Although the number of newly diagnosed patients with chronic hepatitis B (CHB) has reduced due to vaccination administration, approximately 296 million patients still remain and 820,000 CHB-associated deaths happen per year worldwide.¹ In Korea, patients with CHB make up 3% of the total population.² A cure for CHB has not yet been developed, unlike that for chronic hepatitis C, in which case many patients have been cleared of the virus via direct-acting agents. Thus, use of nucleos(t)ide analogs, that inhibit the replication of the hepatitis B virus (HBV) remain the key therapeutic strategy for treating CHB.

Entecavir (ETV) and tenofovir disoproxil fumarate (TDF) are nucleoside and nucleotide analogs, respectively. They show great efficacies and have a high genetic barrier to resistance. Patients treated with ETV may develop antiviral resistance, and long-term use of TDF induces renal injury and bone mineral loss.³ Besifovir dipivoxil maleate (BSV) and tenofovir alafenamide fumarate (TAF) were recently developed and approved, under the same reimbursement conditions, as the first-line treatments for CHB in Korea to obviate these limitations. BSV and TAF are nucleotide analogs with a high genetic barrier to antiviral resistance. Their efficacies are comparable to those of ETV and TDF, with better renal and bone safety profiles than those of TDF.^4-9

However, only few studies have directly compared the efficacy and safety profiles of BSV and TAF. Therefore, in this multicenter cohort study, we compared the real-world clinical trajectories of patients with CHB who were administered BSV or TAF as the first-line antiviral agents.

1. Study population

The Institutional Review Board of the Korea University Medical Center approved this retrospective cohort study (IRB number: 2021AS0201) and waived the requirement for informed consent. We investigated the medical records of 1,129 patients with CHB who were initially prescribed BSV or TAF at the Korea University Anam, Guro, and Ansan Hospitals between May 2017 and March 2021. Patients with a history of (1) treatment with other antiviral agents; (2) initiating BSV or TAF at another hospital; (3) coexisting or newly diagnosed hepatocellular carcinoma (HCC) or another uncured malignancy within 3 months; (4) transplantation of liver or kidney; or (5) hepatitis C or human immunodeficiency virus coinfection at baseline were excluded. All patients underwent blood tests to assess liver and kidney functions, prothrombin time, complete blood counts, hepatitis B e antigen (HBeAg) status, hepatitis B e antibody status, and HBV DNA levels at 1, 3, and 6 months during the initial 6 months and every 3 to 6 months after that. In addition, the hepatitis B surface antigen (HBsAg) titer was quantified every year.

2. Data collection

The patients’ demographic and clinical data and the results of liver and kidney tests were collected when treatment with BSV or TAF was started. Liver cirrhosis was diagnosed based on histological findings or compatible laboratory and radiologic findings.^10,11 Alcohol overuse was defined as a history of >40 g/day in men and >20 g/day in women.¹² Chronic kidney disease (CKD) was defined as a reduced estimated glomerular filtration rate (eGFR) of <60 mL/min/1.73 m² for more than 3 months.¹³ The primary outcome was defined as achieving of a virological response (VR). The secondary outcomes included safety profiles and achievement of additional laboratory endpoints, including those of biochemical and serologic responses. We also assessed, as key clinical endpoints, liver transplantation-free survival and HCC development during treatment with BSV or TAF until September 2021.

A VR was defined as an undetectable serum HBV DNA level (<20 IU/mL) by real-time PCR (Roche Diagnostics, Indianapolis, IN, USA).¹¹ VR60, defined as HBV DNA <60 IU/mL, was also investigated.¹⁴ A biochemical response was defined as the first occurrence of normalization in alanine aminotransferase levels according to two criteria: (1) ≤40 IU/L in both sexes (the standard reference) and (2) ≤34 IU/L in males and ≤30 IU/L in females (the Korean Association for the Study of the Liver Criteria).^15,16 In patients who were HBeAg-positive, the loss of HBeAg was defined as an HBeAg seroclearance.¹⁶ A virologic breakthrough was defined as a minimum 10-fold increase in the serum HBV DNA level compared with the lowest HBV DNA level on therapy or detection of the serum HBV DNA at levels of >200 IU/mL after achieving a VR.¹⁶

The eGFR was calculated using the modification of diet in renal disease formula.^13,17 Acute kidney injury (AKI) was defined as a >0.3 mg/dL or 50% increase in the serum creatinine level compared with the baseline creatinine level.^18,19 Hypophosphatemia was defined as a serum phosphorous level of less than 2.5 mg/dL.

3. Statistical analyses

Statistical analyses were performed using the R software (version 4.0.3; http://cran.r-project.org/). The data are expressed as mean±standard deviation, median (interquartile range), or number (%). The Student t-test and Mann-Whitney U test were used to compare continuous variables. The chi-square test and Fisher exact test were used to compare categorical variables, as appropriate. The probabilities for clinical outcomes including VR, HCC development, and liver transplantation-free survival were estimated using the Kaplan-Meier method. The differences between the groups of patients according to the administered antiviral agent were assessed using the log-rank test. Patients who were lost to follow-up and discontinued the nucleotide analogs treatment were censored on the date of the respective observation. A Cox proportional hazard regression model was established to analyze factors associated with the primary clinical outcomes. Significant factors (p<0.10) in the univariate analyses were subjected to multivariable analysis to determine the independent predictive factors. We used the linear mixed model to find the statistical significance of the time taken for the antiviral agents to induce change in eGFR.

In addition, to reduce selection bias and the effect of potential confounders, we performed propensity score (PS) matching based on age, sex, cirrhosis status, HBeAg status, HBV DNA level, platelet count, total bilirubin level, albumin level, and prothrombin time.

We conducted a noninferiority analysis comparing the efficacies of BSV and TAF. Considering 15% of follow-up loss rate with alpha-error set to 0.05, power to 0.8, and noninferiority margin to 0.15, the minimum number of patients needed was derived from VR rate (81%) of TAF at 2 years as 100 in each group.⁴ We used the PS-matched cohort for this analysis because the cohort was composed of 200 well-balanced subjects per group. If the absolute value of the lower bound of the interval was lower than the noninferiority margin, we concluded that BSV treatment is noninferior to TAF treatment.²⁰

1. Baseline characteristics

The baseline patient characteristics are shown in Table 1. Among the 1,129 patients, 592 were excluded; the reasons are provided in Fig. 1. Finally, 537 treatment-naïve patients were included. Among them, 202 and 335 participants were administered BSV and TAF, respectively. The mean age was 48.8±11.1 years, and 309 of the patients (57.5%) were men. There were no significant differences between the parameters of patients treated with BSV and TAF, except in serum alpha-fetoprotein levels.

Figure 1. Flow diagram of patient inclusion.
BSV, besifovir dipivoxil maleate; TAF, tenofovir alafenamide fumarate; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HIV, human immunodeficiency virus; PS, propensity score.

Table 1 . Baseline Characteristics of the Patients in the Two Antiviral Agent Treatment Groups.

Variable	Besifovir (n=202)	Tenofovir alafenamide (n=335)	p-value
Age, yr	49.8±9.6	48.1±11.8	0.07*
Male sex	116 (57.4)	193 (57.6)	0.99†
BMI, kg/m2	23.9 (21.8–26.0)	24 (21.8–26.3)	0.92‡
Hypertension	35 (17.3)	50 (14.9)	0.54†
Diabetes mellitus	21 (10.4)	35 (10.5)	0.99†
Alcohol overuse	30 (14.9)	49 (14.6)	0.99†
CKD	2 (1.0)	8 (2.4)	0.41†
Cirrhosis	50 (24.8)	86 (25.7)	0.95†
HBeAg positivity	112 (55.4)	191 (57.0)	0.84†
Platelets, ×103/mm3	176 (139–218)	171 (139–219)	0.85‡
Albumin, g/dL	4.2 (3.9–4.4)	4.1 (3.8–4.3)	0.16‡
AST, IU/L	68 (47–126)	71 (44.5–137)	0.62‡
ALT, IU/L	89 (50–160)	100 (53–189)	0.18‡
Bilirubin, mg/dL	0.85 (0.60–1.13)	0.79 (0.58–1.12)	0.56‡
Prothrombin time, INR	1.04 (0.99–1.13)	1.04 (0.99–1.11)	0.19‡
Creatinine, mg/dL	0.78 (0.67–0.90)	0.77 (0.63–0.89)	0.12‡
Sodium, mmol/L	140 (139–142)	140 (139–141)	0.48‡
Phosphate, mg/dL	3.5 (3.1–3.8)	3.5 (3.1–3.8)	0.55‡
Ascites	2 (1.0)	5 (1.5)	0.25†
Child-Pugh score	5 (5–5)	5 (5–5)	0.77‡
Child-Pugh class, B	5 (2.5)	13 (3.9)	0.53†
MELD score	7 (7–8)	7 (7–8)	0.49‡
HBV DNA, log IU/mL	6.53 (5.37–7.96)	6.64 (5.48–8.23)	0.41‡
AFP, ng/dL	7.00 (3.90–14.30)	5.18 (3.00–12.10)	0.01‡

Data are presented as the mean±SD, number (%), or median (interquartile range)..

BMI, body mass index; CKD, chronic kidney disease; HBeAg, hepatitis B e antigen; AST, aspartate aminotransferase; ALT, alanine aminotransferase; INR, international normalized ratio; MELD, Model for End-Stage Liver Disease; HBV, hepatitis B virus; AFP, alpha-fetoprotein..

*Student t-test; ^†Chi-square test; ^‡Mann-Whitney test..

2. Virological responses

A total of 444 patients (169 BSV and 275 TAF) experienced VR at the rates of 69.2%, 85.4%, and 92.1% at years 1, 2, and 3, respectively, during the follow-up period of 4 years (median 28.7 months). For achievement of VR60, the rate was 81.3%, 91.8%, and 95.3% at years 1, 2, and 3. In addition, both groups exhibited an average HBV DNA reduction of at least 4 log₁₀ IU/mL after 3 months (Supplementary Fig. 1). Multivariable Cox regression analysis revealed that HBeAg negativity (hazard ratio [HR]=1.98), a lower logarithm of the HBV DNA level (HR=1.54), a lower platelet count (HR=1.01), and a higher albumin level (HR=1.32) were independently associated with VR achievement. The type of antiviral agent used did not affect VR achievement (p=0.73) (Table 2).

Table 2 . Univariate and Multivariable Cox Regression Analyses for the Achievement of Virological Responses.

Variable	Univariate			Multivarable
	HR (95% CI)	p-value		HR (95% CI)	p-value
Age, yr	1.00 (1.00–1.01)	0.28
Male sex	0.98 (0.82–1.19)	0.87
BMI, kg/m2	1.00 (0.97–1.02)	0.78
Alcohol abuse, presence	1.19 (0.91–1.54)	0.20
Drug, TAF	1.03 (0.85–1.25)	0.73
Diabetes, presence	1.28 (0.95–1.72)	0.10
Hypertension, presence	1.16 (0.90–1.49)	0.25
CKD, presence	0.96 (0.48–1.93)	0.91
Liver cirrhosis, presence	1.64 (1.33–2.02)	<0.001		1.08 (0.86–1.35)	0.50
HBeAg, positive	0.24 (0.20–0.29)	<0.001		0.51 (0.40–0.64)	<0.001
HBV DNA, log IU/mL	0.57 (0.54–0.61)	<0.001		0.65 (0.60–0.71)	<0.001
Platelets, ×109/L	0.99 (0.99–1.00)	<0.001		0.99 (0.99–1.00)	0.009
Prothrombin time, INR	1.21 (0.76–1.92)	0.42
AST, IU/L	1.00 (0.99–1.01)	0.55
ALT, IU/L	0.99 (0.99–1.01)	0.17
Bilirubin, mg/dL	1.02 (0.96–1.07)	0.59
Albumin, g/dL	1.50 (1.19–1.90)	0.001		1.32 (1.04–1.66)	0.02
Creatinine, mg/dL	1.08 (0.64–1.80)	0.78
Sodium, mmol/L	1.03 (0.99–1.08)	0.13
Phosphate, mg/dL	0.91 (0.76–1.10)	0.34
Ascites, presence	0.88 (0.39–1.97)	0.75
Child-Pugh score	0.93 (0.77–1.11)	0.41
Child-Pugh class, B	0.95 (0.55–1.61)	0.84
MELD score	1.04 (0.99–1.08)	0.10
AFP, ng/mL	1.00 (0.99–1.00)	0.48

HR, hazard ratio; CI, confidence interval; BMI, body mass index; TAF, tenofovir alafenamide fumarate; CKD, chronic kidney disease; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; INR, international normalized ratio; AST, aspartate aminotransferase; ALT, alanine aminotransferase; MELD, Model for End-Stage Liver Disease; AFP, alpha-fetoprotein..

Accordingly, the cumulative probabilities of VR were not significantly different between BSV and TAF groups (68.6% vs 69.9%, 84.5% vs 85.9%, and 93.0% vs 92.0% at years 1, 2, and 3, respectively; p=0.73) (Fig. 2A). VR achievement in the subgroups according to the HBeAg status, presence of cirrhosis, and HBV DNA level, were not different between the BSV and TAF groups (all p>0.05) (Supplementary Fig. 2). However, more patients with HBeAg negativity at baseline experienced VR than those with HBeAg positivity (94.5% vs 49.5%, 99.0% vs 74.4%, and 99.0% vs 86.0% at years 1, 2, and 3, respectively; p<0.001) (Fig. 2B). Significantly higher number of patients with cirrhosis or a low HBV DNA level (<10⁶ IU/mL) at baseline achieved VR than those without cirrhosis or with a high HBV DNA level (≥10⁶ IU/mL; p<0.001) (Fig. 2C and D).

Figure 2. Kaplan-Meier plots for virological response (VR) achievement for all patients. Plots for all patients according to (A) the antiviral agent administered, (B) HBeAg status, (C) cirrhosis status, and (D) HBV DNA level. Significant differences were observed in VR achievement according to the HBeAg status, cirrhosis status, and HBV DNA levels (p<0.001), but not between the two treatment groups (p=0.73).
BSV, besifovir dipivoxil maleate; TAF, tenofovir alafenamide fumarate; HBeAg, hepatitis B e antigen; LC, liver cirrhosis; HBV, hepatitis B virus.

3. Biochemical response, serologic responses, and safety outcomes

Alanine aminotransferase elevations at baseline according to two criteria were noted in 441 (166 for BSV and 275 for TAF) and 474 patients (176 for BSV and 298 for TAF) for the standard reference and the Korean Association for the Study of the Liver Criteria, respectively. The rates of biochemical response were similar between the two groups (p=0.96 and p=0.24) (Table 3) according to each criterion. The rate of HBeAg seroclearance was significantly higher in the BSV group than in the TAF group (p=0.01).

Table 3 . Comparison of Clinical Outcomes between the Two Antiviral Agent Groups.

Outcome	Besifovir	Tenofovir alafenamide	p-value
Biochemical response within 1/2/3 yr, %*	85.7/93.0/95.4	87.0/93.1/95.4	0.96
Biochemical response within 1/2/3 yr, %†	63.6/83.1/87.7	73.9/86.3/90.0	0.24
HBeAg seroclearance within 1/2/3 yr, %	19.8/33.3/47.4	11.3/21.7/35.1	0.01
Virologic breakthrough	5	15	0.34
AKI episode in patients without CKD	0	0	0.99
eGFR <60 mL/min/1.73 m2, No. (%)	2 (1.0)	1 (0.3)	0.83
Phosphate <2.5 mg/dL, No. (%)	0	6 (2.1)	0.07

HBeAg, hepatitis B e antigen; AKI, acute kidney injury; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate..

*Biochemical response was defined as alanine aminotransferase normalization ≤40 IU/L in both sexes; ^†Biochemical response was defined as alanine aminotransferase normalization ≤34 IU/L in males and ≤30 IU/L in females..

The HBsAg titer was examined in 133 patients at baseline and re-examined in 94 patients at 1 year after antiviral agent administration. The median change in the logarithm of HBsAg titer during the 1-year follow-up period was –0.14 (interquartile range, –0.38 to –0.02) in the BSV group and –0.12 (interquartile range, –0.43 to 0.02) in the TAF group, with no significant difference observed between the two groups (p=0.73) (Supplementary Fig. 3).

During the follow-up, virologic breakthroughs developed in 20 patients (5 BSV and 15 TAF, p=0.34). However, these could be attributed to poor adherence to drugs. The patients achieved VR again receiving the same drug after giving education for medication compliance. No clinical evidence of drug resistance was observed.

AKI was not observed in patients without CKD. However, eGFR of three patients without CKD at baseline decreased to under 60 mL/min/1.73 m² but greater than 50 mL/min/1.73 m². The incidences were not significant between two groups (Supplementary Fig. 4A). Among patients without CKD at baseline, the linear mixed model revealed that eGFR was more decreased in the TAF group than in the BSV group (group-by-time effect p=0.02) (Fig. 3). Among 10 patients with CKD, eGFR of two patients reduced to 30 mL/min/1.73 m² or lower. Hypophosphatemia occurred in six patients who were administered TAF, but the phosphate levels for these patients were above 2.0 mg/dL (Table 3 and Supplementary Fig. 4B).

Figure 3. Changes in eGFR according to treatment groups without chronic kidney disease at baseline. (A) Mean eGFR levels at baseline and after antiviral treatment. (B) eGFR change from baseline.
eGFR, estimated glomerular filtration rate; BSV, besifovir dipivoxil maleate; TAF, tenofovir alafenamide fumarate.

A total of 20 patients had ascites or Child-Pugh class B liver function at baseline; among them, eight patients had cirrhosis, and two of eight patients had decompensated cirrhosis. After antiviral therapy, 17 patients were recovered to Child-Pugh score 5 status while three patients including two decompensated cirrhotics were not recovered to that until the last follow-up date (Supplementary Fig. 5).

During the follow-up period, only one patient, who had cirrhosis and took BSV, died. The cause of death was cerebrovascular disease (Supplementary Fig. 6).

4. Development of HCC

A total of 15 patients were diagnosed with HCC during the follow-up period (seven administered BSV and eight administered TAF). Multivariable Cox regression analysis revealed that older age (HR=1.08), the status of alcohol abuse (HR=4.51), the presence of cirrhosis (HR=4.39) and ascites (HR=9.57), and a lower logarithm of the HBV DNA level (HR=1.67) were independently associated with HCC development. The type of antiviral agent did not affect HCC development (p=0.31) (Table 4).

Table 4 . Univariate and Multivariable Cox Regression Analyses for Hepatocellular Carcinoma Development.

Variable	Univariate			Multivariable
	HR (95% CI)	p-value		HR (95% CI)	p-value
Age, yr	1.07 (1.02–1.12)	0.007		1.08 (1.01–1.15)	0.02
Male sex	3.05 (0.86–10.82)	0.08		2.70 (0.58–12.58)	0.21
BMI, kg/m2	1.00 (0.87–1.15)	0.98
Alcohol abuse, presence	7.20 (2.61–19.87)	<0.001		4.51 (1.34–15.25)	0.02
Drug, TAF	0.59 (0.21–1.65)	0.31
Diabetes, presence	2.16 (0.61–7.65)	0.23
Hypertension, presence	2.80 (0.96–8.20)	0.06		1.54 (0.45–5.21)	0.49
CKD, presence	5.18 (0.67–39.85)	0.11
Liver cirrhosis, presence	13.29 (3.75–47.13)	<0.001		4.39 (1.02–18.94)	0.04
HBeAg, positive	0.20 (0.06–0.71)	0.01		0.41 (0.09–1.81)	0.24
HBV DNA, log IU/mL	0.55 (0.40–0.76)	<0.001		0.60 (0.37–0.97)	0.04
Platelets, ×109/L	0.99 (0.98–1.00)	0.08		0.99 (0.98–1.01)	0.80
Prothrombin time, INR	1.79 (0.13–25.14)	0.67
AST, IU/L	0.99 (0.98–1.00)	0.18
ALT, IU/L	0.99 (0.98–1.00)	0.13
Bilirubin, mg/dL	1.06 (0.86–1.30)	0.58
Albumin, g/dL	0.55 (0.19–1.62)	0.28
Creatinine, mg/dL	2.40 (0.34–16.97)	0.38
Sodium, mmol/L	0.85 (0.70–1.04)	0.13
Phosphate, mg/dL	1.33 (0.51–3.45)	0.56
Ascites, presence	26.74 (7.45–95.99)	<0.001		9.57 (1.85–49.46)	0.007
Child-Pugh score	1.40 (0.76–2.93)	0.24
MELD score	1.14 (0.94–1.37)	0.19
AFP, ng/mL	1.00 (1.00–1.00)	0.26

HR, hazard ratio; CI, confidence interval; BMI, body mass index; TAF, tenofovir alafenamide fumarate; CKD, chronic kidney disease; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; INR, international normalized ratio; AST, aspartate aminotransferase; ALT, alanine aminotransferase; MELD, Model for End-Stage Liver Disease; AFP, alpha-fetoprotein..

Accordingly, the cumulative probabilities of HCC at years 1, 2, and 3 were 2.1% versus 1.3%, 3.9% versus 1.6%, and 3.9% versus 3.0% were not significantly different between the patients in the BSV and TAF groups (p=0.31) (Fig. 4A). Significantly more number of patients with cirrhosis were diagnosed with HCC than those without cirrhosis (5.8% vs 0.3%, 8.7% vs 0.6%, and 13.4% vs 0.6% at years 1, 2, and 3, respectively; p<0.001) (Fig. 4B). In the subgroups of patients with and without cirrhosis, and the cumulative incidence rates of HCC were not significantly different between the patients in the BSV and TAF groups (p=0.22 and p=0.91) (Fig. 4C and D).

Figure 4. Kaplan-Meier plots for HCC development. All patients according to (A) the antiviral agent and (B) their cirrhosis status. Subgroup analysis comparing the cumulative incidence of HCC in the (C) patients with cirrhosis and (D) those without cirrhosis according to the antiviral agents. There was a significant difference between the patients with cirrhosis and those without cirrhosis (p<0.001). There was no difference between the BSV and TAF treatment groups under various conditions (all p>0.05).
HCC, hepatocellular carcinoma; BSV, Besifovir dipivoxil maleate; TAF, tenofovir alafenamide fumarate; LC, liver cirrhosis.

5. Noninferiority test in the PS-matched cohort

The 1:1 PS-matched analysis constructed 200 pairs; the baseline characteristics of the BSV and TAF groups did not differ significantly (Supplementary Table 1). Further, no significant difference was observed between the rates of VR achievement of the BSV and TAF groups (p=0.28) (Fig. 5A). At year 2, the rates of VR achievement were 85.0% and 88.7% in the BSV and TAF groups, and the absolute value of 95% confidence interval of difference (–0.04 to 0.12) satisfied the 0.15 a priori limit of a noninferiority, suggesting that BSV is noninferior to TAF in terms of VR. Similar biochemical responses were achieved between the two groups within years 1 and 2 (85.5% vs 87.2% and 93.0% vs 92.2%, respectively; p=0.88). BSV was noninferior to TAF for biochemical responses at year 2 (95% confidence interval of difference, –0.06 to 0.07). The rates of HBeAg seroclearance were significantly higher in the BSV group than TAF group (20.1% vs 13.1%, 33.7% vs 23.3%, and 47.7% vs 37.5% within 1-, 2-, and 3-year, respectively; p=0.03), and were similar for all patients within the same group. In addition, cumulative HCC development did not differ between the BSV and TAF groups (p=0.44) (Fig. 5B).

Figure 5. Kaplan-Meier plots for clinical outcomes in the propensity score-matched cohort. (A) VR and (B) HCC development according to the antiviral agents.
BSV, besifovir dipivoxil maleate; TAF, tenofovir alafenamide fumarate; VR, virological response; HCC, hepatocellular carcinoma.

Many antiviral agents are available to treat CHB, but new agents need to be developed. The existing drugs may have excellent efficacy but also possess several weaknesses. In addition, the safety concerns cannot be overlooked as nucleos(t)ide analogs require long-term use as a viral replication inhibitor rather than a virus eliminator.³ Two recently developed antiviral agents, BSV and TAF received National Health Insurance approval as the first-line therapy for CHB in Korea. The efficacies of new antiviral agents have been compared in clinical trials and real-world clinical practice with ETV and TDF, conventional first-line antiviral agents.^4-9 These studies demonstrated that the new drugs are comparable to the conventional ones in terms of efficacy and offer improved safety. In addition, a recent study revealed that liver histology improved significantly under BSV treatment than under TDF treatment.²¹ In the present study, we compared various treatment outcomes in treatment-naïve patients with CHB from multiple centers and found no difference in the efficacy and safety between the BSV and TAF. VR and biochemical responses were achieved in more than 70% of patients within 12 months of treatment. These results were comparable with those of previous clinical trials.^4,22 For achieving an early VR, we demonstrated that there are four independent predictors: the presence of cirrhosis, HBeAg status, platelet count, and serum albumin and HBV DNA levels. BSV and TAF showed no significant differences in their ability to induce VR in patients under various conditions. In the PS-matched cohort, VR rates were 85.0% and 88.7% for the BSV and TAF groups at 2 years. It was confirmed that BSV is noninferior to TAF in terms of VR achievements and biochemical response, representing the efficacy of antiviral agents.

Murata et al.²³ reported that the inhibition of HBsAg production might be different because the inductive effects of interferon-λ3 between nucleoside and nucleotide analogs are different. The present study showed that the HBsAg titer decreased in most patients during the first year after starting the drug, and there was no significant difference between the BSV and TAF treatments outcomes. The results of a recent study support this, showing no difference in intrahepatic covalently closed circular DNA reduction between the participants treated with BSV and TDF.²¹ This may be because BSV, TAF and TDF are nucleotide analogs.

Recent prospective studies have demonstrated the renal safety of BSV and TAF. Yim et al.²² showed that the reduction in the eGFR during 48 weeks of TDF treatment reversed within 12 weeks after switching to BSV. Toyoda et al.²⁴ also demonstrated that the eGFR reduced by a previous TDF therapy in patients with a CKD stage ≥2 was recovered in 25% of the patients after switching to TAF. In a recent study by Jung et al.,²⁵ BSV and TAF exhibited similar risk of renal function decline, but also exhibited renal protective effects compared to patients with CHB who did not received antiviral agents. On the other hand, our study presented that the BSV group showed less eGFR decline than TAF group (Fig. 3). Although direct comparison is not possible, the median change in eGFR at 96-week of participants treated with BSV and TAF were –0.7 and –1.2 in the two previous clinical trials which compared with TDF, supporting less renal injury in patients with BSV treatment.^4,22 This difference might be attributed to L-carnitine co-administered with BSV. L-carnitine has an anti-oxidant effect that can alleviate renal injury. In addition, many studies have shown a renal protective effect of L-carnitine.^26,27 Even if there were significant differences in degrees of decline in the renal function, neither of the two drugs is expected to have specific renal toxicity because there was no development of AKI in patients without CKD.

We identified five independent predictors of HCC development: age, status of alcohol abuse, the presence of cirrhosis and ascites, and serum HBV DNA level. Age and the presence of cirrhosis have been identified as risk factors in previous studies (HR, 1.04 to 1.06 and HR, 1.80 to 5.27, respectively).^6,8,25,28-32 Alcohol use and ascites were also mentioned as risk factors in a study by Yang et al.³³ The results support the need for an early antiviral treatment and alcohol abstinence in patients with CHB. Especially, low HBV DNA was an independent predictor of HCC development, but this result could have been affected by patients with cirrhosis who might have relatively low viral load. The debate over the difference in the protective effect of antiviral agents on HCC development has continued. However, BSV and TAF showed no significant differences under various conditions and a PS-matched cohort in this study.

This study has several limitations. First, this was a retrospective study with inevitable selection bias. However, we attempted to avoid this by conducting multiple subgroup and PS-matched cohort analyses. In addition, since BSV and TAF were approved under the same reimbursement criteria simultaneously, we could avoid the selection bias that was raised in comparative studies of antiviral agents approved at other times. Second, most patients were Korean, and hence, our results cannot be generalized for worldwide population. Third, because only one patient died, we could not further analyze the factors related to mortality in this study. Fourth, we could not include bone mineral density as a parameter in the study because it was examined only in some patients who met the reimbursement criteria of National Health Insurance standards test. Instead, we compared the occurrence of hypophosphatemia during antiviral therapy and found no significant differences between patients treated with BSV and TAF. In summary, VR, biochemical response, and a reduction in the HBsAg titer were also statistically similar between the BSV and TAF groups. Moreover, there was no episode of AKI in patients without CKD at baseline. Interestingly, the BSV group showed a higher HBeAg seroclearance and a lower degree of decline in eGFR than the TAF group did. These findings need to be confirmed in a larger study.

The PS-matched cohort demonstrated noninferiority of BSV to TAF regarding VR and biochemical response. Progression to CKD and hypophosphatemia occurred in less than 3% of the patient population, and no significant difference was observed in these parameters between the BSV and TAF groups. In addition, no significant difference was seen in the survival rate or HCC development between the two groups. The results were similar in the subgroups of cirrhosis and non-cirrhotic chronic liver disease.

Thus, the efficacy and safety profiles of BSV and TAF as the first-line antiviral agents in treatment-naïve patients with CHB were comparable. Additionally, this study warrants a prospective comparison on a larger scale and a longer follow-up study.

This work was supported by Korea University Research Grant.

No potential conflict of interest relevant to this article was reported.

Study concept and design: H.J.Y. Data acquisition: all authors. Data analysis and interpretation: T.H.K., H.J.Y. Drafting of the manuscript: T.H.K. Critical revision of the manuscript for important intellectual content: J.H.K., Y.S.S. Statistical analysis: T.H.K., H.J.Y. Obtained funding: H.J.Y., Y.S.S. Administrative, technical, or material support; study supervision: H.J.Y., Y.S.S. Approval of final manuscript: all authors.

Supplementary materials can be accessed at https://doi.org/10.5009/gnl220390.