Sonohysterography Compared With Endometrial Biopsy for Evaluation of the Endometrium in Tamoxifen-Treated Women

Article Sonohysterography Compared With Endometrial Biopsy for Evaluation of the Endometrium in Tamoxifen-Treated Women Lucy E. Hann, MD, Cathleen M. Kim, MD, Mithat Gonen, PhD, Richard Barakat, MD, Patricia H. Choi, MD, Ariadne M. Bach, MD Objective. This study was performed to compare endometrial biopsy and sonohysterography for evaluation of the endometrium in tamoxifen-treated women. Methods. Medical records were retrospectively reviewed to identify 51 consecutive tamoxifen-treated women who had sonohysterography and correlative endometrial biopsy for evaluation of postmenopausal bleeding or thickened endometrium of greater than 8 mm. Endometrial biopsy and sonohysterographic results were compared in all women, and for 27 (53%) women who had hysteroscopy with dilation and curettage, endometrial biopsy and sonohysterographic findings also were compared with surgical pathologic findings. Results. Thirty-two (63%) of 51 sonohysterograms revealed endometrial polyps; 4 (8%) showed endometrium of greater than 5 mm; 14 (27%) showed endometrium of less than 5 mm; and 1 (2%) was inadequate. Endometrial biopsy findings were benign endometrium in 42 (82%), polyps in 4 (8%), and insufficient samples in 5 (10%). Among the adequate sonohysterograms, 64% (32 of 50) resulted in a diagnosis of polyps (95% confidence interval, 49% 77%) whereas the corresponding proportion for endometrial biopsy was 9% (4 of 46; 95% confidence interval, 2% 21%). For the group with hysteroscopy, 24 (92%) of 26 polyps were confirmed histopathologically; 1 polyp had complex hyperplasia. Polyps were present in 23 (89%) of 26 women with benign endometrium or insufficient samples by endometrial biopsy, and only 1 confirmed polyp was identified by endometrial biopsy. The sensitivity of sonohysterography for diagnosis of endometrial polyps (100%) was significantly higher than for endometrial biopsy (4%; P <.01). Conclusions. In tamoxifen-treated women, sonohysterography provides a significant improvement in sensitivity for diagnosis of endometrial polyps compared with endometrial biopsy. Key words: endometrial biopsy; endometrial polyp; sonohysterography; tamoxifen. Received April 19, 2003, from the Departments of Radiology (L.E.H., C.M.K., P.H.C., A.M.B.), Epidemiology and Biostatistics (M.G.), and Surgery (R.B.), Memorial Sloan-Kettering Cancer Center, New York, New York USA. Revision requested May 19, 2003. Revised manuscript accepted for publication May 27, 2003. Address correspondence and reprint requests to Lucy E. Hann, MD, Department of Radiology, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY 10021 USA. E-mail: hannl@mskcc.org. The benefits of tamoxifen as adjuvant therapy for reduction of recurrence and mortality in women with estrogen-positive breast cancer are well established. 1,2 In breast cancer prevention trials for high-risk women, tamoxifen also has been shown to reduce the rate of development of breast cancer by 50% compared with control subjects. 3 Because tamoxifen has proven efficacy for both adjuvant treatment and prevention of breast cancer, it is widely used, and approximately 80,000 of the 182,000 women diagnosed yearly with breast cancer will receive tamoxifen therapy. 4 2003 by the American Institute of Ultrasound in Medicine J Ultrasound Med 22:1173 1179, 2003 0278-4297/03/$3.50

Sonohysterography of the Endometrium in Tamoxifen-Treated Women Tamoxifen treatment is also associated with endometrial abnormalities because the drug has weakly estrogenic effects. Endometrial abnormalities such as polyps, carcinoma, and hyperplasia are reported to occur in approximately 40% to 50% of tamoxifen-treated women. 1,2,5 13 Most tamoxifen-treated women have benign endometrial abnormalities such as polyps, estimated to occur in 23% to 62% of tamoxifen-treated women. 5,7,9,11,12,14 Endometrial carcinoma is of concern, but it occurs in only 2 per 1000 tamoxifen-treated women, and most of these women have vaginal bleeding. 10,15 Because endometrial carcinoma is so infrequent, routine screening of the endometrium in tamoxifen-treated women has been reported to be unnecessary in the absence of abnormal bleeding. 15 Many tamoxifen-treated women still require endometrial evaluation for abnormal uterine bleeding or a thickened endometrium on transvaginal sonography because these clinical findings are known independent predictors of endometrial disease. 16 Both endometrial biopsy and sonohysterography are commonly performed to evaluate the endometrium in tamoxifen-treated women, but the optimal method has not been determined. 5,11,14,17 26 Endometrial biopsy has reported sensitivities as high as 85% for identification of endometrial hyperplasia and carcinoma, but it is reportedly less sensitive for detection of endometrial polyps, the most frequent abnormality in tamoxifen-treated women. 14,21,26 We performed this study to compare endometrial biopsy and sonohysterography for diagnosis of endometrial abnormalities in tamoxifen-treated women. Materials and Methods Sonographic reports and medical records of all women who underwent sonohysterography from March 1997 to April 2002 were reviewed retrospectively to identify tamoxifen-treated women who had endometrial biopsy and saline infusion sonohysterography for evaluation of abnormal uterine bleeding or thickened endometrium of greater than 8 mm. The study was approved by the Institutional Review Board. Of 259 sonohysterographic examinations performed during this study period, 113 (44%) were in tamoxifen-treated women, and 51 (45%) of these studies had correlative endometrial biopsy within a mean of 1.6 months (range, 0 9 months). Sonohysterographic findings were compared with endometrial biopsy results, and for the subset of patients who had hysteroscopy with dilation and curettage, endometrial biopsy and sonohysterographic findings also were compared with surgical pathologic findings, which were considered the standard of reference. Fifty-one tamoxifen-treated women met study inclusion criteria. Their mean age was 58 years (range, 39 81 years). Forty-five women were postmenopausal, and 6 women were premenopausal. Thirty women (59%) were referred for postmenopausal bleeding. Transvaginal sonograms obtained before sonohysterography revealed endometrial thickness of less than 5 mm in 8 women, 5 to 8 mm in 8 women, and greater than 8 mm in 38 women (75%). Current tamoxifen therapy was confirmed in all patients, but the duration of tamoxifen treatment was not considered. Sonohysterography was performed with an HDI 5000 system (Philips Medical Systems, Bothell, WA) or an Acuson Sequoia system (Siemens Medical Solutions, Mountain View, CA) and endovaginal 4- to 10-MHz transducers. Informed consent was obtained in all cases. The sonohysterographic technique was as described previously. 14 A 5F Ackrad H/S catheter (Ackrad Laboratories, Cranford, NJ) or 5F Uni- Sem catheter (Cooper Surgical, Inc, Trumbull, CT) was used for uterine cannulation. An experienced radiologist (L.E.H., A.M.B., or P.H.C.) performed the study or directly supervised a radiology fellow during the procedure. Sonohysterographic reports were reviewed retrospectively by 2 radiologists (C.M.K. and L.E.H.), who recorded endometrial thickness, endometrial polyps, and any other endometrial abnormalities that were described in the original reports. Subendometrial cysts and submucous leiomyomas were also noted, but only the endometrial findings were considered for analysis. Sonohysterographic findings were then compared with histopathologic findings from correlative endometrial biopsy. The presence of endometrial polyps, hyperplasia, carcinoma, or insufficient tissue from endometrial biopsy was recorded. Sonohysterographic and endometrial biopsy results also were correlated with surgical histopathologic findings in 27 (53%) of 51 tamoxifen-treated women who had hysteroscopy with dilation and curettage within a mean of 1.7 months (range, 0.2 10.6 months). 1174 J Ultrasound Med 22:1173 1179, 2003

Hann et al The proportion of polyps detected by sonohysterography and endometrial biopsy was determined. The sensitivities of sonohysterography and endometrial biopsy were calculated with the use of pathologic findings as the standard of reference, ignoring the verification bias resulting from patients who did not have pathologic analysis, and compared by the McNemar test. All confidence intervals and P values were computed by exact methods. Results Fifty sonohysterographic examinations were completed, and 1 was unsuccessful. Thirty-two (63%) of 51 sonohysterograms revealed endometrial polyps; 4 (8%) showed thickened endometrium (>5 mm); and 14 (27%) showed normal endometrial thickness (<5 mm; Table 1). Fifteen (29%) of 51 sonohysterograms revealed subendometrial cysts; 8 of these cases had coexistent polyps. Submucous leiomyomas were present in 2 cases with normal endometrial thickness. Endometrial polyps detected by sonohysterography had a mean maximal dimension of 16 mm (range, 3 40 mm). Seven patients had more than 1 polyp; the other polyps were solitary. Pathologic analysis of the endometrial biopsy specimens revealed benign endometrium in 42 women (82%) and polyps in 4 (8%), and 5 women (10%) had insufficient samples. Correlation of Sonohysterography With Endometrial Biopsy Results of sonohysterography compared with endometrial biopsy findings are shown in Table 1. There was 1 incomplete sonohysterogram, and 5 endometrial biopsies were insufficient. Among the adequate sonohysterograms 64% (32 of 50) revealed polyps (95% confidence interval, 49% 77%), whereas the corresponding proportion for endometrial biopsy was 9% (4 of 46; 95% confidence interval, 2% 21%). In the 32 women with endometrial polyps on sonohysterography, endometrial biopsy revealed benign endometrium in 26 (81%), insufficient samples in 4 (13%), and polyps in only 2 (6%). Four patients with thickened endometrium of greater than 5 mm on sonohysterography had normal endometrium on endometrial biopsy. Two of the 4 patients with polyps on endometrial biopsy had concordant sonohysterographic findings. The remaining 2 patients had normal endometrium on sonohysterography. Table 1. Sonohysterographic Findings Compared With Endometrial Biopsy Findings Endometrial Biopsy Findings Sonohysterographic Benign Findings Insufficient* Polyp* Endometrium* Total Inadequate study 0 (0) 0 (0) 1 (100) 1 (2) Polyp 4 (13) 2 (6) 26 (81) 32 (63) Thickening 0 (0) 0 (0) 4 (100) 4 (8) Normal endometrium 1 (7) 2 (14) 11 (79) 14 (27) Total 5 (10) 4 (8) 42 (82) 51 (100) *Numbers in parentheses are percentages for row totals. Numbers in parentheses are percentages for column total. Correlation of Sonohysterography and Endometrial Biopsy With Surgical Pathologic Findings From Hysteroscopy With Dilation and Curettage Twenty-seven patients (53%) had surgical pathologic correlation from hysteroscopy with dilation and curettage. Results of sonohysterography and endometrial biopsy are compared with surgical pathologic findings from hysteroscopy with dilation and curettage in Table 2. Twenty-four (92%) of 26 polyps revealed by sonohysterography were confirmed histopathologically (Fig. 1), and 2 (8%) had normal endometrium (Fig. 2). One polyp contained complex endometrial hyperplasia without cytologic atypia (Fig. 3). Histopathologic analysis from subsequent hysterectomy revealed no evidence of malignancy. There were no other cases of endometrial hyperplasia or endometrial carcinoma. Table 2. Sonohysterographic and Endometrial Biopsy Findings Compared With Surgical Pathologic Findings From Hysteroscopy With Dilation and Curettage Surgical Pathologic Findings Imaging/Biopsy Normal Polyp With Findings Endometrium* Polyp* Hyperplasia* Total Sonohysterography Unsatisfactory 1 (100) 0 (0) 0 (0) 1 (4) Polyp 2 (8) 23 (88) 1 (4) 26 (96) Normal endometrium 0 (0) 0 (100) 0 (0) 0 (0) Endometrial biopsy Insufficient 1 (33) 2 (67) 0 (0) 3 (11) Polyp 0 (0) 1 (100) 0 (0) 1 (4) Benign endometrium 2 (9) 20 (87) 1 (4) 23 (85) Total 3 (11) 23 (85) 1 (4) 27 (100) *Numbers in parentheses are percentages of row totals. Polyp had complex hyperplasia without cytologic atypia. Numbers in parentheses are percentages of column total. J Ultrasound Med 22:1173 1179, 2003 1175

Sonohysterography of the Endometrium in Tamoxifen-Treated Women The sensitivity (true-positive fraction) of endometrial biopsy was only 4% (1 of 24) with a confidence interval of 0% to 21%, whereas the sensitivity of sonohysterography was estimated to be 100% (24 of 24) with a confidence interval of 86% to 100%. These results suggest that the true sensitivity is likely to be less than 21% for endometrial biopsy and more than 86% for sonohysterography; hence, sonohysterography provides a significant improvement over endometrial biopsy with regard to sensitivity (P <.01, McNemar test). Specificities were not well estimated in this data set because there were only 3 patients who had normal pathologic findings. Figure 1. Sonohysterogram showing an endometrial polyp (arrow) in a 46-year-old tamoxifen-treated woman who had postmenopausal bleeding. Endometrial biopsy 3.5 months before sonohysterography yielded normal findings. Hysteroscopy confirmed an endometrial polyp. Endometrial biopsy detected only 1 (4%) of 24 polyps confirmed by surgical pathologic analysis from hysteroscopy with dilation and curettage. Polyps were present on surgical pathologic specimens from dilation and curettage in 21 (91%) of 23 patients with benign endometrium on endometrial biopsy and in 2 (67%) of 3 with insufficient endometrial biopsy samples. Discussion Our findings confirm the high frequency of endometrial polyps in tamoxifen-treated women, but polyp detection rates differed significantly between sonography and endometrial biopsy. Polyps were evident in 63% of sonohysterograms but were found in only 8% of endometrial biopsies performed in the same patients. The high frequency of endometrial polyps revealed by sonohysterography supports prior studies. Hulka and Hall 8 found 9 endometrial polyps (81%) confirmed by surgical pathologic analysis in 11 patients treated with tamoxifen. Other investigators have reported polyps on 49% to 62% of sonohysterograms from tamoxifen-treated women. 11,14,25 A lower prevalence of polyps was Figure 2. Sonohysterograms from a 40-year-old tamoxifen-treated woman with postmenopausal bleeding. Sonohysterography revealed a submucous leiomyoma and polyps, but endometrial biopsy results were negative, and findings from hysteroscopy with dilation and curettage were normal, although the examination was limited. A, Longitudinal image showing a large submucous leiomyoma (m) that distorts the anterior endometrium (arrowheads). B, Transverse image showing several small endometrial polyps (open arrows). A B 1176 J Ultrasound Med 22:1173 1179, 2003

Hann et al reported by Deligdisch et al, 12 who retrospectively reviewed the histopathologic slides from 700 women treated with tamoxifen who had postmenopausal bleeding, abnormal sonographic findings, or both and who underwent hysterectomy (134) or endometrial biopsy (566); they revealed polyps in 23%. 12 Endometrial biopsy was not sensitive for the detection of endometrial polyps in this study. In the 32 women with polyps revealed by sonohysterography, endometrial biopsy showed benign endometrium in 26 (81%), insufficient samples in 4 (13%), and polyps in only 2 (6%). When endometrial biopsy was compared with surgical pathologic findings from hysteroscopy with dilation and curettage, 89% of patients with benign endometrium or insufficient samples on endometrial biopsy had endometrial polyps on surgical pathologic analysis. The sensitivity of endometrial biopsy for diagnosis of endometrial polyps was only 4% (95% confidence interval, 0% 21%) compared with 100% sensitivity of sonohysterography (confidence interval, 86% 100%; P <.01). Studies that rely on endometrial biopsy for detection of endometrial abnormalities underestimate the frequency of endometrial polyps. In a study using endometrial biopsy for screening of tamoxifen-treated women, Barakat et al 15 reported that 9 (1.4%) of 635 endometrial biopsies performed in 111 tamoxifen-treated women had abnormal findings. In that study, 14 patients (12.6%) had dilation and curettage with or without hysteroscopy, and polyps were detected in only 4. 15 Reasons for the disparity in the detection rates of endometrial polyps by the two methods include incomplete endometrial sampling during biopsy and the possibility that pedunculated masses may move away from the instrument. Limitations of endometrial biopsy have been reported by Dubinsky et al, 22 who found that the method did not reveal endometrial abnormalities causing postmenopausal bleeding in 45 (30%) of 148 women and that it missed 4 carcinomas. The importance of endometrial polyps is debated because most are benign. Endometrial polyps are a cause of postmenopausal bleeding and abnormal endometrial thickness on endovaginal sonograms. Endometrial polyps in tamoxifen-treated patients differ from polyps in the general population. 12 Tamoxifen-related polyps tend to be larger (mean diameter, 5 cm) and histologically are associated with metaplasia of the epithelium, cystic glandular dilatation, and periglandular stromal condensation. 5,27 Tamoxifen-associated polyps also are reported to have an increased rate of malignant changes. 27 29 Cohen et al 29 reported that 2 (3%) of 67 endometrial polyps from tamoxifen-treated women who had had postmenopausal breast cancer had malignant changes, compared with Figure 3. Sonohysterograms from a 70-year-old tamoxifen-treated woman with thickened endometrium and normal endometrial biopsy findings. Sonohysterography revealed a large endometrial polyp. Histopathologic analysis from hysteroscopy with dilation and curettage revealed an endometrial polyp with complex hyperplasia. A, Longitudinal image showing a large endometrial polyp (calipers) with shadowing. The fluid-filled catheter balloon (curved arrow) is in the lower uterine segment. B, Transverse image showing the polyp (arrows) with acoustic shadowing and multiple small cystic areas. A B J Ultrasound Med 22:1173 1179, 2003 1177

Sonohysterography of the Endometrium in Tamoxifen-Treated Women 5 (0.48%) of 1034 polyps from a control group. Ramondetta et al 28 had 15 patients with endometrial cancer who had been treated with tamoxifen; 5 patients had cancer in the endometrial polyps, and 4 of these 5 had cancer confined only to the polyps. Schlesinger et al 27 reported a 10.7% rate of malignant change in polyps of tamoxifen-treated patients. It is suggested that polyps found in tamoxifen-treated women should be removed surgically. One patient in our study had complex endometrial hyperplasia within a polyp that was revealed initially by sonohysterography but not by endometrial biopsy. The importance of endometrial hyperplasia was described well by Ascher et al. 5 Hyperplasia is divided into 2 categories: atypia or no cytologic atypia, and each group may be subcategorized into simple or complex. Progression to carcinoma is documented in 23% of patients with atypical hyperplasia compared with 2% of patients without atypia. 5 It is noteworthy that there were no cases of endometrial carcinoma or atypical hyperplasia in our study. This underscores the fact that most tamoxifen-related endometrial abnormalities are benign polyps, and the prevalence of carcinoma (0.002%) or hyperplasia (1.3% 20%) is extremely low in tamoxifen-treated women. 5 A potential limitation of this study was that the duration of tamoxifen therapy was not documented. Prior reports have shown that the risk of endometrial abnormalities, including cancer, is correlated with the duration of tamoxifen treatment. 6,10,12 Other study limitations were the small patient population, retrospective design, and lack of surgical pathologic correlation in all patients. There also was a relatively long period between correlative studies, but this factor was unlikely to affect the results because in all cases both endometrial biopsy and sonohysterography were performed for the same indications, and the abnormal bleeding, thickened endometrium shown on transvaginal sonography, or both persisted during the interval between procedures. In conclusion, we found that sonohysterography provided a significant improvement in sensitivity for diagnosis of endometrial polyps compared with endometrial biopsy. Because endometrial polyps are so frequent in tamoxifentreated women, sonohysterography should be considered for evaluation of abnormal uterine bleeding or thickened endometrium even if endometrial biopsy results are negative. References 1. Early Breast Cancer Trialists Collaborative Group. Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy: 133 randomized trials involving 31,000 recurrences and 24,000 deaths among 75,000 women. Lancet 1992; 339: 1 15, 71 85. 2. Osborne CK. Drug therapy: tamoxifen in the treatment of breast cancer. N Engl J Med 1998; 339: 1609 1618. 3. Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 study. J Natl Cancer Inst 1998; 90:1371 1388. 4. Barakat RR. Screening for endometrial cancer in the patient receiving tamoxifen for breast cancer. J Clin Oncol 1999; 17:1967 1968. 5. Ascher SM, Imaoka I, Lage JM. Tamoxifen induced uterine abnormalities: the role of imaging. Radiology 2000; 214:29 38 6. Bergman L, Beelen M, Gallee M, Hollema H, et al. Risk and prognosis of endometrial cancer after tamoxifen for breast cancer. Lancet 2000; 356: 881 887. 7. Berliere M, Charles A, Galant C, Donnez J. Uterine side effects of tamoxifen: a need for systematic pretreatment screening. Obstet Gynecol 1998; 91:40 44. 8. Hulka CA, Hall DA. Endometrial abnormalities associated with tamoxifen therapy for breast cancer: sonographic and pathologic correlation. AJR Am J Roentgenol 1993; 160:809 812. 9. Kedar R, Bourne TH, Collins WP, et al. Effects of tamoxifen on uterus and ovaries of postmenopausal women in a randomised breast cancer prevention trial. Lancet 1994; 343:1318 1321. 10. Fisher B, Costantino JP, Redmond CK, Fisher ER, Wickerham DL, Cronin WM. Endometrial cancer in tamoxifen-treated breast cancer patients: findings from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B14. J Natl Cancer Inst 1994; 86:527 537. 11. Fong K, Kung R, Lytwyn A, et al. Endometrial evaluation with transvaginal US and hysterosonography in asymptomatic postmenopausal women with breast cancer receiving tamoxifen. Radiology 2001; 220:765 773. 1178 J Ultrasound Med 22:1173 1179, 2003

Hann et al 12. Deligdisch L, Kalir T, Cohen C, de Latour M, Le Bouedec G, Penault-Llorca F. Endometrial histopathology in 700 patients treated with tamoxifen for breast cancer. Gynecol Oncol 2000; 78: 181 186. 13. Hann L, Geiss C, Bach A, Tao Y, Baum H, Barakat R. Endometrial thickness in tamoxifen-treated patients: correlation with clinical and pathologic findings. AJR Am J Roentgenol 1997; 168:657 661. 14. Hann L, Gretz E, Bach A, Francis S. Sonohysterography for evaluation of the endometrium in women treated with tamoxifen. AJR Am J Roentgenol 2001; 177:337 342. 15. Barakat RR, Gilewski TA, Almadrones L, et al. Effect of adjuvant tamoxifen on the endometrium in women with breast cancer: a prospective study using office endometrial biopsy. J Clin Oncol 2000; 18:3459 3463. 16. Franchi M, Ghezzi F, Donadello N, Zanaboni F, Beretta P, Bolis P. Endometrial thickness in tamoxifen-treated patients: an independent predictor of endometrial disease. Obstet Gynecol 199; 93:1004 1008. 17. Schwartz LB, Snyder J, Horan C, Porges RF, Nachtigall LE, Goldstein SR. The use of transvaginal ultrasound and saline infusion sonohysterography for the evaluation or asymptomatic postmenopausal breast cancer patients on tamoxifen. Ultrasound Obstet Gynecol 1998; 11:48 53. 18. Laifer-Narin SL, Ragavendra N, Lu DS, Sayre J, Perrella RR, Grant EG. Transvaginal saline hysterosonography: characteristics distinguishing malignant and various benign conditions. AJR Am J Roentgenol 1999; 172:1513 1520. 19. Achiron R, Lipitz S, Sivan E, Goldenberg M, Mashiach S. Sonohysterography for ultrasonographic evaluation of tamoxifen-associated cystic thickened endometrium. J Ultrasound Med 1995; 14: 685 688. 22. Dubinsky T, Stroehlein K, Abu-Ghazzeh Y, Parvey HR, Maklad N. Prediction of benign and malignant endometrial disease: hysterosonographic-pathologic correlation. Radiology 1999; 210:393 397. 23. Tepper R, Beyth Y, Altaras M, et al. Value of sonohysterography in asymptomatic postmenopausal tamoxifen-treated patients. Gynecol Oncol 1997; 64:386 391. 24. Cohen I, Rosen D, Tepper R, et al. Ultrasonographic evaluation of the endometrium and correlation with endometrial sampling in postmenopausal patients treated with tamoxifen. J Ultrasound Med 1993; 12:275 280. 25. Timmerman D, Deprest J, Bourne T, Van den Berghe I, Collins WP, Vergote I. A randomized trial on the use of ultrasonography or office hysteroscopy for endometrial assessment in postmenopausal patients with breast cancer who were treated with tamoxifen. Am J Obstet Gynecol 1998; 179:62 70. 26. Shipley CF III, Simmons CL, Nelson GH. Comparison of transvaginal sonography with endometrial biopsy in asymptomatic postmenopausal women. J Ultrasound Med 1994; 13:99 104. 27. Schlesinger C, Kamoi S, Ascher SM, Kendell M, Lage JM, Silverberg SG. Endometrial polyps: a comparison study of patients receiving tamoxifen with two control groups. Int J Gynecol Pathol 1998; 17:302 311. 28. Ramondetta LM, Sherwood JB, Dunton CJ, Palazzo JP. Endometrial cancer in polyps associated with tamoxifen use. Am J Obstet Gynecol 1999; 180:340 341. 29. Cohen I, Bernheim J, Azaria R, Tepper R, Sharony R, Beyth Y. Malignant endometrial polyps in postmenopausal breast cancer tamoxifen-treated patients. Gynecol Oncol 1999; 75:136 141. 20. Lev-Toaff A, Toaff M, Liu JB, Merton D, Goldberg B. Value of sonohysterography in the diagnosis and management of abnormal uterine bleeding. Radiology 1996; 201:179 184. 21. O Connell LM, Fries MH, Zeringue E, Brehm W. Triage of abnormal postmenopausal bleeding: a comparison of endometrial biopsy and transvaginal sonohysterography versus fractional curettage with hysteroscopy. Am J Obstet Gynecol 1998; 178: 956 961. J Ultrasound Med 22:1173 1179, 2003 1179