The optimal endometrial thickness threshold for prediction of endometrial cancer in postmenopausal women without bleeding remains uncertain-Systematic review and meta-analysis

The incidental finding of endometrial thickness (ET) >4 mm in the absence of postmenopausal bleeding (PMB) is a common cause of referring women to secondary care. However, there is lack of consensus amongst gynecologists as regards the management. It is estimated that up to 15 % of endometrial cancers occur in women without PMB. The aim this study was to determine the optimal ET threshold, on trans-vaginal ultrasound scan, that discriminates normal endometrium from endometrial hyperplasia and cancer in this cohort. On using a thorough search strategy, a total 16 studies including 4088 women were deemed eligible. However, the data were not amenable to meta-analysis. There were wide variations in the thresholds reported with potential bias given the retrospective nature of the majority of the studies. Despite contacting authors, we could not obtain the primary data to generate a Receiver Operating Characteristic (ROC) Curve. No linear or curvilinear association was found between ET thresholds and the percentage of women diagnosed with endometrial hyperplasia and cancer using either Pearson's correlation, linear or curvilinear regression, or a simple visual scan/scatter diagram. The result of this study reveals the lack of evidence to inform clinical practice in this area, and there is a need for a well-designed multi-center prospective study.