Efficacy of entecavir versus tenofovir disoproxil fumarate in preventing hepatocellular carcinoma—The jury is out

Chronic hepatitis B (CHB) is a progressive disease and leads to cirrhosis, cirrhotic complications, hepatocellular carcinoma (HCC), and death in a significant proportion of patients. Long-term treatment with nucleos(t)ide analogues (NUCs) significantly reduces the incidences of adverse liver-related events and improves survival in patients with CHB.¹ Although NUC therapy reduces the incidence of HCC, it does not completely eliminate this risk. Entecavir (ETV), tenfovir disoproxil fumarate (TDF), and tenofovir alafenamide are the recommended first-line NUCs for patients with CHB.¹ Whether the effectiveness for preventing HCC differs among these drugs remains unclear. Choi et al.² first reported that TDF treatment was associated with a significantly lower risk of HCC compared with ETV treatment in a population-based cohort, which was validated in a hospital-based cohort. Subsequently, some retrospective observational studies were conducted to compare the risk of HCC between patients treated with TDF versus those treated with ETV but contradictory results have been obtained.^3-7 As such, meta-analyses have been conducted to resolve the discrepancies.^3-7 Despite similar primary studies being included, the statistical significance of the difference between these two drugs varies among these meta-analyses, although more studies reported a lower risk of HCC with TDF treatment.^3-7 Furthermore, whether TDF treatment confers a lower risk of HCC in patients with cirrhosis than ETV treatment remains controversial.^{8, 9} The biological mechanisms underlying this differential effect remain unclear.

In this issue, Lin et al. compared the risk of HCC in a consecutive cohort of patients with hepatitis B virus (HBV)-related liver cirrhosis who received ETV (n = 198) versus TDF (n = 88) treatment.¹⁰ There were no significant differences in demographics or baseline characteristics between the ETV and TDF groups. During a median follow-up of 57.5 months, 25 (12.6%) patients in the ETV group developed HCC compared to 12 (13.6%) patients in the TDF group. The 5-year cumulative incidence of HCC was comparable between the ETV and TDF groups (6.57% vs. 9.09%, p = .242). They further calculated the standardized incidence ratios of HCC in both groups by comparing the observed incidence with that predicted by the REACH-B model. The observed incidence was not significantly different from the predicted incidence in either group. Multivariable Cox regression analysis identified age, male, diabetes, and platelet as the independent predictors for HCC development. They concluded that ETV and TDF provided comparable preventive effects on HCC development in patients with HBV-related liver cirrhosis.

Although ETV and TDF have been shown to reduce the risk of HCC in patients with HBV-related liver cirrhosis compared to untreated controls, their relative effectiveness remains controversial.^{11, 12} Lin et al.¹⁰ demonstrated that the incidences of HCC were similar across the ETV and TDF groups and the observed incidence was not significantly different than expected when compared to the incidence predicted by the REACH-B model. First, the small number of patients enrolled in this study is probably statistically underpowered to address the question. Second, the REACH-B model was derived from patients without cirrhosis.¹³ Application of this model for the risk prediction of HCC in cirrhotic patients may have underestimated the risk and thus negated the preventive effect of NUC therapy. A prospective, randomized controlled study is ideally needed to compare ETV and TDF for the efficacy of HCC prevention but such trial is unlikely in real-world settings. Multiple factors can confound HCC development in patients with CHB, including age, sex, fibrosis stage, viral load, HBeAg status, alanine transaminase, comorbidities, and concurrent medications.¹⁴ Furthermore, ETV was introduced earlier than TDF in East Asia. Patients with severe liver disease may have been preferentially treated and followed up longer with ETV than TDF. Patients with older age, diabetes, and chronic kidney disease are more likely to receive ETV rather than TDF due to renal and bone safety concerns. As such, heterogeneity in patient characteristics is expected across the ETV and TDF groups within the same study or between different studies, which may significantly impact the risk of HCC.¹⁴ Although researchers have used propensity score matching and weighting to adjust the imbalances in baseline patient characteristics across treatment arms, some residual or unmeasured confounding may have remained to impact the estimates, particularly in studies using administrative claims database, which often lacks laboratory results.^{7, 14} Aggregate data meta-analyses of the published observational studies have also been conducted to clarify the controversy. Nonetheless, differences in eligibility criteria, patient populations, durations of treatment and follow-up, and statistical analysis methods may account for the discrepant results. A meta-analysis of 11 studies involving 42 939 patients (n = 35 960 for ETV vs. n = 6979 for TDF) using individual patient data was recently conducted to overcome the limitations with aggregate data meta-analyses.¹⁵ Patients receiving TDF had a significantly lower risk of HCC compared with ETV (adjusted hazard ratio [HR] 0.77; 95% confidence interval [CI] 0.61–0.98; p = .03). Although lower HCC risk with TDF was consistently observed in propensity score matching or weighting analysis and all subgroup analyses, not all differences across treatment arms were statistically significant. Of note, the difference was not significant in the propensity score weighting analysis (HR 0.83; 95% CI 0.67–1.03; p = .10), in patients initiating treatment after 2011 (adjusted HR 0.83; 95% CI 0.66–1.05; p = .11), or in cirrhotic (HR 0.81; 95% CI 0.65–1.01; p > .05) and non-cirrhotic (HR 0.73; 95% CI 0.49–1.09; p > .05) subgroups.¹⁵ Despite this tremendous effort in clarifying the relative effectiveness of these two drugs for preventing HCC, the question remains open.

In conclusion, the choice between ETV and TDF should be guided by patient factors until more robust evidence regarding their relative effectiveness in HCC prevention becomes available. Prospective enrollment of these patients with comprehensive characterization of relevant confounding factors and adequate long-term follow-up may help clarify the controversy.

The author declares no conflicts of interest.