Evaluation of response to Targeted therapy and immunotherapy for advanced HCCs: A retrospective, controlled study

Background: To validate the correlation between radiomics extracted from magnetic resonance imaging (MRI) and pathologic alterations in advanced hepatocellular carcinoma (HCC) treated with targeted therapy and immunotherapy, and to compare the accuracy of necrosis evaluation by radiologists and subtraction technology. Methods: Thirty-five patients diagnosed unresectable HCCs (n = 39) who underwent curative resection at two centers after receiving combined treatment with tyrosine kinase inhibitors and programmed death ligand-1 antibodies were included. Using linear regression model to verify the consistency between investigator-based imaging evaluation and three-dimensional (3D) subtraction-based imaging analysis and their capability to distinguish major pathological response (MPR) and pathological complete remission (PCR) were compared using receiver operating characteristic (ROC) curve analysis with pathological outcome as the gold standard.The sensitivity, specificity, and area under the ROC curve (AUC) were compared. Results: 3D subtraction-based imaging analysis suggested a strong correlation with histopathologic alteration (R ² = 0.81) and was superior in classifying MPR or PCR comparing to the use of investigator-based imaging evaluation (area under the receiver operator characteristic curve: MRP, 0.80 vs. 0.52; PCR, 0.69 vs. 0.54). Conclusion: 3D subtraction-based imaging analysis based on tumor characteristics showed a stronger correlation with pathology, and 3D subtraction-based imaging analysis was significantly superior to investigator-based imaging evaluation in classifying MPR or PCR.