A nomogram for the prediction of endometrial cancer in postmenopausal women with a thickened endometrium

The aim of this study was to develop a nomogram model for predicting the occurrence of endometrial cancer (EC) in postmenopausal women with a thickened endometrium. The clinical data of postmenopausal women with endometrial thickening who were diagnosed via transvaginal ultrasound and who underwent hysteroscopic endometrial pathology examinations at the Obstetrics and Gynecology Department of Xuanwu Hospital, Capital Medical University, were retrospectively collected from January 2018 to March 2023. Univariate and multivariate logistic regression analyses were carried out to determine the independent risk factors for the development of EC in postmenopausal women with a thickened endometrium. We subsequently constructed a predictive nomogram model for the risk of developing EC in postmenopausal women with endometrial thickening by including significant independent risk factors for the development of EC. Additionally, internal and external validation were performed to evaluate the EC prediction model. In total, 550 postmenopausal women with a thickened endometrium were included in this study. Older age (odds ratio OR: 1.06, 95% confidence interval CI: 1.01–1.12, P = 0.02), postmenopausal bleeding (OR: 12.34, 95% CI: 4.44–34.30, P < 0.01), history of irregular menstruation (OR: 19.20, 95% CI: 1.94–190.43, P = 0.01), greater endometrial thickness (ET) (OR: 5.03, 95% CI: 2.45–10.32, P < 0.01), and the presence of polypoid mass-like lesions (OR: 30.33, 95% CI: 8.97–102.51, P < 0.01) were identified as independent risk factors associated with the occurrence of EC. A nomogram was established according to the abovementioned independent risk factors, with satisfactory discrimination being observed in the training set (AUC: 0.897, 95% CI: 0.858–0.935) and in the validation set (AUC: 0.941, 95% CI: 0.886–0.996). The calibration plots indicated favorable agreement between the predictions of the nomogram model and the actual observations in both the training and validation sets. Furthermore, decision curve analysis indicated that the nomogram model was clinically useful and had good discriminative ability. In this study, we established a predictive nomogram model for the risk of developing EC that will help clinical physicians in developing more accurate and personalized diagnostic and treatment strategies for postmenopausal women with a thickened endometrium. Future multicenter studies are still needed in the future to increase the universality of prediction models.