Search

To the Editor: The study by Suh and coworkers [1] addresses a hot topic, namely the possibility to predict prognosis (recurrence and survival) in patients submitted to curative treatments for early hepatocellular carcinoma (HCC). The authors conclude that patients submitted to radiofrequency ablation are at higher risk of local intrahepatic recurrence in comparison to surgery and that survival is compromised in case of increased production of a-fetoprotein (AFP) by prothrombin, induced by vitamin K absence-II (PIVKA-II), compared to all other groups (not increased AFP by PIVKA-II product or surgical resection). We agree with the conclusion of the higher risk of recurrence, as we recently demonstrated [2], but in our view the second conclusion is not sufficiently supported by evidence. Moreover, some incorrect reporting of data prevent a full understanding of the study. Regarding the latter point, results about AFP and PIVKA-II were reported as mean values. This is unacceptable, as such data are not expected to follow a Gaussian distribution, thus, mean values are poorly or totally non-informative or may even be misleading under these circumstances. Median values have to be reported. Moreover, readers cannot understand how many patients had normal values of these two parameters. Secondly, the authors themselves point out that patients submitted to ablation had a slightly more compromised liver function than surgical patients. It is evident that cirrhotic patients with poorer liver function have a worse overall survival, regardless of the tumours status. We have recently shown in this Journal that resection may apparently provide better survival in early HCC, but only when entire patient series are analysed [2]. Such difference disappeared in very early HCC when patients were balanced in terms of background liver function and tumour features [2]. Suh and colleagues [1] did not show whether the four subgroups were superimposable in terms of highly relevant factors, such as liver function variables, tumour size and number, thus making the issue of survival questionable. This would be recommendable, especially in the two groups of patients, submitted to ablation, for which the information, derived from histology on which prognosis was designed, was not available. Thus, a re-analysis of the Korean data appears to be well-deserved before the conclusions, proposed by the authors, can be adopted.

Background & Aims Aim of the study was to assess the clinical impact of conventional transarterial chemoembolization (cTACE) repeated "on demand" on HCC outcome. Outcome measures were: response rate to first and repeated cTACE, recurrence rates and overall survival. Methods The outcome of 151 consecutive HCC patients submitted to a first cTACE from January 2004 to December 2005 was retrospectively analyzed. Results Complete radiological response (CR) was observed in 72/151 (48%), 34/60 (52%) and 12/22 (55%) patients after first, second and third cTACE, respectively. Recurrence rates at 6 and 12months were 37% and 61% after the first cTACE, and 40% and 59% after the second cTACE, respectively. Patients not achieving CR or with a recurrence after CR not treated with curative therapies were 94 and 84 after first and second cTACE, respectively. Of these, 60/94 (64%) and 22/84 (26%) were submitted to a second and third cTACE, respectively. Median overall survival was 32.0months but 25.0months excluding transplanted patients. Factors at the time of first cTACE associated with overall shorter survival at multivariate analysis were higher bilirubin, higher AFP and not achieving CR. Conclusions CR and recurrence rates after first and second cTACE were similar. About 64% of patients were submitted to second cTACE, while only few patients (26%) were submitted to third cTACE using an "on demand" policy. These figures may be also useful for planning trials for the evaluation of the efficacy of repeated TACE vs . TACE combined with adjuvant treatments or vs . systemic treatments.