Potential conflict of interest: Dr. Wong is a consultant for Novartis. Drs. Sung and Chan are on the speakers' bureau of GlaxoSmithKline and Roche. They are also on the speakers&#146; bureau of and received grants from AstraZeneca.
Approximately 30%-40% of patients with hepatitis B e antigen (HBeAg)-positive chronic hepatitis B treated with peginterferon and/or lamivudine achieve HBeAg seroconversion 6 months after the end of treatment. The durability and long-term effect of treatment are unknown. In this study, 85 HBeAg-positive patients who received peginterferon alfa-2b 1.5 μg/kg/week for 32 weeks and lamivudine 100 mg/day for 52 or 104 weeks were prospectively followed for 6.1 ± 1.7 years posttreatment. Twenty-five (29%) patients had virologic response (HBeAg seroconversion and HBV DNA <10,000 copies/mL) at 5 years. The rate of HBeAg seroconversion rose progressively from 37% at the end of treatment to 60% at 5 years. Twenty-seven (32%) and 11 (13%) patients had undetectable HBV DNA (<100 copies/mL) at the end of peginterferon treatment and at 5 years, respectively. Two (2.4%) patients achieved hepatitis B surface antigen (HBsAg) seroclearance at 2.6 and 84 months posttreatment. Among virologic responders at the end of treatment, 82% and 57% and sustained HBeAg seroconversion and virologic response at 5 years. End-of-treatment serum quantitative HBsAg was significantly lower in patients with sustained virologic response at 5 years (median 1,431 IU/mL versus 2,689 IU/mL [P = 0.041]). At the last follow-up, the liver stiffness measurement by transient elastography was 5.8 ± 2.7 kPa. Only two patients had liver stiffness suggestive of advanced fibrosis. Week 16 HBV DNA, end-of-treatment HBeAg seroconversion, and undetectable HBV DNA were independent factors associated with virologic response at 5 years. The duration of concomitant lamivudine treatment had no impact on any long-term response. Conclusion: Peginterferon has high durability in HBeAg-positive chronic hepatitis B patients with end-of-treatment virologic response. (Hepatology 2010;)
Chronic hepatitis B is one of the most common causes of liver cirrhosis and hepatocellular carcinoma worldwide.1 It is estimated that over 350 million people are chronically infected with hepatitis B virus (HBV) globally. High viral load, positive hepatitis B e antigen (HBeAg) and cirrhosis are independent factors associated with the development of hepatocellular carcinoma.2–4 Correspondingly, effective viral suppression and histologic improvement may reduce the risk of hepatocellular carcinoma and other liver-related complications.5–7
At present, two types of antiviral treatment are available for the treatment of chronic hepatitis B. Oral nucleoside/nucleotide analogs effectively inhibit HBV polymerase, suppress the viral load, and reduce hepatic necroinflammation and fibrosis. However, many patients develop virologic relapse after cessation of treatment.8, 9 Some patients may develop hepatic decompensation if treatment cessation is premature.10 Long-term treatment is also limited by the emergence of drug resistance and potential late adverse effects.11, 12
In contrast, interferon has both direct antiviral and immunomodulatory actions. Although interferon carries more side effects and requires subcutaneous administration, it has the advantage of a finite duration of treatment and no risk of drug resistance. Peginterferon is produced by the addition of a polyethylene glycol moiety to interferon, thus prolonging the half-life of the drug and the average drug concentration in the body. After a course of peginterferon treatment, 30%-40% of the recipients develop HBeAg seroconversion, a rate significantly higher than that of lamivudine users.13 However, long-term data are lacking. In Caucasians treated with peginterferon and/or lamivudine, sustained HBeAg and hepatitis B surface antigen (HBsAg) seroclearance was observed in 81% and 30%, respectively, of those with initial response.14 Because most Asians acquire HBV infection perinatally, it is unclear whether the same favorable outcomes apply to Asians.
In this study, we aimed to study the long-term durability of peginterferon alfa-2b in HBeAg-positive chronic hepatitis B patients who participated in two previously conducted clinical trials. Factors associated with long-term virologic response were also identified.
ALT, alanine aminotransferase; anti-HBe, antibody against hepatitis B e antigen; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; IQR, interquartile range; qHBsAg, quantitative hepatitis B surface antigen.
Patients and Methods
The details of the two clinical trials were previously reported.15, 16 In the first study, patients were randomized to receive peginterferon alfa-2b for 32 weeks and lamivudine for 52 weeks versus lamivudine monotherapy for 52 weeks.15 In the second study, patients were randomized to receive different combination schedules of peginterferon and lamivudine.16 Study participants were chronic hepatitis B patients 18-65 years of age recruited at the hepatology clinic of the Prince of Wales Hospital, Hong Kong. All patients had positive HBsAg for at least 6 months, positive HBeAg, serum HBV DNA of at least 500,000 copies/mL, and serum alanine aminotransferase (ALT) levels between 1.3 and 5 times the upper limit of normal. Patients with liver decompensation or prior use of interferon or antiviral therapy were excluded. Other exclusion criteria were coinfection with hepatitis C virus, delta virus, or human immunodeficiency virus; history of hepatocellular carcinoma; coexistence of autoimmune hepatitis, Wilson disease, hemochromatosis or α1-antitrypsin deficiency; serious medical or psychiatric illness; concurrent use of corticosteroid or immunosuppressive agents; and pregnancy. The study protocols were approved by the local ethics committee. All patients provided informed written consent.
Study Intervention and Follow-up.
In the long-term study, we included only subjects in the peginterferon arms. All patients received peginterferon alfa-2b at a dosage of 1.5 μg/kg/week (up to a maximum of 100 μg per week) subcutaneously for 32 weeks. They also received oral lamivudine 100 mg daily for 52 weeks15 or 104 weeks.16 Most of the patients who received lamivudine monotherapy in the original clinical trial received alternative antiviral therapy, which precluded meaningful comparison with those treated with peginterferon and lamivudine. Thus, the lamivudine monotherapy arm was not included in the current analysis.
At the end of the clinical trials, the patients were prospectively followed up at the outpatient clinic every 3-6 months as clinically indicated. The follow-up time was calculated from the end of peginterferon treatment in the original studies. For patients who did not complete the full course of treatment, the follow-up time was calculated from the date of the intended final dose of peginterferon. During each follow-up visit, liver biochemistry, serum alpha-fetoprotein, HBeAg, and antibody against HBeAg (anti-HBe) were checked. HBsAg and HBV DNA were measured once a year. Patients were offered oral nucleoside/nucleotide analogs in case of virologic and/or biochemical relapse.
The primary efficacy end point of the long-term follow-up study was virologic response at 5 years, defined as HBeAg seroconversion and HBV DNA reduction to below 10,000 copies/mL. HBeAg seroconversion was defined as negative HBeAg and positive anti-HBe for at least two consecutive visits 6 months apart. Other secondary endpoints included serum HBV DNA reduction to below 10,000 copies/mL and undetectable level (<100 copies/mL), HBsAg seroclearance, normalization of ALT level, and need for retreatment. Patients who defaulted follow-up before year 5 or received retreatment were considered nonresponders for all categorical endpoints except for HBeAg response, for which we adopted the last observation carried forward method at the time of retreatment.
Enzyme-linked immunosorbant assay kits were used to test HBsAg (Roche Diagnostics Corp., Indianapolis, IN), HBeAg, and anti-HBe (Sanofi Diagnostics, Pasteur, France). TaqMan real-time polymerase chain reaction assay was used to measure HBV DNA levels.17 The detection range was 102-109 copies/mL. HBV genotyping was performed by restriction fragment length polymorphism at the initial visit as described.18 Quantitative HBsAg (qHBsAg) was quantified using Architect HBsAg QT (Abbott Diagnostic, Wiesbaden, Germany).19 The sensitivity of the Architect assay ranged from 0.05 to 250 IU/mL. Samples with an HBsAg titer >250 IU/mL were diluted to 1:500 to 1:999 to bring the reading to the range of the calibration curve.
Liver biopsy was performed at the baseline and end of per-protocol lamivudine treatment.15, 16 Liver biopsies were evaluated by two pathologists who were blinded to the treatment assignment (P. C.-L. C., A. W.-H. C.). Ishak scoring system was used to score the histologic activity index (0-18) and liver fibrosis (0-6).20
Liver Stiffness Measurement.
Liver stiffness measurement using transient elastography (Fibroscan; Echosens, Paris, France) was performed on patients who did not receive retreatment until the final follow-up visit. The procedure was performed by a single experienced investigator (K. K.-L. Y.) according to the instructions and training provided by the manufacturer. Details of the technical background and examination procedure have been described.21 Ten successful acquisitions were performed on each patient. The median value represented the liver elastic modulus. The success rate was calculated as the number of successful measurements divided by the total number of measurements. The operators were blinded to all clinical data and the diagnosis of the patients. Only cases with 10 successful acquisitions, an interquartile range (IQR) to median ratio below 0.3, and a success rate of ≥60% were evaluated. According to our previous ALT-based algorithm with histologic validation, probable advanced fibrosis was defined as liver stiffness ≥9.0 kPa in patients with normal ALT and ≥12.0 kPa in patients with elevated ALT.22
All statistical analysis was performed using SPSS version 16.0 software (SPSS, Chicago, IL). Continuous variables are expressed as the mean ± standard deviation or median (IQR). Group comparison was performed using the Student t test, Mann-Whitney U test, chi-square test, or Fisher's exact test as appropriate. The cumulative HBeAg seroconversion and reversion was performed by way of Kaplan-Meier analysis. The accuracy of qHBsAg in predicting combined response at 5 years was assessed by way of receiver operating characteristics curve analysis. Binary logistic regression analysis was performed to identify independent factors associated with virologic response at 5 years. Statistical significance was taken as a two-sided P value of less than 0.05.
Eighty-five patients who received combination therapy of peginterferon alfa-2b and lamivudine were included in this study. Their clinical characteristics are shown in Table 1. Three (3.5%) patients stopped peginterferon prematurely after 11, 17, and 22 doses, respectively. Two patients stopped treatment because of side effects, and one died from variceal bleeding during the study period. The posttreatment follow-up duration was 6.1 ± 1.7 years (range, 1.0-8.3 years). Five (5.9%) patients defaulted follow-up 1.0-2.0 years after stopping peginterferon treatment. Twenty-five (29%) patients received retreatment by nucleoside/nucleotide analogs due to disease relapse at a median of 2.3 years (range, 0.7-6.0 years).
Table 1. Clinical Characteristics of Chronic Hepatitis B Patients With and Without Virologic Response at 5 Years
All (N = 85)
Virologic Response at 5 Years (n = 25)
No Virologic Response at 5 Years (n = 60)
Excludes patients on retreatment.
Continuous variables are expressed as the mean ± standard deviation or median (IQR).
Overall, virologic response was achieved by 28 (33%) patients at the end of peginterferon treatment and 25 (29%) patients at 5 years. Sixteen of 28 (57%) patients with end-of-treatment virologic response had sustained virologic response at 5 years. In comparison, nine of 57 (16%) patients who did not achieve end-of-treatment virologic response developed virologic response at 5 years (P < 0.001 versus patients with end-of-treatment response) (Fig. 1).
At the end of peginterferon treatment, 31 (37%) patients achieved HBeAg seroconversion. The overall rate of HBeAg seroconversion rose progressively to 60% by year 5 posttreatment (Fig. 2). Among patients with HBeAg seroconversion at the end of treatment, seven (23%) developed HBeAg reversion. The cumulative incidence of HBeAg reversion at 6 months, 1 year, 2 years, and 3 years posttreatment was 7%, 10%, 19%, and 23%, respectively by way of Kaplan-Meier analysis (Fig. 3A). HBeAg reversion occurred at a median of 1.2 years posttreatment (range, 0.2-2.7). No more HBeAg reversion was observed from 3 to 8 years. At 5 years, 24 of 31 (77%) patients had sustained HBeAg seroconversion.
Among the 54 (64%) patients who remained HBeAg-positive at the end of peginterferon treatment, 37 (69%) developed HBeAg seroconversion after stopping peginterferon. The cumulative incidence of spontaneous posttreatment HBeAg seroconversion at 6 months, 1 year, 2 years, 3 years, 4 years, and 5 years posttreatment was 25%, 33%, 49%, 65%, 67%, and 71%, respectively (Fig. 3B). Spontaneous HBeAg seroconversion occurred at a median of 1.5 years posttreatment (range, 0.1-7.1). Twenty-seven of 37 (73%) patients who had spontaneous HBeAg seroconversion remained HBeAg-negative and anti-HBe–positive at 5 years, whereas 10 (27%) developed HBeAg reversion afterwards.
At year 5 after peginterferon treatment, HBeAg seroconversion was sustained in 36 of 55 (67%) patients who received concomitant lamivudine for 1 year and 15 of 30 (50%) patients who received concomitant lamivudine for 2 years (P = 0.17).
At the end of peginterferon treatment, 27 (32%), 31 (36%), and 55 (65%) patients had HBV DNA below 100, 400, and 10,000 copies/mL, respectively. At 5 years, 11 (13%), 11 (13%), and 26 (31%) patients had HBV DNA below 100, 400, and 10,000 copies/mL, respectively. Patients with end-of-treatment virologic response were more likely to have sustained viral suppression at 5 years. At 5 years, HBV DNA suppression to below 10,000 copies/mL was found in 16 of 28 (57%) patients with end-of-treatment virologic response, compared with 10 of 57 (18%) patients without end-of-treatment response (P < 0.001) (Fig. 1).
On the other hand, the duration of concomitant lamivudine treatment had no impact on long-term viral suppression. At 5 years, HBV DNA was below 10,000 copies/mL in 19 of 55 (35%) patients who received concomitant lamivudine for 1 year and seven of 30 (23%) patients who received concomitant lamivudine for 2 years (P = 0.28).
Two (2.4%) patients achieved HBsAg seroclearance. Both occurred after stopping peginterferon treatment. One case of HBsAg seroclearance occurred 2.6 months after the end of treatment, whereas the second case occurred 84 months after stopping treatment. At the end of treatment, both patients achieved HBeAg seroconversion and had undetectable HBV DNA.
The median qHBsAg level was 6,419 IU/mL (IQR, 2,284-22,942) at baseline and 2,389 IU/mL (IQR, 1,193-8,536) at the end of peginterferon treatment. The baseline qHBsAg level was 4,402 IU/mL (IQR, 1,828-11,525) among patients achieving virologic response at 5 years and 9,127 IU/mL (IQR, 2,523-30,643) among those who did not achieve virologic response (P = 0.064). Patients who achieved virologic response at 5 years also had significantly lower qHBsAg at the end of treatment (1,431 IU/mL; IQR, 605-3,846) than those without virologic response (2,689 IU/mL; IQR, 1,416-10,944) (P = 0.041). The area under the receiver operating characteristics curve of qHBsAg at the end of treatment to predict virologic response at 5 years was 64.3% (95% confidence interval, 51.8%-76.8%). Using end-of-treatment qHBsAg below 1,500 IU/mL to predict virologic response at 5 years, the sensitivity, specificity, positive predictive value, and negative predictive value were 52%, 73%, 46%, and 77%, respectively.
At 5 years, 48 (57%) patients had normal ALT levels. Among 57 patients with normal ALT levels at the end of treatment, 34 (60%) had normal levels at 5 years. In comparison, 14 of 28 (50%) patients with abnormal ALT levels at the end of treatment also had ALT normalization at 5 years (P = 0.40 versus biochemical responders). On the other hand, virologic response at the end of treatment predicted biochemical response at 5 years. Twenty-two (79%) patients with end-of-treatment virologic response had normal ALT levels at 5 years, compared with 26 of 57 (46%) patients without virologic response (P = 0.004). In contrast, long-term biochemical response was not affected by the duration of concomitant lamivudine treatment. Thirty-three of 55 (60%) patients receiving concomitant lamivudine for 1 year and 15 of 30 (50%) patients receiving concomitant lamivudine for 2 years had normal ALT levels at 5 years (P = 0.37).
Liver Stiffness Measurement
Transient elastography was performed in 53 of 60 (88%) patients who did not receive retreatment at a median duration of 7.0 years (range, 4.8-8.5 years) after the end of peginterferon treatment. Liver stiffness measurement was reliable in 46 (87%) patients, including 25 (54%) patients with end-of-treatment virologic response and 21 (46%) patients without virologic response. Their baseline and follow-up (end of lamivudine treatment) fibrosis stage by histology was 1 (IQR, 1-2) and 1 (IQR, 1-2), respectively. Stage 4 to 6 fibrosis was found in five (11%) and four (9%) patients at baseline and follow-up liver biopsies, respectively. There was no significant difference in the follow-up fibrosis stage between patients with and without end-of-treatment virologic response (median fibrosis stage 1 [IQR, 1-4] versus 1 [IQR, 1-2]; P = 0.088).
The median liver stiffness at the last visit was 5.8 ± 2.7 kPa. Only two (4%) patients had liver stiffness suggestive of advanced fibrosis at 9.7 kPa and 16.8 kPa. The first patient was a 54-year old man who achieved HBeAg seroconversion and low HBV DNA (541 copies/mL) at the end of peginterferon treatment. However, his ALT level remained high at 159 IU/L. Baseline and end-of-treatment fibrosis stage was 5 and 4, respectively. At the time of transient elastography, his ALT level was 55 IU/L and HBV DNA was 4,309 copies/mL. The second patient was a 41-year old man who remained HBeAg-positive with HBV DNA at 12,227 copies/mL at the end of peginterferon treatment. His ALT level was high at 89 IU/L. He had stage 1 fibrosis at both baseline and end of treatment. At the last follow-up visit at 6 years, he remained HBeAg-positive. The ALT and HBV DNA levels were 136 IU/L and 1.37 × 109 copies/mL, respectively.
Among 42 patients without histologic advanced fibrosis at the end of peginterferon treatment, only one (2%) had liver stiffness suggestive of advanced fibrosis at the last visit. Among four patients with histologic advanced fibrosis at the end of peginterferon treatment, one (25%) still had advanced fibrosis at the last visit, whereas three (75%) had liver stiffness not suggestive of advanced fibrosis.
Predictors of Response at 5 Years
Patients with virologic response at the end of peginterferon treatment were more likely to have normal ALT levels, sustained HBeAg seroconversion, and HBV DNA suppression to below 10,000 copies/mL at 5 years (Fig. 1).
Patients who achieved virologic response at 5 years had lower HBV DNA levels and a higher histologic activity index at baseline (Table 1). On the other hand, age, sex, body mass index, ALT, viral genotype, and fibrosis stage at baseline did not influence the long-term virologic response. Patients with virologic response at 5 years also had significantly lower HBV DNA levels at week 16 and 24. In addition, patients who achieved virologic response at 5 years were more likely to have achieved HBeAg seroconversion and HBV DNA suppression at the end of peginterferon treatment (Table 1). They also had lower end-of-treatment qHBsAg.
On multivariate analysis, low week 16 HBV DNA remained an independent factor associated with virologic response at 5 years after adjusting for age, sex, viral genotype, baseline ALT, HBV DNA, and histologic activity index (Table 2). Six of 33 (18%) patients with week 16 HBV DNA >10,000 copies/mL and 19 of 52 (37%) patients with week 16 HBV DNA <10,000 copies/mL achieved virologic response at 5 years (P = 0.070). In comparison, only one of 20 (5%) patients with week 16 HBV DNA above 100,000 copies/mL and 24 of 65 (37%) patients with week 16 HBV DNA below 100,000 copies/mL achieved virologic response at 5 years (P = 0.005). HBeAg seroconversion and HBV DNA suppression to <100 copies/mL were independent end-of-treatment factors associated with virologic response to peginterferon at 5 years (Table 2).
Table 2. Multivariate Analysis of Factors Associated With Virologic Response at 5 Years
Odds Ratio (95% Confidence Interval)
Baseline, on-treatment variables, and end-of-treatment variables were evaluated in separate multivariate models.
Baseline and on-treatment variables
Baseline ALT (IU/L)
Baseline HBV DNA (log copies/mL)
Histologic activity index
Week 16 HBV DNA (log copies/mL)
HBV DNA <100 copies/mL
This study provided data on the long-term response to peginterferon treatment in Chinese patients with HBeAg-positive chronic hepatitis B. The majority of patients who achieved virologic response at the end of treatment had a durable response. At 5 years, around 80% of initial responders had sustained HBeAg seroconversion and ALT normalization, and 60% had HBV DNA <10,000 copies/mL. Spontaneous HBeAg seroconversion was commonly observed even after stopping peginterferon treatment. Up to 60% of the overall population developed HBeAg seroconversion by 5 years. Progression to cirrhosis or advanced fibrosis was rare in treatment responders. End-of-treatment virologic response was the strongest predictor of long-term sustained virologic response.
In patients who achieve HBeAg seroconversion after conventional interferon-α, HBeAg reversion is observed in ≈10%-20% after 4-14 years.23–26 A multicenter European study showed that 81% of responders to peginterferon alfa-2b had sustained HBeAg loss after a mean follow-up of 3 years.14 Similarly, in our study, 77% of the initial responders continued to have HBeAg seroconversion at 5 years. These rates compare favorably with the durability of oral nucleoside analogs. In patients who achieved HBeAg seroconversion after lamivudine treatment, nearly half develop HBeAg reversion by 2 years.8 Although longer consolidation therapy may enhance the durability of lamivudine, a recent study showed that 82% of patients still developed virologic rebound at 4 years, and 16% also had ALT increased to 2 times the upper limits of normal or more.9 The poor durability of lamivudine may result in hepatic decompensation and deaths.10 With the development of oral antiviral drugs with high genetic barrier to resistance such as entecavir and tenofovir, the problem of drug resistance is much reduced.27, 28 Nevertheless, cost-effectiveness, long-term safety, and durability after treatment cessation remain important questions.
As an immunomodulatory agent, peginterferon may have an effect on HBeAg seroconversion and HBsAg seroclearance long after the end of treatment.29 Because the current study did not include untreated controls, we could not confirm whether the high rate of HBeAg seroconversion was due to the delayed effect of peginterferon or merely the rate of spontaneous HBeAg seroconversion with time. However, in patients with HBeAg-negative chronic hepatitis B treated with peginterferon alfa-2a, HBsAg seroclearance occurs in 3% at month 6 posttreatment and 9% at year 3.30 The high rate of HBsAg seroclearance was not observed in control patients who received lamivudine monotherapy. In Caucasians with HBeAg-positive chronic hepatitis B treated with peginterferon alfa-2b, the rate of HBsAg seroclearance rose from 7% at week 78 to 11% at about 3 years, although the study did not show a parallel increase in the rate of HBeAg seroclearance.14 In our study, progressive increase in HBeAg seroconversion was observed in patients without initial virologic response even after 5 years.
Only two patients in this study achieved HBsAg seroclearance; this rate was lower than those reported by Buster et al.14 and Marcellin et al.30 The difference could not be explained by baseline ALT and HBV DNA levels, which were similar among all three studies. On the other hand, in the study by Marcellin et al., high rate of HBsAg seroclearance was mainly observed in patients with genotype A HBV infection. In randomized controlled trials, patients with genotype A HBV infection had the best response to peginterferon and had the highest rates of HBeAg seroconversion and HBV DNA suppression.31, 32 Apart from the influence of genotypes, the difference may also be explained by the age of infection. In Asia, most patients acquire HBV infection through perinatal transmission. This is characterized by a long immunotolerant phase and a lower rate of spontaneous HBsAg seroclearance.33
The pivotal trials on peginterferon treatment in patients with chronic hepatitis B used virologic response 6 months after treatment as a study endpoint.15, 31, 34 In chronic hepatitis C, negative hepatitis C virus RNA 6 months after interferon and/or ribavirin treatment is a robust endpoint indicative of viral clearance.35 On the other hand, HBV cannot be eradicated using currently available treatments. In this study, HBeAg reversion occurred in up to 20% of patients with initial HBeAg seroconversion. Because no more HBeAg reversion occurred after 3 years posttreatment, sustained HBeAg seroconversion for 3 years should be present before the disease can be considered to be in long-term remission.
On-treatment HBV DNA suppression was predictive of long-term outcome. After adjusting for baseline factors, week 16 HBV DNA remained independently associated with virologic response at 5 years. This confirmed the results of our previous report, which showed that on-treatment HBV DNA levels could predict virologic response at 1 year.36 Because only 5% of patients with week 16 HBV DNA >100,000 copies/mL could achieve virologic response at 5 years, this may be considered as an early stopping rule. Moreover, unlike the multicenter studies, baseline ALT and HBV DNA levels were less important to predict long-term response.14, 30 This might be due to our smaller sample size, difference in the definition of treatment response, duration of follow-up, ethnic differences, and genotype distribution.
Our study had a few limitations. First, the sample size was relatively small. However, our cohort was unique in that over 90% of the study population underwent long-term scheduled follow-up. In addition, our cohort represents the largest Asian study population on the long-term outcome of HBeAg-positive patients. Second, for ethical reasons, it was inevitable that some patients with virologic and biochemical relapse required retreatment with oral nucleoside/nucleotide analogs. However, this issue was addressed cautiously by the intention-to-treat design. Lastly, liver biopsy was not performed at the final visit to confirm the favorable long-term histologic response. Nevertheless, transient elastography represents one of the most accurate noninvasive tests for liver fibrosis and has been validated in different chronic liver diseases.21, 37 In addition, transient elastography has a high negative predictive value and is useful in excluding advanced fibrosis.
In conclusion, peginterferon alfa-2b has high durability in HBeAg-positive chronic hepatitis B patients with end-of-treatment virologic response. Virologic relapse is rare after sustained response for 3 years. Spontaneous HBeAg seroconversion may occur after the end of treatment.