THE COMPARISON OF THE METHODS FOR TUMOR RESPONSE ASSESSMENT IN PATIENTS WITH HEPATOCELLULAR CARCINOMA RECEIVING TRANSARTERIAL CHEMOEMBOLIZATION

초록

Backgrounds: WHO or RECIST have been considered as suboptimal methods for tumor response assessment in hepatocellular carcinoma (HCC), especially receiving transarterial chemoembolization (TACE). Recently, new methods including concept of viable tumor such as modified WHO (EASL) and RECIST (mRECIST) has been proposed. Herein we evaluated whether there was differences on the efficacy to assess treatment response and to predict prognosis between WHO, RECIST, EASL, mRECIST in HCC patients receiving TACE. Methods: We retrospectively reviewed 141 HCC patients who were newly diagnosed and received TACE as initial treatment without additional treatment between August 2005 and November 2006. Among them, 98 patients who had adequate and measurable pre and post CT/MRI for assessing treatment response were analysed. Response assessment was made at 1 month after 2nd TACE We evaluated intermethod agreement (kappa values) between methods and assessed the efficacy to predict time to progression (TTP) and overall survival (OS). Results: The results of tumor response (number of CR/ PR/ SD/ PD) by WHO/ RECIST/ EASL/ mRECIST was 1/ 14/ 74/ 9, 1/ 12/ 77/ 8, 34/ 34/ 25/ 5, 34/ 28/ 31/ 5, respectively. The kappa value of intermethod agreement for EASL vs. WHO, EASL vs. RECIST, mRECIST vs. WHO, and mRECIST vs. RECIST was 0.102, 0.088, 0.112 and 0.122, respectively. The kappa value for WHO vs. RECIST and EASL vs. mRECIST was 0.883 and 0.759 (p=0.000). The median TTP was 8.8, 9.4, 9.8, and 9.8 months for WHO, RECIST, EASL and mRECIST respectively. The median OS was 24.2 months. Hazard ratios for TTP in responders compared with nonresponders were 0.313 (95%CI; 0.190-0.517, p=0.000) for EASL and 0.335 (95% CI; 0.206-0.544, p=0.000) for mRECIST. Hazard ratios for survival in responders compared with nonresponders were 0.208(95% CI; 0.116-0.373, p=0.000) for EASL and 0.269 (95%CI; 0.151-0.480, p=0.000) for mRECIST. In the univariate analysis, tumor size, infiltrative type, PVT, EASL responder, and mRECIST responder were significant risk factors for survival. Multivariate analysis showed EASL responder (HR 0.165, 95%CI 0.056-0.486, p=0.001) and infiltrative type (HR 4.504, 95% CI 1.579-12.848, p=0.005) were independent predictable factors for survival. Conclusions: The recent proposed methods (EASL and mRECIST) showed different characteristics from preexisting methods (WHO and RECIST) in assessing treatment response and furthermore, presence or absence of response by EASL methods was independent predictor for survival in HCC patients receiving TACE. EASL method should be applied as response assessment method for HCC patients receiving TACE.