A Novel Nomogram for Predicting Endometrial Malignancy in Postmenopausal Women
ABSTRACT
Aim
To identify clinical risk factors for endometrial malignancy in postmenopausal women and develop a predictive model for early detection and personalized intervention.
Methods
This study analyzed 1146 postmenopausal women undergoing diagnostic hysteroscopy. Inclusion required: confirmed menopause (age ≥ 40) with recent endometrial thickness measurement, successful hysteroscopy, histopathological verification, and complete records. Exclusions involved incomplete data, type II carcinoma, hormonally active tumors, or prior hysteroscopy indications. Demographics, clinical features, comorbidities, imaging data, and biomarkers were analyzed. Histology was confirmed via standard pathology. Risk factors were identified through univariate and multivariate logistic regression. The resultant predictive nomogram for endometrial malignancy risk stratification underwent rigorous validation using: (1) receiver operating characteristic curve analysis (discriminative power); (2) calibration plotting (prediction accuracy); and (3) decision curve analysis (clinical net benefit).
Results
Among 1146 postmenopausal women undergoing diagnostic hysteroscopy, histopathological analysis identified 69 cases (6.0%) of endometrial carcinoma (EC) and 15 cases (1.3%) of atypical endometrial hyperplasia, with the remaining cases (92.7%) demonstrating benign pathology. Multivariate analysis identified seven independent risk factors for EC: elevated fibrinogen and D‐dimer levels, hypertriglyceridemia, decreased high‐density lipoprotein, postmenopausal bleeding, ultrasonography blood‐flow signals, and increased endometrial thickness. The predictive nomogram incorporating these parameters demonstrated outstanding diagnostic performance, with area under the curve values of 0.955 in the training cohort and 0.960 in the validation cohort, indicating excellent discriminative ability for EC risk stratification.
Conclusion
We developed and validated a novel 7‐indicator prediction model for assessing endometrial malignancy risk in postmenopausal women undergoing hysteroscopy biopsy.