Prognostic Value Of Tumor Deposits In Patients With Stage Ⅲ Colorectal Cancer: A Population Based Study

Purpose:In this study,based on the data of patients with stage III colorectal cancer in SEER database,the independent prognostic factors of colorectal cancer were screened after statistical analysis,and the influence of tumor deposition on prognosis was analyzed emphatically,and then the prediction of 1-year,3-year and 5-year cancer-specific survival of colorectal cancer patients was constructed(Cancer-specific survival,CSS)and Overall survival(OS),and verify and evaluate the model.The constructed model can provide clinical workers and patients with an economical,simple and accurate survival prediction tool to help formulate individualized treatment scheme and long-term follow-up plan.Methods: The clinical data of patients diagnosed with stage III colorectal cancer in SEER database from 2010 to 2015 were collected for analysis,and a total of 22529 eligible patients with colorectal cancer were included.All data were randomly divided into training and validation sets in a 7:3 ratio.Univariate and multivariate Cox regression models were used to screen independent risk factors for stage Ⅲ colorectal cancer.Finally,the survival prediction model of the line graph is constructed.The data of 158 patients with stage Ⅲcolorectal cancer in Shaoxing City People’s Hospital were collected as external validation of the model.The accuracy of the nomogram model was evaluated by the C-index,the area Under the curve(AUC)of the Receiver operating Characteristic(ROC),and the calibration curves,and the value of the model in clinical application was evaluated by the decision curve analysis(DCA).Finally,based on the total score of the nomogram survival prediction model,a risk stratification system was established to distinguish patients with different risks.Results: In this study,according to univariate and multivariate Cox regression analysis,the results showed:Age,race,tumor pathological grade,tumor histological type,T stage,N stage,carcinoembryonic antigen,peripheral nerve invasion,tumor deposition,radiotherapy and chemotherapy are independent prognostic factors for patients with stage III colorectal cancer.A survival prediction model was constructed based on multivariate Cox regression analysis.The c-index for predicting CSS was 0.715(95%CI:0.710-0.723)in the training set and 0.721(95%CI:0.709-0.733)in the validation set.The c-index for predicting OS was0.717(95%CI:0.711-0.723)in the training set and 0.719(95%CI:0.709-0.729)in the validation set.The results show that the model has good discriminant ability.Both AUC and calibration curve showed good accuracy and discrimination.DCA shows that nomograms of CSS and OS have good clinical potential.Finally,all included patients were divided into high and low risk strata according to the total score of the nomographic survival prediction model.Conclclusions: This study identified survival risk factors for patients with stage III colorectal cancer and developed nomograms to predict cancer-specific and overall survival in these patients.The model has good clinical application value and can be used as a clinical decision-making tool for doctors and patients.Secondly,the prognostic value of the number of tumor deposits is equivalent to that of lymph node metastasis,which provides a new idea for TNM staging.