Impact of tenofovir vs. entecavir treatment on progression of chronic hepatitis B: A nationwide cohort study

Abstract

Background & aims: Conflicting evidence exists on hepatocellular carcinoma (HCC) risk in patients with chronic hepatitis B (CHB) receiving tenofovir vs. entecavir. We assessed the impacts of the two drugs on the clinical trajectory of CHB at a population level.

Methods: We conducted a retrospective nationwide cohort study using data from Taiwan's National Health Insurance Research Database, including 55,885 patients with CHB who were treatment-naïve aged 30-75 years receiving tenofovir (n = 17,137) or entecavir (n = 38,748) monotherapy for ≥3 months between November 2009 and December 2020, and followed until December 2022. Multi-state modeling was applied to evaluate treatment impacts on disease progression trajectory, comprising CHB (with/without compensated cirrhosis), hepatic decompensation, HCC, death, and a switching treatment state. Propensity score matching/weighting and subgroup analyses were used to confirm the robustness of findings.

Results: During a median 6.3 years of follow-up, 1,524 patients developed hepatic decompensation, 3,591 developed HCC, and 3,436 died. Tenofovir compared with entecavir was associated with lower risk of CHB progression, with adjusted-hazard ratios (95% CIs) of 0.84 (0.75-0.95) and 0.76 (0.70-0.83), respectively, for transitions to hepatic decompensation and HCC from baseline, and 0.65 (0.48-0.89) for HCC risk from decompensation. Propensity score matching/weighting analyses yielded similar results. Among patients without experiencing decompensation, a significantly lower HCC risk with tenofovir was observed across multiple subgroups, including age, sex, diabetes, and cirrhosis, and by sensitivity analyses. The 5-year risk of major adverse liver-related outcomes (decompensation, HCC, and liver-related deaths) was 5.5% and 7.5% for tenofovir and entecavir, respectively (adjusted-hazard ratio 0.80; 95% CI 0.74-0.85).

Conclusions: Using multi-state modeling on the temporal evolution of CHB severity, long-term tenofovir treatment compared with entecavir was associated with a lower risk of severe liver outcomes.

Impact and implications: The comparative effectiveness of tenofovir vs. entecavir treatment in preventing hepatocellular carcinoma (HCC) for patients with chronic hepatitis B (CHB) remains controversial. Using a nationwide cohort study involving 55,885 patients to investigate CHB as a multi-state disease over 13 years, we show that tenofovir (vs. entecavir) treatment is associated with lower risks of HCC, hepatic decompensation, and a composite endpoint of adverse liver-related outcomes (hepatic decompensation, HCC, and liver-related deaths), regardless of age, sex, diabetes, alcohol-related disorders, and cirrhosis state. Our results provide new evidence justifying the use of tenofovir or entecavir to prevent the evolution of CHB severity.

Keywords: Ascites; Esophageal varices; Non-alcoholic fatty liver disease; Severe decompensated cirrhosis; Variceal bleeding.

Graphical abstract

Fig. 1

Flowchart of study patient selection. ∗A case can be excluded because of more than one criterion. Therefore, the total excluded cases in each step can be less than the sum of case numbers excluded by individual criterion. NHIRD, National Health Insurance Research Database.

Fig. 2

Flow diagram illustrating the multi-state trajectory analysis of progression from baseline to hepatic decompensation, then to hepatocellular carcinoma (HCC) or death. HCC, death, and treatment switching are absorbing states. ∗Deaths after HCC were not modeled as a separate outcome, and were not taken into the calculation of the number of liver-related deaths in the figure (details of causes of deaths are shown in Table S3). State-specific number of events is reported in boxes, and transition-specific number of events is reported on arrows.

Fig. 3

Cumulative risk of transition from baseline to major adverse liver-related outcomes between the tenofovir and entecavir groups. The plots show the probabilities estimated by the Aalen-Johansen estimator and multi-state modeling in total patients (A) and by baseline compensated cirrhosis (B). The risk of hepatic decompensation presented in the figure refers to ‘the risk of developing decompensation without having HCC’. Composite endpoint: HCC, hepatic decompensation, and liver-related deaths. HCC, hepatocellular carcinoma.

Fig. 4

Hazard ratios (95% CIs) for tenofovir vs. entecavir for each transition, stratified by age, sex, and comorbidity in the total patients with CHB. Hazard ratios (HRs) were derived from multi-state models adjusted for age, sex, and baseline comorbidities (compensated cirrhosis, diabetes, and alcohol-related disorders). (A) Baseline → hepatic decompensation. (B) Baseline → HCC. (C) Hepatic decompensation → HCC. CHB, chronic hepatitis B; HCC, hepatocellular carcinoma.

Fig. 5

Hazard ratios (95% CIs) for tenofovir vs. entecavir for each transition by length of follow-up in total patients with CHB. Composite endpoint: HCC, hepatic decompensation, and liver-related deaths. Hazard ratios (HRs) were derived from multi-state models adjusted for age, sex, and baseline comorbidities (compensated cirrhosis, diabetes, and alcohol-related disorders). CHB, chronic hepatitis B, HCC, hepatocellular carcinoma.