Blinded independent central response assessment using RECIST, mRECIST, and CHOI criteria in patients treated with sorafenib for advanced hepatocellular carcinoma (HCC).

172 Background: Despite survival benefit in patients with advanced HCC, sorafenib has been infrequently associated with changes in tumors dimensions, challenging standard RECIST criteria. Experts have suggested new response criteria based on changes in tumor vascularization and density (mRECIST and Choi). Aim is to evaluate response in patients treated with sorafenib using three different methods and their correlation with overall survival (OS) Methods: Radiological evaluations were performed from consecutive patients treated with sorafenib from January 2007 to February 2009. Patients were considered evaluable if CT scans were performed within 6 weeks before sorafenib and in the three months following treatment initiation at the first evaluation. A radiologist blinded for clinical data performed tumor evaluations according to RECIST, mRECIST, and Choi criteria. Best response was correlated with OS Results: Among 82 patients, 60 were evaluable (median age 63, male/female 52/8). The majority of patients were Child A (80%). Tumor staging showed BCLC stage B and C in 33% and 67%, respectively. The median duration of treatment with sorafenib was 6.6 months (range 1.2-16.5). Tumor evaluations by CT scan were performed in a median of 2.1 months (range 1.5-3.2). For the entire cohort, median OS was 9.9 months, and survival rates at 6, 12, and 24 months were 69%, 39% and 22%, respectively. The table summarizes the best responses measured using RECIST, mRECIST, and Choi criteria. CONCLUSIONS Better OS was observed in patients with complete/partial response than in patients with stable disease and progression, regardless the type of radiological criteria. However, mRECIST and Choi criteria identified more adequately than RECIST subgroups of patients with complete/partial response who also experienced longer survival [Table: see text].