Page 104 - The diagnostic work-up of women with postmenopausal bleeding

Chapter 5
An unexpected finding of our study is the relatively high rate of (pre) cancer after a benign endometrial sampling.Two previous meta-analyses showed the sensitivity of Pipelle®tobe97%and99.6%,respectively.22,23Althoughbothmeta-analysesincluded pre- and postmenopausal women, the number of women with postmenopausal bleeding was limited (Clark described 314 postmenopausal women, of whom 14 had endometrial cancer. Dijkhuizen described 319 postmenopausal women, of whom 52 had endometrial cancer).All included studies used dilatation and curettage (D&C) as the reference standard,24-28 and therefore are expected to miss 50-85% of all focal intracavitary pathology.29,30 At present, hysteroscopy with guided biopsies is the gold standard to diagnose endometrial abnormalities.11
Of major importance is the clinical relevance of the diagnosis of endometrial cancer. Mingels et al recently performed histological assessment of the whole endometrium in a cohort of 48 postmenopausal women without bleeding who had hysterectomy because of a prolapse.34 Four (8.3%) of 48 women had atypical hyperplasia while two women (3%) had a small focal endometrial cancer and in 27% an endometrial polyp was found.This suggests a higher prevalence of endometrial pathology in (asymptomatic) postmenopausal women then assumed.35 This underscores that the relation between intracavitary pathology and postmenopausal bleeding is debatable. Although the women in our study diagnosed with atypical hyperplasia or carcinoma had hysterectomy with BSO, the clinical course of these cancers, when left untreated, is unclear.The fact that overall in this study eight women in the hysteroscopy group and one woman in the expectant group were diagnosed with a (pre) cancer, suggests that due to randomisation, we missed a few (pre) cancers in the expectant management group.This strengthens the indication for further diagnostic work-up in women with a benign result of endometrial sampling to exclude focal (pre) cancers, because endometrial sampling can miss these.
Conclusion
In women with PMB, a thickened endometrium and benign endometrial sampling, operative hysteroscopy does not reduce recurrent bleeding. However, the finding of a 6 % prevalence of (pre) cancer in an endometrial polyp, not diagnosed by endometrial sampling, indicates intracavitary diagnostics as standard procedure in these women. In alignment with other studies,31,32 we found SIS to be accurate in the diagnosis endometrial polyps, indicating that a strategy starting with SIS triaging for hysteroscopy or not would be reasonable.Whether the initial diagnostic work- up should start with SIS, followed by endometrial sampling, or vice versa, needs
102