Construction of a Risk Model for Colon Cancer Prognosis Based on Ubiquitin-Related Genes

BACKGROUND: Colon cancer is a frequently developed malignancy from the digestive system that leads to poor prognosis of patients due to its high recurrence and high metastasis. Dysregulation of ubiquitin-mediated signaling can result in tumor formation and metastasis. We aimed to develop prognostic markers related to ubiquitination in colon cancer and a risk assessment model based on these markers to improve the prognosis of colon cancer patients. METHODS: We constructed a prognosis-related model by performing differential expression analysis on ubiquitin-related genes in colon cancer patients based on public data and then undertaking Cox analysis, which selected 7 ubiquitin-related prognostic genes (TRIM58, ZBTB7C, TINCR, NEBL, WDR72, KCTD9, and KLHL35). The samples were divided into high and low RiskScore groups according to the risk assessment model, and as Kaplan–Meier suggested, the overall survival of patients with high RiskScore was prominently lower than that of patients with low RiskScore. The accuracy of RiskScore was assessed by receiver operating characteristic curves. Accordingly, the area under the curve values of 1-, 3-, and 5-year were 0.76, 0.74, and 0.77 in the training set and 0.67, 0.66, and 0.74 in the validation set, respectively. RESULTS: These data confirmed the preferable performance of this prognostic model in predicting colon cancer patients’ prognoses. The relationship between this RiskScore and clinicopathological factors of colon cancer patients was analyzed via stratification. Univariate and multivariate Cox regression analyses were performed to determine whether this RiskScore could be applied as an independent prognostic factor. Finally, in order to better apply the prognostic model in clinical practice, we constructed an overall survival nomogram for colon cancer patients’ prognoses based on clinical factors and RiskScores, which has preferable prediction accuracy and is better than the traditional tumor, node, and metastasis (TNM) staging system. CONCLUSIONS: The overall survival nomogram for prognosis can assist clinical oncologists to make a more accurate evaluation of patients’ prognosis, as well as the implementation of individualized diagnosis and treatment for colon cancer patients.