The most common etiology of hepatocellular carcinoma (HCC) in China is hepatitis B virus infection, which is in marked contrast to the United Kingdom with common etiologies like alcoholism, non-alcoholic fatty liver disease, and hepatitis C virus infection. There is no doubt that a randomized comparative trial (RCT), involving a large number of patients with HCC, would provide a clearer answer to the point raised by Metussin et al., in their Letter to the Editor. Unfortunately, such a study would be difficult to carry out. Although in our study there were more Child-Pugh B patients in the TACE group, there was no significant difference between the 2 groups. In response to the criticisms about the exclusion from our study of patients 70 years or older, this RCT was conducted on a selected group of patients, who were suitable to undergo both partial hepatectomy (PH) and TACE, and chronological age of over 70 years alone has never been a contraindication to these treatments in our center.Clinical practice guidelines for HCC state that drug-eluting beads (DEB) had similar response rates when compared with conventional TACE [1]. However, DEB is recommended for its less systemic adverse events. Although many clinical studies on DEB have been reported, the benefits on survival of DEB, when compared with conventional TACE, remain controversial. Until now, there has not been any clinical comparative trial which compares DEB with PH on survival in patients with BCLC B stage HCC. A study [2] showed a mean survival of 21.7 months in patients treated with DEB-TACE, which was less than the 41 months in the PH group in our study. The two studies used similar criteria to enroll patients. Our study suggested that PH would produce better overall survival than DEB.It is controversial to perform partial hepatectomy (PH) on hepatocellular carcinoma (HCC) patients with severe cirrhosis. Thus, at the beginning of the study, severe cirrhosis and a low platelet (PLT) were considered as exclusion criteria of the study.In our study, the receiver operating characteristic curve for PLT showed an optimal cut-off value of 121 Â 10 9 /L in detecting death (area under the ROC curve = 0.415). In our study, univariate analysis using the Log-rank test showed the variable for PLT <121 Â 10 9 /L could not be entered into multivariate analysis, according to our pre-set criteria (v 2 = 3.287, p = 0.07). In response to the criticism by Pang et al., in their Letter to the Editor, we carried out another multivariate analysis to include PLT. The result was still the same as with our previous conclusions: the type of treatment, number of tumor, and gender were independent risk factors associated with overall survival (OS). PLT was not an independent risk factor (HR, 0.743; 95% CI, 0.485 to 1.136, p = 0.170).Dr. Qing Pang and his associates stated in their letter that patients with a high preoperative PLT had significantly worse OS and recurrence free survival (RFS). This statement was based on a small sample size, retrospective clinical study, which was rec...