Number of Nodules but not Size of Hepatocellular Carcinoma Can Predict Refractoriness to Transarterial Chemoembolization and Poor Prognosis

BACKGROUND: To determine whether response to transarterial chemoembolization (TACE) predicts survival and to identify pretreatment factors associated with TACE response and prognosis. METHODS: Between April and September 2010, 50 patients underwent TACE for hepatocellular carcinoma. Response to TACE was assessed using post-treatment computed tomography (CT) and magnetic resonance imaging (MRI) scans and tumor marker levels and classified as Response Poor (P) and Non-poor (NP). Time zero was set to September 30, 2010, and survival rates were analyzed by landmarking. Cumulative survival rates were calculated using the Kaplan-Meier method and compared according to grades using the log-rank test; contributing factors to survival were analyzed using a Cox proportional hazards model. Pretreatment factors were analyzed for 109 TACE sessions performed until October 2017, using a multiple logistic regression model. Receiver operating characteristic (ROC) curves were generated to determine the best tumor number for predicting response P. RESULTS: Response P patients showed significantly lower cumulative survival rates than Response NP patients (P < 0.001). On multivariate analysis, tumor number (hazard ratio (HR), 1.475), protein-induced vitamin-K absence-II (HR, 4.539), and the number of previous TACE sessions (HR, 1.472) were identified as pretreatment factors contributing to Response P. Further, pre-treatment platelet count (HR, 0.876) and tumor number (HR, 1.330) were factors contributing to survival in multivariate analysis. ROC curve analysis revealed that the optimal cut-off value to discriminate Response P was 7.5. CONCLUSIONS: Response to TACE can predict survival. Pretreatment tumor number is a useful factor for predicting both TACE response and prognosis.