Development and Internal Validation of a Nomogram for Predicting Recurrence in Endometrial Cancer Based on Pathological and Histological Indicators

Background: Endometrial cancer (EC) is a common malignant tumor in the female reproductive system. Identifying patients with a high risk of recurrence is beneficial for formulating personalized follow-up and treatment plans. This study aims to develop a prediction model for evaluating the risk of recurrence of EC after treatment. Methods: This study conducted a retrospective analysis on 486 patients with EC and randomly divided them into a training group (n = 389) and a validation group (n = 97). A nomogram was constructed after identifying predictors. The concordance index (C-index), receiver operating characteristic (ROC) curve, calibration plots, net reclassification index (NRI), integrated discrimination improvement (IDI), decision curve analysis (DCA) and Kaplan-Meier curves were used to evaluate the predictive model for EC recurrence. Results: A predictive nomogram was constructed based on the six selected predictors. The ROC curve (area under the curve = 0.890) and calibration curve indicate that the model has high discrimination and calibration capabilities. The NRI in the training set was 0.321 (95% CI: 0.031– 0.438), and the IDI was 0.133 (95% CI: 0.052– 0.215), indicating a significant improvement compared to the ESGO-ESTRO-ESP pattern. The DCA curves indicated that this model exhibited excellent discriminative performance and clinical application value. Conclusion: A nomogram based on pathological factors and immunohistochemical indicators was constructed and validated for predicting the recurrence of EC. Its predictive performance was superior to the ESGO-ESTRO-ESP pattern, and it can be used as a prognostic tool for clinical risk stratification.