Applied Radiology

Dr. Salkowski is an Assistant Professor of Diagnostic Radiology in the Departments of Radiology and Obstetrics & Gynecology at the Medical College of Wisconsin in Milwaukee, WI.

T he technique of hysterosonography (HSG) was introduced in 1987, and became used clinically in the early 1990s. Hysterosonography is the technique whereby sterile saline is instilled into the uterine cavity, thus allowing improved transvaginal ultrasound (TVUS) imaging of the uterus and endometrium. The clinical applications of HSG are for the evaluation of postmenopausal bleeding, further evaluation of an abnormal endometrial interface seen on transvaginal sonography, and to assess possible causes of infertility.

Background

Traditionally, in patients requiring evaluation of the uterine cavity, there had been two mainstay methods. It could be evaluated indirectly by hysterosalpingography with the use of radiographic contrast and x-rays, or by direct visualization with hysteroscopy.

In the late 1980s, transvaginal sonography began to be used clinically in addition to traditional transabdominal sonography in the evaluation of the female pelvis. The use of TVUS significantly improved the ability to examine the structures of the female pelvis. Transvaginal ultrasound became the imaging method of choice for evaluating endometrial abnormalities, particularly in patients with abnormal vaginal bleeding. It is a highly sensitive technique for detecting small uterine masses and endometrial abnormalities, but it is nonspecific. ¹ The disadvantage of TVUS, however, is that it cannot reliably distinguish between lesions of myometrial origin located centrally in the uterus and those of endometrial origin. Endometrial polyps are more evident during the proliferative phase of the menstrual cycle by transvaginal sonography because of the inherent physiologic changes of the endometrium that can be seen on ultrasound, but can be easily shown by HSG at any stage of the menstrual cycle. ²

Hysterosonography technique

Initially a brief patient history is obtained, regarding symptoms, duration of symptoms, and any history suggesting active pelvic infection. Also, it is important to inquire if the patient has had a recent endometrial biopsy. If the patient is premenopausal, she should also have a negative pregnancy test.

A baseline transvaginal ultrasound is performed to assess the uterus/endometrium and adnexal structures. Using aseptic technique, a sterile speculum is placed in the vagina with direct visualization of the cervix. The cervix is cleansed with betadine. The sterile 5-French hysterosonography catheter (Lyco HSG Catheter; Model 06-105F) is then placed into the uterine cavity by direct visualization of the external cervical os. The speculum is removed carefully, leaving the catheter in place, and the 5-7 MHz multifrequency transvaginal ultrasound transducer is inserted into the vagina. Under ultrasound visualization, the balloon of the catheter is inflated with saline. The balloon is then drawn gently toward the internal os to anchor it in place and also to avoid the instilled saline from draining out through the cervix during the exam. Sterile saline is then instilled through the catheter to distend the uterine cavity during TVUS imaging. Approximately 10 cc of saline is usually sufficient to achieve adequate uterine distention. Sagittal and coronal imaging of the uterus is performed. It is common to see fluid in the cul-de-sac after the procedure. This infers patency of at least one fallopian tube.

Normal HSG

In the normal exam, the uterine cavity is distended symmetrically. The endo-metrium appears symmetric and uniformly thick. On conventional TVUS, a uniform endometrial thickness of the two coapted layers of <5 mm is considered normal in the postmenopausal woman. ^1,3-5 The measurement of the endometrial thickness on HSG is obtained by adding the thickness of the opposite endometrial linings. Note that the total added thickness should not exceed 5 mm. No masses distort the uterine cavity or are present within the uterine cavity (figure 1).

Abnormal HSG

Several abnormalities can be detected by HSG; the most commonly found include adhesions, polyps, leiomyomas, endometrial tumors, and tamoxifen-related changes.

Adhesions --By conventional TVUS, the endometrial lining may appear normal. Adhesions can be one of the etiologies of infertility. On HSG, bridging bands of tissue can be seen which distort the uterine cavity. Thick, broad bands may make it difficult to distend the uterine cavity (figures 2 and 3).

Endometrial polyps-- Patients with uterine polyps can present with unexplained infertility and abnormal vaginal bleeding. On conventional TVUS, a thickened endometrium can often be seen. On HSG, an intracavitary polyp outlined by the anechoic saline can be demonstrated (figure 4). The endometrial lining is normal. Often the point of attachment, as well as the thickness of the stalk, can be identified on HSG. Color Doppler can be useful in helping to identify the stalk of the polyp. Usually, polyps are seen as homogenous, smooth hyperechoic masses with well-defined margins. They may be sessile or pedunculated, solitary or multiple. Occasionally, small cystic spaces are visible within the polyps. Polyps are associated with a three-fold increase in endometrial cancer. ⁶

Uterine leiomyomas --Leiomyomas are common gynecological abnormalities. They can be clinically silent, or can present with abnormal vaginal bleeding, pain, or unexplained infertility. Conventional TVUS can be used to identify the leiomyoma, its relationship with the endometrial interface, and any distortion of the myometrial/endometrial interface. HSG can be helpful for further localization of the leiomyoma, whether it is submucosal or intramural (figure 5). This information can affect the patient's management and also help to differentiate if the fibroid is the etiology of the abnormal vaginal bleeding.

Intramural leiomyomas are located within the myometrium and do not distort the endometrial cavity. Submucosal leiomyomas often distort the endometrial cavity and have little or no intervening myometrium between the fibroid and endometrium. There is a normal layer of endometrium present over the fibroid (figure 6).

Endometrial abnormalities --Basically, endometrial abnormalities can occur in two groups of patients, perimenopausal and postmenopausal women with bleeding. The major concern with the perimenopausal patient with unexplained vaginal bleeding is endometrial carcinoma; however, the most common etiology is atrophic endometrium. Therefore, demonstrating an endometrial stripe by ultrasound of <5 mm would be very important in differentiating these two groups of patients and their subsequent treatment.

The majority of postmenopausal women with bleeding will have benign etiologies for the bleeding. Exogenous estrogen is a cause in 30% of patients; atrophic endometritis/vaginitis in 30%; endometrial hyperplasia in 10%; endometrial polyps in 10%; submucosal fibromyomas in 10%; and endometrial carcinoma in approximately 10% of these patients. ^3,4 Approximately 75% of all cases of endometrial carcinoma occur in postmenopausal women. ⁷ Because this cancer often presents with abnormal vaginal bleeding, most cases are detected at an early stage when they are highly curable. Early detection can offer a 5-year survival rate of 83%. ⁵ These cancers also tend to be early-stage, low-grade, minimally invasive lesions that have a favorable prognosis. ⁷ Patients who are taking unopposed estrogen therapy have a four- to eight-fold increase in risk of developing endometrial carcinoma. This risk can be nearly eliminated by combining progesterone with estrogen therapy.

On HSG, endometrial carcinoma can appear as an irregularly thickened endometrium with variable echogenicity, with an endometrial-myometrial junction that may or may not be intact. ¹ An endometrial thickness >5 mm in a postmenopausal woman should be further evaluated for the possibility of endometrial carcinoma.

Tamoxifen-related changes --Tamoxifen is widely used as part of the treatment of breast cancer. It has an antiestrogenic effect on the breast and a weak estrogenic effect on the uterus. An increase in endometrial cancer has been reported with the use of tamoxifen as a finding of the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14 trial. ⁷ The final conclusion with regard to this patient population was that the benefit of tamoxifen treatment for breast cancer outweighs the potential increase in endometrial cancer.

Other findings that can be seen in the endometrium of patients taking tamoxifen include polypoid lesions (figure 7), endometrial hyperplasia, and subendometrial cystic lesions. ^8,9 These microcystic changes are located in the proximal myometrium while the endo-metrium remains thin and atrophic. These changes are thought to represent reactivation of foci of adenomyosis due to the estrogen effects of tamoxifen. ¹⁰

Endometrial biopsy vs. HSG in the postmenopausal woman

Hysteroscopy is an accurate method for the diagnosis and treatment of endometrial abnormalities. However, not every patient with postmenopausal bleeding will require such an invasive and high-cost intervention for diagnosis and treatment. The cost of performing and interpreting the results of a blind endometrial biopsy is comparable to the cost incurred for performing a TVUS. Based on the findings of the TVUS and the finding of thickened endometrial lining, the additional cost of performing a hysterosonogram justifies the added information it provides and thus results in decreased total patient care cost.

According to a prospective double-blind study of 47 postmenopausal women comparing HSG with hysteroscopy, it was found that HSG was 86% sensitive and 100% specific in the diagnosis of endometrial pathology. ² An endometrial thickness >5 mm is a sensitive finding on TVUS for the presence of endometrial abnormalities. Sensitivities as high as 97% have been reported in detecting endometrial cancer in patients with endometrial thicknesses >5 mm. ²

Variable results in sensitivity for detecting endometrial tumors with blind endometrial biopsies range from 33% to 85%. ³ Endometrial biopsy samples only a small portion of the endometrium, and the endometrial lesion may be easily missed when performed blindly. Therefore, endometrial biopsy alone is not sufficient to evaluate the postmenopausal patient with bleeding. The added information attained with TVUS increases the sensitivity of detection of endometrial abnormalities. Which, if found, could be further evaluated with HSG and then be used as a guide for biopsy by hysteroscopy.

Clinical indications for HSG

Hysterosonography can be useful in several clinical situations. It is effective in the postmenopausal woman with vaginal bleeding who is found to have an abnormal endometrial thickness. It can be used as a screening tool to evaluate for endometrial changes related to tamoxifen. Uterine malformations as a possible etiology of infertility can be easily evaluated with HSG. The uterine cavity can also be assessed after myomectomy, metroplasty, and lysis of synechiae with HSG.

Absolute contraindications to performing HSG are pregnancy, stenotic endocervix, active pelvic inflammatory disease, and recent pelvic surgery and/or biopsy. The major risks are infection and perforation. There also are a few limitations of HSG. If the patient has cervical stenosis and the cervix can not be cannulated, the procedure can not be performed. Inadequate distention of the uterine cavity can occur under several conditions. Primarily, this occurs in patients with a patulous cervix, but can also occur if fluid easily leaks out of the fallopian tubes. Bilateral tubal occlusion may result in difficult and painful distension of the endometrial cavity. DeVore also described the theoretical risk of spread of infection and dissemination of endometrial carcinoma by this technique. ¹² However, this is unlikely as such a small amount of fluid is instilled and it is instilled under low pressure. ^4,10-12 Finally, it is sometimes difficult to visualize abnormalities just above the internal os. The balloon of the catheter may obscure these lesions. To help with visualization of this region, saline is instilled through the catheter while watching with ultrasound as the catheter is being withdrawn.

Conclusion

Hysterosonography is well tolerated by patients. It does not require anesthesia or analgesia. It can be performed in an outpatient setting and does not incur the expense of the operating room. It allows improved depiction of the endometrial cavity without the use of ionizing radiation or iodinated contrast. It allows differentiation of intracavitary, endometrial, and submucosal abnormalities with a greater degree of confidence. AR

References

1. Lev-Toaff AS: Sonohysterography: Evaluations of endometrial and myometrial abnormalities. Sem Roent 31:288-298, 1996.

2. Wolman I, Jaffa AJ, Hartoov J, et al: Sensitivity and specificity of sonohysterography for the evaluation of the uterine cavity in perimenopausal patients. J Ultrasound Med 15:285-288, 1996.

3. Dubinsky TJ, Parvey HR, Makland N: The role of transvaginal sonography and endometrial biopsy in the evaluation of peri- and postmenopausal bleeding. AJR 169:145-149, 1997.

4. Dubinsky TJ, Parvey HR, Gormaz G, et al: Transvaginal hysterosonography: Comparison with biopsy in the evaluation of postmenopausal bleeding. J Ultrasound Med 14:887-893, 1995.

5. Kasakes CJ, Meilstrp JW, Harris RD: Common imaging problems in the postmenopausal pelvis: The endometrial stripe, interuterine fluid, and adnexal cyst. Sem Roent 41:279-287, 1996.

6. Pettersson B, Adami HO, Lindgren A, et al: Endometrial polyps and hyperplasia as risk factors for endometrial cancer: A case control study of curettage specimens. Acta Obstet Gynaecol Scand 64:653-659, 1985.

7. Brakat RR: Contemporary issues in the management of endometrial cancer. CA Cancer J Clin 48:299-314, 1998.

8. Anteby EY, Yagel S, Weissman A, et al: Sonographic evaluation of the uterus in postmenopausal women receiving tamoxifen: Characterization of mid-uterine abnormalities. Eur J Ob Gyn Rep Bio 69:115-119, 1996.

9. Achiron R, Lipitz S, Sivan E, et al: Sonohysterography for ultrasonographic evaluation of tamoxifen-associated cystic thickened endometrium. J Ultrasound Med 14:685-688, 1995.

10. Goldstein SR: Saline infusion sonohysterography for the patient with abnormal bleeding. Appl Radiol 26(10):33-36, 1997.

11. Dubinsky TJ, Parvey HR, Gormaz G, Makland N: Transvaginal ultrasonography in the evaluation of small endoluminal masses. J Ultrasound Med 14:1-6, 1995.

12. DeVore GR, Schwartz PE, Morris J: Hysterography: A 5-year follow-up in patients with endometrial carcinoma. Obstet Gynecol 60:369-372, 1982.