized as representing endometriosis. In the mid-1980s, we recognized that different forms of nonpigmented lesions also contained the required histologic features of endometriosis. Deep infiltrating endometriosis2,3 is recognized as a separate clinical entity that might be easily missed at laparoscopy and may have a different etiology than the classic peritoneal and ovarian lesions.
Typical brown/black lesions are now believed to be the end result of subtle lesions that have evolved over a number of years.4 New lesions appear red and then turn white as they shed and grow, and they ultimately turn black as a result of both the hemosiderin deposition and intraluminal debris. This theory of lesion evolution is supported by the average age at which these lesions are found in patients. The red lesions are found mainly in younger patients; the black and white lesions are generally observed in older patients.5 The black lesions may be of different dimension and depth of penetration than the red and white lesions. The red lesions are the most metabolically active, whereas the white lesions are the least active.
Ovarian Endometriosis
Ovarian endometriosis, or endometriomas, increase with age and are generally associated with a more advanced stage of the disease.6 This form of endometriosis can be diagnosed with a high level of accuracy by serial ultrasounds.7 The main differential diagnosis is a hemorrhagic corpus luteum, which will disappear over the course of a few months. These ovarian forms of endometriosis often have associated peritoneal implants, although only a small percentage of patients with peritoneal implants will eventually develop an endometrioma. Endometriomas can be uniloculated or multiloculated. They are more commonly localized in the left ovary8 because peritoneal implants are more common in the left hemipelvis.
Deep Endometriosis
Invasion of endometriosis deeper than 5 mm has been associated with increased pain.9 These lesions have been grouped with the following classification: large area surrounded by sclerotic white lesions (Type I), area of retraction of the bowel (Type II), a nodule of the rectovaginal septum (Type III), and sclerosing endometriosis invading the sigmoid colon (Type IV). These lesions can be difficult to classify during laparoscopy due to the retroperitoneal nature of the disease. A rectovaginal examination during the menstrual period in the office setting or a thorough examination under anesthesia before laparoscopy may alert the surgeon to the presence of these types of lesions.
PREVALENCE OF ENDOMETRIOSIS
Over the past three decades, the incidence of endometriosis in patients who present with either pelvic pain or infertility has been increasing and has been reported to be as high as 80%.10 This increase has been proportional to the surgeon’s awareness of subtle lesions and deep endometriosis. If microscopic endometriosis is included, virtually all women with pelvic pain or infertility have some form of endometriosis. The classic lesions generally appear later in life and will have an incidence as high as 70% in patients with pelvic pain or infertility.11
The incidence of ovarian cystic endometriosis (endometriomas) increases with age,6 but its incidence and progression seem to stabilize in patients who are older than age 35. The prevalence of endometriosis in the general population is estimated to range from 1% to 3%. Approximately 10% to 20% of patients undergoing laparoscopy for pelvic pain or infertility are found to have deep endometriosis.6
CLASSIFICATION OF ENDOMETRIOSIS
The classification of endometriosis has been an evolving process. The first classification system to gain widespread acceptance was that of Acosta and colleagues in 1973,12 which classified endometriosis as mild, moderate, or severe and showed an inverse relationship to pregnancy rates. In 1978, the American Fertility Society (AFS, now called the American Society for Reproductive Medicine [ASRM]) convened a committee to design a classification system, which was published in 1979.13 This classification included four stages and used a weighted pain score that included assessment of the extent of endometriosis in two dimensions and the presence and extent of adhesions in the peritoneum, ovaries, and tubes. It also took into account whether the endometriosis was unilateral or bilateral. The size of the endometrioma was also taken into account, along with the presence of filmy versus dense adhesions.
In 1985,14 the original AFS classification was revised because of its inability to predict pregnancy. It now includes a three-dimensional assessment of the disease differentiating superficial from invasive disease (Figs. 49-1 and 49-2). It also quantifies the extent of adhesion around the tubes and ovaries and distinguishes between incomplete and complete enclosure of the fimbria, the latter of which automatically places the disease in the moderate category. Cul-de-sac adhesions are weighted heavily, with complete obliteration, which is characteristic of severe disease, resulting in a stage IV assignment. An endometrioma that is larger than 3 cm in diameter is at least stage III disease.
In 1996,15 a further revision was introduced in the AFS classification (ASRM) in an attempt to correlate the results with pain. This included a description and photography of the lesions. The lesions are categorized as red (red, red-pink, clear, white, yellow-brown, and peritoneal defects) and black (blue and black) (Figs. 49-3 and 49-4). The percentage of each of these lesions is also documented. Despite these improvements, the correlation between the stage of endometriosis and the likelihood of pregnancy is poor, and future improvement in the classification is likely.16
Figure 49-3 Typical black and white lesions of the pelvic peritoneum with dense fibrosis and retraction.
Another limitation of the ASRM classification stems from the fact that this system was mainly designed to assess a patient’s fertility potential and correlates poorly with the pain level. Many authors17 have found no correlation between the degree of pain and the ASRM classification system. However, a correlation between pain and the depth of penetration (5 mm or more) does seem to exist.9
PREDISPOSING FACTORS FOR ENDOMETRIOSIS
There is a correlation between the volume of menstrual flow and the prevalence of endometriosis. This may lead to an increase in menstrual tubal reflux. Clinical conditions that are associated with an increase in menstrual reflux—a noncommunicating uterine horn, for example—are associated with a virtually 100% incidence of endometriosis.2 Increased exposure to menstrual flow, which occurs with early menarche and nulliparity, has been correlated with a high incidence of the disease.18 In a recent study of 2777 women with endometriosis,19 the authors reported that the proportion of patients who had a uterine fibroid was significantly higher than those without fibroids. This finding suggests a causal relationship between endometriosis and the presence of uterine fibroids, which might be due to an increased menstrual flow reflux in these patients.
There appears to be some correlation between a higher degree of education, social stress, and the prevalence of disease.19 The Nurses’ Health Study prospectively assessed predisposing factors for endometriosis and observed an association with early age of menarche, shorter length of menstrual cycles during late adolescence, and nulliparity. In parous women, a decreased risk was found with the number of liveborn children and duration of lactation.20 In the best animal model for endometriosis, the stress of captivity was associated with a higher incidence of endometriosis.21
Heredity is an important predisposing factor for endometriosis because the prevalence is increased sevenfold among first-degree relatives. In monozygotic twins, the prevalence increased 15-fold.22 Exposure to pollutants, especially endocrine-disrupting compounds such as dioxins or polychlorinated biphenyls, might also play a role in the predisposition to endometriosis. In the baboon, for instance, exposure to increasing amounts of dioxin correlated with the incidence and extent of endometriosis.23 In vitro data has also shown that these dioxins affect the expression of stromal progesterone receptor expression.24 Some human data also support a role for these toxins in the pathophysiology of endometriosis.25
There is a correlation between infertility, pelvic pain, dysmenorrhea, dyspareunia, and endometriosis. Although stage I endometriosis may not be more common in infertile women than in fertile women, stage II is significantly more common in infertile patients.26
Although some authors have found a correlation between alcohol use and endometriosis, others have not been able to confirm these results. The use of oral contraception has been linked to an increased risk27 of endometriosis, but other studies either found a protective effect28 or no correlation at all.19 The use of an intrauterine contraceptive device is not associated with increased incidence of endometriosis.19 Body mass index has not been correlated with the incidence of endometriosis. It is unknown whether diet and exercise can halt or limit the evolution of endometriosis. Early exposure to pregnancy may help protect against endometriosis. Indeed, symptoms associated with endometriosis such as pelvic pain are often decreased for a number of years after pregnancy.
The difficulty with assessing risk is often due to the common presence of endometriosis in the pelvis that may not be pathologic. Based on a study that involved repetitive laparoscopy in baboons, we know that endometriosis appears spontaneously29 and that new lesions seem to appear and disappear spontaneously. Thus, endometriotic lesions in humans most likely are in continuous evolution as well.
PATHOPHYSIOLOGY
Transplantation Theory
The basic components of the disease process are an abnormal endometrium, invasion of the peritoneum, angiogenesis, and an immune response and inflammation (Table 49-1; Fig. 49-5). The most accepted general concept of the etiology of this disease is the transplantation theory of Sampson. In 1927, Sampson proposed that endometrial cells regurgitated through the fallopian tubes and implanted on different peritoneal surfaces, leading to endometriosis. However, three major elements must be present to support the Sampson theory: first, endometrial cells must be refluxed through the fallopian tubes at the time of menstruation; second, these cells must be viable; and third, they should implant anatomically based on the distribution of reflux cells in the peritoneal cavity.
Table 49-1 Important Components for Implantation of Endometrium
MMP = matrix metalloproteinases; VEGF = vascular endothelial growth factor
An intact peritoneal surface does not act as a barrier to endometriosis formation.30 However, certain conditions need to be present for endometrial tissue to invade and persist. Recent data from the literature suggest that the endometrium of patients who are in the process of developing endometriosis is intrinsically different from that of patients without endometriosis.
Intrinsic Abnormalities of the Endometrium
Apoptosis
Apoptosis plays a critical role in the normal function of the endometrium.31 This process helps eliminate senescent cells from the functional layer of the late secretory and menstrual endometrium. Apoptosis in the eutopic and ectopic endometrium is controlled by the Bcl-2 and Fas/FasL systems. The Bcl-2 system diminishes apoptosis, whereas the Fas/FasL system promotes apoptosis. These processes are altered in the endometrium of patients with endometriosis in a way that decreases apoptosis, leading possibly to an increase of viable regurgitated endometrial cells.
Decreased Progesterone Sensitivity
Matrix metalloproteinases promote tumor-cell invasion of tissues. Alterations in these enzymes are therefore likely to increase the invasive potential of refluxed endometrium.32,33 Furthermore, altered expression of these enzymes during the luteal phase may be associated with implantation disorders in patients with endometriosis.
For endometrial cells to attach to the peritoneal surface, they must bind to the extracellular matrix (ECM). After attachment of endometrial cells to the peritoneal surface, endometrial glands must invade the peritoneal surface. This tissue degrades the ECM between the mesothelial cells, possibly via the persistent expression of MMP in refluxed endometrial tissue. MMP expression has been found in red lesions of endometriosis32 and endometriomas independent of the menstrual cycle phase. Type III procollagen—a byproduct of ECM metabolism—has been detected in the peritoneal fluid of patients with endometriosis.34
Angiogenic Factors
Angiogenic factors may play a role in the survival of the implants. These angiogenic factors are present in the peritoneal fluid of 58% of patients with endometriosis.35 Vascular endothelial growth factors (VEGF) can be produced by activated peritoneal macrophages, refluxed endometrial cells, and endometriotic lesions. VEGF36 is responsible for creating a new blood supply to the newly implanted endometrial cells. VEGF levels are elevated in the peritoneal fluid of women with endometriosis, and they correlate with the stage of the disease. Levels are higher around red lesions than around more mature black lesions. Other angiogenic factors such as interleukin 1 (IL-1) and IL-6 are also expressed by endometriotic tissue. Once scar tissue forms around the implants, neovascularization is limited.37
Abnormalities of the Immune System
Immune Cells
Natural killer cells are an important component of the immune system. Because they are cytotoxic, they have antitumor effects. In patients with endometriosis, this intraperitoneal cytotoxicity is decreased.38 T lymphocytes, which are responsible for cellular immunity, are increased in the peritoneal fluid of patients with endometriosis.39 Both helper and suppressor subtypes of T lymphocytes seem to be increased as well. Their role in the development of endometriosis remains unclear. Considerable evidence also suggests that endometriosis is associated with an increased incidence of autoantibodies of the immunoglobulin G (IgG) and IgA class. Autoantibodies to ovarian and endometrial cells have been reported in patients with endometriosis.40
Inflammatory Cells and Cytokines
A complex cascade of immune events results in an impaired peritoneal environment (Table 49-2). Macrophages, which are the most abundant nucleated cells in the peritoneal fluid, are increased in patients with endometriosis. The increase in number does not correlate with the stage of endometriosis.41 Increased activation of these macrophages is associated with increased secretion of cytokines but not phagocytic activity. The scavenging capacity of these macrophages appears to be impaired, resulting in decreased phagocytosis of endometrial cells.42 A protein similar to haptoglobin on endometriotic epithelial cells may decrease macrophage phagocytic capacity. Impaired clearance supports the pathogenetic hypothesis of implantation. The cytokines that are secreted play a role in the implantation of endometrial tissue as well as in the resulting infertility and pain. Macrophages and monocytes in peritoneal fluid of patients with endometriosis seem to increase the growth and maintenance of endometriotic implants.43
| Increased activated macrophages |
The observed increase in peritoneal fluid macrophages in patients with endometriosis can be linked to the increased expression of monocyte chemoattractant protein-1 (MCP-1).44 MCP-1 is produced by endometrial cells, mesothelial cells, macrophages, and endometriotic implants. The presence of endometrial tissue from patients with endometriosis in the peritoneal cavity increases expression of MCP-1 by the mesothelium.45 Oxidative stress is an activator of MCP-1 expression through nuclear factor κB (NFκB). MCP-1 is the most critical cytokine in the macrophage-related pathophysiology of endometriosis. The concentration of MCP-1 in the peritoneal fluid correlates well with the stage of the disease.
Cytokines and growth factors are found in higher concentrations in patients with endometriosis.46 Cytokines are small proteins that act as communicators between cells, especially those involved in inflammation and the immune system. These cytokines act in autocrine or paracrine loops to magnify their effect. They promote implantation, proliferation, and survival of the endometriotic tissue and are thought to be mediators in the clinical manifestation of the disease. Levels of several interleukins have been reported to be elevated in the peritoneal cavity of patients with endometriosis. They are secreted by the endometriotic lesions and recruited macrophages. Most interleukins are linked to stimulation of prostaglandins, fibroblast proliferation, and altered immune function. IL-1, which originates from activated macrophages and endometriosis implants, stimulates VEGF47 and IL-6. Although IL-6 has not consistently been shown to be increased in peritoneal fluid, the serum level has been shown to be more predictably elevated. IL-6 is produced by endothelial cells, fibroblasts, and monocytes as well as by eutopic and ectopic endometrium. It has a large number of physiologic reproductive functions, such as a promoter of endometrial proliferation and macrophage activator.
Interleukin-8 has also been found to be elevated in peritoneal fluid of patients with endometriosis. IL-8 can promote the implantation and growth of ectopic endometrium in the peritoneal cavity. Because it can also stimulate the adhesion of endometrial stromal cells to fibronectin, it can increase endometrial cell attachment to the mesothelium. IL-8 can also promote the invasion of endometrial cells by increasing the secretion of metalloproteinase,48 and it has been shown to stimulate endometrial cell proliferation in vitro and angiogenesis.
Another cytokine chemoattractant for monocytes called RANTES (Regulated on Activation, Normal T-cell Expressed and Secreted)49 is expressed in high concentrations in the peritoneal fluid of patients with endometriosis, and its level correlates with the stage of the disease. It seems to be produced in increased amounts by ectopic endometrial implants in response to chemokines. Tumor necrosis factor (TNF) is another cytokine that is produced in increased amounts and found in higher concentrations in the peritoneal fluid of patients with endometriosis.50 It is recognized as the most commonly elevated cytokine in peritoneal fluid.50 It causes an increase in the adherence of endometrial cells to mesothelial cells and may be involved in infertility because it adversely affects sperm motility and is toxic to embryos.36
The soluble form of intercellular adhesion molecule-1 (sICAM-1) is shed in higher amounts from the endometrium of patients with endometriosis, and this has been shown to correlate with the number of endometriotic implants in the pelvis.51 A prospective cohort study found that sICAM-1, when used in combination with CA-125, was a good marker for deep peritoneal endometriosis.52
Leptin is an adipocyte hormone that is produced by endometriotic tissues.53 Its peritoneal concentration is inversely correlated with the stage of endometriosis.54
Macrophage migration inhibitory factor is a pro-inflammatory cytokine produced by macrophages and endometriotic stromal and glandular cells. Its production is increased by TNF-α. Macrophage migration inhibitory factor was found recently to be significantly elevated in the peritoneal fluid of patients with endometriosis, but its levels did not correlate with the extent of disease or depth of penetration.55
Oxidative Stress in Peritoneal Endometriosis
Macrophages in the peritoneal fluid of patients with endometriosis produce large amounts of reactive oxygen species,56 which may lead to the expression of genes that allow cell adhesion.57 Oxidative stress occurs when levels of reactive oxygen species overwhelm the body’s antioxidant defense system. The concept of peritoneal cavity oxidative stress in patients with endometriosis is an extension of the chronic inflammatory nature of the disorder. Several studies have shown that antioxidant defenses may be altered in endometriosis, leading to an increased risk of oxidative stress. The peritoneal fluid from endometriosis patients may contain low levels of vitamin E (an antioxidant). Mifepristone (RU 486) has been showed to inhibit endometrial cell growth through its antioxidant effects.58 Oxidative stress may contribute in part to infertility through its effects on gametes or gamete interaction.36
Genetics of Endometriosis
Comparing the prevalence of the disease between monozygotic and dizygotic twins, it is estimated that 51% of the prevalence of the disease is attributed to genetic factors. The influence of specific environmental factors in a population of women make it difficult to separate out the genetic contribution to the disease. Therefore, the use of animal models such as the monkey may allow a more controlled assessment of such factors. In monkey colonies, there was evidence of familial aggregation63—the risk for full siblings was higher than that for half siblings (maternal or paternal). Population studies in Iceland64 from all identified cases of endometriosis showed a higher degree of common relatives than a control group of unaffected women, which confirms that the endometriosis group originated from a small number of founder women. A Norwegian study65 showed a greater familial association with women having severe endometriosis. Thus, the genetic influence may be more pronounced in relatives with severe disease. This was also demonstrated in a UK study.61 These findings are typical of diseases that are inherited as complex genetic traits, such as diabetes, asthma, and hypertension.
Researchers have suggested that many genes are altered with a higher frequency in patients with endometriosis, including the GALT gene and ICAM-1—an adhesion molecule encoding gene. Genes involved in ovarian steroidogenesis have also been implicated, including the gene encoding for the estrogen receptor and the gene encoding for the progesterone receptor. Aromatase enzyme gene expression seems to be increased within endometriosis lesions. This enzyme is usually not expressed in eutopic endometrium, and increased expression may lead to higher local concentrations. This might explain rare cases of postmenopausal women with active endometriotic lesions. Some of these rare cases have been reported to respond to the administration of aromatase inhibitors. A higher prevalence of the 4 G allele of the plasminogen activator inhibitor-1 (PAI-1) is found in patients with endometriosis.62 This results in hypofibrinolysis and persistence of the fibrin matrix that could support the initiation of endometriotic lesions.
Potential Alternative Hypothesis: Coelomic Metaplasia Theory
Endometriomas
The etiology of endometrioma formation is still the cause of many debates (Fig. 49-6). After analyzing serial sections of ovaries with endometriomas, Hughesdon66 suggested that 90% of endometriosis resulted from the adhesion of ovaries to the surrounding structures, which leads to progressive invagination of the ovarian cortex after repetitive bleeding from the endometrial implants on the surface of the ovaries. They would, in fact, constitute a pseudocyst. Some investigators also suggested that endometriomas might be the result of infiltration of functional cysts by endometriosis. Donnez and colleagues67 have suggested an alternative pathophysiology, which is based on coelomic metaplasia of invaginated epithelial inclusions.
Rectovaginal Endometriosis
Endometrial glands and stroma are often found in the retrocervical or rectovaginal space. The histology of these lesions differs from that of superficial endometriotic lesions.67 Smooth muscle proliferation and fibrosis are constantly present, are similar to adenomyotic nodules, and are responsible for the nodular aspect of these lesions often found on palpation of the rectovaginal wall. These lesions will typically lead to long-term pain and deep dyspareunia. They can be found in the uterosacral ligament, the posterior vaginal wall, or sometimes in the anterior rectal wall.68
