Advertisement
Mayo Clinic Proceedings Home

Diagnostic Options for Assessment of Postmenopausal Bleeding

  • Andrew E. Good
    Correspondence
    Address reprint requests to Dr. A. E. Good, Department of Obstetrics and Gynecology, Mayo Clinic Rochester, 200 First Street SW, Rochester, MN 55905
    Affiliations
    Department of Obstetrics and Gynecology, Mayo Clinic Rochester, Rochester, Minnesota
    Search for articles by this author
      Postmenopausal women should undergo endometrial assessment when bleeding occurs in the absence of replacement hormone therapy, after a 1-year regimen of continuous-combined hormone replacement therapy, or at an unexpected time during cyclic replacement therapy. Endometrial assessment can be accomplished by endometrial biopsy, transvaginal ultrasonography, or hysteroscopy. Dilation and curettage should be reserved for those rare cases when the other methods are infeasible. Each method of assessment has advantages and disadvantages, and the choice of method should be made accordingly.

      Abbreviations:

      HRT (hormone replacement therapy)
      As the post-World War II generation ages, the number of women entering menopause will increase, and primary-care physicians who treat such women will encounter concerns about hormone replacement therapy (HRT) and its complications. A common complaint is postmenopausal bleeding, which is not dependent on whether the woman is taking HRT. This article examines the various methods available for assessment of postmenopausal bleeding and details the associated advantages and disadvantages.

      Definition

      The mean age at which North American women reach menopause is about 51 years. Menopause, the process of slowly diminishing ovarian hormone production and the resultant physiologic changes, is defined as the absence of menses for 6 months or more. Obviously, symptoms and physiologic changes can occur before then. Thus, postmenopausal bleeding is defined as bleeding that occurs after 1 year of amenorrhea if the woman is not receiving HRT. Bleeding that accompanies the systematic withdrawal of replacement progestins is not considered abnormal postmenopausal bleeding.
      The prevalence of serious pathologic problems—that is, endometrial carcinoma or complex endometrial hyperplasia with atypia (a premalignant condition)—is estimated to be 5% or less in all women with unexplained postmenopausal bleeding. Thus, about 95% of bleeding is due to benign causes. One study in which endometrial sampling was used showed that, in 801 asymptomatic women, cancer or atypical hyperplasia was found in only 0.76% of the specimens.
      • Archer DF
      • Mclntyre-Seltman K
      • Wilborn Jr, WW
      • Dowling EA
      • Cone F
      • Creasy GW
      • et al.
      Endometrial morphology in asymptomatic postmenopausal women.
      Of note, stage I, grade 1 endometrial cancer has a 98% 5-year survival rate. Because this type of cancer is associated with bleeding during the early course of disease, vigilant surveillance leads to early discovery and improved chances for cure. Therefore, all women with unexplained bleeding should undergo endometrial sampling.
      In women who are taking progestin cyclically, either for 10 days or for 12 days each month, withdrawal bleeding is expected, and endometrial assessment is unnecessary if bleeding occurs exclusively during that time frame. Women who are receiving a continuous-combined regimen have a high rate of breakthrough bleeding during the first 6 months of therapy. One study revealed a 60% rate of spotting or bleeding during the first 6 months. This rate decreases to 5 to 20% after 1 year of therapy. The consensus was that any bleeding after 6 months of continuous-combined therapy should be investigated; however, because of the decreased incidence of bleeding after 1 year, some clinicians delay endometrial assessment until 1 year has passed.
      In summary, the following women should undergo endometrial assessment: those who experience bleeding after 1 year of amenorrhea if they are not receiving HRT, those who have bleeding after 1 year while receiving a continuous-combined HRT regimen, and those who experience bleeding at an unexpected time while receiving cyclic HRT.

      Methods Of Endometrial Assessment

      Endometrial Biopsy.—Endometrial biopsy is an office procedure that can be performed by using various techniques; anesthesia is unnecessary. With any of the techniques, a small tube is inserted through the cervix into the endometrial cavity. Then, tissue or cells are collected by brushing, washing, scraping, or vacuum. The sensitivity of these sampling tubes for detecting pathologic tissue ranges from 88 to 97%, the same as for dilation and curettage. Associated cost and patient discomfort differ with the various methods, but the accuracy of each method is comparable.
      A widespread method of endometrial sampling involves use of a flexible tube (3 mm in diameter) that is inserted through the cervical canal. Examples of such tubes are the Pipelle sampler and Z-Sampler. With these tubes, an obturator is withdrawn, and then suction is created in the tube, and tissue can be obtained. This type of sample is more easily interpreted by the pathologist than that obtained by simply collecting cells.
      The technique for endometrial sampling is straightforward. (1) Perform a pelvic examination to determine whether the uterus is anteverted or retroverted. (2) Insert the speculum, and swab the cervix with antiseptic. (3) Attempt to advance the catheter. If the catheter cannot be advanced easily, grasp the cervix with a tenaculum and use this approach for countertraction. If the procedure is still unsuccessful, insert a probe or a small dilating sound into the uterus. Putting a “bend” in the catheter may be necessary if the uterus is severely flexed. (4) Quickly pull the obturator as far out as possible and move the catheter in and out while rotating it around its axis. Perform this maneuver for at least 10 seconds. (5) Withdraw the catheter and expel the sample onto blotter paper and put the sample in a bottle of fixative.
      The sensitivity and specificity of this method vary among studies. One study showed a 97% sensitivity when specimens obtained with the Pipelle sampler were compared with those obtained at hysterectomy in patients with known endometrial cancer.
      • Stovall TG
      • Photopulos GJ
      • Poston WM
      • Ling FW
      • Sandles LG
      Pipelle endometrial sampling in patients with known endometrial carcinoma.
      In another study in which the same type of comparison was made, only a 67% sensitivity was found.
      • Ferry J
      • Farnsworth A
      • Webster M
      • Wren B
      The efficacy of the Pipelle endometrial biopsy in detecting endometrial carcinoma.
      Several reasons could explain this difference, including variations in sampling technique.
      One other study showed an 83% sensitivity for the diagnosis of endometrial cancer when present. This study had 11 false-negative results; however, in 5 of these patients, the tumor was confined to a polyp, and in 3, the tumor filled less than 5% of the endometrial surface. Thus, these were very early-stage lesions.
      In a study in which follow-up biopsies were performed at least 2 years after the initial biopsy, 2% of patients had cancer detected on the repeated biopsy. Half of these patients were in the group that had “insufficient tissue” obtained at the initial biopsy, and half were from the group that had a “benign tissue” diagnosis.
      • Chambers JT
      • Chambers SK
      Endometrial sampling: When?.
      Thus, if bleeding continues after a biopsy shows normal results, resampling or assessment by another method is reasonable. The frequency of rebiopsy depends on the clinical situation. Endometrial sampling methods are least sensitive with early lesions (which should continue to expand and bleed) and with polyps (which will continue to bleed).
      An endometrial biopsy is not possible if the patient has cervical stenosis, is pregnant, or has acute pelvic inflammatory disease. An endometrial biopsy should be performed with caution in a woman with a severe bleeding disorder. Subacute bacterial endocarditis prophylaxis is unnecessary, although no studies have specifically analyzed the risks. Because subacute bacterial endocarditis prophylaxis is not recommended for dilation and curettage,
      • Durack DT
      Prevention of infective endocarditis.
      it seems logical that such prophylaxis would be unnecessary for an endometrial biopsy, a less traumatic procedure.
      Another advantage of endometrial sampling is its convenience. No special expertise is needed, and the procedure can be performed quickly. In addition, it has no substantial equipment costs.
      • Grimes DA
      Diagnostic dilation and curettage: a reappraisal.
      The cost of endometrial biopsy, including the pathologist's fee, varies by institution. The sampling devices are inexpensive, physician time to perform the procedure is less than 10 minutes, and tissue sampling is straightforward. The charge is usually between $125 and $300.
      • Grimes DA
      Diagnostic dilation and curettage: a reappraisal.
      In summary, endometrial sampling is a quick, convenient, inexpensive, and accurate method of assessing postmeno-pausal bleeding. It should be used by those who manage menopausal patients.
      Ultrasonography.—Transvaginal ultrasonography is an accurate method for measuring endometrial thickness because the transducer is close to the target area—the en-dometrium. A strong correlation exists between increased endometrial thickness and pathologic involvement, especially in light of abnormal postmenopausal bleeding.
      On the basis of ultrasonography, the thickness of the endometrium depends on the hormonal milieu. The endometrial stripe is an area of differing echogenicity in the middle of the uterine shadow, examined lengthwise. The endometrium is measured from the outer edge of the endometrial stripe to the opposite outer edge. The accuracy of the measurements is within 2 mm; thus, reproducibility is achieved.
      • Parsons AK
      Detection and surveillance of endometrial hyperplasia/carcinoma.
      Atrophie endometrium has a thickness of about 3 mm. The stripe varies between 4 and 8 mm in women who are receiving cyclic HRT and is about 5 mm if they are receiving combined HRT. Reports of carcinomas are extremely rare when the stripe is 4 mm or less. One study found one early-stage carcinoma and one complex hyperplasia in stripes of 5 mm.
      • Varner RE
      • Sparks JM
      • Cameron CD
      • Roberts LL
      • Soong SJ
      Transvaginal sonography of the endometrium in postmenopausal women.
      Thus, the range of thickness considered to be “free” of cancer is from 4 to 6 mm, depending on the investigator of the study. Most consider 4 mm to be the cutoff point indicating that no cancer is present.
      One study found that the mean endometrial thickness was 17 mm for the presence of carcinoma, 10 mm for hyperplasia, and 5 mm for control specimens.
      • Levine D
      • Gosink BB
      • Johnson LA
      Change in endometrial thickness in postmenopausal women undergoing hormone replacement therapy.
      The sensitivity of transvaginal ultrasonography varies from 81 to 100% (mean, 91%), and the specificity varies from 61 to 100% (mean, 84%). These diagnostic sensitivities compare favorably with those from studies using endometrial biopsies and dilation and curettage. Ultrasonography can provide reassurance if the endometrial stripe is 4 mm or less. Nevertheless, a narrow stripe can possibly contain an early pathologic change; thus, repeated scans may be necessary. If bleeding continues, even in the presence of a thin stripe, other diagnostic modalities may be necessary.
      The presence of fluid in the endometrial cavity has been associated with carcinoma of the endometrium; thus, its presence necessitates investigation. In most cases, however, this fluid has collected because of partial cervical stenosis that prevents the normal egress of endometrium. Biopsy is advisable if possible. If it is not possible and the endometrial stripe is 4 mm or less, a follow-up ultrasound study in 3 months is appropriate.
      Patients who are taking tamoxifen may have atypical ultrasonographic findings. In such women, a common finding is an endometrial stripe that appears to be extremely thick. This is due to a change in the myometrium under the endometrium that mimics the ultrasonographic appearance of thickened endometrium. This may lead to an unnecessary biopsy.
      • Bornstein J
      • Auslender R
      • Pascal B
      • Gutterman E
      • Isakov D
      • Abramovici H
      Diagnostic pitfalls of ultrasonographic uterine screening in women treated with tamoxifen.
      • Tsuda H
      • Kawabata M
      • Kawabata K
      • Yamamoto K
      • Hidaka A
      • Umesaki N
      Comparison between transabdominal and transvaginal ultrasonography for identifying endometrial malignancies.
      If bleeding occurs in a woman who is taking tamoxifen, however, a biopsy should be performed.
      The advantage of transvaginal ultrasonography is that it can be performed in almost every woman. An ultrasound machine and the expertise to analyze the images obtained are necessary. The charge for a scan is similar to that for an office endometrial biopsy. The time needed to obtain the measurement is also comparable.
      Measurements of the endometrial stripe with use of the abdominal approach are not as accurate as those obtained with the transvaginal approach.” If, however, the patient is unable to tolerate the vaginal probe, the abdominal approach can be used.
      Sonohysterography.—Sonohysterography is performed by instilling 10 mL of sterile saline into the endometrial cavity through a Soules catheter or similar device while a transvaginal scan is being obtained. This technique separates the two walls of the endometrium and produces a sonolucent window through which endometrial thickness can be completely examined. Specifically, if endometrial polyps are the cause of the thickening, their presence becomes obvious as they float in the saline. They appear as projections in the fluid.
      Sonohysterography is also useful for the detection of submucous fibroids, which can create the appearance of a thickened endometrium until saline is infused. After saline infusion, the stripe covering the fibroid can be clearly delineated and measured. The fibroids can also be measured serially for growth. One study found sonohysterography to be superior to routine transvaginal scanning and hysteroscopy for assessing size, site, and growth of fibroids.
      • Goldstein S
      Sonohysterography as an office procedure.
      Furthermore, sonohysterography can detect focal areas of endometrial thickening, which raise the index of suspicion that an actual pathologic condition is present. Endometrial sampling is necessary in such cases. Sampling can be done with directed endometrial biopsy, hysteroscopy, or dilation and curettage.
      Hysteroscopy.—Two types of hysteroscopy are available. One uses a rigid scope with a fluid distending medium, and at least heavy sedation is necessary. The other type uses a flexible fiberoptic cable and carbon dioxide as distending medium. The former is used for diagnosis and for procedures such as endometrial ablation and resection of fibroids and uterine septa. The latter is much more suited for diagnosis because only simple biopsies and removal of polyps are possible. Flexible hysteroscopy can also be used in an office setting because no anesthesia or sedation is needed.
      Office hysteroscopy is ideal for direct visualization of the endometrial cavity. If polyps are suspected, hysteroscopy can easily be used to confirm the presence of a polyp and possibly even remove it. Suspicious endometrial areas can be detected, and directed endometrial biopsy can be done.
      Office hysteroscopes are available in several diameters; the most common are 3 and 4 mm. The 4-mm scopes have an operating port through which biopsy forceps or scissors can be advanced. Antiseptic solution is used to prepare the patient, and a cotton swab dipped in lidocaine is put into the cervical canal. The scope is gently inserted into the uterine cavity, and carbon dioxide is used as a distending medium. Usually, countertraction on the cervix is necessary to help guide the scope through the canal. Inspection of the cavity can be done quickly and usually does not cause much, if any, patient discomfort. Biopsy specimens can be obtained from suspicious areas, either under direct vision or with an endometrial biopsy cannula guided by knowledge of the site of the lesion.
      The chief advantage of hysteroscopy is direct visualization of the pathologic tissue; thus, an appropriate biopsy specimen can be obtained. Hysteroscopy, especially operative hysteroscopy, is the “gold standard” against which all other methods of endometrial assessment are compared. It has supplanted dilation and curettage in this respect because of the ability to visualize the biopsy site.
      • Altaras MM
      • Aviram R
      • Cohen I
      • Markov S
      • Goldberg GL
      • Beyth Y
      Microhysteroscopy and endometrial biopsy results following failed diagnostic dilatation and curettage in women with postmenopausal bleeding.
      A major disadvantage of office hysteroscopy is that it cannot be used in patients with cervical stenosis—these are the same patients who cannot undergo biopsy. It also cannot be used in patients who are actively bleeding, which impairs the ability to visualize the cavity. Office hysteroscopy can be moderately expensive because of the cost of the equipment. Nonetheless, it is still less expensive than dilation and curettage.
      In summary, office hysteroscopy can serve as the final “arbiter” when neither biopsy nor ultrasound examination offers a suitable explanation for bleeding.
      Dilation and Curettage.—Dilation and curettage was long considered the “gold standard” for the diagnosis and treatment of postmenopausal bleeding but is no longer because of the “blind nature” of the procedure. It is now usually done in conjunction with hysteroscopy in order to visualize the endometrial cavity both before and after curettage, to assess the site of possible pathologic tissue, and to evaluate the effectiveness of curettage.
      Dilation and curettage is a relatively simple procedure to perform. Because the patient has undergone anesthesia, stenotic cervices can be dilated. It is also often therapeutic, and no additional bleeding episodes occur after the operation. These advantages are negligible, however, in light of the cost and inconvenience of the procedure. Thus, dilation and curettage should be reserved for those patients with a suspiciously thick endometrial stripe in which tissue cannot be obtained with any other modality or when the cause of bleeding cannot be determined with use of any of the less expensive methods. Even then, operative hysteroscopy is preferable if equipment and expertise are available. Operative hysteroscopy is substantially more accurate than dilation and curettage in the assessment of the endometrial cavity and any possible lesions.
      • Grimes DA
      Diagnostic dilation and curettage: a reappraisal.

      Algorithm

      An algorithm for diagnosing postmenopausal bleeding is presented in Figure 1. Endometrial biopsy is the first choice because of its convenience and lower cost. If ultrasonography is available, it is a suitable alternative in terms of expense and accuracy. If neither of these provides satisfactory answers, then hysteroscopy, especially operative hysteroscopy, is a reasonable next step. Dilation and curettage should be reserved for those patients in whom none of the other methods is possible.
      Figure thumbnail gr1
      Fig. 1Algorithm for assessment of postmenopausal bleeding. HRT = hormone replacement therapy; D & C = dilation and curettage.

      References

        • Archer DF
        • Mclntyre-Seltman K
        • Wilborn Jr, WW
        • Dowling EA
        • Cone F
        • Creasy GW
        • et al.
        Endometrial morphology in asymptomatic postmenopausal women.
        Am J Obstet Gynecol. 1991; 165: 317-320
        • Stovall TG
        • Photopulos GJ
        • Poston WM
        • Ling FW
        • Sandles LG
        Pipelle endometrial sampling in patients with known endometrial carcinoma.
        Obstet Gynecol. 1991; 77: 954-956
        • Ferry J
        • Farnsworth A
        • Webster M
        • Wren B
        The efficacy of the Pipelle endometrial biopsy in detecting endometrial carcinoma.
        Aust N Z J Obstet Gynaecol. 1993; 33: 76-78
        • Chambers JT
        • Chambers SK
        Endometrial sampling: When?.
        Where? Why? With what? Clin Obstet Gynecol. 1992 Mar; 35: 28-39
        • Durack DT
        Prevention of infective endocarditis.
        N Engl J Med. 1995; 332: 38-44
        • Grimes DA
        Diagnostic dilation and curettage: a reappraisal.
        Am J Obstet Gynecol. 1982; 142: 1-6
        • Parsons AK
        Detection and surveillance of endometrial hyperplasia/carcinoma.
        in: Lobo RA Treatment of the Postmenopausal Woman: Basic and Clinical Aspects. Raven Press, New York1994: 385-395
        • Varner RE
        • Sparks JM
        • Cameron CD
        • Roberts LL
        • Soong SJ
        Transvaginal sonography of the endometrium in postmenopausal women.
        Obstet Gynecol. 1991; 78: 195-199
        • Levine D
        • Gosink BB
        • Johnson LA
        Change in endometrial thickness in postmenopausal women undergoing hormone replacement therapy.
        Radiology. 1995; 197: 603-608
        • Bornstein J
        • Auslender R
        • Pascal B
        • Gutterman E
        • Isakov D
        • Abramovici H
        Diagnostic pitfalls of ultrasonographic uterine screening in women treated with tamoxifen.
        J Reprod Med. 1994; 39: 674-678
        • Tsuda H
        • Kawabata M
        • Kawabata K
        • Yamamoto K
        • Hidaka A
        • Umesaki N
        Comparison between transabdominal and transvaginal ultrasonography for identifying endometrial malignancies.
        Gynecol Obstet Invest. 1995; 40: 271-273
        • Goldstein S
        Sonohysterography as an office procedure.
        Contemp Ob/Gyn. 1995 Apr 15; : 9-17
        • Altaras MM
        • Aviram R
        • Cohen I
        • Markov S
        • Goldberg GL
        • Beyth Y
        Microhysteroscopy and endometrial biopsy results following failed diagnostic dilatation and curettage in women with postmenopausal bleeding.
        Int J Gynaecol Obstet. 1993; 42: 255-260