| Objective: The age of onset of colorectal cancer tends to be younger,and some patients have metastases to the brain,liver,lung,bone and other organs.Among them,the liver is one of the most common metastatic sites of colorectal cancer,and the prognosis is not optimistic.This study aims to use the Surveillance Epidemiology and End Results(seer)database to investigate the incidence and associated factors of early-onset colorectal cancer,construct a nomogram based on prognostic-related variables to predict the risk of liver metastasis of early-onset colorectal cancer and predict the overall survival,and guide individualized treatment,which is helpful in the management of early-onset colorectal cancer and improvement of survival.Methods:The patients diagnosed with colorectal cancer from 2010 to 2016 were retrieved from the Surveillance Epidemiology and End Results database,and the incidence rates of different age groups,genders,and different distant metastatic organs(bone,brain,liver,lung)after age standardization are analyzed and calculated.We select patients with early-onset colorectal cancer for further study: retrieving and extracting data from a total of 29,459 patients with early-onset colorectal cancer diagnosed in 2010-2016 from the Surveillance,Epidemiology and End Results database,and finally included a total of 16,915 cases in the study,randomly divided according to the ratio of 7:3 for the training cohort(11840 cases)and the validation cohort(5075 cases).validation cohort is used for internal verification.Univariate and multivariate logistic regression analyses are used to examine risk factors of liver metastasis.The result of multivariable analysis is used to construct a nomogram to predict the risk of liver metastasis of early-onset colorectal cancer.The univariable and multivariable cox regression analyses were conducted to determine the statistically significant variables for overall survival.Then,we establish a nomogram to predict their 1-,3-and 5-year overall survival probabilities.The predictive nomogram performance is estimated by the receiver operating characteristics cure,the concordance index,and calibration curve.In addition,Kaplan-Meier method is used to classify patients into high-risk and low-risk according to the best cut-off risk score of the nomograms.The risk stratification effectively avoides the survival paradox.Results: The incidence of colorectal cancer decreases year by year from2010 to 2016,and the incidence increases with age,starting from 35 years,the incidence continues to increase.The incidence of colorectal cancer by gender and distant metastatic organs is stable,and the incidence of men is higher than that of women.The most common distant metastatic organ is the liver,followed by the lungs,and the least common is the brain.Through univariate and multivariate logistic regression analysis,it is found that the grade,N stage,treatment methods(primary tumor resection,radiotherapy,chemotherapy),distant metastasis(bone,lung),CEA level,tumor deposits,perineural invasion is significantly related to liver metastasis of early-onset colorectal cancer.The best cut-off value,specificity,sensitivity of the total score of risk nomogram for liver metastasis of early-onset colorectal cancer in the training cohorts are-1.627,0.801,0.754,respectively.The best cut-off value,the specificity,sensitivity of the total score of the nomogram in the validation cohorts are-1.903,0.763,0.763,respectively.ROC curves presente good discrimination in the training cohort [area under the curve(AUC)0.848] and the validation cohort(AUC0.839).Univariate and multivariate cox regression analysis find that race,primary tumor location,grade,N stage,primary tumor resection,chemotherapy,tumor size,distant metastasis(bone,brain,liver,lung),CEA level,tumor deposits and perineural invasion are independent prognostic factors for the overall survival rate of patients with early-onset colorectal cancer.However,the differences in gender and T stage are not significant.The 1-,3-,5-year OS AUC are 0.845,0.837 and 0.815 in the training cohorts,and 0.85,0.838 and 0.814 in the validation cohorts.In addition,using the optimal cut-off value 201.43,all patients are stratified into high risk and low risk,Kaplan-Meier curve indicated that patients with higher risk had worse survival outcomes.The calibration curves exhibite good consistency between the predicted and actual survival rates.Conclusions: The nomogram is an effective clinical tool.This article studies and analyzes the relevant epidemiological information,clinicopathological and molecular characteristics of early-onset colorectal cancer,and uses the nomogram to stratify the risk of patients with early-onset colorectal cancer,which will help Clinicians manage patients and formulate more precise individualized treatment strategies. |
PDF Full Text Request