Construction of prognosis prediction model for patients undergoing surgery for intrahepatic cholangiocarcinoma

SUN Yushu; WEI Xing; DAI Fengxue; MU Yongping; YANG yang

doi:10.3969/j.issn.1009-0460.2025.02.007

Abstract：Objective To explore the factors affecting the prognosis of patients undergoing surgery for intrahepatic cholangiocarcinoma and construct a predictive model.Methods A total of 158 patients with intrahepatic cholangiocarcinoma underwent surgical resection in Inner Mongolia Hospital,Peking University Cancer Hospital from January 2015 to June 2021 were retrospectively collected.Clinicopathological characteristics of the patients were analyzed,and the overall survival time(OS)was followed up.The median OS of patients in different clinicopathological subgroups was compared.Cox proportional regression model was used to analyze the factors affecting OS in patients undergoing surgery for intrahepatic cholangiocarcinoma.The nomogram prediction model was constructed and verified according to the variables screened by multi-factor analysis.Results Follow-up until September 2023,the median OS of 158 patients was 19.0 months,and the 1-year and 3-year overall survival rates after surgery were 64.7％and 32.3％,respectively.There were statistically significant differences in median OS among subgroups with or without intrahepatic bile duct calculus,number of tumor lesions,scope of liver resection,degree of tumor differentiation,tumor cell type,maximum tumor diameter,presence or absence of vascular invasion,TNM stage,presence or absence of lymph node metastasis,presence or absence of satellite lesion,incisal border status,and CA19-9 level(P＜0.05).Multivariate Cox proportional regression model showed that intrahepatic bile duct calculi,number of lesions,scope of liver resection,tumor differentiation degree,tumor cell type and CA19-9 level were all independent factors influencing OS in patients undergoing intrahepatic cholangiocarcinoma surgery(P＜0.05).The nomogram risk model was constructed based on the selected variables from the multi-factor analysis,and the consistency index(C-index)was 0.713.Bootstrap self-sampling method was used for internal verification,and the calibration curve was obtained by repeated self-sampling 500 times.The average absolute error of the calibration curve was 0.016.Conclusion The prognosis of patients undergoing intrahepatic cholangiocarcinoma surgery is affected by many factors,such as the presence or absence of intrahepatic cholangiolithiasis,the number of tumor lesions,the scope of liver resection,the degree of tumor differentiation,tumor cell type and CA19-9 level.The constructed nomogram model can obtain better prediction effect.