Construction of Nomogram-Based Prediction Model for Clinical Prognosis of Patients with Stage II and III Colon Cancer Who Underwent Xelox Chemotherapy after Laparoscopic Radical Resection

Abstract

Objective. To construct a nomogram-based prediction model for the clinical prognosis of patients with stage II and III colon cancer who underwent Xelox chemotherapy after laparoscopic radical resection based on large data sets. Methods. A total of 7,832 patients with colorectal cancer who received postoperative Xelox-based chemotherapy were screened from the Surveillance, Epidemiology, and End Results database (USA) as the training data set. In addition, 348 domestic patients were screened as the validation data set. Multivariate Cox regression analysis was performed to identify variables for inclusion in the nomogram-based prediction model. The predictive accuracy of the model was assessed using C-index and calibration curve. Results. Age, cell differentiation, nerve invasion, T and N stages of tumours, number of dissected lymph nodes, and carcinoembryonic antigen (CEA) level were found to influence the efficacy of postoperative chemotherapy. The nomogram-based prediction model was successfully constructed. The C-index of both the training set and validation set were higher than those of the 7th edition of TNM staging system published by the American Joint Commission on Cancer ( $\text{C}-\text{index}\text{of}\text{training}\text{data}\text{set}=0.728$ , $\text{C}-\text{index}\text{of}\text{validation}\text{data}\text{set}=0.734$ ). The prediction results of the model in the calibration curve showed a good fit with the actual situation. Conclusion. We successfully constructed a nomogram-based model to predict the clinical prognosis of patients with colorectal cancer receiving postoperative Xelox-based chemotherapy after laparoscopic radical resection, which showed good clinical application value for predicting the efficacy of postoperative Xelox-based chemotherapy in patients with colorectal cancer.