Clinical Study On Response Evaluation Criteria For The Assessment Of Tumor Response To Sorafenib In Combination With Transarterial Chemoembolization In Hepatocellular Carcinoma

Background&Aims There was no study investigating the use of RECIST, EASL andmRECIST for assessing treatment response to combination therapy of sorafenib andtransarterial chemoembolization in patients with hepatocellular carcinoma (HCC). Theaim of this study was to explore the earliest time point for accurate assessment of responseto combination therapy in HCC patients, as well as to compare the three criteria and tovalidate the prognostic value of them when applied at this early time point post-therapy.Methods114consecutive patients, receiving combination therapy during the period ofJune,2008to July,2011,with sufficient imaging follow-up entered this study. RECIST,EASL and mRECIST were applied for evaluation of response. Kaplan–Meier survivalanalyses were used to explore differences in overall survival (OS) between responders andnon-responders, and the earlist time for response assessment was found out. Responsesassessed by different mothod were compared at this early evaluation time points. Cox regression and Kaplan–Meier survival analyses were used to validate the prognostic valueof tumor response assessed by each method to predict long-term survival in patients withHCC underwent combination treatment of sorafenib and TACE.Results At1-2-and2-3-month post-therapy, no difference in OS between responders andnonresponders were observed regardless of which criterion was used. At3-4-month,Kaplan–Meier survival analyses demonstrated a significantly prolonged survival in EASLand mRECIST responders, compared with nonresponders. However, there was still nosignificant association between treatment response and survival based on RECIST.Response rate at3-4-month obtained using EASL (50.6%) or mRECIST (51.6%) wasmuch higher than that assessed with RECIST (16.5%). The intermethod agreementbetween mRECIST and EASL was perfect (k=0.9), while it was slight (k=0.3) betweenmRECIST and RECIST. According to Cox proportional hazard model, a52%and a50%risk reduction were observed for EASL and mRECIST responders respectively, comparedto nonresponders.Conclusions3-4-month post-treatment is considered to be the earliest time point forevaluation of response to combination therapy. EASL and mRECIST responses areindependent predictors for OS when assessed at this early time point.