Nomograms to predict cancer-specific mortality in colon adenocarcinoma with different types of villous architecture.

ABSTRACT

PURPOSE: The role of villous architecture in the prognosis of colon adenocarcinoma remains unclear. This study aimed to investigate the prognostic factors of colon adenocarcinoma with different types of villous architecture and to establish nomograms for predicting cancer-specific mortality. METHODS: This retrospective study included 10,427 patients with colon adenocarcinoma arising in adenomas with villous architectures. The patients were stratified into the tubulovillous adenocarcinoma cohort and villous adenocarcinoma cohort. The prognostic risk factors, which were incorporated into nomograms for survival prediction, were determined by the log-rank test and Cox hazard models. The Harrell's Concordance Index (C-index) and calibration curve were utilized to evaluate the prediction accuracy. RESULTS: The pathological type of villous architecture was independently associated with the mortality of the entire population. Age, race, tumor size, T/N/M stage, and chemotherapy were independent risk factors of mortality in both cohorts. Interestingly, tumor differentiation was a prognostic factor for tubulovillous adenocarcinoma rather than villous adenocarcinoma, while the retrieved lymph node number was a prognostic factor for villous adenocarcinoma rather than tubulovillous adenocarcinoma. Survival analysis showed that the mortality rate of villous adenocarcinoma was higher than that of tubulovillous adenocarcinoma (HR 1.361, P < 0.001). We then established nomograms to predict the mortality of both cohorts and found excellent discrimination and predictive accuracy (C-index 0.842 and 0.821). CONCLUSION: Villous architecture is a determinant of colon adenocarcinoma outcomes, which might prompt reports of villous architecture in colon adenocarcinoma specimens by pathologists. Our population-based nomograms could be useful for predicting the survival of patients with colon adenocarcinoma and guiding individualized treatments.