Trans-arterial Chemo-embolization Is Safe and Effective for Elderly Advanced Hepatocellular Carcinoma Patients

Results From an International Database

Matan J. Cohen; Izhar Levy; Orly Barak; Allan I. Bloom; Mario Fernández-Ruiz; Massimo Di Maio; Francesco Perrone; Ronnie T. Poon; Daniel Shouval; Thomas Yau; Oren Shibolet


Liver International. 2014;34(7):1109-1117. 

In This Article


We performed a literature search of the medical literature reporting prognosis following TACE among patients with HCC, which included data stratification according to age groups. We did not limit our search to the English language, nor did we search for a specified age cut-off for defining elderly patients. Relevant publications were reviewed to confirm that TACE was provided as sole treatment for HCC and that elderly patients were part of the data set.

We contacted all corresponding authors and proposed collaboration and sharing of the databases to generate one structured database. These data were combined to create a metadatabase. Because the studies were designed in different countries at different times and with different parameters, there were only a few overlapping variables collected in all studies. The available data sets, together, allowed for the generation of a larger data set and provided insight into differences between the published series with regard to demographics, disease stage, outcome and side effects. We hypothesized that there would not be significant variability among study outcomes, confirming the role of TACE for elderly patients with advanced HCC.

All data sets collected have no patient-unique identifying features (personal identification numbers, social security number etc.). Data included the following variables: age at diagnosis, patient sex, date of diagnosis, HBV status, HCV status, cirrhosis at diagnosis, Child-Pugh score (CPT) at diagnosis, Okuda stage and Cancer of the Liver Italian Program (CLIP) score at diagnosis, survival time until censorship or death, mortality status. Each participating group had local institutional board approval for their cohort study. Two data sets also had available data regarding the number of TACE performed per patient and post-TACE complications: vascular damage, significant haemorrhage, acute kidney injury and deterioration in liver function. All data were collected into a single data set file and the geographical origin was also included as a variable. In all data sets, HCC diagnosis was determined in accordance with established guidelines published by both the EASL and the AASLD.[4,3] We note that IRB approval limited the Chinese cohort from providing data for patients younger than 65.

TACE Technique

The CLIP multicentre trials (CLIP-01 and CLIP-03) did not impose a specific TACE procedure and the technique was left at each centre's discretion according to their best clinical practice. In the Spanish study, after arteriography and portography, a mixture of lipiodol and doxorubicin (40–60 mg) or cisplatin (50–100 mg) and iodized oil (5–30 ml) was injected into the hepatic artery until stasis within the tumour vessels occurred. The feeding arteries to the HCC were embolized with gelatin sponge particles. In the Israeli center, after vascular imaging, whenever possible, super selective TACE was attempted using a co-axial microcatheter and a mixture of 50 mg doxorubicin with lipiodol oil and gelfoam slurry or powder. If dictated by tumour burden and feasibility (hepatic reserve, Child Pugh status), segmental or lobar TACE was performed. Percutaneous vascular closure devices were routinely employed from 2007 onward. Finally, the procedure in Hong-Kong included arteriography and superior mesenteric arterial portovenography. The right or left hepatic artery feeding the tumour was superselectively catheterized. Using the pumping method an emulsion was prepared by mixing cisplatin (1 mg/ml) with Lipiodol in a volume ratio of 1 to 1. Various amounts of the emulsion, up to a maximum of 60 ml (containing 30 mg of cisplatin) were injected slowly under fluoroscopic monitoring according to the size of the tumour and the arterial blood flow. This was followed by embolization with small gelatin sponge (Spongostan; Ferrosan, Johnson & Johnson Medical Ltd., Skipton, England) pellets of 1 mm diameter mixed with 40 mg of gentamicin.


Descriptive statistics were employed to present age groups characteristics. Categorical variables are presented as percentages and distributions, and continuous variables are presented with mean and standard deviation (SD). Associations between categorical variables were assessed with the Fisher exact test, and comparison of continuous variables was performed with the Student's t-test or with the Mann–Whitney U test. Survival charts were generated employing the Kaplan–Meier method and survival curves were compared with the log-rank test. Follow-up times were estimated using the reverse Kaplan–Meier methods.[23] After confirming that the proportional hazard assumptions were met, we determined the association of age group and potential confounders with mortality using the Cox proportional hazard regression analysis. To this end, we assessed several models; firstly, we included age at diagnosis, sex, year of diagnosis, HBV and HCV status, cirrhosis on diagnosis and disease stage. Not all data sets had all these variables available, and therefore, we repeated the models without the missing variables. All analyses were stratified per study of origin. We performed several sensitivity analyses. We repeated analysis without stratifying per study of origin. In addition, we assessed age at diagnosis as a continuous variable and also as a categorical variable in two strategies: (i) either above 80 years of age or lower; (ii) In three categories – younger than 65, 65–75 and above 75. We repeated the analyses with various risk stratification/staging scores. Because of the absence of patients younger than 65 from the Chinese study, we additionally analysed all included data for patients 65 years of age at diagnosis and older and stratified age group to patients between 65 and 70, 70–75 and older than 75. In all analyses, two tailed P-values below 0.05 were considered statistically significant.