Abstract
BACKGROUND: A prognostic model to predict the individual disease-specific survival (DSS) rates of non-metastatic esophageal cancer (nMEC) patients after preoperative radiotherapy (pRT) has not been established. In the current study, we aimed to establish a survival nomogram for nMEC patients after pRT. METHODS: We identified 2,424 nMEC patients who underwent pRT from the Surveillance, Epidemiology, and End Results database. Approximately, 80% (n=1,948) of the included patients were randomly selected and designated as training data set, and the remaining patients (n=476) were defined as external validation set. Nomogram was established by the training set and validated by the validation set. RESULTS: According to the results of the multivariate analysis, a nomogram combined with age at diagnosis, sex, tumor location, yp-T stage, yp metastatic lymph node ratio stage (yp-mLNRS), and grade was developed. The C-index of the model was significantly higher than that of yp-TNM staging system (0.62, 95% CI, 0.58 to 0.66 vs 0.55, 95% CI, 0.51 to 0.60; p<0.001). Calibration plots of the nomogram showed that the probability of DSS rates optimally corresponded to the survival rates were observed. CONCLUSION: The proposed nomogram resulted in more reliable DSS prediction for nMEC patients in general population, regardless of the patient's histological type. Upon validation, it will aid in individualized survival prediction and prove useful in clinical decision making in nMECs after pRT.