Identifying risk factors for cancer-specific early death in patients with advanced endometrial cancer: A preliminary predictive model based on SEER data

Abstract

Objective To identify risk factors associated with cancer-specific early death in patients with advanced endometrial cancer and to develop a preliminary nomogram prediction model based on these factors, with an emphasis on the potential implications for clinical practice. Methods Patients from the Surveillance, Epidemiology, and End Results (SEER) database in the United States from 2018 to 2021 were included in the study. The study data was randomly divided into a training cohort and a validation cohort at a ratio of 7:3. Multivariate logistic regression analysis was performed in the training cohort to screen for risk factors for cancer-specific early mortality in advanced endometrial cancer patients, and a preliminary nomogram prediction model was further constructed. The results of the Receiver Operating Characteristic (ROC) curve, calibration analysis, and clinical decision curve analysis (DCA) were presented for transparency. Results Significant risk factors for cancer-specific early death were identified, including tumor size (≥101 mm, OR = 2.11, P < 0.001), non-endometrioid histology (OR = 3.11, P < 0.001), high tumor grade (G3, OR = 2.68, P = 0.007), advanced tumor stages (T3-T4, OR = 1.84, P = 0.004), and metastatic stage (M1, OR = 2.05, P < 0.001), as well as the presence of liver metastases (OR = 2.21, P = 0.005) and brain metastases (OR = 8.08, P < 0.001). Protective factors that were significantly associated with a reduced risk of early death included hysterectomy (OR = 0.13, P = 0.012), radical surgery (OR = 0.21, P < 0.001), radiation therapy (OR = 0.40, P < 0.001), and chemotherapy (OR = 0.31, P < 0.001). A preliminary nomogram model was demonstrated adequate predictive performance with AUC values of 0.89 (95% CI 0.87 to 0.91) in the training cohort and 0.88 (95% CI 0.84 to 0.91) in the validation cohort. The model’s predictive performance was further supported by the calibration and DCA analyses, suggesting its potential clinical utility. Conclusion This study identified key risk factors for early cancer-specific mortality in patients with advanced endometrial cancer. The preliminary nomogram model holds promise for predicting early death risk and could be valuable in clinical practice. Future work may explore its performance with additional data to ensure broad applicability.