Superior Efficacy of Besifovir in Reducing Hepatocellular Carcinoma Risk Compared to Tenofovir in Chronic Hepatitis B Patients: A Nationwide Cohort Study Authors: Yoon E Shin, Jae Young Kim, Hyuk Kim, Jeong-Ju Yoo, Sang Gyune Kim, Young Seok Kim Published: 2026 ABSTRACT BackgroundBesifovir dipivoxil maleate (BSV) is a novel therapeutic option for the treatment of chronic hepatitis B (CHB). However, its long-term effectiveness in reducing the risk of hepatocellular carcinoma (HCC) remains unclear.MethodsWe conducted a retrospective cohort study using the Korean Health Insurance Review and Assessment Service database from 2018 to 2022. A total of 32,383 patients with CHB were included (BSV, n = 2,358; tenofovir disoproxil fumarate [TDF], n = 30,025). Propensity score matching was conducted at a 1:3 ratio to minimize selection bias. The primary outcome was the incidence of HCC, with cirrhosis development as the secondary endpoint.ResultsIn the overall population, 57 cases of HCC were reported in the BSV group (2,358 individuals, 5,918 person-years), while 1,268 cases of HCC were reported in the TDF group (30,025 individuals, 83,587 person-years). After propensity score matching, the incidence of HCC was significantly lower in the BSV group (9.63 per 1,000 person-years; 95% confidence interval [CI], 7.43–12.49) than in the TDF group (15.36 per 1,000 person-years; 95% CI, 13.72–17.18), with an incidence rate ratio of 1.59 (P = 0.001). This protective effect remained consistent in both cirrhotic (hazard ratio [HR], 1.609; P = 0.028) and non-cirrhotic subgroups (HR, 1.831; P = 0.002).ConclusionBSV treatment demonstrated superior efficacy in reducing the risk of HCC compared to TDF, regardless of baseline cirrhosis status. These findings suggest that BSV may provide enhanced protection against HCC in patients with CHB. INTRODUCTION Chronic hepatitis B (CHB) remains a major global health concern, affecting approximately 296 million people worldwide and contributing substantially to liver-related morbidity and mortality.12 Current first-line treatments for CHB include high-genetic-barrier nucleos(t)ide analogues, including entecavir (ETV), tenofovir disoproxil fumarate (TDF), and tenofovir alafenamide (TAF).34 These agents have demonstrated remarkable efficacy in viral suppression, reducing the risk of progression to cirrhosis and hepatocellular carcinoma (HCC) and improving survival rates.56 However, antiviral therapy requires indefinite treatment, which carries a risk of drug resistance and long-term side effects, especially those related to renal function and bone mineral density.7 Besifovir dipivoxil maleate (BSV), a novel acyclic nucleotide phosphonate derivative, has emerged as a promising therapeutic option for CHB.8 BSV is oxidized in the liver to its active form, LB80317, which selectively inhibits hepatitis B virus (HBV) DNA polymerase through competitive binding.9 Phase 3 clinical trials have demonstrated BSV’s non-inferiority to TDF in achieving a virologic response, with the added benefit of an improved safety profile, particularly with respect to renal and bone toxicity.1011 Furthermore, in vitro studies have shown BSV’s effectiveness against both adefovir- and tenofovir-resistant viral strains, suggesting its potential utility in the management of drug-resistant cases.12 Since its approval as a first-line treatment for CHB in South Korea in 2017, BSV has been increasingly accepted in clinical practice and adopted in the Korean clinical practice guidelines.13 Despite these favorable characteristics, data on the long-term clinical outcomes of BSV, particularly in relation to HCC prevention, remain limited. A multicenter cohort study by Lee et al.14 showed a lower risk of HCC in patients treated with BSV than ETV, TAF, and TDF; the study had the limitation of a relatively small sample size. Using nationwide data from the Korean Health Insurance Review and Assessment Service (HIRA), our group previously compared BSV with TAF and demonstrated no significant difference in HCC incidence (1.6 vs. 2.2 per 1,000 person-years; P = 0.284).15 However, the comparative effect against other antiviral agents could not be evaluated. This study aimed to fill the knowledge gap by comparing the incidence of HCC between the BSV and TDF treatment groups using comprehensive clinical data from the HIRA database. This large-scale analysis sought to provide robust evidence of BSV’s effectiveness in reducing HCC risk, a key outcome in CHB management. METHODS Data source This retrospective cohort study utilized data from the HIRA. South Korea’s healthcare system operates a government-led, single-payer insurance model with mandatory enrollment, resulting in a highly reliable and representative database that covers approximately 98% of the national population. Patient data, including personal information, diagnoses, procedures, test records, and prescribed medications, are managed using standardized codes, making the data highly usable. Diagnoses were defined using the International Classification of Diseases, 10th Revision (ICD-10) codes. Study population Eligible participants were adults (≥ 18 years) with CHB who initiated treatment with either BSV or TDF during the study period. CHB diagnosis was confirmed using ICD-10 codes (B18.0 and B18.1) and prescription records. A total of 119,122 treatment-naïve patients who received antiviral therapy (BSV, TDF, TAF, or ETV) for over three months between January 2018 and December 2022 were initially screened. Among these, 93,586 patients met the inclusion criteria. Patients were excluded under the following conditions: 1) a prior diagnosis of HCC or history of liver transplantation; 2) concurrent infection of hepatitis C virus or human immunodeficiency virus; 3) less than 12 months of follow-up duration; 4) prior exposure to other antiviral medications within 12 months; or 5) a diagnosis of any malignancy within five years before enrollment (Fig. 1). Study population flowchart. TDF = tenofovir disoproxil fumarate, BSV = besifovir dipivoxil maleate. Outcome and definition The primary outcome of this study was the incidence of HCC among CHB patients treated with BSV or TDF. Additionally, a subgroup analysis examined the development of cirrhosis among patients without pre-existing cirrhosis. CHB was defined using ICD-10 codes B18.0 or B18.1, while cirrhosis was defined using codes K74, K70.2, K70.3, K76.6, and K76.7. Antiviral drug exposure was measured using the Korea Drug (KD) code; TDF was represented by KD codes 493901ATB and 686500ATB, and BSV by KD code 665401ATB. HCC was defined by ICD-10 code C22.0 (Supplementary Table 1). Statistical analysis Baseline characteristics were assessed using Student’s t-tests for continuous variables and chi-square tests for categorical variables. To minimize potential confounding variables between the BSV and TDF groups, propensity score matching was performed at a 1:3 ratio. The matching criteria included age, sex, duration of medication use, presence of decompensation, and the Charlson Comorbidity Index (CCI) score. The CCI score, which was used to predict mortality by reflecting patients’ comorbid conditions, included the underlying conditions outlined in Supplementary Table 2. Incidence rates were calculated per 1,000 person-years with corresponding 95% confidence intervals (CIs). Kaplan-Meier curves and log-rank tests were used to evaluate differences over time. Additionally, multivariable Cox regression analysis was conducted. A two-sided P value < 0.05 was considered statistically significant. All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA). Ethics statement This study was conducted in accordance with the principles of the 1964 Declaration of Helsinki and was approved by the Institutional Review Board of Soonchunhyang University Bucheon Hospital under number SCHBC 2023-07-016-001, registered on September 23, 2023. The requirement for informed consent was waived due to the use of de-identified data. RESULTS Baseline characteristics of the study population A total of 32,383 individuals were included in the study, comprising 2,358 patients in the BSV group and 30,025 in the TDF group, resulting in the TDF group being approximately ten times larger. The overall mean age was 48.9 ± 12.1 years, and patients in the BSV group had a higher mean age of 50.7 ± 11.0 years. The CCI score, which reflects comorbid conditions, was similar between groups at around 2.5 ± 1.3. The prevalence of cirrhosis was higher in the BSV group (20.2%) compared to the TDF group (13.7%), while decompensation was more prevalent in the TDF group (10.3%) than in the BSV group (7.5%). After propensity score matching, 9,432 patients were included, comprising 2,358 patients in the BSV group and 7,074 patients in the TDF group. The matching process successfully balanced baseline characteristics between the two groups, with no significant differences in key demographic and clinical parameters. The mean age was comparable (BSV: 48.3 ± 12.7 years; TDF: 48.7 ± 12.4 years; P = 0.342), as was the gender distribution, with males comprising approximately 58% of each cohort. Notably, the prevalence of baseline cirrhosis was also comparable (BSV: 24.3%; TDF: 24.7%; P = 0.721) (Table 1). Table 1Baseline characteristics before and after propensity score matchingVariablesBefore PSMAfter PSMTotal (N = 32,383)BSV (n = 2,358)TDF (n = 30,025)P valueSMDTotal (N = 9,432)BSV (n = 2,358)TDF (n = 7,074)P valueSMDAge48.9 ± 12.150.7 ± 11.048.8 ± 12.1< 0.0010.16150.5 ± 11.350.7 ± 11.050.4 ± 11.40.2380.028SexFemale13,372 (41.3)951 (40.3)12,421 (41.4)0.3240.0213,888 (41.2)951 (40.3)2,937 (41.5)0.3100.024Male19,011 (58.7)1,407 (59.7)17,604 (58.6)0.3245,544 (58.8)1,407 (59.7)4,137 (58.5)0.310Duration of antiviral medication839.1 ± 519.7863.1 ± 480.8837.2 ± 522.60.0130.050858.4 ± 515.4863.1 ± 480.8856.8 ± 526.50.5910.012Comorbidity (CCI)2.5 ± 1.32.4 ± 0.92.5 ± 1.3< 0.001−0.1032.4 ± 0.92.4 ± 0.92.4 ± 0.90.7140.009Presence of cirrhosis4,586 (14.2)476 (20.2)4,110 (13.7)< 0.0011,423 (15.1)476 (20.2)947 (13.4)< 0.001Presence of decompensation3,255 (10.1)176 (7.5)3,079 (10.3)< 0.001703 (7.5)176 (7.5)527 (7.4)0.9820.001Data were presented as means ± standard deviations for continuous variables and as numbers (percentages) for categorical variables. Continuous variables were tested using the t-test, and categorical variables were tested using the χ2 test. An SMD ≤ 0.1 indicates adequate balance between the two groups.PSM = propensity score matching, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, SMD = standardized mean difference, CCI = Charlson Comorbidity Index. Baseline characteristics before and after propensity score matching Data were presented as means ± standard deviations for continuous variables and as numbers (percentages) for categorical variables. Continuous variables were tested using the t-test, and categorical variables were tested using the χ2 test. An SMD ≤ 0.1 indicates adequate balance between the two groups. PSM = propensity score matching, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, SMD = standardized mean difference, CCI = Charlson Comorbidity Index. Incidence of HCC The BSV group was observed for a total of 5,918 person-years, during which 57 cases of HCC were reported. The TDF group was observed for a total of 83,587 person-years, during which 1,268 cases of HCC were reported. The median follow-up period was 3.2 years (interquartile range, 2.4–3.9 years). The BSV group demonstrated significantly lower HCC incidence, with an incidence rate of 9.63 per 1,000 person-years (95% CI, 7.43–12.49), compared with 15.36 per 1,000 person-years (95% CI, 13.72–17.18) in the TDF group. The incidence rate ratio (IRR) of HCC between the two groups was 1.59 (95% CI, 1.20–2.12; P = 0.001), indicating substantially better protection against HCC development in the BSV group (Table 2). The Kaplan-Meier plot and log-rank test further confirmed this finding, showing significantly better HCC-free survival in the BSV group (log-rank P < 0.001) (Fig. 2). Table 2Incidence ratio of hepatocellular carcinoma in patients treated with BSV or TDFVariablesBefore PSMAfter PSMIR (95% CI)IRR (95% CI)P valueIR (95% CI)IRR (95% CI)P valueTotalBSV9.63 (7.43–12.49)Ref9.63 (7.43–12.49)RefTDF15.17 (14.36–16.03)1.57 (1.21–2.05)< 0.00115.36 (13.72–17.18)1.59 (1.20–2.12)0.001Non-cirrhosisBSV6.30 (4.40–9.00)Ref6.30 (4.40–9.00)RefTDF11.07 (10.33–11.86)1.76 (1.22–2.53)0.00211.16 (9.69–12.85)1.77 (1.21–2.60)0.004CirrhosisBSV23.42 (16.06–34.15)Ref23.42 (16.06–34.15)RefTDF42.43 (38.74–46.47)1.81 (1.23–2.67)0.00344.40 (36.86–53.47)1.90 (1.24–2.89)0.003BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, PSM = propensity score matching, IR = incidence ratio, CI = confidence interval, IRR = incidence rate ratio. Incidence ratio of hepatocellular carcinoma in patients treated with BSV or TDF BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, PSM = propensity score matching, IR = incidence ratio, CI = confidence interval, IRR = incidence rate ratio. Kaplan-Meier plots for HCC incidence by antiviral agents in all patients. (A) Before PSM. (B) After PSM. PSM = propensity score matching, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, HCC = hepatocellular carcinoma. The protective effect of BSV was consistent across different patient subgroups. Among patients with cirrhosis, BSV treatment was associated with a lower HCC incidence than the TDF treatment (18.24 vs. 29.37 per 1,000 person-years; hazard ratio [HR], 1.609; 95% CI, 1.054–2.456; P = 0.028). Similarly, in non-cirrhotic patients, BSV treatment demonstrated a superior protection compared with TDF treatment (6.83 vs. 12.51 per 1,000 person-years; HR, 1.831; 95% CI, 1.252–2.678; P = 0.002) (Table 2, Fig. 3). Kaplan-Meier plots for HCC incidence by antiviral agents in subgroup analyses. (A) Cirrhotic patients, Before PSM. (B) Cirrhotic patients, After PSM. (C) Non-cirrhotic patients, Before PSM. (D) Non-cirrhotic patients, After PSM. PSM = propensity score matching, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, HCC = hepatocellular carcinoma. Risk factors for HCC Multivariable Cox regression analysis identified TDF use, older age, male sex, cirrhosis, and decompensated liver disease as significant risk factors for HCC. In the overall population, the HR for HCC in the TDF group indicated a significantly higher risk compared to that in the BSV group (HR, 1.703; 95% CI, 1.282–2.262; P < 0.001). This trend was consistent across both non-cirrhotic and cirrhotic subgroups. In the non-cirrhotic subgroup, the HR for HCC was 1.831 (95% CI, 1.246–2.690; P = 0.002), while in the cirrhotic subgroup, the HR was 1.609 (95% CI, 1.054–2.455; P = 0.028) (Table 3, Supplementary Table 3). Table 3Cox regression analysis for hepatocellular carcinomaVariablesUnivariateMultivariateHR (95% CI)P valueHR (95% CI)P valueTotal patientsAntiviral medicationBSV1 (ref)1 (ref)TDF1.628 (1.227–2.161)< 0.0011.703 (1.282–2.262)< 0.001Age1.045 (1.035–1.054)< 0.0011.036 (1.025–1.046)< 0.001SexFemale1 (ref)1 (ref)Male2.025 (1.598–2.566)< 0.0012.182 (1.717–2.775)< 0.001CCI score1.177 (1.082–1.280)< 0.0010.990 (0.899–1.090)0.838Presence of cirrhosis3.722 (3.010–4.602)< 0.0012.057 (1.622–2.608)< 0.001Presence of decompensation6.323 (5.072–7.883)< 0.0013.810 (2.959–4.906)< 0.001Non-cirrhotic patientsAntiviral medicationBSV1 (ref)1 (ref)TDF1.810 (1.232–2.660)0.0031.831 (1.246–2.690)0.002Cirrhotic patientsAntiviral medicationBSV1 (ref)1 (ref)TDF1.933 (1.269–2.945)0.0021.609 (1.054–2.455)0.028HR = hazard ratio, CI = confidence interval, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, CCI = Charlson Comorbidity Index. Cox regression analysis for hepatocellular carcinoma HR = hazard ratio, CI = confidence interval, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, CCI = Charlson Comorbidity Index. Incidence of liver cirrhosis We examined additional benefits of BSV treatment by analyzing the incidence of cirrhosis. In the BSV group, 1,625 non-cirrhotic patients were observed over a total observation period of 3,916 person-years, during which 151 cases of cirrhosis were reported. In the TDF group, 24,093 non-cirrhotic patients were followed for a total observation period of 63,581 person-years, during which 2,168 cases of cirrhosis were reported. The cirrhosis IRR for the TDF group compared to the BSV group was 0.90 (95% CI, 0.75–1.07; P = 0.238), indicating no significant difference between the two groups (Supplementary Table 4). The Kaplan-Meier curve showed a similar incidence rate of cirrhosis between the groups (Fig. 4). However, multivariable Cox regression analysis revealed a significantly lower risk of cirrhosis in the TDF group than in the BSV group (HR, 0.670; 95% CI, 0.597–0.753; P < 0.001) (Table 4). Kaplan-Meier plots for cirrhosis incidence by antiviral agents in non-cirrhotic patients. (A) Before PSM. (B) After PSM. PSM = propensity score matching, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate. Table 4Cox regression analysis for cirrhosis in non-cirrhotic patientsVariablesUnivariateMultivariateHR (95% CI)P valueHR (95% CI)P valueAntiviral medicationBSV1 (ref)1 (ref)TDF0.945 (0.789–1.133)0.5430.670 (0.597–0.753)< 0.001Age1.034 (1.028–1.041)< 0.0011.033 (1.028–1.038)< 0.001SexFemale1 (ref)1 (ref)Male1.564 (1.331–1.838)< 0.0011.482 (1.324–1.659)< 0.001CCI score1.050 (0.969–1.137)0.231Medication duration1.000 (1.000–1.000)0.292HR = hazard ratio, CI = confidence interval, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, CCI = Charlson Comorbidity Index. Cox regression analysis for cirrhosis in non-cirrhotic patients HR = hazard ratio, CI = confidence interval, BSV = besifovir dipivoxil maleate, TDF = tenofovir disoproxil fumarate, CCI = Charlson Comorbidity Index. Sensitivity analysis To address potential selection bias due to differential treatment indications, a sensitivity analysis was conducted after excluding patients with 1) end-stage renal disease (ESRD), 2) decompensated cirrhosis, or 3) malignancy. In this restricted cohort, the incidence of HCC remained lower in the BSV group than in the TDF group (IRR, 2.01; 95% CI, 1.29–3.15; P = 0.002), consistent with the main analysis (Supplementary Table 5). Antiviral switching An additional analysis was performed to evaluate antiviral switching patterns in patients initially treated with BSV or TDF. Among the 2,358 patients who initiated BSV therapy, 165 patients (7.0%) switched to another antiviral agent during follow-up. The most frequent switching destination was TAF (43.6%), followed by TDF (34.6%). Among the 30,025 patients initially treated with TDF, 1,894 patients (6.3%) switched to a different therapy, most commonly to TAF (75.6%) and ETV (21.5%) (Supplementary Table 6). DISCUSSION This study provides robust evidence of BSV’s superior efficacy in reducing the risk of HCC compared with TDF in patients with CHB. These findings align with the growing body of literature suggesting that antiviral therapies with improved safety profiles and targeted mechanisms of action may offer significant long-term benefits beyond viral suppression. The primary outcome revealed a significantly lower incidence of HCC in the BSV group than in the TDF group, with an IRR of 1.59 (P = 0.001). This effect was consistently observed in both cirrhotic and non-cirrhotic subgroups, indicating that the protective effect of BSV against HCC is evident regardless of baseline liver disease status. Furthermore, the association remained consistent in sensitivity analyses excluding patients with ESRD, decompensated liver cirrhosis, or malignancy. Along with HCC risk reduction, the study also examined the incidence of liver cirrhosis as a secondary outcome. Kaplan–Meier curves and log-rank tests demonstrated no significant difference in the cumulative incidence of cirrhosis between the BSV and TDF groups. However, multivariable Cox regression analysis, which accounts for potential confounding variables, revealed a significantly lower risk of cirrhosis in the TDF group compared with the BSV group (HR, 0.670; P < 0.001). This apparent discrepancy may reflect differences in outcome definitions and susceptibility to coding bias, as HCC is generally captured more robustly than cirrhosis in claims-based datasets. Based on our analysis, BSV demonstrated superiority over TDF in reducing HCC incidence, with benefits observed in both cirrhotic and non-cirrhotic groups. Several factors may explain the improved outcomes associated with BSV. First, BSV may reduce the risk of cirrhosis more effectively than TDF, a major risk factor for HCC development.16 A phase 3 study comparing BSV and TDF in HCC patients (46 participants) reported more favorable outcomes with BSV compared with TDF, including improved liver histology and intrahepatic covalently closed circular DNA (cccDNA) quantification (77.8% vs. 36.4%).17 Another potential contributor is the co-administration of L-carnitine with BSV. L-carnitine depletion is a known adverse effect of BSV. Patients are typically advised to take 660 mg of L-carnitine supplements, which may reduce liver fatty acid levels and inflammation.181920 However, a recent study found no significant improvement in hepatic steatosis with the combination, underscoring the need for further research to determine its clinical impact.21 Finally, BSV is metabolized to LB80317, which selectively inhibits HBV DNA polymerase with a high genetic barrier to resistance. This targeted mechanism likely contributes to the reduction in HCC risk by minimizing hepatic inflammation and preserving liver function. BSV’s favorable safety profile, especially regarding renal and bone toxicity, may also improve treatment adherence and long-term outcomes, as demonstrated in previous trials.10 Additionally, we examined whether antiviral switching might have contributed to the observed differences in HCC incidence. Switching rates were comparable between the two groups (7.0% vs. 6.3%), suggesting that treatment continuation was generally stable across both agents. Although the claims database did not allow identification of the clinical reasons for switching, the predominant transition from TDF to TAF is consistent with real-world practice, where TAF is preferred for patients with renal or bone-related comorbidities. Several limitations of this study should be considered. First, the median follow-up duration of 3.2 years may be insufficient to fully evaluate the long-term protective effects of antiviral therapy on HCC development. The flattened incidence curve observed in the BSV group beyond two years likely reflects the limited number of patients with extended observation due to the recent introduction of BSV. Second, the retrospective design and reliance on administrative claims data may introduce the potential for selection bias and residual confounding, despite the use of propensity score matching. Several important clinical and behavioral variables, such as hepatic fibrosis stage, viral activity, and lifestyle factors, were not available in the database. Additionally, there may be potential confounding related to the distribution of prescriptions across hospital levels. BSV is mainly prescribed in tertiary care hospitals, whereas TDF is used across all levels of healthcare. As a result, unmeasured bias associated with hospital-level prescribing patterns cannot be excluded completely. Third, the study population was exclusively derived from Korea, which may limit the generalizability of the findings to regions with different HBV genotypes and diverse ethnic backgrounds. Despite these limitations, the implications of this study are significant for clinical practice. The demonstrated efficacy of BSV in reducing HCC risk underscores its potential role as a first-line treatment option for CHB, especially in patients at high risk for HCC. Additionally, the findings highlight the importance of individualized treatment strategies that balance antiviral efficacy with long-term safety. Further research is needed to elucidate the mechanisms underlying BSV’s protective effects and to explore its role in combination therapies aimed at achieving a functional cure for HBV. This study highlights the clinical benefits of BSV over TDF in reducing the risk of HCC and mitigating liver disease progression in patients with CHB. By addressing a critical endpoint in CHB management, these findings add to the growing body of evidence supporting BSV as a safe and effective therapeutic option. Ongoing research into its long-term outcomes and broader applicability will further inform treatment guidelines and enhance patient care in CHB.