Clinicopathological characteristics and treatment of patients with colorectal neuroendocrine carcinoma: A retrospective study of 157 cases

The present study sought to explore the clinicopathological characteristics, and construct a prognostic nomogram predicting disease-free survival (DFS) and overall survival (OS) in patients with colorectal neuroendocrine carcinoma (CRNEC). Data of patients diagnosed with CRNEC from Harbin Medical University Cancer Hospital (Harbin, China) between January 2000 and December 2019 were retrospectively collected. These patients were randomly divided into different groups by histological type, differentiation grade and AE1/AE3 expression. Subsequently, χ² test and Fisher's exact test were employed to assess associations between categorical variables and multiple groups, Kaplan-Meier estimates were generated to analyze survival, and univariate and multivariate Cox proportional hazards models were fitted to identify the potential prognostic factors. The multivariate analyses formed the basis for constructing prognostic nomograms. Nomogram validation comprised calculation of the C-index and generation of calibration curves. As a result, three variables were identified as the potential independent predictors for DFS (body mass index, age and differentiation) and OS (sex, age and differentiation). Nomograms were established using these variables based on the multivariate analyses. The C-index for the nomogram prognostic models predicting DFS and OS were 0.843 (95% CI, 0.665-0.936) and 0.819 (95% CI, 0.644-0.919), respectively. The predicted and observed survival probabilities showed concordance across different time points (1, 3 and 5 years), as evidenced by the calibration curves overlapping with the reference lines. Patients with neuroendocrine tumor-G1/G2 pathological subtypes, low grade differentiation or AE1/AE3(+) expression had longer DFS and OS compared with other groups. Overall, the prognostic nomograms established could effectively predict the prognosis of patients with CRNEC. High-grade CRNEC represents a highly aggressive malignancy associated with an unfavorable clinical outcome. In addition, AE1/AE3 expression may be a helpful biomarker for improving the diagnostic accuracy of patients with CRNEC.