Nomogram for predicting cancer-specific mortality risk in endometrial cancer after postoperative radiotherapy

Endometrial cancer (EC) incidence keeps rising amid population aging and obesity. This study aimed to investigate independent prognostic factors for the overall survival (OS) and cancer-specific survival (CSS) in postmenopausal EC patients treated with surgery and radiotherapy, and to construct a nomogram to predict survival rate.

We analyzed postmenopausal EC patients who had undergone surgery and radiotherapy between 2010 and 2015, using data from the Surveillance Epidemiology and End Results (SEER) database. The patients were divided into the training and validation cohorts using a ratio of 7:3. Prognostic factors were identified using univariate and multivariate analyses, and a nomogram was developed to predict OS and CSS using Cox proportional hazards and Fine-Gray compete-risk models. We evaluated the nomogram's performance using C-index, receiver operating characteristic (ROC) curve, and calibration plot, and compared it with American Joint Committee on Cancer (AJCC) staging model to assess the impact of surgical modality, radiotherapy modality, chemotherapy or not, and pathological types on OS in different risk stratifications.

A total of 10,906 EC patients were enrolled in the study and randomly assigned to the training (n=7,633) and validation (n=3,273) cohorts. The competing risk nomogram finalized 13 prognostic variables including age, tissue type, brain metastasis, lung metastasis, marital status, T-stage, N-stage, stage, grade, radiotherapy modality, tumor size, surgical site, and regional lymphadenectomy. The nomogram showed good performance in predicting CSS with C-indexes of 0.800 and 0.755 for the training and validation sets, respectively. The nomogram performed significantly better than AJCC staging model. In addition to the above variables, the Cox proportional hazards model incorporated bone metastasis, chemotherapy, and the total number of in situ/malignant tumors, with the C-index of the training set being 0.753 and 1-, 3-, and 5-year areas under the curve (AUCs) being 0.825, 0.785, and 0.772, respectively. The calibration plots showed good consistency. Subgroup analysis showed that the low-risk and high-risk groups showed that patients receiving radiotherapy had longer survival times compared to those receiving chemoradiotherapy, and patients undergoing brachytherapy had longer survival times compared to those receiving external beam radiotherapy (EBRT).

Our study developed and validated a nomogram for predicting OS and CSS in postmenopausal EC patients treated with surgery and radiotherapy. These findings provide valuable insights for individualized risk assessment and treatment decision.