Hepatitis B virus (HBV) infection affects 240 million people worldwide and causes 686 000 deaths per year.1 For patients with chronic hepatitis B, hepatocellular carcinoma is a major cause of mortality.1, 2, 3, 4 Therefore, reducing hepatocellular carcinoma risk with long-term antiviral therapy is one of the cornerstones in the management of chronic hepatitis B, for which the first-line drugs are entecavir and tenofovir disoproxil fumarate.1, 2, 5
Although both entecavir and tenofovir disoproxil fumarate reduce hepatocellular carcinoma risk, the comparative effectiveness of these two drugs remains unclear.6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 Between 2019 and 2020, several meta-analyses20, 21, 22, 23, 24, 25, 26, 27 synthesised the evidence, but findings comparing the two drugs are still conflicting. The pooled results reported to date were often drawn from studies with heterogeneous populations, making results difficult to interpret. In addition, previous meta-analyses have included only a few comparative studies, pooled unadjusted data with adjusted data, did not exclude articles with overlapping populations, analysed hepatocellular carcinoma data as a dichotomous outcome as opposed to time-to-event data, and lacked detailed subgroup analysis, all of which have restricted the study conclusions.20, 21, 22, 23, 24, 25, 26, 27
To clarify this controversy, we critically appraised the existing literature in an updated systematic review and meta-analysis, which included several newly published articles that were not included in previous meta-analyses. We also evaluated heterogeneity, which was not adequately addressed in previous publications. Our analysis includes studies that investigated hepatocellular carcinoma incidence with the use of tenofovir disoproxil fumarate or entecavir in patients with chronic hepatitis B from both comparative and non-comparative studies. By including non-comparative studies we could elucidate heterogeneity in the risk for hepatocellular carcinoma, which might not be equal between patients who are given entecavir and tenofovir disoproxil fumarate. Furthermore, we determined pooled hazard ratios (HRs) for the risk of hepatocellular carcinoma in patients treated with tenofovir disoproxil fumarate versus entecavir, taken from comparative studies with covariate adjustment and subgroup analyses. Lastly, we estimated the pooled HRs for mortality and outcomes following a liver transplant, when data were available.
Research in context
Evidence before this study
It remains unclear whether tenofovir disoproxil fumarate and entecavir differ in their association with the risk for hepatocellular carcinoma in patients with chronic hepatitis B. Before undertaking this study, we searched PubMed, Embase, Web of Science, and the Cochrane Library using the key search terms including “hepatocellular carcinoma”, “hepatitis B virus”, and “tenofovir disoproxil fumarate or entecavir”, without language restrictions. We identified eight existing meta-analyses with inconsistent conclusions, possibly because of the heterogeneity that was not fully addressed and study limitations which were overlooked.
Added value of this study
In this systematic review and meta-analysis, we identified 31 studies including several new studies that were not included in previous meta-analyses. First, we showed that the pooled 5-year cumulative hepatocellular carcinoma incidence with tenofovir disoproxil fumarate and with entecavir were similar (p=0·87) from all eight studies with matched populations using propensity score matching. Second, we showed there was a similar association between hepatocellular carcinoma incidence and the use of tenofovir disoproxil fumarate or entecavir. Lastly, there was no difference between tenofovir disoproxil fumarate and entecavir in studies with no or minimal disparity in the follow-up duration, but tenofovir disoproxil fumarate was associated with a significantly lower risk of hepatocellular carcinoma compared with entecavir in studies in which the follow-up duration for entecavir was longer than that of tenofovir disoproxil fumarate by at least 1 year.
Implications of all the available evidence
Tenofovir disoproxil fumarate and entecavir should be considered equally effective for the prevention of hepatocellular carcinoma; therefore, the choice of treatment should not be made by an assumed difference in treatment efficacy but will depend on factors such as cost and tolerability.