3D MR Elastography-based Stiffness as a Marker for Predicting Hepatocellular Carcinoma Refractory to Transarterial Chemoembolization

Motivation: HCC exhibits diverse responses to TACE due to tumor heterogeneity. It has been previously demonstrated that MRE-based stiffness is promising in prediction of the treatment outcomes of HCC. Goal(s): To evaluate the value of 3D MRE-based stiffness preoperatively for predicting HCC refractory to TACE. Approach: We assessed the clinical-radiological features and stiffness of MRE, developed a model for predicting TACE refractoriness in training cohorts, and validated its performance in validation cohort. Results: 3D MRE-based tumor stiffness is an independent predictor for predicting TACE refractoriness. The model based on tumor stiffness achieved AUCs of 0.887 and 0.833 in the training and validation cohorts. Impact: 3D MRE-based tumor stiffness is a promising biomarker for predicting TACE refractory HCC. The MRE-based model might be a useful tool for evaluating TACE refractory HCC.