This chapter deals primarily with APGO Educational Topic Area:
TOPIC 54 ENDOMETRIAL HYPERPLASIA AND CARCINOMA
Students should be able to outline a basic approach to the causes and diagnosis of endometrial hyperplasia or cancer, particularly in the postmenopausal patient who presents with bleeding. They should identify risk factors, common presenting signs and symptoms, and physical examination findings.
Clinical Case
A 42-year-old patient comes to see you because of irregular bleeding. She reports that she has had regular menstrual cycles until 3 months ago, when she started noticing that she was having two “periods” a month. She denies any lightheadedness, pain, bloating, or fever.
Approximately 2% to 3% of women will develop uterine cancer during their lifetime. Ninety-seven percent of all uterine cancers arise from the glands of the endometrium and are known as endometrial carcinomas. The remaining 3% of uterine cancers arise from mesenchymal uterine components and are classified as sarcomas.
Endometrial carcinoma is the most common genital tract malignancy and the fourth most common cancer after breast, bowel, and lung carcinoma. In 2012, over 47,000 new cases were diagnosed, resulting in over 8,000 deaths. Fortunately, patients with this disease usually present early in the disease course with some form of abnormal uterine bleeding (AUB), particularly postmenopausal bleeding. With early diagnosis, survival rates are excellent.
ENDOMETRIAL HYPERPLASIA
Endometrial hyperplasia is the most common precursor to endometrial adenocarcinoma. In 1994, the World Health Organization defined a classification system of endometrial hyperplasia based on four types of simple and complex hyperplasia, with or without atypia (Table 49.1). This system is still used today though it is limited by interobserver variability.
Types
Simple Hyperplasia
Simple hyperplasia is the least significant form of endometrial hyperplasia and is not commonly associated with progression to endometrial carcinoma. In this type of hyperplasia, both glandular elements and stromal cell elements proliferate excessively. Histologically, glands vary markedly in size, from small to cystically enlarged (the hallmark of this hyperplasia). Cystic glandular hyperplasia should not be confused with a normal postmenopausal variant—cystic involution of the endometrium—which is histologically not a hyperplastic condition.
Complex Hyperplasia
Complex hyperplasia represents an abnormal proliferation of primarily glandular elements without concomitant proliferation of stromal elements. This increased gland-to-stroma ratio gives the endometrium a “crowded” picture, frequently with glands appearing almost back-to-back. As the severity of the hyperplasia increases, the glands become more crowded and more structurally bizarre. It is thought that complex hyperplasia represents a true intraepithelial neoplastic process, and it is occasionally found coexisting with areas of endometrial adenocarcinoma.
Hyperplasia (Simple or Complex) with Cytological Atypia
Hyperplasia characterized by significant numbers of glandular elements that exhibit cytological atypia and disordered maturation (loss of cellular polarity, nuclear enlargement with increased nucleus-to-cytoplasm ratio, dense chromatin, and prominent nucleoli) is considered a precursor lesion to endometrial carcinoma (Fig. 49.1).
Pathophysiology and Risk Factors
The primary process central to the development of endometrial hyperplasia (and most endometrial cancer) is overgrowth of the endometrium in response to excess unopposed estrogen. Sources of estrogen may be endogenous(ovarian; peripheral conversion of androgenic precursors) or exogenous (Box 49.1). Endometrial proliferation represents a normal part of the menstrual cycle and occurs during the follicular, or estrogen-dominant, phase of the cycle. With continued estrogen stimulation through either endogenous mechanisms or by exogenous administration, simple endometrial proliferation will become endometrial hyperplasia. Research suggests that this transformation may be time and dose dependent. When proliferation becomes hyperplasia is not clear, although studies showing sequential change suggest it requires 6 months or longer of stimulation without progesterone opposition. The risk factors for hyperplasia and endometrial cancer are identical (Table 49.2). The risk of progression to cancer with the various forms of hyperplasia are detailed in Table 49.1 In one study, more than 42% of women with endometrial atypia had invasive endometrial cancer demonstrated when hysterectomy was performed within 3 months.
BOX 49.1 Estrogen Sources
Endogenous
Glandular
Estradiol (ovary)
Estrone (ovary)
Peripheral
Estrone (fat, conversion of androstenedione)
Tumor
Granulosa cell of ovary (an uncommon tumor and source)
Exogenous
Conjugated estrogen (mostly estrone)
Lyophilized estradiol
Cutaneous patches
Vaginal creams
Patient History
AUB is the hallmark symptom of both endometrial hyperplasia and cancer. Further evaluation to rule out underlying carcinoma is warranted in two general scenarios: 1) any patient with AUB over age 35 years and 2) a woman with AUB less than or equal to age 35 years in the presence of additional risk factors (family history of breast, colon, or gynecologic cancer; obesity; prior endometrial hyperplasia; chronic anovulation; tamoxifen use; or estrogen therapy).
FIGURE 49.1. Complex hyperplasia with severe nuclear atypia of endometrium. (A) The proliferative endometrial glands reveal considerable crowding and papillary infoldings. The endometrial stroma, although markedly diminished, can still be recognized between the glands. (B) Higher magnification demonstrates disorderly nuclear arrangement and nuclear enlargement and irregularity. Some contain small nucleoli. (Provided by Gordana Stevanovic, MD, and Jianyu Rao, MD, Department of Pathology, UCLA; Berek JS, Berek & Novak’s Gynecology. 14th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2007:1347, figure 33.1.)
Evaluation
Endometrial Evaluation
Histologic evaluation of a sample of the endometrium establishes the diagnosis of endometrial hyperplasia or carcinoma. Endometrial biopsy is most easily accomplished by any number of different atraumatic aspiration devices used in the office. The diagnostic accuracy of office endometrial biopsy is 90% to 98%, compared with dilation and curettage (D&C) or hysterectomy.
The routine Pap smear is not reliable in diagnosing endometrial hyperplasia or cancer, because only 30% to 40% of patients with endometrial carcinoma have abnormal Pap test results. On the other hand, endometrial carcinoma must be considered, and endometrial sampling obtained, when atypical endometrial cells or atypical glandular cells of undetermined significance are found on the Pap smear.
The most common indication for endometrial sampling is abnormal bleeding. After ruling out pregnancy in premenopausal women, an adequate tissue sample can be obtained with relatively little discomfort. Further management is usually dictated by the results of the biopsy specimen. D&C or hysteroscopy with directed endometrial biopsy may be undertaken when outpatient sampling is not possible (e.g., because of a stenotic cervical os or a patient who cannot tolerate the outpatient procedure) or when the outpatient sampling has been nondiagnostic.
Sometimes the office endometrial biopsy will be reported as having “insufficient tissue for diagnosis.” In a postmenopausal woman who is not taking hormone therapy, this result is compatible with the normal atrophic condition of the endometrium. In other cases, the clinical suspicion of a possible hyperplastic endometrial process may be high enough to warrant hysteroscopic evaluation with directed sampling, which allows more complete evaluation of the endometrium as well as direct visualization of polyps, myomas, and structural abnormalities (Fig. 49.2).
FIGURE 49.2. Diffuse hyperplasia as seen by hysteroscopy. The thickened walls of the uterus are closely opposed. (Baggish MS, Valle RF, Guedj H. Hysteroscopy: Visual Perspectives of Uterine Anatomy, Physiology & Pathology. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2007:233, figure 18.19.)
Ultrasound
Transvaginal ultrasound (with or without the instillation of fluid for contrast, sonohysterography) may be used as an adjunct means of evaluation for endometrial hyperplasia as well as for polyps, myomas, and structural abnormalities of the uterus. An endometrial thickness of >4 mm in a postmenopausal patient, a polypoid mass, or fluid collection is considered an indication for further evaluation and obtaining histologic samples. (Fig. 49.3) It is also useful in patients who have multiple medical problems, to help determine if the risks of endometrial sampling are less than the risk of not sampling. Multiple metaanalyses have demonstrated that the probability of endometrial cancer, though not zero, is low, approximately 1%, with an endometrial stripe thickness of 4 mm or less. The value of measuring the endometrial stripe in a premenopausal woman is less significant, given day-to-day variations throughout the menstrual cycle.
Management
The primary goals of treating endometrial hyperplasia are to reduce the risk of malignant transformation and to control the presenting symptoms.
FIGURE 49.3. Endometrial stripe (calipers) as seen on ultrasound.
In women with breast cancer treated with tamoxifen, the optimal manner of monitoring the endometrium is unclear. Tamoxifen acts as a weak estrogen and is associated with increased risk of endometrial hyperplasia and carcinoma. Most agree that, unless the patient has been identified to be at high risk for endometrial cancer, a routine ultrasonography and endometrial biopsy in asymptomatic women are not necessary. Endometrial abnormalities should be excluded in the presence of new symptoms, such as bloody vaginal discharge, spotting, and AUB.
Medical Therapy
Synthetic progesterones or other progestins are central in the medical treatment of endometrial hyperplasia. They act through a number of pathways. First, they work to alter the enzymatic pathways, which eventually convert endogenous estradiol to weaker estrogens. Second, they decrease the number of estrogen receptors in the endometrial glandular cells, rendering them less susceptible to exogenous stimulation. Finally, the stimulation of progesterone receptors results in thinning of the endometrium and stromal decidualization. With time, this results in a decrease in endometrial glandular proliferation, which renders the endometrium atrophic.
In cases of hyperplasia without atypia, medical therapy is utilized first. The mean duration of progression from endometrial hyperplasia to carcinoma in those that do progress is relatively long: perhaps 10 years for those without atypia and 4 years for those with atypia. The most common treatment regimen is cyclic medroxyprogesterone acetate, which is administered for 10 to 14 days each month for at least 3 to 6 months. Continuous administration is equally effective and may aid with compliance in patients who have an irregular cycle length.
Beyond Medical Therapy
Many view hyperplasia with atypia as a continuum with endometrial cancer. Hence, aggressive therapy for these patients is warranted, given the increased likelihood of progression to endometrial cancer. After initial diagnosis, D&C is indicated to better sample the endometrium and exclude the possibility of a coexistent endometrial cancer. In young women who desire future fertility, long-term, high-dose progestin therapy may be used in an attempt to avoid a hysterectomy. As an alternative to oral therapy, the levonorgestrel intrauterine contraceptive has been reported to have response rates ranging from 58% to 100%. Definitive therapy by hysterectomy is recommended after a patient has completed childbearing. Patients who are treated medically for atypical hyperplasia should also be followed with periodic endometrial sampling (every 3 months after therapy), so treatment response can be monitored.
ENDOMETRIAL POLYPS
Most endometrial polyps represent focal, accentuated, benign hyperplastic processes. Their histologic architecture is characteristic and may commonly be found in association with other types of endometrial hyperplasia or even carcinoma. Polyps occur most frequently in perimenopausal or immediately postmenopausal women, when ovarian function is characterized by persistent estrogen production due to chronic anovulation. The most common presenting symptom is abnormal bleeding. Small polyps may often be incidentally found as part of endometrial sampling or curettage done for evaluating AUB. Rarely, a large polyp may begin to protrude through the cervical canal. Such cases present with bleeding irregularities, and low, dull, midline pain, as the cervix is slowly dilated and effaced. In these cases, surgical removal is necessary to reduce the amount of bleeding and to prevent infection developing within the exposed endometrial surface. Less than 5% of polyps show malignant change. When they do, however, they may represent any endometrial histologic variant. Polyps in postmenopausal women or women taking tamoxifen are more likely to be associated with endometrial carcinoma than those found in reproductive-age women.
ENDOMETRIAL CANCER
Endometrial carcinoma is typically a disease of postmenopausal women. Between 15% and 25% of postmenopausal women with bleeding have endometrial cancer. A majority of cases are diagnosed while in stage I (72%). Despite recognition at early stages, endometrial cancer is the eighth leading site of cancer-related mortality among women in the United States.
Most primary endometrial carcinomas are adenocarcinomas (Fig. 49.4). Because squamous epithelium may coexist with the glandular elements in an adenocarcinoma, descriptive terms that include the squamous element may be used. In cases where the squamous element is benign and makes up less than 10% of the histologic picture, the term adenoacanthoma is used. Uncommonly, the squamous element may appear malignant on histologic assessment and is then referred to as adenosquamous carcinoma. Other descriptions, such as clear cell carcinoma and papillary serous adenocarcinoma, may be applied, depending on the histologic architecture. All of these carcinomas are considered under the general category of adenocarcinoma of the endometrium and are treated in a similar manner.
FIGURE 49.4. Types of endometrial tumors. (A) Endometrial adenocarcinoma. This polypoid exophytic tumor has invaded the outer third of the myometrium. (B) Serous carcinoma of the endometrium. The tumor is a polypoid mass arising in an atrophic uterus. Extensive myometrial lymphatic spread and involvement of the ovary are evident. (From Berek JS, Hacker NF. Practical Gynecologic Oncology. 4th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2005:figures 6.42, 6.49.)
Pathogenesis and Risk Factors
Two kinds of endometrial carcinoma have been identified. Type I endometrial carcinoma is “estrogen dependent” and accounts for approximately 90% of cases. It is most commonly caused by an excess of estrogen unopposed by progestins. These cancers tend to have low-grade nuclear atypia, endometrioid cell types, and an overall favorable prognosis. The second type, type II or “estrogen-independent” endometrial carcinoma, occurs spontaneously, characteristically in thin, older postmenopausal women without unopposed estrogen excess, arising in an atrophic endometrium rather than a hyperplastic one. These cancers tend to be less differentiated, with a poorer prognosis. Estrogen-independent cancer is less common than the estrogen-dependent cancer. Unusual histologic subtypes, including papillary serous adenocarcinoma and clear cell adenocarcinoma of the endometrium, tend to be more aggressive than the more common adenocarcinoma (Table 49.3).
Endometrial carcinoma usually spreads throughout the endometrial cavity first and then begins to invade the myometrium; endocervical canal; and, eventually, the lymphatics. Hematogenous spread occurs with endometrial carcinoma more readily than in cervical cancer or ovarian cancer. Invasion of adnexal structures may occur through lymphatics or direct implantation through the fallopian tubes. After extrauterine spread to the peritoneal cavity, cancer cells may spread widely in a fashion similar to that of ovarian cancer.
As mentioned earlier, the risk factors for developing endometrial cancer are identical to those for endometrial hyperplasia (see Table 49.2).
Diagnosis
Endometrial sampling, prompted by vaginal bleeding, most frequently establishes the diagnosis of endometrial cancer. Vaginal bleeding or discharge is the only presenting complaint in 80% to 90% of women with endometrial carcinoma. In some, often older patients, cervical stenosis may sequester the blood in the uterus, with the presentation being hematometra or pyometra and a purulent vaginal discharge. In more advanced disease, pelvic discomfort or an associated sensation of pressure caused by uterine enlargement or extrauterine disease spread may accompany the complaint of vaginal bleeding or even be the presenting complaint. Fewer than 5% of women found to have endometrial carcinoma are actually asymptomatic.
Special consideration should be given to the patient who presents with postmenopausal bleeding (i.e., bleeding that occurs after 12 months of amenorrhea in a patient who has been diagnosed as menopausal). In this group of patients, it is mandatory to assess the endometrium histologically because the risk of endometrial carcinoma is approximately 10% to 15%, although other causes are more common (Table 49.4). Other gynecologic assessments should also be made, including careful physical and pelvic examination, as well as a screening Pap smear. Preoperative measurement of the CA-125 level may be appropriate, because it is frequently elevated in women with advanced-stage disease. Elevated levels of CA-125 may assist in predicting treatment response or in posttreatment surveillance.
Prognostic Factors
The current International Federation of Gynecology and Obstetrics (FIGO) staging of endometrial cancer (adopted in 1988) lists three grades of endometrial carcinoma:
• G1 is well-differentiated adenomatous carcinoma (less than 5% of the tumor shows a solid growth pattern).
• G2 is moderately differentiated adenomatous carcinoma with partly solid areas (6%–50% of the tumor shows a solid growth pattern).
• G3 is poorly differentiated or undifferentiated (greater than 50% of the tumor shows a solid growth pattern).
Most patients with endometrial carcinoma have G1 or G2 lesions by this classification, with 15% to 20% having undifferentiated or poorly differentiated G3 lesions.
The FIGO staging system incorporates elements correlated with prognosis and risk of recurrent disease—histologic grade, nuclear grade, depth of myometrial invasion, cervical glandular or stromal invasion, vaginal and adnexal metastasis, cytology status, disease in the pelvic and/or periaortic lymph nodes, and presence of distant metastases (Box 49.2). The single most important prognostic factor for endometrial carcinoma is histologic grade. Histologically, poorly differentiated or undifferentiated tumors are associated with a considerably poorer prognosis because of the likelihood of extrauterine spread through adjacent lymphatic and peritoneal fluid. Depth of myometrial invasion is the second most important prognostic factor.
Survival rates vary widely, depending on the grade of tumor and depth of penetration into the myometrium. A patient with a G1 tumor that does not invade the myometrium has a 95% 5-year survival rate, whereas a patient with a poorly differentiated (G3) tumor with deep myometrial invasion may have a 5-year survival rate of only 20%.
Treatment
Hysterectomy is the primary treatment of endometrial cancer. The addition of complete surgical staging with an assessment of retroperitoneal lymph nodes is not only therapeutic but also associated with improved survival.Complete surgical staging includes pelvic washings, bilateral pelvic and periaortic lymphadenectomy, and complete resection of all disease. Sampling of the common iliac nodes, regardless of depth of penetration or histologic grade, may provide further information about the histologic grade and depth of invasion. Palpation of lymph nodes is equally inaccurate and should not be used as a substitute for surgical resection of nodal tissue for histopathology.
Exceptions to the need for surgical staging include young or perimenopausal women with G1 endometrioid adenocarcinoma associated with atypical endometrial hyperplasia and women at increased risk for mortality secondary to comorbidities. Women in the former group who desire to maintain their fertility may be treated with high-dose progestin monitored by serial endometrial sampling. Women in the latter group may be treated with vaginal hysterectomy. In ultrahigh–risk surgical patients, therapeutic radiation may be used as primary treatment, although results are suboptimal.
Postoperative radiation therapy should be tailored to known metastatic disease or used in cases of recurrence. In patients with surgical stage I disease, radiation therapy may reduce the risk of recurrence, but does not improve the survival. For women with positive lymph nodes (stage IIIc), radiation therapy is critical in improving survival rates. Women with intraperitoneal disease are treated with surgery, followed by systemic chemotherapy or radiation therapy or both.
Recurrent Endometrial Carcinoma
Postoperative surveillance for women who have not received radiation therapy involves speculum and rectovaginal examinations every 3 to 4 months for 2 to 3 years, and then twice a year to detect pelvic recurrent disease, particularly in the vagina. Women who have received radiation therapy have a decreased risk of vaginal recurrence as well as fewer therapeutic options to treat recurrence. Therefore, these women benefit less from frequent surveillance with cervical cytology screening and pelvic examinations for detection of recurrent disease.
Recurrent endometrial carcinoma occurs in about 25% of patients treated for early disease, one half of those within 2 years, increasing to three fourths of those within 3 to 4 years. In general, those with recurrent vaginal disease have a better prognosis than those with pelvic recurrence, who, in turn, fare better than those with distant metastatic disease (lung, abdomen, lymph nodes, liver, brain, and bone).
Recurrent estrogen-dependent or progestin-dependent cancer may respond to high-dose progestin therapy. A major advantage of high-dose progestin therapy is its minimal complication rate. Systemic chemotherapy may produce occasional favorable short-term results, but long-term remissions with these therapies are rare.
Hormone Therapy after Treatment for Endometrial Carcinoma
The use of estrogen therapy in patients previously treated for endometrial carcinoma has long been considered contraindicated because of the concern that estrogen might activate occult metastatic disease. For women with a history of endometrial carcinoma, hormone therapy can potentially be used for the same indications as for any other woman, except that the selection of appropriate candidates for estrogen treatment should be based on prognostic indicators, and the patient must be willing to assume the risk. Cautious individualized assessment of risks and benefits should, therefore, be made on a case-by-case basis.
UTERINE SARCOMA
Uterine sarcomas represent an unusual gynecologic malignancy, accounting for approximately 3% of cancers involving the body of the uterus, and only about 0.1% of all myomas. Progressive uterine enlargement occurring in the postmenopausal years should not be assumed to be the result of simple uterine leiomyomata, because appreciable endogenous ovarian estrogen secretion is absent, thereby minimizing the potential for growth of benign myomas. Even postmenopausal women on low-dose hormone therapy are not at risk for stimulation of uterine enlargement. When progressive growth is present in postmenopausal women, uterine sarcoma should be strongly considered. Other symptoms of uterine sarcoma include postmenopausal bleeding, unusual pelvic pain coupled with uterine enlargement, and an increase in unusual vaginal discharge. Surgical removal is the method of most reliable diagnosis. Accordingly, hysterectomy is usually indicated in patients with documented and, especially, progressive, uterine enlargement (Fig. 49.5).
Sarcomas originate from the myometrium or the stromal components of the endometrium. There are many types of uterine sarcoma. They are histologically divided into nonepithelial type and mixed epithelial–nonepithelial type. Nonepithelial type uterine sarcomas include endometrial stromal sarcoma, undifferentiated endometrial sarcoma, leiomyosarcoma, and mixed endometrial stromal and smooth muscle tumors. Mixed epithelial–nonepithelial tumors include adenosarcomas, which contain a benign epithelial component and a malignant stromal element.
The virulence of uterine sarcomas is directly related to the number of mitotic figures and degree of cellular atypia as defined histologically. These tumors are very aggressive and are more likely to spread hematogenously than endometrial adenocarcinoma. When uterine sarcoma is suspected, patients should undergo a tumor survey to assess for distant metastatic disease. At the time of hysterectomy, it is necessary to thoroughly explore the abdomen and sample commonly affected node chains, including the iliac and periaortic areas. The staging for uterine sarcoma is surgical and identical to that for endometrial adenocarcinoma.
Unfortunately, the 5-year survival of patients with a uterine sarcoma is worse than those with similar stage endometrial carcinoma. Survival ranges from 29% to 76%. Radiation and chemotherapy provide little survival advantage as adjuvant therapy to hysterectomy but can decrease recurrence rate. Adjuvant therapy in the form of hormonal antagonists and analogues, such as aromatase inhibitors, progestins, and gonadotropin-releasing hormone agonists, have demonstrated similar results.
FIGURE 49.5. Uterine sarcoma. This hysterectomy specimen shows a large, partially necrotic polypoid mass filling the endometrial cavity and extensively invading the uterine wall. (From Berek JS, Hacker NF. Practical Gynecologic Oncology. 4th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2005:figure 6.47.)
Clinical Follow-Up
You perform an endometrial biopsy, which demonstrates simple hyperplasia without atypia. You explain that this represents abnormal cells with a heightened potential for endometrial cancer. She opts for progesterone treatment, which she continues for 1 year. Follow-up biopsies of the endometrium are normal.
thePoint Visit http://thepoint.lww.com/activate for an interactive USMLE-style question bank and more!