Postmenopausal bleeding is a symptom that can announce the presence of a gynecologic malignancy. In this column, we will discuss the important considerations to make in the work-up of this symptom.
Roughly 10% of women will present for evaluation of postmenopausal bleeding.1 More than a third of these women will have benign pathology, with the incidence of endometrial cancer in this group at only about 5%.2 Other gynecologic malignancies should be considered as well, including cervical, vaginal, vulvar, and more rarely, those of the fallopian tubes or ovaries.
Visible cervical lesions should be sampled with a biopsy rather than a cytologic test. Pap testing is designed to evaluate asymptomatic women for occult dysplasia rather than diagnostic confirmation of symptomatic visible lesions. Inappropriate use of cervical cytology in such patients can expose them to false-negative results or delays in definitive diagnosis. It should be noted that it is very uncommon for cervical dysplasia to present with bleeding, and therefore, the presence of bleeding should heighten suspicion for an underlying malignancy.
Use of ultrasound
Ultrasound is a commonly performed initial approach to work-up because of its noninvasive nature. Transvaginal ultrasound has a high negative predictive value of 99.4%-100% in ruling out malignancy.3 Among women with postmenopausal bleeding, the risk of cancer is 7.3% if their endometrial lining is 5 mm or greater and less than 0.07% risk if their lining is 4 mm or less. Therefore, this cutoff dimension is typically used to triage patients to additional sampling.
If ultrasound is performed on postmenopausal women who are asymptomatic (no bleeding), then an endometrial stripe of greater than 11 mm is considered justification for further work-up and is associated with a 6.7% risk of endometrial cancer.4 If the ultrasound reveals intracavitary lesions, a sonohysterogram would be preferred to characterize intrauterine pathology. In fact, sonohysterography is superior to transvaginal ultrasound (with a sensitivity of 80% vs. 49%, respectively) in detecting endometrial polypoid lesions.5 Preoperative identification of an intracavitary lesion may assist in selecting the best sampling technique (blind vs. hysteroscopy-guided approach).
If an ultrasound reveals a thickened or unevaluable endometrial stripe or if the clinician chooses to proceed directly with diagnostic confirmation, several options for endometrial sampling exist, including office-based or operative procedures, as well as blind or visually guided ones. Endometrial pipelle biopsy, D&C without hysteroscopy, endometrial lavage, and endometrial brush biopsy all constitute “blind” sampling techniques. Targeted biopsy techniques include hysteroscopy D&C and saline infusion sonohysterography–guided biopsy.
Although D&C may be considered the gold standard of diagnostic sampling techniques, it should be noted that 60% of these procedures sample less than half of the endometrium.6 When used in conjunction with hysteroscopy, the sensitivity in detecting cancer is high at 97% with a specificity of 93%-100%.7
While some patients are candidates for office-based procedures, D&C often requires regional or general anesthesia and is frequently performed in a hospital-based environment or surgical center. This may be most appropriate for patients who have had failed office attempts at sampling, have multiple medical comorbidities that limit the feasibility of office-based procedures (such as morbid obesity), or have severe cervical stenosis. D&C is associated with an increased risk for uterine perforation, compared with outpatient sampling procedures.
The need to go to the operating room rather than to an ambulatory setting also may increase the costs borne by the patient. The advantages of D&C include the potential for large-volume sampling and the potentially therapeutic nature of the procedure in cases of benign pathology.
Office-based sampling techniques include those using a pipelle, those employing an endometrial brush, and those guided by saline infusion sonohysterography. If performed in the office, they require minimal or no cervical dilation, are associated with a lower risk of perforation or adverse reaction to anesthesia, and usually have lower costs for patients.
Endometrial pipelle biopsies are a very effective diagnostic tool when there is global, endometrial pathology; they have a sensitivity of 83% in confirming cancer.8 It is an inexpensive and technically straightforward technique that can be easily performed in an office setting.
However, when the endometrial lining is atrophied, alternative tools may provide superior results. Endometrial brushes have been shown to be 33% more successful in collecting adequate samples,compared with pipelles, because they sample a larger endometrial surface area.9
There is ongoing development of sampling techniques, such as endometrial lavage or the combination of saline infusion sonohysterography and endometrial biopsy.10 However, future studies regarding accuracy, cost, and patient acceptability are needed before these techniques are translated to the clinical setting.
Targeted endometrial sampling
Targeted or visually guided sampling, such as hysteroscopy, has been shown to be very accurate in identifying benign pathology, although the sensitivity of hysteroscopic diagnosis of cancer is significantly lower at approximately 50%.11 Therefore, the benefit of hysteroscopy is in complementing the blind nature of D&C by guiding sampling of intracavitary lesions, should they exist.
Hysteroscopy is safe in endometrial cancer and is not associated with upstaging the cancer from transtubal extirpation of malignant cells.12
The addition of hysteroscopy contributes some cost and equipment to the blind D&C procedure; therefore, it might be best applied in cases where there is known intracavitary pathology or inadequate prior sampling. In well-selected patients, hysteroscopy often can be used in an office setting, which improves the practicality of the procedure. Smaller and, in some cases, disposable equipment aids in the feasibility of adding visual guidance to office sampling.
Postmenopausal women have a higher risk for sampling failure, compared with younger women. Obesity also is a risk for failed sampling.13 Cervical ripening with misoprostol may increase access to the endometrial cavity, and ultrasound guidance may decrease the risk of uterine perforation in a stenotic cervix.
Clinicians should ensure that histology results are concordant with clinical data. Discordant results should be reevaluated. For example, if an ultrasound demonstrates a thickened endometrial stripe, but the sampling reveals “scant atrophic tissue,” then there is unexplained pathology to address. Further work-up, such as more comprehensive sampling with hysteroscopy, should be considered in such cases. Additionally, persistent postmenopausal bleeding, despite a benign endometrial biopsy, should be reevaluated over time to rule out occult disease missed during prior sampling.
Appropriate strategies for the work-up of postmenopausal bleeding should be tailored to each patient and their risk factors. Clinicians are now equipped with multiple ways of obtaining clinical data, and patients have options that may decrease barriers to their care. Hysteroscopy does not improve upon D&C in the diagnosis of endometrial cancer, although it may be helpful in distinguishing and treating nonmalignant lesions.
Dr. Cotangco is a resident in the department of obstetrics and gynecology at the University of Illinois, Chicago. Dr. Rossi is an assistant professor in the division of gynecologic oncology at the University of North Carolina, Chapel Hill. They reported having no relevant financial disclosures.
10. PLoS Med. 2016 Dec. doi: 10.1371/journal.pmed.1002206.