Construction And Verification Of A Nomogram For Pred Icting The Overall Survival Of Patients With Intrahepatic Cholangiocarcinoma After Radical Resection

Background and aimIntrahepatic cholangiocarcinoma(ICC)is the second most common liver malignant tumor after hepatocellular carcinoma.Radical tumor resection is still the best treatment method.At present,the 8th edition AJCC TNM staging system is most commonly used to predict patient prognosis.However,due to individual differences,the 8th edition AJCC TNM staging system is considered applicable to the whole patient population.It remains to be further explored for individual predictions.The purpose of this study is to establish an effective nomogram to predict the overall survival(OS)of ICC patients by collecting patient clinical information.Final individualized prognostic prediction system is established by comparing with the 8th edition AJCC TNM staging system in terms of prediction accuracy and clinical applicability.MethodsPatient data for nomogram construction was acquired from the Surveillance,Epidemiology,and End Results(SEER)database.Patients pathologically diagnosed of ICC from 2005 to 2016 were collected.All data were randomly divided into training cohort and internal verification cohort with the radio of 7:3.COX univariate and multivariate analyses were used to determine independent risk factors affecting the patients’ prognosis,and therefore used to construct a nomogram.The prediction accuracy and discriminative ability of the nomogram are evaluated by the concordance index(C-index)and the calibration curve.And area under receiver operating characteristic(ROC)curve(AUC)and the decision curve analysis(DCA)threshold area were used to compare the prediction accuracy of 1-year,3-year,and5-year OS and clinical applicability between the 8th edition AJCC TNM staging system and the present constructed nomogram.Finally,patients pathologically diagnosed of ICC from January 2013 to May 2020 in the Second Affiliated Hospital of Nanchang University were retrospectively collected and used for external verification.ResultsThe study contained two parts of data.We established a nomogram using the training cohort data,and conducted internal and external verifications.The present nomogram showed good prediction accuracy which was constructed based on independent risk factors including age,T stage,N stage,M stage,tumor differentiation and chemotherapy.The C-index was 0.724(95% CI,0.712 to 0.736)for OS prediction.The AUC of the nomogram predicting 1-,3-and 5-year OS rates were 0.75,0.721 and 0.761,respectively,which were higher than the AUC value of the 8th edition AJCC TNM staging system.Meanwhile,the area under the DCA threshold of the nomogram was also higher than the 8th edition AJCC TNM staging system.The calibration curve showed good agreement between the predicted and the actual observed value of the nomogram.ConclusionOur nomogram constructed based on clinically accessible clinical indicators provided more accurate prognostic predictions for patients with ICC.