Thickened endometrium in asymptomatic postmenopausal women - determining an optimum threshold for prediction of atypical hyperplasia and cancer

There is a lack of consensus about the significance and the management of an incidental finding of endometrial thickness (ET) > 4 mm on transvaginal ultrasound scans in postmenopausal women without postmenopausal bleeding (PMB). The data of 1995 consecutive women attending PMB clinic were collected prospectively; of them 81 (4.1%) were referred because of ET >4 mm without PMB. The prevalence of endometrial atypical hyperplasia and cancer was 4/81 (4.9%), and polyp was 20/81 (24.7%). On using a receiver operator characteristic curve, the diagnosis of endometrial atypical hyperplasia and cancer using the ET threshold of ≥10 mm had a sensitivity of 100% (95% CI = 40-100%), a specificity of 60% (95% CI = 48-71%) with AUC = 0.8 (95% CI = 0.66-0.93), p = .04. For the 35 women with an ET ≥10 mm, the prevalence of endometrial atypical hyperplasia and cancer was 4/35 (11.4%) and benign endometrial polyps was 16/35 (45.7%). The use of ≥10 mm ET threshold to prompt investigations did not miss any case of endometrial atypical hyperplasia or cancer. Impact Statement What is already known on this subject? Unlike women with PMB in whom an endometrial thickness (ET) ≤ 4 mm is considered as low risk for endometrial hyperplasia and cancer, in postmenopausal women without PMB, the threshold that separates normal from a pathologically thickened endometrium has not been standardised. A decision-analysis study in a hypothetical cohort found that ET threshold of 11 mm yields a similar separation as ≤4 mm in those with PMB. What the results of this study add? The study uses prospectively collected data from consecutive patients using a standardised format, thus minimising bias from incomplete data. This study is the third prospective series in the literature to address the dilemma of the management of asymptomatic postmenopausal women with an incidental finding of a thickened endometrium. It showed that the prevalence of endometrial atypical hyperplasia and cancer is high enough to justify investigation and for the clinical problem not to be trivialised. All cases of endometrial atypical hyperplasia and cancer had ET of ≥10 mm. What are the implications of these findings for clinical practice and/or future research? Our data strengthen the current body of literature to help the development of clinical practice guidelines about the management work-up. However, a well-designed multi-centre large prospective study is required to confirm the findings since most studies in the literature are either retrospective or small.