Although postmenopausal bleeding (PMB) accounts for only 5% of all gynecologic office visits, its evaluation is a source of consternation for many providers. 1 The top three diagnoses associated with PMB are endometrial atrophy (40% to 50%), endometrial cancer (10%), and polyps (3%). 2 Because 90% of PMB is associated with a benign condition, the ideal diagnostic method is noninvasive. This reduces individual risk and increases accurate diagnosis of malignancy.

Previously, the available diagnostic methods for evaluation of PMB were endometrial biopsy (EMB), dilation and curettage (D & C), or hysteroscopy. These methods are all invasive or surgical in nature. Ultrasonography is nonsurgical and a relatively noninvasive diagnostic method for evaluation of PMB, and its use is now widespread.

Diagnostic Tools

The most invasive diagnostic method, hysteroscopy, is considered the gold standard in the evaluation of PMB. A gynecologist performs this procedure with the patient under anesthesia. Hysteroscopy entails a visual examination of the uterine cavity using endoscopic equipment.

Dilation and curettage entails dilation of the cervix and gentle scrapping of the uterine lining to obtain an endometrial sample for histologic evaluation. Endometrial biopsy, although less invasive than D & C or hysteroscopy, is more invasive than ultrasonography. A flexible catheter is introduced into the uterus and the endometrial sample is obtained by gentle suction. Many family practice clinicians routinely perform EMB.

Transvaginal ultrasound (TVUS) is the least invasive diagnostic tool for the evaluation of PMB, and is fairly accurate in ruling out malignancy. It is typically used in the initial evaluation of PMB and after an unsuccessful EMB to confirm the result and to guide further diagnostic tests. 3,4

The following definitions are valuable when evaluating diagnostic tests:

Sensitivity-the proportion of persons predicted to have a disease to those that actually have it.

Specificity-the proportion of persons who do not have the disease, those who are predicted not to have the disease, as well as the proportion of persons correctly identified with the disease.

Accuracy-the degree to which a measurement represents the true value of the attribute that is being measured.

Positive predictive value-the probability that a person with a positive test result has the condition.

Negative predictive value-the probability that a person with a negative test result does not have the condition. 5

Hysteroscopy

Although hysteroscopy can be performed in the outpatient setting, it requires anesthesia and a gynecologist skilled in the procedure. The accuracy of hysteroscopy is most useful in making the diagnosis of cancer when compared to other types of endometrial disease. In one case series of 181 patients, the sensitivity was 96.6% and the specificity 100% when hysteroscopy was used in conjunction with EMB. 6 However, even hysteroscopy has limitations. Although efficient in the detection of pathological intrauterine lesions, it is only moderately successful in determining physiological changes such as proliferative endometrium or endometrial hyperplasia. 7 This underscores the importance of tailoring the evaluation of PMB to the individual patient, as well as combining diagnostic methods when appropriate.

The widespread use of hysteroscopy, and the ability to perform it on an outpatient basis, has resulted in a reduction in the use of D & C for the evaluation of PMB. Indeed, some authors suggest that D & C should not be used as a diagnostic or therapeutic option for patients with PMB. 8

EMB

In a recent systematic quantitative review, 9 EMB was only moderately accurate in diagnosing endometrial hyperplasia. With a positive EMB test, the probability of endometrial hyperplasia on endometrial tissue sampling obtained by hysteroscopy or D & C under anesthesia was 57.7%. A positive test also increased the probability of carcinoma from 14% to 31.3%, while a negative test decreased it to 2.5%. 10 Additional endometrial testing should be performed when symptoms persist, or when intrauterine abnormalities are suspected, even in the presence of a negative result. 11

Another concern with the use of EMB is its negative predictive value. A positive test with EMB is a more definitive test of endometrial cancer than a negative test is for ruling it out. 8

The specificity of EMB is 99.1%, the sensitivity 84.2%, the accuracy 96.9%, and the positive and negative predictive values 94.1% and 93.7%, respectively. 12 This demonstrates that a diagnosis of endometrial cancer on biopsy is definitive and should lead to treatment. A negative biopsy, however, may require further evaluation, especially if symptoms persist.

Lastly, 16.1% of the time, the sample obtained is "insufficient for diagnosis." This is perhaps due to lack of provider expertise in the technique or insufficient endometrial tissue in women with an atrophic endometrium. Should this occur, further evaluation is needed.

Ultrasonographic Measurement

Ultrasonographic measurement of the endometrium, a commonly used diagnostic tool for PMB, entails inserting a probe attached to a transducer into the vagina. The transducer then measures the endometrium to the nearest millimeter. The endometrium looks like a stripe under sonography, hence the term "endometrial stripe" (EMS). Endometrial stripe is commonly used when referring to endometrial measurements. An EMS measurement > 5 mm is considered abnormal. This was determined by multiple studies of EMS measurements. Measurements > 5 mm were highly correlated with a histologic diagnosis of endometrial cancer. 13-14

Sensitivity of the EMS test was recently reported to be 91% and specificity was 58%. 15 Using pretest and posttest probabilities in a hypothetical patient, the pretest probability of endometrial cancer in the presence of abnormal bleeding was calculated to be 10%. With a positive test result (EMS > 5 mm), the probability of cancer increases to 19%. However, with a negative test result (EMS <=5 mm), the probability decreases to 1.7%. These results indicate that a normal finding is as valuable as an abnormal finding in ultrasonographic screening for endometrial cancer.

A recent, prospective, blinded study 16 concluded that repeat EMS measurements by different sonographers correlated closely, which reflects the accuracy of the EMS test. However, this study took place in a university setting. It is therefore necessary to exercise caution when generalizing these results to non-university settings, or settings in which the skill levels of the ultrasonographers differ.

Although TVUS measurement of the EMS is a useful, highly sensitive and noninvasive test, it has its limitations. Some investigators have determined that EMS has a low positive predictive value for cancer. This is especially true in women taking hormone therapy (HT) and tamoxifen (Nolvadex), or women with recurrent or postmenopausal bleeding that occurs long after menopause. 17

Although the EMS measurement offers a noninvasive and potentially cost-effective method of evaluating PMB, there are factors associated with its use that are problematic. If the EMS is inaccurately measured > 5 mm (a false positive result), the patient will be subjected to further invasive tests such as D & C or hysteroscopy in order to obtain endometrial tissue for histological assessment. Similarly, if the EMS is inaccurately measured <=5 mm (a false negative result), then the diagnosis and treatment of endometrial cancer may be delayed.

Endometrial stripe measurement should not be used to evaluate premenopausal women with abnormal bleeding. An EMS measurement > 5 mm in premenopausal women may be normal due to hormonal influences. Likewise, patients with multiple risk factors for endometrial cancer should have further evaluation of PMB, even if their EMS measurement is <=5 mm.

It is difficult to ascertain the sensitivity and specificity of EMB, compared to ultrasonography because an accurate reference standard for EMB does not exist. The main limitation of EMB is a high rate of inadequate sampling, making additional tests necessary. 2

Sonohysterography

Another important diagnostic tool that has recently come into use is saline sonohysterography. This out-patient procedure allows a visual evaluation of the uterine cavity. It is useful in diagnosing intracavitary lesions such as polyps, leiomyoma, and masses. It does not, however, provide a histologic diagnosis. This test is highly sensitive (97%), and also has a high negative predictive value (94.3%) when combined with EMB. 18 It is important, however, to recognize the limitations of this test. Additional diagnostic testing, such as EMB and/or hysteroscopy, is needed when there is persistent bleeding or a need for a histologic diagnosis based on the patient's history.

In summary, a negative EMS measurement (<=5mm) should be followed by additional tests such as EMB or hysteroscopy in high-risk patients or patients with persistent symptoms. A positive EMS measurement (> 5 mm) should be followed by additional tests to identify a cause for the ultrasonographic abnormality such as polyps, hormone use, leiomyoma, or endometrial cancer. The EMS should not be used as a screening tool in premenopausal women. It is important to use the correct diagnostic tools to assess each patient, according to their individual needs and histories.

Case Studies

Case 1:This case illustrates the correct plan of action for this patient with PMB.

A 55-year-old postmenopausal woman presented with a 3-year history of PMB. She was not on hormone therapy (HT). Past medical history was significant for obesity and hypertension treated with hydrochlorothiazide. Her gynecologic history was significant for heavy irregular menses for many years prior to menopause, which occurred at the age of 52. She had a TVUS that revealed an EMS measurement of 8 mm. Endometrial biopsy was significant for hyperplasia. She underwent hysteroscopy with visualization of the entire cavity and a D & C. Histology from the procedure was significant for complex hyperplasia without atypia. She underwent progesterone therapy, with plans for reevaluation in 3 months.

Case 2:This case illustrates inappropriate management of a patient with PMB.

A 54-year-old patient complained of PMB for a "few months." A TVUS was performed and the EMS measured 2 mm. Because the EMS measured < 5 mm, she was reassured that the likelihood of endometrial cancer was quite small. She was instructed to return if any further bleeding occurred. She continued to have spotting, however, and returned for a follow-up. An EMB was performed for a histological assessment of endometrial tissue. The results were scant, "fragments of endometrium and mucus, insufficient for diagnosis." The patient was reassured that she did not have endometrial cancer although her spotting continued and plans were made for evaluation only if bleeding increased.

Case 3:This case illustrates inappropriate management of a patient with PMB. She had multiple risk factors for endometrial cancer, including tamoxifen use, history of breast cancer, continued bleeding and spotting. An endometrial stripe less than 5 mm did not rule out pathology in a patient with multiple risk factors and continued clinical symptoms.

A 53-year-old patient, not on HT, presented with a 3-month history of PMB. She experienced menopause at age 50. Her past medical history included breast cancer, which was treated by modified radical mastectomy and tamoxifen. Her gynecologic history was significant for two normal, spontaneous vaginal deliveries at term. She had not had any gynecologic surgeries. She had a TVUS that revealed an EMS of 4 mm. The patient continued to have bleeding and spotting. Her provider reassured her that an EMB was not necessary since the EMS was normal. The provider then planned to repeat the TVUS in 3 to 6 months.

REFERENCES