We read with great interest the article by Jung et al., which evaluated the association between telomere maintenance gene polymorphisms and the development and clinical outcomes of hepatitis B virus (HBV)-associated hepatocellular carcinoma (HCC). They suggest that telomere maintenance gene polymorphisms may act as a prognostic factor in predicting clinical outcomes of Korean HCC patients with chronic HBV infection. Nevertheless, close inspection of the study revealed some minor issues that are worth mentioning and clarifying.
First, in Jung et al.'s study, genotype identification was performed with the GenomeLab SNPstream (Ultra-high throughput; UHT), which uses the multiplexed polymerase chain reaction in conjunction with tag array single-base extension genotyping. However, the measures for genotyping quality control, such as positive and negative controls, blindness to the case-control status, a different genotyping assay to confirm the data, and/or random repetition of a portion of samples were less described in the paper. Did the authors perform these procedures? If the authors did not make quality control for genotyping, a bias could not be avoided because the genotyping data by the UHT method could not achieve a 100% accuracy rate except for DNA direct sequencing.
Second, when testing telomere maintenance gene polymorphism and HCC risk, the odds ratios (ORs) and their 95% confidence intervals (CIs) were only adjusted for age, gender, Child-Pugh class, alpha-fetoprotein (AFP) level, and HBV-DNA level. As is well known to us, other factors, such as alcohol consumption, smoking, and family history of HCC are also risk factors for HCC. Why were the ORs and their 95% CIs not adjusted for these factors? Were these factors in the studied population similar between HCC cases and controls?
Third, the authors found that advanced tumor staging and tumor postoperative recurrence were independent predictors of reduced survival. After careful analysis, we found that the tumor postoperative recurrence was the most important factor that predicted reduced survival in the paper (as shown in table 7 of Jung et al.'s study, HR for tumor postoperative recurrence was 5.629, while the HR for the combined high-risk genotype was only 1.542). Therefore, when testing telomere maintenance gene polymorphism and overall survival of HCC, tumor staging and tumor postoperative recurrence were the most important confounding factors, although these factors were adjusted for in the multivariate analysis. In view of this, Kong et al. performed survival analysis according to significant genotype stratified by tumor stage when they investigated the association between vascular endothelial growth factor gene polymorphisms and HCC patient prognosis. If the authors could also perform survival analysis according to significant genotype stratified by tumor staging and tumor postoperative recurrence, the results would be more convincing.
Fei Liu, M.D.
Yong-Gang Wei, M.D.
Wen-Tao Wang, M.D.
Bo Li, M.D.
Department of Liver Surgery
Liver Transplantation Center
West China Hospital of Sichuan University
Chengdu, Sichuan Province, China