Comparison of clinicopathological features and survival analysis between esophageal neuroendocrine carcinoma and esophageal squamous cell carcinoma based on the SEER database, alongside nomogram analysis for esophageal neuroendocrine carcinoma.

ABSTRACT

Background: Esophageal neuroendocrine carcinoma (ENEC) is a rare subtype of esophageal cancer (EC). It presents distinctive clinical and pathological features in comparison to esophageal squamous cell carcinoma (ESCC). To better elucidate the disparities between the two and establish a prognostic prediction model for ENEC, we conducted this study. Methods: Data of ENEC and ESCC patients (1975 to 2016) were extracted from the Surveillance, Epidemiology and End Results (SEER) database. Patients with a confirmed pathological diagnosis of ENEC and ESCC were enrolled in the study. The Chi-square test was employed to compare categorical variables, and the median survival time was analyzed using the Kaplan-Meier curve. Training and validation groups were randomly assigned at a ratio of 73. Factors with a significance level of <0.05 in the multifactor regression model as well as age were integrated into the nomogram model. Concordance index (C-index), calibration curves, and decision curve analyses (DCA) were generated for model validation. Results: This study encompassed a total of 737 ENEC patients and 29,420 ESCC. Compared to ESCC, ENEC patients had higher probability of liver metastasis (13.8% vs. 1.9%, P<0.001), poor differentiation (68.0% vs. 37.1%, P<0.001), and late SEER stage (52.8% vs. 26.9%, P<0.001). Patients who received either surgery, radiotherapy (RT), or chemotherapy had a significantly longer disease-specific survival (DSS) and overall survival (OS) (all P<0.001). After propensity score matching (PSM), ENEC patients were associated with shorter DSS (7.0 months vs. not reached, P<0.0001) and OS (7.0 vs. 12.0 months, P<0.0001) compared to ESCC. Race, SEER stage, surgery, RT, and chemotherapy were identified as predictors of DSS and were incorporated into the nomogram model together with age. The validation of the model using C-index (0.751 and 0.706, respectively) and calibration curves reflected the better discrimination power of the model. In addition, DCA supported the favorable potential clinical effect of the predictive model. Lastly, a risk classification based on the nomogram also verified the reliability of the model. Conclusions: ENEC and ESCC exhibit distinct clinicopathological features. Patients with ENEC experience significantly poorer survival outcomes compared to those with ESCC. Surgical intervention, radiation therapy, and chemotherapy significantly improve OS and DSS for ENEC patients. The nomogram prediction model, constructed based on age, race, stage, and treatment regimen, demonstrates accurate and effective predictive capabilities for prognostic factors in ENEC patients.