Presented in part at the National Institute of Health Liver Meeting 2006, April 6–8, 2006; the 12th International Symposium on Viral Hepatitis and Liver Disease, July 4, 2006; the 56th Annual Meeting of the American Association for the Study of Liver Diseases, October 30 2006; the 42nd Annual Meeting of the European Association for the Study of the Liver, April 11-15, 2007.
Clinical trial registration number: NCT00096785.
Potential conflict of interest: Dr. Hann is on the speakers' bureau of and received grants from Bristol-Myers Squibb, Gilead, and Novartis. Dr. Sollano advises Eisai. He is also on the speakers' bureau of GlaxoSmithKline and Novartis. Dr. Lao-Tan advises AstraZeneca. She is on the speakers' bureau of Sanofi-Aventis. She also received grants from LSO. Dr. Yuen is on the speakers' bureau of and received grants from Bristol-Myers Squibb. Dr. Jeffers advises, is on the speakers' bureau of, and received grants from Bristol-Myers Squibb. He is on the speakers' bureau of and received grants from Roche and Gilead. Dr. Sherman is a consultant for, advises, and is on the speakers' bureau of Bristol-Myers Squibb and Gilead Sciences. Dr. Min is a consultant for, advises, is on the speakers' bureau of, and received grants from Bristol-Myers Squibb, Gilead, and Novartis. Drs. Mencarini, Diva, and Wilber own stock in Bristol-Myers Squibb.
This study was undertaken to compare the early antiviral activity and viral kinetic profiles of entecavir (ETV) versus adefovir (ADV) in hepatitis B e antigen positive nucleoside-naïve adults with chronic hepatitis B (CHB). Sixty-nine nucleoside-naïve CHB patients with baseline HBV DNA of 108 copies/mL or more were randomized 1:1 to open-label treatment with entecavir 0.5 mg/day or adefovir 10 mg/day for a minimum of 52 weeks. The primary efficacy analysis compared mean reduction in HBV DNA at week 12 adjusted for baseline levels using linear regression. Entecavir was superior to adefovir for mean change from baseline in HBV DNA at week 12 (−6.23 log10 copies/mL versus −4.42 log10 copies/mL, respectively; mean difference −1.58 log10 copies/mL; P < 0.0001). Both drugs demonstrated biphasic viral kinetics, with a first phase of rapid decline lasting 10 days. A significant difference favoring ETV was reached at day 10 (day 10 ETV−ADV difference estimate: −0.66 log10 copies/mL; 95% CI [−0.30, −0.01]). Early virological response was found to be predictive of subsequent virological response, with those having lower HBV DNA levels at day 10 being more likely to achieve HBV DNA of less than 300 copies/mL at week 48. In addition, there was considerably less variability in the extent of HBV DNA reductions in patients treated with entecavir versus adefovir. Both the mean decrease in serum HBV DNA and the proportion of patients achieving HBV DNA less than 300 copies/mL were greater in entecavir-treated than adefovir-treated patients at weeks 2, 4, 8, 12, 24, and 48. At week 48, one (3%) ETV-treated versus 15 (47%) ADV-treated patients had HBV DNA of 105 copies/mL or more. Both antivirals were well tolerated. Conclusion: Entecavir therapy resulted in earlier and superior reduction in HBV DNA compared with adefovir in nucleoside-naïve HBeAg-positive patients with CHB. (HEPATOLOGY 2009;49:72-79.)
Each year, more than 50 million people are infected with hepatitis B virus (HBV), and more than one million deaths are attributed to the complications of chronic HBV infection (CHB), including cirrhosis and hepatocellular carcinoma.1, 2 Worldwide, CHB infection results in approximately one third of all cases of cirrhosis and more than three quarters of all cases of hepatocellular carcinoma.3 It is not possible to predict which chronically infected patients will develop these complications. However, large long-term studies have shown that the risk of developing cirrhosis and hepatocellular carcinoma is directly proportional to the serum HBV DNA level.4, 5 By implication, treatments that reduce HBV DNA may prevent progression of liver disease in patients with CHB. This effect has been confirmed in patients both with and without advanced hepatic fibrosis, in whom the rate of liver-related complications was significantly decreased by treatment with lamivudine.6, 7 However, prolonged treatment with lamivudine is limited by high rates of resistance and loss of therapeutic effect. More potent agents with lower rates of resistance were sought to provide sustained long-term suppression of viral replication and prevent progression of liver disease.
Adefovir (ADV) is not cross-resistant with lamivudine and was developed for the treatment of CHB; however, in two pivotal phase III clinical trials of ADV, among subjects receiving the approved dose of 10 mg once daily, only 21% of hepatitis B e antigen (HBeAg)-positive patients, and 51% of HBeAg-negative patients, had achieved serum HBV DNA of less than 400 copies/mL at 48 weeks.8, 9 In an analysis of five studies (N = 1210) of ADV treatment from 48 through 192 weeks, higher serum HBV DNA at week 48 was predictive of the development of ADV resistance at 4 years.10
Entecavir (ETV) (Baraclude, Bristol-Myers Squibb) is a potent guanosine analog with a high genetic barrier to resistance and rapid virological suppression.11 Therapy with ETV resulted in superior histological, biochemical, and virological efficacy after 1 year of treatment compared with lamivudine in HBeAg-positive and HBeAg-negative patients with CHB.12-16 Follow-up of patients treated in clinical trials beyond 48 weeks demonstrated durable viral suppression.17-20
Viral kinetic studies of nucleoside analogs have shown a biphasic decay in HBV DNA.21-23 A 12-week study showed that a higher dose (30 mg/day) than the currently approved 10-mg daily dose of ADV suppressed HBV DNA replication to a greater extent.21 In the same study,21 the half-life of circulating virus was estimated to be 26.4 hours during treatment with ADV. Another study showed that the half-life of circulating virus was 24 hours during treatment with lamivudine.22 In contrast, in a study of ETV, the median half-life of viral turn-over was 16 hours.23
The early antiviral efficacy of ETV has not been directly compared with ADV in patients with CHB infection. We designed this prospective, randomized study to compare the early antiviral activity and viral kinetic profiles of ETV with that of ADV in HBeAg-positive, nucleoside-naïve adults who are chronically infected with HBV.
ADV, adefovir; ALT, alanine aminotransferase; CHB, chronic hepatitis B; ETV, entecavir; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; PCR, polymerase chain reaction.
Patients and Methods
This was an open-label phase IIIb study in which eligible patients were prospectively randomized (1:1) to open-label treatment with either oral ETV 0.5 mg or oral ADV 10 mg for a minimum duration of 52 weeks. The study was conducted in accordance with the ethics principles of the Declaration of Helsinki and was consistent with Good Clinical Practice Guidelines and the requirements of the local regulatory authorities in each country. All patients provided written informed consent.
The study was designed by the sponsor (Bristol-Myers Squibb), who collected the data, monitored the conduct of the study, and performed the statistical analyses. All authors had access to the complete study reports, and were actively involved in the data analysis and the interpretation of the results.
Patients eligible for the trial were aged 16 years or older, had HBeAg-positive CHB infection, compensated liver disease with a serum alanine aminotransferase (ALT) level between 1.3 and 10 times the upper limit of normal, and had never received treatment with nucleosides or nucleotides with activity against HBV. Because viral kinetics were to be evaluated, patients were required to have a serum HBV DNA level of 108 copies/mL or greater at screening, so that viral levels would not be truncated at the level of quantitation during the first 12 weeks. Women of childbearing potential were required to use an acceptable method to avoid pregnancy during treatment and for 8 weeks after completion of the study.
Patients were excluded if they had evidence of infection with human immunodeficiency virus, hepatitis C or D viruses, or a recent history of acute pancreatitis. Patients with a serum creatinine greater than 1.5 mg/dL, hemoglobin less than 10 g/L, or platelet count less than 70,000/mm3 were also excluded.
Outcomes and Assessments.
The primary objective of this randomized, international study was to compare the early antiviral efficacy, as measured by the mean reduction in serum HBV DNA by polymerase chain reaction (PCR) assay at week 12 and adjusted for baseline level between ETV and ADV.
Secondary efficacy end points included an assessment of viral kinetics over the first 12 weeks of therapy; mean change in HBV DNA from baseline to weeks 24 and 48; proportion of patients with undetectable serum HBV DNA (<300 copies/mL) at weeks 12, 24, and 48; the proportion of patients with normalization of serum ALT; HBe seroconversion at week 48; and safety. Safety was assessed by reviewing adverse events, clinical laboratory abnormalities, and discontinuations because of adverse events.
Serum HBV DNA was measured by quantitative PCR assay (Roche Amplicor, limit of quantification 300 copies/mL) before administration of the first dose of study medication on day 1 and then daily until day 14 and at weeks 3, 4, 6, 8, 10, 12, 24, and 48 of treatment. HBV serologic tests (hepatitis B surface antigen, HBeAg, anti-hepatitis B surface, anti-HBe) were conducted on day 1 and at weeks 12, 24, and 48 of treatment.
The target sample size of at least 26 patients per treatment group provided 90% power to demonstrate superiority in the mean reduction of serum HBV DNA at week 12, assuming a difference between treatment groups of 1 log10 copies/mL, a standard deviation of 0.8 for ETV and 1.3 for ADV, and a two-sided significance level of 0.05.
The efficacy analysis included all evaluable patients: those who received assigned therapy for a minimum of 12 weeks and had both baseline and week 12 HBV DNA PCR measurements. Estimates of mean reductions in HBV DNA by PCR from baseline were based on patients with measurements. Estimates of proportions were based on all evaluable patients, with missing data treated as failure (Noncompleter = Failure).
The viral kinetic assessment was based on estimates of biphasic decay kinetics, which have been previously described.21, 23 The first-phase and second- phase slopes for each therapy were estimated using spline regression (in other words, 2 lines joined at a knot or change point) applied to viral load data.24 The initial choice of the knot (day 7 or 10) was based on previous literature concerning other anti-HBV compounds,21, 25 with the final knot (day 10) selected based on cross-validation.26 A spline was fitted to the HBV DNA by PCR measurements for each individual patient, and the parameter estimates were then averaged for patients within each treatment group. Comparisons of the two slopes were based on t tests using the Satterthwaite method for unequal variances. Significance of the separation of the two fitted lines was determined based on 95% confidence intervals by study day. Exploratory analyses for predicting HBV DNA less than 300 copies/mL at weeks 24 and 48 based on HBV DNA categories at day 10 (<106, 106 to <107, 107 to <108, and ≥108 copies/mL) used the Cochrane-Armitage trend test. We employed the standard four-parameter exponential decay model to estimate the following viral kinetic parameters: effectiveness in blocking virus production by infected cells (ϵ), effectiveness in blocking de novo infection of susceptible cells (η), death rate of virus-producing infected cells (δ), and clearance rate of free virus (c).21, 23, 25, 27
A total of 132 adults were enrolled in 26 centers from eight countries (Canada, Hong Kong, Indonesia, Philippines, Singapore, Taiwan, Thailand, and the United States), 69 of which were randomized to receive either ETV or ADV. The first patient was enrolled in January 2005, and the last patient completed 12 weeks of treatment in January 2006.
The flow of patients through the trial is shown in Fig. 1. A total of 33 patients randomized to ETV and 32 patients randomized to ADV completed 12 weeks of treatment and were included in the primary efficacy analysis. One patient was randomized to ADV but received treatment with ETV and was excluded from the efficacy analysis. Two patients treated with ETV and one patient treated with ADV discontinued therapy before week 12. The characteristics of the two treatment groups were well matched at baseline (Table 1). Because of the inclusion criterion requiring HBV DNA of 108 copies/mL or greater, patients in both groups had high levels of HBV DNA at baseline (mean HBV DNA, 10.26 log10 copies/mL for ETV and 9.88 log10 copies/mL for ADV).
Table 1. Baseline Characteristics
Entecavir 0.5 mg/day
Adefovir 10 mg/day
(n = 33)
(n = 32)
ULN, upper limit of normal.
Mean age, years (SE)
Male sex, n (%)
Race, n (%)
Hawaiian or other Pacific Islander
Mean serum HBV DNA, log10 copies/mL (SE)
Mean serum ALT activity, U/L (SE)
Serum ALT distribution, n (%)
>1× ULN to <2×ULN
≥2× ULN to ≤5×ULN
Primary Efficacy End Point.
The mean reduction in serum HBV DNA level at week 12 was significantly greater in patients randomized to ETV compared with ADV (−6.23 versus −4.42 log10 copies/mL, mean difference estimate −1.58 log10 copies/mL; 95% confidence interval −2.28 to −0.88 log10 copies/mL, P < 0.0001).
Secondary Efficacy End Points.
The mean decrease in serum HBV DNA levels was also greater with ETV than ADV in patients randomized and treated at weeks 2, 4, 8, 24, and 48 (Table 2). The magnitude of the difference increased at each subsequent time point. Consistent with these results, the proportion of patients with HBV DNA of less than 300 copies/mL was higher in patients treated with ETV than in those treated with ADV at weeks 12, 24, and 48 (Fig. 2). At week 24, 15 ETV-treated patients (45%) and four ADV-treated patients (13%) achieved HBV DNA less than 300 copies/mL. At week 48, 19 ETV-treated patients (58%) and six ADV-treated patients (19%) achieved HBV DNA less than 300 copies/mL. Three percent of patients who received ETV had HBV DNA of 105 copies/mL or more at weeks 24 and 48, compared with 50% and 47%, respectively, of ADV-treated patients.
Table 2. Mean (SE) HBV DNA Reductions (log10 copies/mL) from Baseline During Treatment
Entecavir 0.5 mg/day
Adefovir 10 mg/day
Entecavir − Adefovir Difference; 95% CI
(n = 33)
(n = 32)
−0.82; −1.31 to −0.33
−1.00; −1.53 to −0.47
−1.26; −1.92 to −0.61
−1.58; −2.28 to −0.88
−1.86; −2.51 to −1.20
−1.86; −2.69 to −1.03
Normalization of serum ALT (≤1 × upper limit of normal) was documented in 25 (76%) ETV-treated patients and 20 (63%) ADV-treated patients at week 48. Furthermore, at week 48, another seven (21%) ETV-treated and eight (25%) ADV-treated patients achieved ALT levels greater than 1 to less than 2 × upper limit of normal.
HBeAg loss and HBe seroconversion rates were similar for both ETV-treated and ADV-treated patients. For ETV-treated patients, HBeAg loss and HBe seroconversion rates were 6 of 33 (18%) and 5 of 33 (15%), respectively, versus 7 of 32 (22%) and 7 of 32 (22%), respectively, for ADV-treated patients (P not significant).
Viral Kinetic Assessment.
The spline-fitting procedure estimated the slopes of both the first and second phases of viral load decline, as well as the change point (“knot”) at which the transition from 1 phase to the other occurs. In the viral kinetic assessments, the knot was placed at day 10 to join the curve for the rapid first-phase decline and the slower second-phase decline in HBV DNA. The slopes of the first-phase decline were −0.391 for ETV and −0.329 for ADV. As seen in Fig. 3, after day 10, the curves describing the phase 2 decline in HBV DNA diverged, and estimates of the slope were significantly different for the two therapies (−0.034 for ETV versus −0.024 for ADV, P < 0.02).
The mean estimate of efficacy (ϵ) based on the four-parameter exponential decay model was significantly greater for ETV than for ADV (99.9% versus 99.5%; mean difference 0.41%; 95% confidence interval, 0.09-0.73). The estimated half-life of circulating virus was significantly shorter in patients treated with ETV than ADV (14.5 hours versus 32.5 hours; mean difference, −17.9 hours; 95% confidence interval, −5.2 to −30.7 hours).
The fitted curves for mean serum HBV DNA through 12 weeks of treatment superimposed on the individual profiles for each patient are shown in Fig. 4. The same figure also reflects the higher variability observed among the individual ADV curves compared with the ETV curves.
Exploratory analysis showed that HBV DNA categories at day 10 were predictive of achieving HBV DNA less than 300 copies/mL at weeks 24 and 48, with trend test P-values of less than 0.0001 for both end points based on the combined treatment groups. At day 10, 17 (26%), 15 (23%), 14 (22%), and 19 (29%) of 65 evaluable patients had HBV DNA <106, 106 to <107, 107 to <108, and ≥108 copies/mL, respectively. The proportion of patients who achieved HBV DNA <300 copies/mL at week 24 for each HBV DNA category at day 10 were 13 of 17 (76%), 4 of 15 (27%), 2 of 14 (14%), and 0 of 19, respectively. At week 48, the corresponding proportions were 14 of 17 (82%), 6 of 15 (40%), 4 of 14 (29%), and 1 of 19 (5%). Similar trends were observed when data were analyzed by treatment group.
Treatment was generally safe and well tolerated. The most common adverse events are presented in Table 3. One ADV patient discontinued treatment because of an adverse event (ALT flare), which was considered by the investigator to be probably related to study drug treatment, and resolved after withdrawal of treatment. No other ALT flares and no deaths were documented during the study.
Table 3. Safety According to Treatment Received to Week 48
Entecavir 0.5 mg/day
Adefovir 10 mg/day
(n = 36)
(n = 33)
One patient in the entecavir group and 2 patients in the adefovir group experienced an elevated alanine transferase level grade 4. One of these elevations, in an adefovir recipient, was considered to be related to treatment and resulted in discontinuation of study treatment. One adefovir recipient fractured a rib in an automobile accident (not likely related to study treatment).
This is a direct comparison of the antiviral activity of ETV and ADV in nucleoside-naïve HBeAg-positive patients with compensated CHB. The study also enhances the understanding of the early viral kinetics of both drugs. Treatment with ETV was superior to treatment with ADV for the mean reduction from baseline in HBV DNA by PCR at week 12. After week 12, continued treatment through week 48 resulted in a more pronounced reduction in HBV DNA in the ETV group compared with the ADV group, with a higher proportion of patients achieving HBV DNA less than 300 copies/mL by PCR at week 48. These results prospectively confirm the results of a cross-study analysis, which proposed that, over a 12-month period, the antiviral efficacy of ETV would be superior to that of both lamivudine and ADV in nucleoside-naïve patients with chronic HBV infection.28
The viral kinetic analyses based on the regression spline and the four-parameter exponential decay models provide consistent evidence of the greater early antiviral activity of ETV compared with ADV. Similar to prior reports, the decline in serum HBV DNA over the first 12 weeks of treatment was biphasic for both drugs.21-23 The initial first phase lasted 10 days, followed by a second slower phase. The rate of biphasic decline showed a statistically significant treatment difference in favor of ETV starting at day 10 and significantly greater rate of reductions in HBV DNA after day 10, during the second phase of viral clearance. Lower (day 10) HBV DNA levels were predictive of achieving HBV DNA less than 300 copies/mL at weeks 24 and 48. The observations that the differential efficacy of two treatment regimens can be detected as early as day 10, and that these early responses are predictive of longer-term virological response at weeks 24 and 48, have important implications about the value of early kinetics in the future assessment of alternative anti-HBV therapies as well as in improving understanding of how these drugs are used in current and future treatment strategies.
The half-life of circulating virus in ETV-treated patients was estimated to be 14.5 hours, which is consistent with the previous estimate for ETV-treated patients of 16 hours.23 The estimate of circulating half-life of virus for ADV-treated patients was 32.5 hours in this study, similar to a prior estimate of 26.4 hours.21 Both the regression spline model and the four-parameter exponential decay model fit the data well. We suggest that in practice these models be used in a complementary manner; the simple regression spline model provides straightforward treatment comparisons of (mean) viral load measurements over time, whereas the four-parameter exponential decay model provides estimates of viral kinetic parameters.
Serum HBV DNA does not rebound during treatment with nucleoside analogs, unless viral resistance develops or treatment is discontinued.27 One potential benefit of rapid and potent suppression of HBV replication is reduced risk of drug resistance. It has been shown that the subsequent chance of lamivudine resistance is directly proportional to the HBV DNA levels at week 24 of lamivudine treatment.29 Another study of telbivudine versus ADV showed that suppression of HBV DNA levels at 24 weeks correlates with efficacy outcomes and viral breakthroughs at 1 year.30 Longer-term studies with ETV in nucleoside-naïve patients have reported that most patients achieve HBV DNA suppression within the first 24 weeks of treatment and is maintained for up to 4 years of treatment31 with minimal resistance at 5 years (1%).11 However, HBV resistance was not monitored in the current study. The week 48 data obtained in this study for ETV are consistent with the results of larger-phase III studies that established the superiority of ETV over lamivudine.12, 13 In nucleoside-treatment naïve HBeAg-positive patients, the 0.5-mg dose of ETV produced a mean reduction in serum HBV DNA of 6.9-log10 copies/mL, and 67% of patients had undetectable HBV DNA (<300 copies/mL) after 48 weeks of treatment.12 Patients in both groups of this study had higher baseline HBV DNA than in other studies, which may explain the lower rates of patients who achieved HBV DNA of less than 300 copies/mL in this study.
There are several potential limitations of the current study. We did not evaluate viral kinetics according to HBV genotype. This parameter should be considered in future studies as knowledge of HBV kinetics matures. The influence of HBeAg status on the viral kinetics of HBV is not well understood and was not considered in this study. This population consisted of HBeAg-positive patients who had high-level viremia (HBV DNA >108 copies/mL), and the viral kinetic parameters of both ETV and ADV may demonstrate different patterns in patients with low-level viremia. The first-phase slope of the decline in HBV DNA has been reported to be steeper in HBeAg-negative patients than in HBeAg-positive patients.25 Future studies of viral kinetics also should incorporate quantitative studies of e and s antigens, as well as ALT levels, to provide a comprehensive understanding of the biological response to potent inhibition of HBV DNA replication.
In conclusion, ETV 0.5 mg daily produced more rapid and significantly greater suppression of HBV DNA than ADV 10 mg daily in nucleoside-naïve patients with HBeAg-positive CHB. Moreover, there was considerably less variability in the extent of HBV DNA reductions in patients treated with ETV, with more patients achieving HBV DNA less than 300 copies/mL at week 48. Early virological response was found to be predictive of subsequent virological response, with those having lower HBV DNA levels at day 10 being more likely to achieve HBV DNA less than 300 copies/mL at weeks 24 and 48. Greater and more rapid reduction in HBV DNA levels may result in more prolonged efficacy and less risk of viral resistance. ETV provides rapid and potent suppression of HBV DNA and has been shown to be a useful agent in the long-term management of patients with CHB infection.