Prognostic prediction model for patients with pathological T1N0 stage esophageal squamous cell carcinoma undergone esophagectomy.

Abstract

There is a shortage of reliable predictive models to provide valuable prognostic information for early esophageal squamous cell carcinoma (ESCC) without lymph node metastasis (LNM). We aimed to develop and validate a nomogram using the prognostic factors in T1N0 ESCC patients. Patients with pathological T1N0 ESCC who underwent esophagectomy between 2014 and 2021 at three institutes were reviewed. The prognostic factors were evaluated by Cox proportional hazards model and a nomogram was developed. Patients were divided into high- and low-risk groups based on cut-off value of total points in the nomogram. Overall survival (OS) was estimated by the Kaplan-Meier method and compared using the log-rank test. A total of 275 patients were included and split into training (n=180) and external validation (n=95) cohorts. In the training cohort, multivariable analysis showed that the surgical approach, T1 substage, and carcinoembryonic antigen (CEA) level were independent prognostic factors. The developed nomogram had relatively high performance, with the area under the receiver operating characteristic (ROC) curve (AUC) of 0.783, 0.711 and 0.612 for 1-, 3-, and 5-year OS, respectively. The calibration curves showed that the predicted probability was in good agreement with the actual probability. Forty-seven was determined as cut-off value of total points. High-risk group (n=148) showed a significant poor OS than low-risk group (n=127) (P<0.001). Left surgical approach, stage T1b, and higher CEA were associated with poorer prognosis in T1N0 ESCC patients. The nomogram demonstrated a good performance to predict the individual survival.