Abstract
BACKGROUND: The aim of the present study was to construct a nomogram model of high-risk stage III/IV colorectal cancer (CRC). METHODS: Gene mutation and clinical information of 251 CRC patients were downloaded from The Cancer Genome Atlas (TCGA). Targeted next-generation sequencing was performed on 44 patients to screen shared mutation genes with frequency >5% between TCGA and clinical cohorts. Univariable and multivariable logistic regression analyses were used to analyze the mutant genes and clinical indexes, and a high-risk stage III/IV nomogram model was constructed. The nomogram model was further validated in the clinical cohort. RESULTS: SMAD family member 4 (SMAD4), zinc finger homeobox 3 (ZFHX3), and phosphatidylinositol 3,4,5-trisphosphate-dependent Rac exchanger 2 (PREX2) mutations; pathological location; and preoperative carcinoembryonic antigen (CEA) value were screened out to compose a high-risk III/IV nomogram model. The nomogram had good calibration and discriminative ability, with an area under the curve of 0.76 [95% confidence interval (CI): 0.69-0.84]. Hosmer-Leme show test indicated that the model had good goodness of fit (P=0.83). The decision curve revealed this a nomogram model was feasible in clinical practice. In our clinical cohort, the calibration curve did not show good calibration and discrimination. CONCLUSIONS: We established a nomogram model, including the mutation status of SMAD4, ZFHX3, and PREX2; pathological location; and preoperative CEA value, which showed accuracy in the risk prediction of stage III/IV CRC patients.