The median age for endometrial cancer is about 58 years. The risk factors associated with the development of carcinoma of the endometrium are listed in Box 41-1. Any factor that increases the exposure to unopposed estrogen increases the risk for endometrial cancer. If the proliferative effects of estrogen are not counteracted by a progestin, endometrial hyperplasia and possibly adenocarcinoma can result.

Obesity results in an increased extraovarian aromatization of androstenedione to estrone. Androstenedione is secreted by the adrenal glands, whereas the increased peripheral conversion occurs predominantly in fat depots but also in the liver, kidneys, and skeletal muscles. Granulosa-theca cell tumors of the ovary produce estrogen, and up to 15% of patients with these tumors have an associated endometrial cancer.

Unopposed estrogenic stimulation from anovulatory cycles occurs in patients who have polycystic ovarian syndrome (Stein-Leventhal syndrome) and in patients with a late menopause. In postmenopausal women taking estrogen replacement without a progestin for menopausal symptoms, the risk for cancer developing appears to be both dose and duration dependent. This increased risk varies from 2-fold to 14-fold compared with nonusers. The addition of progestin in a cyclic fashion for 10 to 14 days of the month or in a continuous fashion daily throughout the month eliminates this increased risk. Women taking tamoxifen for breast cancer have a twofold to threefold increased risk for endometrial cancer. Young women who use oral contraceptives have been shown to have a lower incidence of subsequent endometrial cancer.

About 5% of endometrial cancers occur in women with the hereditary nonpolyposis colon cancer syndrome (HNPCC), which is caused by germ line mutations in the DNA repair genes. Women with the HNPCC syndrome have about a 40% risk for developing endometrial cancer, usually before the menopause.

Population screening for endometrial cancer is not feasible because there is no simple method of cancer detection available. However, screening may be justified for high-risk women, including those with a family history of HNPCC syndrome, those with polycystic ovarian disease, and any woman with an intact uterus taking unopposed estrogen. Only about 50% of women with endometrial cancer will have malignant cells on a Papanicolaou smear.

Since the 1990s, transvaginal ultrasonography has been increasingly used for endometrial evaluation. Almost all women with endometrial hyperplasia or carcinoma will have an endometrial thickness of 5 mm or more. Tamoxifen produces a confusing ultrasonic image, which leads to frequent false-positive reports.

The most common symptom of endometrial cancer is abnormal vaginal bleeding, which is present in 90% of patients. Postmenopausal bleeding is always abnormal and must be investigated. The most common conditions associated with postmenopausal bleeding are listed in Table 41-1. In the premenopausal patient, especially after age 35 years, menorrhagia or intermenstrual bleeding may signal an endometrial malignancy.

TABLE 41-1 ETIOLOGY OF POSTMENOPAUSAL BLEEDING

The general physical examination may reveal obesity, hypertension, and the stigmata of diabetes mellitus. Evidence of metastatic disease is unusual at initial presentation, but the chest should be examined for any effusion and the abdomen carefully palpated and percussed to exclude ascites, hepatomegaly, or evidence of upper abdominal masses.

On pelvic examination, the external genitalia are usually normal. The vagina and cervix are also usually normal but should be carefully inspected and palpated for evidence of involvement. A patulous cervical os or a firm, expanded cervix may indicate extension of disease from the corpus to the cervix. The uterus may be of normal size or enlarged, depending on the extent of the disease and the presence or absence of other uterine conditions, such as adenomyosis or fibroids. The adnexa should be carefully palpated for evidence of extrauterine metastases or an ovarian neoplasm. A granulosa cell tumor or an endometrioid ovarian carcinoma may occasionally coexist with endometrial cancer.

Any woman who presents with postmenopausal bleeding should have a transvaginal ultrasound. If the endometrial thickness is greater than 5 mm, endometrial evaluation is necessary. Outpatient techniques for endometrial sampling include the use of the Kevorkian curette, Vabra aspirator, Gravlee jet washer, and Pipelle cannula. These techniques have a diagnostic accuracy of about 90%. If the endometrial biopsy reveals endometrial cancer, definitive treatment can be arranged. If the endometrial biopsy is negative for cancer or reveals endometrial hyperplasia, a hysteroscopy and fractional dilation and curettage should be performed under general anesthesia. Specimens from the endometrium and endocervix should be submitted separately for histologic evaluation to determine whether the tumor has extended to the endocervix.

In a premenopausal patient with high-risk factors and abnormal uterine bleeding, the endometrium must be sampled. Failure to respond to medical management or a suspicious transvaginal ultrasound is another indication for hysteroscopy and uterine curettage. A grossly obvious lesion of the cervix or vagina should be biopsied directly.

In addition to a thorough physical examination, blood studies should include a complete blood count, determinations of hepatic enzymes, serum electrolytes, blood urea nitrogen, serum creatinine, and a coagulation profile. A routine urinalysis should be performed. The only radiologic study necessary is a chest x-ray. Additional radiographic procedures, including abdominopelvic computed tomography or magnetic resonance imaging, may be performed, particularly for poorly differentiated cancers, to look for metastatic disease.