A nomogram based on radiomics to predict microvascular invasion and overall survival in intrahepatic cholangiocarcinoma

Abstract

Background: Intrahepatic cholangiocarcinoma (ICC) is aggressive and ranks second among primary liver cancer. Microvascular invasion (MVI) is a predictor of poor prognosis in ICC, and preoperative prediction of MVI status can facilitate personalized therapeutic regime. Purpose: To establish a comprehensive model based on MR radiomics for MVI status stratification and overall survival prediction in ICC patients preoperatively. Methods: 249 ICC patients were randomized into training and validation cohort (174:75) and a time-independent test cohort with 47 ICC patients were enrolled. Independent clinical and imaging predictors were identified by univariate and multivariate logistic regression analysis. The radiomic model was based on the robust radiomic features extracted by logistic regression classifier and the least absolute shrinkage and selection operator algorithm. The imaging-radiomics (IR) model integrates the independent predictors and robust radiomics features. The predictive efficacy of models is evaluated by receiver operating characteristic curve, calibration curve and decision curve. Multivariate Cox analysis identified the independent risk factors of overall survival, Kaplan-Meier curves were plotted and nomogram visualized the predictive model. Results: Imaging model comprises tumor size and intrahepatic duct dilatation, radiomics model comprises 25 stable radiomics features. The IR model shows desirable performance (AUCtraining= 0.890, AUCvalidation= 0.885 and AUCtest= 0.815). The calibration curve and decision curve validate the clinical utility. Overall survival predicted by histological and IR model-predicted MVI groups exhibit similar predictive efficacy. Conclusion: IR model and nomogram based on IR model-predicted MVI status may be a potential tool in MVI status stratification and overall survival prediction of ICC patients preoperatively.