Construction Of Prognostic Line Map Model Of Colorectal Cancer Patients With Liver Metastasis By Primary Tumor Resection Combined With Chemotherapy Based On SEER Database

Background and Objective:Colorectal cancer is a common malignant tumor in the world,and its incidence and risk of death are increasing year by year.And some studies have found that about 2/3 of patients have synchronous or metachronous distant metastasis,of which the liver is a common site of metastasis.In recent years,primary tumor resection combined with systemic chemotherapy has been considered as the baseline treatment for colorectal cancer patients with liver metastasis,and most patients can benefit from it,but there are still some patients who are ineffective to this regimen or have II stage recurrence,so it is urgent to explore the factors affecting the prognosis of patients with liver metastasis who receive primary resection combined with chemotherapy.Although some scholars at home and abroad have studied it,there are some deficiencies because of the small sample size or incomplete follow-up data.Therefore,this study will retrospectively analyze the clinicopathological factors of CRCLM patients with primary tumor resection combined with chemotherapy based on SEER database,and explore the relevant risk factors affecting prognosis,and construct a predictive model to evaluate the prognosis of patients based on the above analysis results,in order to provide clinicians with a good predictive model of prognosis,so as to improve the quality of life of patients.Methods:In this study,the clinical data of colorectal cancer patients with liver metastasis who received primary resection combined with chemotherapy from 2010 to 2015 were extracted from SEER database.According to the inclusion and exclusion criteria,3057 cases were randomly divided into model group(nude 2140)and control group(nude 917)according to the 7:3 ratio by SPSS software.The independent prognostic factors affecting OS were screened out according to Cox univariate and multivariate regression analysis.And construct the line chart to evaluate OS by R software.According to the Fine-Gray competitive risk model,the independent prognostic factors affecting CSS were screened,and the line chart was constructed according to the above method.C-index index,calibration chart and AUC of ROC curve were used to evaluate the reliability of OS prediction model and CSS prediction model,and verified by verification group.Finally,the patients were scored according to the constructed line chart,and the prognostic risk was stratified.Results:1.Chi-square test showed that there was no significant difference in clinical baseline data between the model group and the control group(P>0.05).Univariate and multivariate Cox regression analysis showed that age,primary tumor location,tumor size,tumor grade,N stage,number of lymph nodes detected,positive number of lymph nodes,radiotherapy,CEA,perineural infiltration and marriage were independent prognostic factors of OS in CRCLM patients with primary tumor resection combined with chemotherapy.2.Univariate and multivariate analysis was carried out by Fine-Gray competitive risk model.After excluding the confounding factors of non-tumor death,it was found that age,primary location of tumor,tumor grade,N stage,number of lymph nodes detected,number of positive lymph nodes,tumor size,CEA,perineural infiltration and marital status were independent prognostic factors of CSS.3.The C-index index,ROC curve and calibration curve test show that the OS and CSS related prognosis map model of CRCLM patients with primary tumor resection combined with chemotherapy based on SEER database has good predictive ability.4.According to the constructed line chart,the stratification of prognostic risk of the patients included in the study showed that in terms of OS prognosis,the scores of<223,223,335 and>335 were low risk group,middle risk group and high risk group,respectively,and in terms of CSS prognosis,they were low risk group,medium risk group and high risk group.Conclusion:The clinical prediction models of OS and CSS of CRCLM patients with primary tumor resection combined with chemotherapy established by SEER data have high accuracy in evaluating and predicting the prognosis of patients,and can provide reliable reference value and basis for clinicians to carry out personalized diagnosis and treatment.