Chapter 56. Endometriosis

• Endometriosis is a disorder in which abnormal growths of tissue, histologically resembling the endometrium, are present in locations other than the uterine lining.
• Although endometriosis can occur very rarely in postmenopausal women, it is found almost exclusively in women of reproductive age.
• All other manifestations of endometriosis exhibit a wide spectrum of expression.
• The lesions are usually found on the peritoneal surfaces of the reproductive organs and adjacent structures of the pelvis, but they can occur anywhere in the body (Fig. 56–1).
• The size of the individual lesions varies from microscopic to large invasive masses that erode into underlying organs and cause extensive adhesion formation.
• Similarly, women with endometriosis can be completely asymptomatic or may be crippled by pelvic pain and infertility.

Endometriosis is a common and important health problem of women. Its exact prevalence is unknown because surgery is required for its diagnosis, but it is estimated to be present in 6–10% of women in the reproductive age group and 25–35% of infertile women. It is seen in 1–2% of women undergoing sterilization or sterilization reversal, in 10% of hysterectomy surgeries, in 16–31% of laparoscopies, and in 53% of adolescents with pelvic pain severe enough to warrant surgical evaluation. Endometriosis is the commonest single gynecologic diagnosis responsible for hospitalization of women aged 15–44, being found in more than 6% of patients.

The cause of endometriosis is complex, and the leading theories include retrograde menstruation with transport of endometrial cells, metaplasia of coelomic epithelium, hematogenous or lymphatic spread, and direct transplantation of endometrial cells. A combination of these theories is likely to be responsible.

A theory of retrograde menstruation was proposed during the 1920s. It was postulated that endometriosis occurred because viable fragments of endometrium were shed at the time of menstruation and passed through the fallopian tubes. Once in the pelvic cavity, the tissue became implanted on peritoneal surfaces and grew into endometriotic lesions. Subsequent observations have confirmed that some degree of retrograde menstruation normally occurs in women with patent tubes, that outflow tract obstructions (cervical stenosis, transverse vaginal septa) increase the incidence of endometriosis, and that intentional deposition of endometrium onto peritoneum can initiate endometriosis. Also, the risk of developing the disease is higher in women with prolonged menstrual flow and in those with short menstrual cycle lengths (more menses per year). This theory is simple, attractive, and easily explains why endometriosis is most commonly found on the peritoneal surfaces of the ovaries, cul-de-sac, and bladder and why lesions may develop in episiotomies and other incisions. However, it does not explain why all women do not develop endometriosis, nor does it explain the rare cases of endometriosis in the lung, brain, or other soft tissues or in nonmenstruating subjects (women with Turner's syndrome or with absent uteri).

There is evidence that altered humoral and cell-mediated immunity plays a role in the pathogenesis of endometriosis. The activity of natural killer cells may be reduced, and deficient cellular immunity may cause an inability to recognize endometrial tissue in abnormal locations. Endometriosis may occur when the deficiency in cellular immunity allows menstrual tissue to implant and grow on the peritoneum.

Genetic influences in the development of endometriosis have also been described. Studies have found that 7–9% of endometriosis patients' first-degree female relatives are diagnosed with the disease—significantly greater than the control rate of 1–2%. Further investigation has revealed a possible role for the HLA-B7 allele. The expression of HLA-B7 has been shown to inhibit the cytotoxic activity of natural killer-like T lymphocytes, suggesting that the growth of ectopic endometrial cells might be under genetic control.

The gross appearance of endometriosis at operation is usually quite characteristic and, to an experienced surgeon, is sufficient for diagnosis. The smallest (and presumably earliest) implants are red, petechial lesions on the peritoneal surface. With further growth, menstrual-like detritus accumulates within the lesion, giving it a cystic, dark brown, dark blue, or black appearance. The surrounding peritoneal surface becomes thickened and scarred. These "powder burn" implants typically attain a size of 5–10 mm in diameter. With progression of disease, the number and size of lesions increase, and extensive adhesions may develop. When present on the ovary, cysts may enlarge to several centimeters in size and are called endometriomas or "chocolate cysts." Severe disease can erode into underlying tissues and distort the remaining organs with extensive adhesions. In addition to these traditional presentations, endometriosis lesions can have a variety of nonclassical appearances: clear vesicles, white or yellow spots or nodules, circular folds of peritoneum ("pockets"), and visually normal peritoneum (lesions so small they can only be detected microscopically).

The distribution of lesions also exhibits a characteristic pattern. Solitary lesions are possible, but multiple implantations are the rule. The most common site of disease is the ovary (approximately half of all cases), followed by the uterine cul-de-sac, posterior broad ligament, uterosacral ligaments, uterus, fallopian tubes, sigmoid colon, appendix, and round ligaments. Implants may occur over the bowel, bladder, and ureters; rarely, they may erode into underlying tissue and cause blood in the stool or urine, or their associated adhesions may result in stricture and obstruction of these organs. Implants can occur deep in tissue, especially on the cervix, posterior vaginal fornix, or within wounds contaminated by endometrial tissue. Very rarely, endometriosis is found distant from the pelvis, in such sites as the lung, brain, and kidney. Pleural implantations are associated with recurrent right pneumothoraces at the time of menses, termed catamenial pneumothorax. Similarly, lesions in the central nervous system can cause catamenial seizures.

The microscopic finding that these lesions are composed of tissue histologically resembling endometrial glands and stroma gives endometriosis its name (Fig. 56–2). The normal endometrial appearance is best seen in small, early lesions; with advanced disease, cyst formation, and fibrosis, the wall of the implant is lined by a monolayer of cells, if at all. Blood is present inside the cyst, and hemosiderin-laden macrophages are found in the cyst wall.

Overproduction of prostaglandins by an increase in cyclooxygenase-2 activity as well as overproduction of estrogen by increased aromatase activity are considered key factors in the development of endometriosis. Progesterone resistance is also seen, which weakens the antiestrogenic effect of progesterone. This can lead to a chronic inflammatory response, and the most commonly found inflammatory cytokines are interleukin 1,6, and 8, and tumor necrosis factor α.

It is generally agreed that pelvic pain occurs premenstrually in endometriosis patients. Because of this, pain from endometriosis is thought to be due to stimulation from estrogen and progesterone during the menstrual cycle; the tissue of the implant is stimulated to grow in much the same way as is the endometrium. The implants enlarge and may undergo secretory change and bleeding; however, the fibrotic tissues surrounding the implants prevent the expansion and escape of hemorrhagic fluid that occurs in the uterus. With subsequent cycles, this process repeats itself. Pain is produced by pressure and inflammation within and around the lesion, by traction on adhesions associated with the lesions, by the number of implants and their proximity to nerves and other sensitive structures, and by the mass effect of large lesions. Although this sequence of events explains why premenstrual pelvic pain can occur in endometriosis, it is incomplete, because many patients with extensive endometriosis have no pain. It is a common observation that the occurrence and severity of pain from endometriosis bear little relationship to the amount and distribution of the disease. Severe pain in patients with endometriosis is associated with deeply infiltrating lesions, and it is thought that the degree of pain is perhaps determined by the depth of invasion.

The relationship between endometriosis and infertility has been more extensively investigated. Moderate and severe endometriosis is associated with pelvic adhesions that distort pelvic anatomy, prevent normal tubo-ovarian apposition, and encase the ovary. Implants can destroy ovarian and tubal tissue, although occlusion of fallopian tubes is rare.

It is not difficult to understand how advanced disease can result in infertility, but minimal or mild endometriosis, in which pelvic anatomy is entirely normal except for a few peritoneal surface lesions, can also cause infertility. The mechanism by which this occurs is unknown. Various theories have been proposed to explain this phenomenon.

Several investigators have examined peritoneal fluid abnormalities. The peritoneal fluid is an ultrafiltrate of plasma, with <5 mL normally present in the pelvis. After ovulation, a transient rise to approximately 20 mL occurs. The volume of peritoneal fluid and the concentrations of various hormones and other substances in it affect the processes of ovulation, ovum pickup, tubal function, and sperm function.

A marker for ovarian reserve, antimüllerian hormone, has also been found to be decreased in early-stage endometriosis.

Risk factors for endometriosis include family history, early menarche, long duration of menstrual flow, heavy bleeding during menses, and shorter cycles. Regular exercise of >4 hr/wk, higher parity, and longer duration of lactation were all associated with a decreased risk for endometriosis.

Prevention of endometriosis is not currently possible. Traditionally, women with relatives affected by endometriosis—or in whom the diagnosis has recently been made—are advised not to postpone childbearing. The merits of this advice have not been proved. A more thorough understanding of the pathophysiology of endometriosis is required before preventive strategies can be devised.

Endometriosis is common among women of reproductive age, and its prevalence increases to 30–40% among infertile women. Clinical findings vary greatly depending on the number, size, and extent of the lesions and on the patient population being examined.

The diagnosis of endometriosis is often strongly suspected from a patient's initial history. Infertility, dysmenorrhea, and dyspareunia are the main presenting complaints. Most patients complain of constant pelvic pain or a low sacral backache that occurs premenstrually and subsides after menses begins. Dyspareunia is often present, particularly with deep penetration. Lesions involving the urinary tract or bowel may result in bloody urine or stool in the perimenstrual interval. Implantations on or near the external surfaces of the cervix, vagina, vulva, rectum, or urethra may cause pain or bleeding with defecation, urination, or intercourse at any time in the menstrual cycle. Adhesions from endometriosis may cause discomfort at any time during the cycle, and a sensation of pelvic pressure may result if large masses are present. Premenstrual spotting may occur and is more likely to be associated with endometriosis than with luteal phase inadequacy. It must be emphasized, however, that many patients either have no symptoms or have infertility as their only symptom and that the extent of disease often has little correlation with the severity of symptoms.

The physical examination may also be helpful in discerning whether endometriosis is present. Classically, pelvic examination reveals tender nodules in the posterior vaginal fornix and pain upon uterine motion. The uterus may be fixed and retroverted due to cul-de-sac adhesions, and tender adnexal masses may be felt because of the presence of endometriomas. Careful inspection may reveal implants in healed wounds, especially episiotomy and caesarean section incisions, in the vaginal fornix, or on the cervix. Biopsy may be required to prove that the lesions are due to endometriosis. However, many patients have no abnormal findings on physical examination.

For the vast majority of patients, endometriosis is included in the differential diagnosis of infertility or pelvic pain. Endometriosis should be suspected in any patient of reproductive age complaining of pain or infertility. Medical treatment can be given for pelvic pain thought to be due to endometriosis, but the specific diagnosis of endometriosis should not be made unless documented by direct visualization. The final diagnosis of endometriosis can only be made at laparoscopy or laparotomy, by direct observation of the implants. Occasionally, an isolated endometrioma is removed, and the diagnosis must be made histologically by the demonstration of "endometrial" glands and stroma or of hemosiderin-laden macrophages in the cyst wall.

Except for special circumstances, such as urography or sigmoidoscopy for suspected bowel or urinary involvement, ancillary diagnostic studies (ultrasound, x-rays, computed tomography scans) are of little help in diagnosis. CA-125 is often elevated in women with endometriosis; however, it has been shown that this marker is elevated in many other pelvic diseases and therefore has little specificity in the diagnosis of endometriosis. However, an elevated CA-125 that returns to normal levels after medical or surgical treatment can be helpful in the evaluation for recurrences.

The varied presentations of endometriosis mandate that it be considered in the differential diagnosis of virtually all pelvic disease. In particular, the pain, infertility, and adhesions associated with endometriosis must be distinguished from similar symptoms accompanying pelvic inflammatory disease and pelvic tumors. Usually this will require operative evaluation. A patient with a persistent adnexal mass >5 cm should never be presumed to have an endometrioma even if endometriosis has been diagnosed previously. Such masses require surgical diagnosis.

True complications of endometriosis are few. Implants over the bowel or ureters may cause obstruction and silent impairment of renal function. The erosive nature of the lesions in advanced aggressive disease can cause a myriad of symptoms, depending on the tissue damaged. Endometriomas can cause ovarian torsion or can rupture and spill their irritating contents into the peritoneal cavity, resulting in a chemical peritonitis. Excision of endometriosis causing catamenial seizures or pneumothorax may be necessary.

Several classification schemes to assist in describing the anatomic location and severity of endometriosis at operation have been created. Although none is entirely satisfactory, the scoring systems are useful for reporting operative findings and for comparing the results of various treatment protocols. The revised American Fertility Society classification is the most commonly used system and is given in Table 56–1 and Figure 56–3. It should be noted that this system does not correlate well with the symptoms of pain, dyspareunia, or infertility, but is mainly designed for uniform recording of operative findings.

Treatment options are dictated by the patient's desire for future fertility, her symptoms, the stage of her disease, and to some extent her age. It must be emphasized that therapy for endometriosis requires operative inspection of the lesions for correct diagnosis and staging and to be sure that the patient's symptoms are attributable to endometriosis only.

Expectant Management

In asymptomatic patients, those with mild discomfort, or infertile women with minimal or mild endometriosis, expectant management may be appropriate. Although endometriosis is generally felt to be a progressive disease, there is no evidence that treating an asymptomatic patient will prevent or ameliorate the onset of symptoms later. Many reports have found expectant management of infertile women with minimal or mild endometriosis to be as successful as medical or surgical therapies.

Analgesic Therapy

Analgesic treatments include nonsteroidal anti-inflammatory agents and prostaglandin synthetase-inhibiting drugs. These drugs are appropriate sole therapy for endometriosis when the patient has mild premenstrual pain from minimal endometriosis, no abnormalities on pelvic examination, and no desire for immediate fertility.

Hormonal Therapy

The goal of treatment with hormonal therapy is to interrupt the cycles of stimulation and bleeding of endometriotic tissue. This can be achieved with various agents.

Oral Contraceptive Pills (Ocps)

OCPs are a good choice for patients with minimal or mild symptoms. Generally monophasic products are used, which are prescribed either cyclically or continuously for 6–12 months. The continuous exposure to combination oral contraceptive pills results in decidual changes in the endometrial glands. Continuous use of OCPs has been shown to be effective in decreasing dysmenorrhea and may also retard progression of endometriosis.

Progestins

These agents work via a mechanism similar to that of the OCPs, causing decidualization in the endometriotic tissue. Oral medroxyprogesterone acetate can be prescribed as a 10–30-mg dosage daily. An alternative regimen is norethindrone acetate 5 mg daily or megestrol acetate prescribed as a 40-mg daily dose. Depot medroxyprogesterone acetate 150 mg administered intramuscularly can also be given as a single injection every 3 months.

The levonorgestrel-releasing intrauterine device has also been shown to relief dysmenorrheal and pelvic pain. Eighty percent of women treated with progestins have a partial or complete relief of pain.

Danazol

Danazol is a 19-nortestosterone derivative with progestin-like effects. Danazol acts via several mechanisms to treat endometriosis. It acts at the hypothalamic level to inhibit gonadotropin release, inhibiting the midcycle surge of luteinizing hormone and follicle-stimulating hormone. Danazol also inhibits steroidogenic enzymes in the ovary that are responsible for estrogen production. As a result, a hypoestrogenic environment is created. This, in addition to the androgenic effects of danazol, prevents the growth of endometriotic tissue.

The dosage of danazol is 400 to 800 mg/d in divided doses for 6 months. Side effects of danazol include acne, oily skin, deepening of the voice, weight gain, edema, and adverse plasma lipoprotein changes. Most changes are reversible upon cessation of therapy, but some (such as deepening of the voice) may not be.

Pain relief is achieved in up to 90% of patients taking danazol.

GnRH Agonists

Gonadotropin-releasing hormone (GnRH) agonists are analogues of the 10-amino-acid peptide hormone GnRH. With the continuous administration of GnRH analogues, suppression of gonadotropin secretion occurs, resulting in elimination of ovarian steroidogenesis and suppression of endometrial implants. Pain related to endometriosis is relieved in most cases by the second or third month of therapy. GnRH agonists can be administered intramuscularly as leuprolide acetate 3.75 mg once a month, intranasally as nafarelin 400 to 800 μg daily, or subcutaneously as goserelin 3.6 mg once a month.

The use of these agents is generally limited to 6 months because of the adverse effects associated with a hypoestrogenic state, particularly loss of bone mineral density. Other side effects include vasomotor symptoms, vaginal dryness, and mood changes.

Many side effects can be minimized by providing add-back therapy in addition to the GnRH agonists in the treatment of endometriosis. The addition of 2.5 mg of norethindrone or 0.625 mg of conjugated estrogens with 5 mg/d of medroxyprogesterone acetate seems to provide relief of vasomotor symptoms and decrease bone mineral density loss in a 6-month treatment period. The addition of 5 mg of norethindrone acetate alone or in conjunction with low-dose conjugated equine estrogen seems to eliminate the loss of bone mineral density effectively as well. Adding bisphosphonates, parathyroid hormone, or calcitonin can also minimize the bone loss.

Aromatase Inhibitors

Anastrozole (1 mg daily) and letrozole (2.5 mg daily) are the most commonly used aromatase inhibitors. They act by inhibiting the enzyme aromatase, which functions in the conversion of androgens to estrogens. They can be used as an adjuvant treatment combined with other agents such as GnRH analogs.

Surgical Treatment

In women who want to preserve fertility, who have severe disease, or who have adhesions, conservative surgical therapy is the treatment of choice. This surgery attempts to excise or destroy all endometriotic tissue, remove all adhesions, and restore pelvic anatomy to the best possible condition. Conservative surgery has traditionally been performed at laparotomy, but a laparoscopic approach is associated with a shorter hospital stay and less morbidity, and it is more cost effective. This is particularly true in contemporary practice, where this therapy is usually performed at the time of the initial diagnostic laparoscopy. Reported pregnancy rates after conservative surgery are inversely proportional to the severity of disease and vary greatly. In counseling patients, approximate pregnancy rates of 75% for mild disease, 50–60% for moderate disease, and 30–40% for severe disease should be quoted; however, individualization of therapy is stressed.

Presacral neurectomy to relieve pain should be performed only in selected cases, such as women with recurrent endometriosis, severe incapacitating dysmenorrhea, or disease that did not respond to initial treatment, as efficacy of this treatment is controversial.

If the patient does not desire future childbearing and has severe disease or symptoms, definitive surgery is appropriate and often curative. This entails total abdominal hysterectomy, bilateral salpingo-oophorectomy, and excision of remaining adhesions or implants. If endometriosis remains after excision, postoperative medical therapy may be indicated. After this or after complete excision, hormone replacement therapy is indicated. Estrogen-progestin therapy may be used without reactivating the endometriosis, but individualization of therapy is required.

Assisted Reproduction

Infertile women with endometriosis who are older, or who have failed other therapies for infertility, can undergo assisted reproduction, such as ovulation induction with intrauterine insemination or in vitro fertilization (IVF). However, it was found that women with endometriosis undergoing IVF have significantly lower pregnancy rates, fertilization rates, implantation rates, mean number of oocytes retrieved, and peak estradiol concentrations as compared with women with tubal factor infertility. The need to treat women surgically or medically before starting an IVF cycle remains unclear.

Proper counseling of patients with endometriosis requires attention to several aspects of the disorder. Of primary importance is the initial operative staging of the disease to obtain adequate information on which to base future decisions about therapy. The patient's symptoms and desire for childbearing dictate appropriate therapy. Most patients can be told that they will be able to obtain significant relief from pelvic pain and that treatment will assist them in achieving pregnancy.

Long-term concerns must be more guarded in that all current therapies offer relief but not cure. Even after definitive surgery, endometriosis may recur, but the risk is very low (approximately 3%). The risk of recurrence is not significantly increased by estrogen replacement therapy. After conservative surgery, reported recurrence rates vary greatly but usually exceed 10% in 3 years and 35% in 5 years. Pregnancy delays but does not preclude recurrence. Recurrence rates after medical treatment also vary and are similar to or higher than those reported after surgical treatment.

Although many patients are concerned that endometriosis will progress inexorably, experience has been that conservative surgery avoids the necessity for hysterectomy in the great majority of cases. The course of endometriosis in any individual is impossible to predict at present, and future treatment options should greatly improve what can now be offered.