CHAPTER 33 Endometriosis

INTRODUCTION 488

DEFINITION 488

PREVALENCE 488

PATHOGENESIS 489

PERITONEAL FLUID ENVIRONMENT IN ENDOMETRIOSIS 490

PRESENTATION 491

ENDOMETRIOSIS AND INFERTILITY 493

DIAGNOSIS 494

CLASSIFICATION SYSTEMS 500

TREATMENT: GENERAL PRINCIPLES 500

MEDICAL TREATMENTS 501

SURGICAL TREATMENT 504

RECURRENT ENDOMETRIOSIS 505

KEY POINTS 506

Introduction

Endometriosis is one of the most common benign gynaecological conditions. It is second only to uterine fibroids as the most common reason for major surgical procedures in women under 45 years of age. It has been estimated that it is present in between 10% and 25% of women presenting with gynaecological symptoms in the UK and USA (Tyson 1974). These figures are based on the findings of patients who have undergone laparoscopy for diagnostic indications, such as pelvic pain or infertility, or in patients undergoing laparotomy. Although it is such a widespread condition, it is true to say that there is limited understanding regarding its aetiology and pathogenesis, and the condition still arouses much controversy with regard to its diagnosis, treatment and management. Endometriosis varies in severity from minimal disease with a few peritoneal implants to severe disease producing adhesions, deep infiltrating lesions and involvement with ovarian cyst formation.

Definition

Endometriosis is defined as a disease characterized by the presence of tissue that is morphologically and biologically similar to normal endometrium, and contains functional endometrial glands and stroma, in ectopic locations outside the uterine cavity. This ectopic endometrial tissue initially responds to hormones and drugs in a generally similar manner to eutopic endometrium undergoing cyclical changes. Cyclical bleeding from the endometriotic deposit appears to contribute to the induction of a local inflammatory reaction, fibrous adhesion and formation, and, in the case of deep ovarian implants in the ovary, leads to the formation of an endometrioma or chocolate cyst. Endometrial tissue deep within the myometrium of the uterine wall is termed ‘adenomyosis’. This condition is increasingly viewed as a separate pathological entity to that of endometriosis, since it affects a different population of women, probably has a different aetiology and is most often found in the absence of other evidence of pelvic endometriosis.

Prevalence

Continued growth of endometriotic tissue, as with that of the endometrium, is dependent on ovarian steroid hormones, particularly oestrogen. Thus, endometriosis is prevalent in the reproductive years with a peak incidence between 30 and 45 years of age, although it is increasingly being diagnosed in much younger women as the threshold for investigation of gynaecological symptoms utilizing diagnostic laparoscopy has altered. Endometriosis is thus primarily a disease of the reproductive years and is only rarely described in adolescence, when it is associated with obstructing genital tract abnormalities, or in postmenopausal women, where it has been reactivated because of hormone replacement therapy.

No racial differences in the incidence of the disease have been found, except for Japanese women who have been reported to have twice the incidence of Caucasian women (Miyazawa 1976).

The exact prevalence of endometriosis is unknown since precise diagnosis depends on observation of implants, predominantly at the time of laparoscopy or laparotomy. Until simple non-invasive screening tests are developed, the true prevalence will remain unknown. Current prevalence therefore depends upon identification in women who are either symptomatic or undergoing various operative procedures. The incidence is markedly variable as the data in Table 33.1 show, but the prevalence of endometriosis in the reproductive years is estimated to be approximately 10% (Eskenazi and Warner 1997). Endometriosis commonly affects women during their childbearing years. In the main, this is reflected in deleterious sexual, reproductive and social consequences as a result of its associated painful symptoms and often associated infertility. Symptomatology may extend over several decades of a patient’s life because of its often late diagnosis and the recurrent nature of the disease. For individual patients and healthcare systems, it represents a major call upon resource use.

Table 33.1 Prevalence of endometriosis through various presentations

Presentation	Prevalence (%)
Unexplained infertility	70–80*
Infertile women (all causes)	15–20*
At diagnostic laparoscopy	0–53^†
At treatment laparoscopy	0.1–50^†
Women undergoing sterilization	2^‡
In women with diagnosed first-degree relatives	7^§

* Source: Kistner RW 1977 In: Sciarra J (ed) Gynaecology and Obstetrics, Vol. 1. Harper and Row, London.

^† Source: Houston DE 1984 Epidemiology Reviews 6: 167–191.

^‡ Source: Strathy JH, Molgaard GA, Coulam CB, Molton LJ III 1982 Fertility and Sterility 38: 667–672.

^§ Source: Simpson JL, Elias S, Malinak LR, Buttram VC Jr 1980 American Journal of Obstetrics and Gynecology 137: 327–331.

Pathogenesis

The precise aetiology of endometriosis still remains unknown. Indeed, it is often called the ‘disease of theories’ because of the many mechanisms postulated to explain its pathogenesis. It is likely that no single theory can explain all forms of endometriosis. Different types of endometriosis may have different origins (i.e. they are multifactorial in origin). Peritoneal endometriosis could arise as a result of retrograde menstruation, ovarian endometriosis via coelomic metaplasia, and rectovaginal endometriosis from development within Müllerian duct remnants. The major theories of causation of endometriosis are metaplasia (transformation) of coelomic epithelium, and implantation of endometrial fragments through retrograde menstruation.

Transformation of coelomic epithelium

This theory, first described by Meyer (1919), postulated the possibility of differentiation by metaplasia towards an endometrial-like tissue of the original coelomic membrane following prolonged irritation and oestrogen stimulation. It is proposed that these adult cells undergo dedifferentiation back to their primitive origin and then transform to endometrial cells. If this theory is correct, metaplasia should occur wherever coelomic membranes are present. This theory has many attractions which could explain the occurrence of endometriosis in nearly all the ectopic sites in the presence of aberrant Müllerian cells. What induces this transformation — whether it is hormonal stimuli, inflammatory irritation or other processes — is uncertain. If coelomic metaplasia is similar to metaplasia elsewhere, the frequency of the disorder should increase with advancing age. The clinical pattern of endometriosis is distinctly different from this, with an abrupt halt in the disease with the cessation of menses at the menopause and reduced oestrogen production, thus raising some questions over this theory.

Menstrual regurgitation and implantation (metastatic theory)

As early as 1927, Sampson proposed the metastatic theory, postulating that retrograde menstrual flow transported desquamated endometrial fragments through the fallopian tubes into the peritoneal cavity (Sampson 1927). Once there, the still viable cells subsequently implanted and began growth and invasion. In support of this theory, experimental endometriosis has been induced in animals with replacement of menstrual fluid or endometrial tissue in the peritoneal cavity. Supporting this theory in humans is the finding that endometriosis is commonly found in young girls with associated abnormalities in the genital tract causing obstruction to the outflow of menstrual fluid (Schifrin et al 1973). Halme et al (1984) observed bloody fluid at the time of menstruation in the pelvis during laparoscopic assessment, but as this finding occurs in up to 90% of all women, it is regarded as a physiological phenomenon. The high incidence of retrograde menstruation suggests that this phenomenon alone does not give rise to endometriosis, but that some other factor(s) must be involved in development of the disease. These factors could include some alteration in the (uterine) endometrium, altered immune response to retrograde menstruation (hence failure to clear the peritoneal cavity of debris efficiently), or a more favourable peritoneal environment which may stimulate the growth and implantation of ectopic endometrium within the peritoneal cavity itself.

Genetic and immunological factors

Many studies have indicated that there may be a genetic factor related to endometriosis since the disease is more prevalent in certain families. It has been shown that women with an affected first-degree relative have a seven times higher risk of developing endometriosis, which may be severe (Simpson et al 1980, Halme et al 1986). Endometriosis is also more common in monozygotic twin sisters than dizygotic twins, but no association was found with specifically identified tissue types (Simpson et al 1984). Whilst Dmowski et al (1981) demonstrated a decreased cellular immunity to endometriotic tissue in women with endometriosis, no clinically significant immune system abnormality has been observed in women with the disease; hence, the precise genetic or immune components increasing an individual’s potential to develop this disorder are yet to be defined.

Vascular and lymphatic metastasis (Halban’s theory)

Halban’s theory suggests that distant endometriosis (outwith the abdominal cavity) occurs via vascular or lymphatic spread of viable endometrial cells or fragments. The theory would explain the rare occurrence of endometriosis in lung, limbs and brain, but not in more common sites within the peritoneal cavity and pelvis.

Endometrial disease theory

Many view the superficial implants as ‘physiological’ lesions which may regress spontaneously. Deep infiltrating lesions and ovarian endometriotic cysts are pathological and arise from cells that have undergone somatic mutations. Such mutations may have been produced by environmental factors (e.g. pollutants, dioxins). These abnormal cells then develop into a ‘benign tumour’ consisting of endometriotic glands and stroma. For a review on environmental factors in the aetiology of endometriosis, the reader is directed to Missmer and Mohllagee (2008).

Further support of the above theory stems from biochemical differences demonstrated in the endometrium in patients with and without endometriosis (Guidice and Kao 2004). These include differences in the expression of metalloproteinases, tumour susceptibility genes and angiogenic factors. Such changes may allow certain types of retrograde endometrial fragments to implant more readily than others, and this may be due to an underlying inherited genetic tendency.

Iatrogenic dissemination

Transplantation of endometrial cells can occur during gynaecological surgical procedures. This would explain the development of endometriotic nodules in abdominal wall scars following caesarean section, myomectomy and hysterectomy, and in laparoscopic port scars. In addition, implantation at perineal repair can cause endometriosis to develop in episiotomy scars.

Conclusion

The conclusion reached from the above theories is that pelvic endometriosis is probably a consequence of transplantation of viable endometrial cells regurgitated at the time of menstruation from the fallopian tubes into the peritoneal cavity. In addition, transport of endometrial cells may occur by other routes (some iatrogenic). It is unclear whether endometriotic implants are derived from in-situ pluripotential cells generated by metastatic seeding, but it is known that endocrine and immunological factors allow growth and spread within the pelvis and neighbouring organs. Delayed childbearing, either by choice or infertility, has been implicated as a risk factor for the development of endometriosis. The risk of developing endometriosis also corresponds with cumulative menstruation, menstrual frequency and volume. Women with shorter menstrual cycles of less than 27 days and longer flows (more than 7 days) are twice as likely to develop the disease compared with women with longer cycles and shorter flows. Thus, many components are necessary to allow endometriotic deposits to implant and subsequently grow (Figure 33.1).

Figure 33.1 Suggested aetiological factors in the pathogenesis of endometriotic implants.

The cell-mediated efferent arm of the immune system controls implantation or rejection of such fragments. In most women, these fragments are not in a favourable state or in a favourable environment for implantation, and are engulfed by macrophages and thus disposed. In women with deficient or altered cell-mediated immunity, implantation of the ectopic fragments allows implantation to occur with subsequent development of endometriotic implants. Such immune changes could be transmitted genetically and could be both qualitative and quantitative, resulting in variable ages of onset, extent and duration of expression. Autoantibody production would be a secondary phenomenon, developing as the number and extent of endometriotic deposits increase.

Peritoneal Fluid Environment in Endometriosis

It has been suggested that peritoneal fluid volume and its contents may be adversely affected by the presence of endometriotic tissue in the pelvis, with possible consequent interference with tubo-ovarian function and/or fertilization and early implantation. Peritoneal fluid is an ultrafiltrate of plasma, and the volumes are very low in normal women. Endometriosis may alter the peritoneal fluid volume by increasing fluid production in the ovary, altering mesothelial permeability or increasing the hydrostatic pressure as a result of altered protein content.

Peritoneal macrophages

Normal peritoneal fluid contains approximately 10⁶ cells per ml; 90% are macrophages, and the remainder are lymphocytes and desquamated mesothelial cells. The role of peritoneal macrophages in women with endometriosis has been the source of much interest, and women with endometriosis associated with infertility have been reported to have significantly higher concentrations of macrophages in peritoneal fluid than either fertile women or infertile women without endometriosis. Macrophages in patients with endometriosis appeared to be highly phagocytic against spermatozoa in vitro compared with those from fertile women or infertile women without endometriosis (Muscato et al 1982). In addition, they are also able to survive better in vitro than those from fertile controls (Halme et al 1986). Peritoneal fluid from patients with endometriosis has also been shown to have a cytotoxic effect on in-vivo cleavage of mouse embryos. These findings on the quantitative and qualitative properties of macrophages in peritoneal fluid may partially explain the mechanisms of infertility in patients with endometriosis.

Prostaglandins and prostanoids

The role of prostaglandins and their metabolites in peritoneal fluid in the pathogenesis and symptomatology of endometriosis is controversial. It has been reported that increased levels of prostaglandin F_2α (PGF_2α) are found in the peritoneal fluid of patients with the disease (Meldrum et al 1977). In addition, increased peritoneal fluid volume and increased concentrations of the prostaglandin metabolites thromboxane B₂ and 6-keto PGF_2α have been noted. Other investigators have found no increase in either the volume of peritoneal fluid or its concentrations of prostaglandins or metabolites (Rock et al 1982). These conflicting reports may reflect the timing of peritoneal fluid sampling and difficulties in assay measurement of the small quantities of substrates, all of which have very short half-lives.

In recent years, it has been appreciated that more subtle forms of endometriosis may be present with only minimal evidence of visual changes in the peritoneum. However, if there are changes in the prostaglandin content of the peritoneal fluid, the mechanisms by which these changes influence endometriosis or its association with infertility remain unclear.

Cytokines and growth factors

A large number of cells, including macrophages, lymphocytes, fibroblasts and epithelial cells, synthesize a wide range of polypeptides of high biological activity within the cytokine/growth factor group. They have multiple effects including the regulation of differentiation, growth and function of a wide variety of cell types. It was generally believed that cytokines were produced in increased amounts as part of the inflammatory process or as part of a response to infection or immune challenge. Cytokines are very potent and active at picogram to nanogram levels, and are difficult to detect and identify. Research activity has been concentrated on the measurement of various cytokines including various interleukins, colony stimulating factors, human necrosis factor and transforming growth factors. To date, no consistent changes between normal, infertile women with and without endometriosis and patients with endometriosis who are not infertile have been reported to satisfactorily explain the aetiology or pathogenesis of the disease.

Peritoneal environment: conclusions

Peritoneal macrophages play an important role in the peritoneal cavity where endometrial tissue fragments often arrive and are likely to adhere to the mesothelial lining. The adherence may be mediated by cell adhesion molecules and other factors produced by activated macrophages. Endometrial tissue will potentially adhere if the regurgitated amount of the tissue is too great, or if the capacity of the intra-abdominal cells to clear the cellular debris is in any way impaired. Once an endometrial fragment has gained adherence, further tissue growth may be promoted by steroids, cytokines, growth factors and angiogenic factors present in peritoneal fluid in a paracrine and autocrine fashion.

Presentation

Endometriosis commonly presents between the ages of 25 and 35 years, although it can present in early adolescence and in postmenopausal women on hormone replacement therapy. The symptoms of endometriosis are variable and often unrelated to the extent of the disease process as currently quantified. The three most common complaints amongst women with endometriosis are dysmenorrhoea, dyspareunia and pelvic pain. The pain symptoms are often cycle related and increase premenstrually. However, it must be stated that the finding of endometriosis may not conclusively link it with painful symptoms in an individual, since the severity of symptoms is rarely correlated with the extent of the disease, and endometriosis is often found coincidentally (during surgery or investigation for other gynaecological conditions, such as infertility) at similar levels in patients not complaining of pain.

Atypical bleeding patterns are a leading symptom in a variety of gynaecological diseases, but may also characterize patients with endometriosis. Premenstrual spotting and menometrorrhagia are frequently noted. On the other hand, cyclical rectal bleeding or haematuria is pathognomonic of the disease and, although rarely observed (1–2% of cases), these symptoms give strong evidence for bowel or bladder involvement. Painful micturition or defaecation at the time of menstruation may be the first signs of progressing disease. The various symptoms of endometriosis as found in various sites of implantation are shown in Table 33.2.

Table 33.2 Symptoms of endometriosis related to sites of implants

Symptoms	Site
Dysmenorrhoea	Reproductive organs
Lower abdominal pain
Pelvic pain
Low back pain
Menstrual irregularity
Rupture/torsion endometrioma
Infertility
Cyclical rectal bleeding	Gastrointestinal tract
Tenesmus
Diarrhoea/cyclic constipation
Cyclical haematuria	Urinary tract
Dysuria (cyclical)
Ureteric obstruction
Cyclical haemoptysis	Lungs
Cyclical pain and bleeding	Surgical scars/umbilicus
Cyclical pain and swelling	Limbs

Dysmenorrhoea

Dysmenorrhoea is the most common presenting symptom, affecting 70–80% of women with proven endometriosis. Dysmenorrhoea often occurs with pelvic pain, but patients commonly describe a background of constant dragging, aching pain which may be exacerbated by the menses but is often a different type of pain from the typical cramping nature of spasmodic dysmenorrhoea and often in a different site. Typically, pain starts some days before menstruation as an ache or discomfort, similar to congestive pelvic pain. It worsens and becomes spasmodic with the start of the menses and continues throughout days of bleeding, lessening as menstrual flow ebbs. Unlike ‘primary dysmenorrhoea’ which starts soon after the establishment of ovulatory cycles, endometriosis-related dysmenorrhoea often starts in the early 20s, and is usually very severe, incapacitating and unresponsive to simple/mild analgesics or antiprostaglandins. If endometriosis is untreated, dysmenorrhoea progressively increases in duration, but severity tends to remain constant.

Dyspareunia

Another common symptom of endometriosis is deep dyspareunia resulting from stretching at intercourse of the involved pelvic tissues such as a fixed retroverted uterus, the uterosacral ligaments or rectovaginal septum; or pressure on an involved enlarged, often adherent, ovary. The presence of endometriotic tissue within these areas, however, is not always associated with dyspareunia; perhaps less than half of patients who are coitally active admit to this symptom when deposits are found in these areas. The pain continues for a variable time after intercourse, often as a dull ache. Typically, dyspareunia is exaggerated in the phase before menstruation commences. This pain is more severe in women with deep rectovaginal septum involvement and may result in complete apareunia.

Origin of pain

The basis of the pelvic pain and dysmenorrhoea is uncertain but could reflect stretching of tissues by the menstrual process and an effect of the local production of prostaglandins within the endometriotic implants. The pain also relates to tissue damage and fixity of organs from scar and adhesion formation. What is immediately clear when reviewing patients with endometriosis is the huge variation in extent of symptomatic disease, which does not correlate with the extent of the observed disease process. In addition, many patients who are asymptomatic are found to have endometriosis, and in some of these patients, severe disease is discovered following laparoscopy during infertility investigations. One possible explanation for this is that in these individuals, the disease may have disrupted the pelvic sensation altogether. Pain may arise from mediator release from the endometriotic deposits (prostaglandins, inflammatory products from macrophages or interleukins) in the early stages of the disease at menstruation. Pain associated with more advanced disease is likely caused by extensive adhesions, ovarian endometriotic cysts or deep infiltration of disease. The pain may result from mechanical compression or disruption of nociceptors, particularly around the uterosacral ligaments.

Non-cyclic chronic pelvic pain

In approximately 30–50% of women with pelvic endometriosis, chronic pelvic pain (CPP) of varying severity, pattern and location is reported. Pelvic pain is arguably the symptom causing the most misery amongst endometriosis sufferers, and is more distressing than infertility. CPP, as with other forms of pain, often produces a serious long-term detrimental effect on the personality, and affects working ability and social and marital life. Possible causes include adhesions, ovarian cysts, peritoneal inflammatory reaction around implants, or involvement of bladder or bowel. Pain may be triggered by ovulation, bowel evacuation or micturition. Pain resulting from adhesions may be provoked or worsened by certain body positions or movements. A common feature of CPP is worsening before and during menstruation.

Menstrual irregularities

In 16% of women, the disease is associated with heavy and/or irregular periods or premenstrual spotting, the causes of which are undetermined.

Other symptoms associated with the menstrual cycle

Direct involvement of the bladder wall can result in haematuria and dysuria occurring at the time of menstruation. Invasion into the muscular coat of the descending colon and rectum (most commonly) can cause cyclical rectal bleeding and painful defaecation (dyschezia), whilst involvement of the small bowel or colon elsewhere may not present until a narrowing of the bowel lumen has occurred with complete or partial obstruction. Rarely, dysmenorrhoea may be associated with symptoms due to extrapelvic endometriosis (e.g. sciatica, groin pain), or cyclical haemoptysis or pneumoperitoneum with lung/pleural involvement. Other rare associations include bleeding, swelling and/or pain in surgical scars, umbilicus or episiotomy site affected by endometriosis.

Correlation of symptoms and severity of endometriosis

The frequencies of the more common symptoms in endometriosis patients are summarized in Table 33.3. Whilst the symptoms of dysmenorrhoea, dyspareunia and pelvic pain can occur with other gynaecological disorders, it is the combination, cyclical and menstrually related component of several symptoms which should alert the clinician to the potential underlying presence of endometriosis. Many women have delayed diagnosis of their condition, which may mean that it has progressed to a more extensive and potentially less reversible or curable stage at the time of diagnosis.

Table 33.3 Frequency of the more common symptoms in endometriosis patients

Symptom	Likely frequency (%)
Dysmenorrhoea	60–80
Pelvic pain	30–50
Infertility	30–40
Dyspareunia	25–40
Menstrual irregularities	10–20
Cyclical dysuria/haematuria	1–2
Dyschezia	1–2
Rectal bleeding (cyclic)	<1

There appears to be little correlation between sites involved in endometriosis and symptoms. Various ‘types’ of symptoms can, however, be related to some degree to the system involved (see Table 33.2). However, these do not always correlate with the anatomy and type of pain innervation of the pelvis (for review, see MacLaverty and Shaw 1995).

One reason for the apparent lack of correlation between disease severity and symptom severity is that the classification systems for endometriosis thus developed have primarily been directed towards infertility prediction rather than pain symptom severity.

Deeply infiltrating endometriosis is very strongly associated with the presence and severity of pelvic pain. In addition, superficial non-pigmented endometriosis has the capacity to produce more prostaglandin F (PGF) than pigmented classic powder burn lesions. PGF is implicated in pain causation (Vernon et al 1986). Thus, in the early, less florid stages before it has become destructive and more easily recognized, endometriosis may be producing large quantities of PGF and hence possesses greater potential to increase the severity of pain. The type of pain may alter with disease progression, with constant pain and exacerbation at the menses initially in the disease, and pain later becoming continuous due to scar formation and organ fixity.

Endometriosis and Infertility

It is accepted that endometriosis resulting in structural damage to the tubes and ovaries causes infertility. However, what is less clear is whether the milder forms of endometriosis are also the cause of infertility in otherwise asymptomatic patients.

Endometriosis was one of the most frequently made diagnoses in couples undergoing infertility investigation, when routine use of laparoscopy for investigation of such couples was employed, in past years. The assumption was made that endometriotic implants are responsible for the patient’s inability to conceive. Estimates of the incidence of endometriosis in the general population of reproductive age vary between 2% and 10%. From retrospective studies in infertile patients, the incidence has been reported as being between 20% and 40% (Mahmood and Templeton 1990). This increased incidence in infertile patients has led many clinicians to consider the endometriotic implants to be responsible, in some way, for the associated infertility. The question is how, and a number of suggested mechanisms have been reported. These are summarized in Table 33.4. For the majority of these potential causes, there are few or no consistent data to provide a sustainable explanation. Thus, the nature of the relationship between mild endometriosis and infertility remains unresolved.

Table 33.4 Possible mechanisms of causation of infertility with mild endometriosis

Problem area	Mechanism
Ovarian function	Endocrinopathies • Anovulation • Luteinized unruptured follicle syndrome • Altered prolactin release • Altered gonadotrophin midcycle surge
	Luteolysis caused by prostaglandin F_2α
	Oocyte maturation defects
Coital function	Dyspareunia causing reduced penetration and coital frequency
Tubal function	Alterations in tubal and cilial motility by prostaglandins
Tubal function	Impaired fimbrial oocyte pick-up
Sperm function	Phagocytosis by macrophages
Sperm function	Inactivation by antibodies
Endometrium	Interference by endometrial antibodies
Endometrium	Luteal-phase deficiency
Early pregnancy failure	Increased early abortion
Early pregnancy failure	Prostaglandin induced or immune reaction