To the Editor:
We read with interest the study by Kim et al. that was recently published in the Journal of Hepatology.
[1]
The authors concluded that the risk of hepatocellular carcinoma (HCC) did not differ between patients treated with entecavir (ETV) and those treated with tenofovir disoproxil fumarate (TDF). This seems to contradict our earlier publication, which showed a lower risk of HCC with TDF than with ETV.[2]
The contradicting results should be interpreted with caution, taking into account differences in the characteristics of the study population, study design, statistical power, and potential confounding.The most prominent difference between the 2 studies is the number of patients with cirrhosis included in the analyses. Although the 2 studies applied a similar definition for cirrhosis (image diagnosis or platelet count, <150,000/µl), cirrhosis was more prevalent in our study population (58.8%) than in that of Kim et al. (31.4%). The number of patients included in the propensity score-matched analysis was greater in our study (510 vs. 380 pairs). Furthermore, the mean baseline hepatitis B virus (HBV) DNA levels were higher in our study (6.6 vs. 5.6 log10 IU/ml). Lastly, patients with decompensated cirrhosis were included in our study but excluded in the study of Kim et al. All these factors indicate that the patients in our study had a higher risk of developing HCC than those in the study by Kim et al. We are particularly curious why the authors excluded patients with decompensated cirrhosis in their analysis. It is already well known that hepatic decompensation is reversible in most patients by treatment with ETV or TDF.
3
, 4
Patients with decompensated cirrhosis have the highest risk of HCC, which may increase the statistical power of the analyses greatly. It is obvious that with insufficient statistical power, a significant difference in HCC incidence could not be identified between the 2 treatments.Another important issue to be raised is the consideration of treatment modification during follow-up. In our study, treatment was modified in 11.7% of patients in the ETV group, mostly to TDF, during follow-up because of incomplete viral suppression (HBV DNA level, ≥100 IU/ml), virological breakthrough, or drug resistance. We censored the patients at 6 months after the treatment modification. However, Kim, et al. did not report the rate of treatment modification in their study, and it seems that the factor was not adjusted in their analyses. If TDF confers a lower risk of HCC than ETV, the patients in the ETV group with treatment modification from ETV to TDF during follow-up may have a reduced risk of HCC, consequently diluting the statistically significant difference in HCC risk between the 2 original groups.
We also have a concern about the propensity score-matched analysis in the study of Kim et al. They used only 9 variables for the matching, and well-known predictors of HCC, such as HBV DNA levels, alanine aminotransferase levels, and international normalized ratio, were not included in the matching for an unreported reason, despite the availability of data. As many variables are included in the matching process, the risk of confounding can be minimized, and we used 16 matching variables in our study.
The annual incidence of HCC in patients with chronic HBV infection ranges from 0.2% to 1% without cirrhosis and from 1% to 5% with cirrhosis. Comparative effectiveness research that compares these rare outcomes requires sufficient numbers of patients and outcome events to ensure statistical power. Showing non-significance in clinical outcomes between the 2 treatments is easier than identifying statistically significant differences in comparative effectiveness research.
[5]
This may explain why many previous studies, including the study of Kim et al., have shown no significant difference in risk of HCC among drugs.1
, 6
, 7
, 8
Collectively, it is important to note that all the studies that compared the risk of HCC between TDF and ETV therapies have indicated one direction favoring TDF or no direction. No study has shown the opposite direction of favoring ETV over TDF.
6
, 7
, 8
Even the study by Kim et al. also indicated a lower risk of HCC with TDF in patients with cirrhosis (hazard ratio 0.85; HCC incidence at 5 years of treatment, 16.0% with TDF vs. 20.9% with ETV), although the difference was not statistically significant. A meta-analysis consisting of 7 studies (3,698 patients) reported a lower incidence of HCC in patients with TDF than in those with ETV.[9]
Recently, another large historical cohort study from Hong Kong showed a significantly lower risk of HCC in TDF than in ETV.[10]
Given that a randomized clinical trial, which is the optimum for this topic, cannot be conducted in the future, we have to depend heavily on the results of observational studies. Accordingly, caution is required in interpreting the results from observational studies, considering whether they have sufficient numbers of patients and outcome events, with high internal validity in study design and analysis.
Conflict of interest
The authors declare no conflicts of interest that pertain to this work.
Please refer to the accompanying ICMJE disclosure forms for further details.
Supplementary data
The following are the Supplementary data to this article:
- Supplementary Data
References
Author names in bold designate shared co-first authorship
- A multi-center study of entecavir vs. tenofovir on prognosis of treatment-naive chronic hepatitis B in the Republic of Korea.J Hepatol. 2019; 71: 456-464
- Risk of hepatocellular carcinoma in patients treated with entecavir vs tenofovir for chronic hepatitis B: a Korean Nationwide cohort study.JAMA Oncol. 2019; 5: 30-36
- Efficacy of entecavir in treatment-naive patients with hepatitis B virus-related decompensated cirrhosis.J Hepatol. 2010; 52: 176-182
- Long-term effect of antiviral therapy on disease course after decompensation in patients with hepatitis B virus-related cirrhosis.Hepatology. 2015; 61: 1809-1820
- Tenofovir vs entecavir for hepatocellular carcinoma prevention in patients with chronic hepatitis B: one of these things is not like the other.JAMA Oncol. 2019; 5: 17-18
- Mortality, liver transplantation and hepatic complications in patients with treatment-naive chronic hepatitis B treated with entecavir vs tenofovir.J Viral Hepat. 2018; 25: 1565-1575
- Real-world single-center experience with entecavir and tenofovir disoproxil fumarate in treatment-naive and experienced patients with chronic hepatitis B.Saudi J Gastroenterol. 2018; 24: 326-335
- Long-term efficacy and safety of lamivudine, entecavir, and tenofovir for treatment of hepatitis B virus-related cirrhosis.Clin Gastroenterol Hepatol. 2013; 11: 88-94
- The effectiveness of TDF versus ETV on incidence of HCC in CHB patients: a meta analysis.BMC Cancer. 2019; 19: 511
- LBO-03-Tenofovir treatment has lower risk of hepatocellular carcinoma than entecavir treatment in patients with chronic hepatitis B.J Hepatol. 2019; 70e128
Article info
Publication history
Published online: July 24, 2019
Identification
Copyright
© 2019 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
