Treatment of patients with chronic hepatitis B who have failed previous antiviral treatment with pegylated interferon α2a (40 kda; PEGASYS )

Original article Antiviral Therapy 13:555 562 Treatment of patients with chronic hepatitis B who have failed previous antiviral treatment with pegylated interferon α2a (40 kda; PEGASYS ) Henry L-Y Chan*, Vincent W-S Wong, Angel M-L Chim, Grace L-H Wong, Hoi-Yun Chan and Joseph J-Y Sung Department of Medicine and Therapeutics and Institute of Digestive Disease, The Chinese University of Hong Kong, Hong Kong SAR, China *Corresponding author: E-mail: hlychan@cuhk.edu.hk Background: Although nucleot(s)ide analogues can effectively suppress hepatitis B virus (HBV) replication, many patients experience relapse of hepatitis after cessation of treatment. We aimed to investigate the efficacy of pegylated interferon α2a (PEG-IFN-α2a) in these difficult-to-treat patients. Methods: Chronic hepatitis B patients who have received antiviral drugs for 12 months and stopped for 6 months were treated by 48-week PEG-IFN-α2a. Virological response was defined as HBV DNA <10,000 copies/ml and hepatitis B e antigen (HBeAg) seroconversion (for HBeAg-positive patients). Results: A total of 40 patients, 29 HBeAg-positive and 11 HBeAg-negative, with median log 10 HBV DNA 7.3 copies/ml and alanine aminotransferase 110 IU/ml were studied. The last antiviral treatment was given for 92 ±61 weeks and stopped for 176 ±88 weeks. At the end of treatment, 22 (12 HBeAg-positive and 10 HBeAg-negative; 55%) patients had virological response and 16 (7 HBeAg-positive and 9 HBeAg-negative; 40%) patients had undetectable HBV DNA (<100 copies/ml). At 24 weeks post-treatment, 14 (8 HBeAg-positive and 6 HBeAgnegative; 35%) patients had virological response and 9 (5 HBeAg-positive and 4 HBeAg-negative; 23%) patients had undetectable HBV DNA. Two (5%) patients had lost hepatitis B surface antigen. HBV DNA levels at week 24 best predicted sustained virological response (area under curve 0.76, 95% confidence interval 0.60 0.92, P=0.007). At HBV DNA cutoffs of 3 logs and 5 logs at week 24, the sensitivity/specificity for sustained virological response were 50%/85% and 86%/62%, respectively. Conclusions: PEG-IFN-α2a was effective in the treatment of chronic hepatitis B patients who have failed previous antiviral treatment. Introduction Chronic hepatitis B virus (HBV) infection is the most common cause of liver cirrhosis and hepatocellular carcinoma in Asia. Persistent high viraemia and hepatic necroinflammation is associated with a higher risk of liver-related complications including hepatocellular carcinoma [1,2]. Nucleos(t)ide analogues such as lamivudine have potent antiviral activities that can effectively suppress the replication of HBV. In hepatitis B e antigen (HBeAg)-positive patients, HBeAg seroconversion is usually regarded as an indicator of successful viral clearance. However, up to 50% of patients had HBV reactivation and hepatitis relapse after stopping lamivudine treatment despite HBeAg seroconversion [3,4]. There is increasing evidence that extended antiviral treatment after HBeAg seroconversion can consolidate the viral suppression and reduce the risk of post-treatment relapse [5]. Therefore, most of the current treatment guidelines recommend treatment cessation after patients have developed HBeAg seroconversion for over 6 months [6 8]. In a placebo-controlled, randomized trial of 2-year lamivudine treatment among HBeAg-negative patients, approximately two-thirds of patients had HBV DNA >10,000 copies/ml and 40% of patients had elevated alanine aminotransferase (ALT) within 6 months after stopping lamivudine [9]. The failure of viral clearance by antiviral agents might be related to their weak efficacy to clear the covalently closed circular (ccc) DNA inside the hepatocytes [10,11]. Pegylated interferon α2a (PEG-IFN-α2a), either used alone or in combination with lamivudine, has been shown to induce virological response in approximately one-third of HBeAg-positive patients and 43% of HBeAg-negative patients at 6 months post-treatment [12 16]. The virological response tends to be sustained 2008 International Medical Press 1359-6535 555

HL-Y Chan et al. in the majority of patients for at least up to 3 years [17,18]. The efficacy of PEG-IFN-α2a to clear cccdna seems to be superior to that of nucleos(t)ide analogues [19,20]. In a subgroup analysis of a previous clinical trial, patients who have previous exposure to antiviral agents responded equally well to PEG-IFN-α2a as the treatment-naive patients [21]. However, the previous antiviral treatments in these patients were not clearly documented and it was uncertain whether the treatment was adequate. In this study, we aimed to investigate the efficacy of PEG-IFN-α2a among a cohort of chronic hepatitis B patients who have failed previous antiviral treatment. These patients must have received a minimum of 1 year treatment but still have active viraemia and disease activity for at least 6 months after stopping treatment. In other words, these patients were unlikely to develop spontaneous disease remission and were difficult-totreat. Predictors of virological response to PEG-IFN-α2a were also investigated in this study. Methods Patients Chronic hepatitis B patients were recruited from the Hepatitis Clinic of the Prince of Wales Hospital (Shatin, Hong Kong) between April 2005 and October 2006. Eligible patients must be aged 18 70 years and have positive hepatitis B surface antigen (HBsAg) for at least 6 months. The serum HBV DNA must be at least 100,000 copies/ml and ALT must be 1.5- to 10- fold higher than the upper limit of normal at the screening visit. Both HBeAg-positive and HBeAgnegative patients were recruited. As this study was designed to evaluate patients who did not respond to an adequate trial of antiviral treatment, patients must have a documented history of previous treatment by nucleos(t)ide analogues for at least 1 year and the treatment must have been stopped for at least 6 months before the screening visit. The reason for previous treatment cessation could be treatment failure, part of the clinical trial or mutual agreement between the physician and the patient. Patients were excluded if they had coinfection with hepatitis C virus (HCV) or human immunodeficiency virus (HIV), decompensated liver disease, history of hepatocellular carcinoma, other causes of liver disease, serious medical or psychiatric illness, or concurrent use of corticosteroid or immunosuppressive agents. Women of child-bearing age must have negative urine pregnancy test before recruitment. The study was conducted in accordance with the guidelines of the Declaration of Helsinki. The protocol was approved by the local ethics committee. All patients gave witnessed written informed consent before enrolment. Study design This study was a single-centred, open-labelled study of PEG-IFN-α2a (Pegasys ; F Hoffmann-La Roche, Basal, Switzerland) for the treatment of chronic HBV infection that failed previous antiviral treatment. PEG-IFN-α2a was prescribed as subcutaneous injections at a dose of 180 μg/week for 48 weeks. Treatment was commenced within 4 weeks of the screening visit. HBV genotype and genotypic resistance to lamivudine and adefovir were checked at the baseline visit. Patients were monitored at weeks 1, 2, 4, 8, 12, 16, 24, 32, 40 and 48 for liver biochemistry, virology tests, adverse events and drug compliance. Patients were followed-up at weeks 52, 56, 60 and 72 after cessation of treatment. The investigators interviewed the patients for symptomatic adverse effects and monitored laboratory tests closely at each follow-up visit. The severity of adverse events were assessed according to a preset table and classified into mild (grade 1), moderate (grade 2), severe (grade 3) and life-threatening (grade 4). The dosage of PEG-IFN-α2a was reduced from 180 μg/week to 135 μg/week for grade 3 adverse events (or sometimes grade 2 adverse events at the discretion of the investigator). The dosage could be further reduced to 90 μg/week and then to 45 μg/week if the adverse events persisted or recurred despite initial dose reduction. PEG-IFN-α2a was stopped in cases of grade 4 adverse events. Laboratory assays Serological assays HBsAg was tested using commercially available ELISA kits (COBAS CORE HBsAg IIEIA, Roche Diagnostics Corporation, Indianapolis, IN, USA) and confirmed by a neutralization assay. Antibodies against HCV (anti- HCV, third generation assay) and HIV (anti-hiv) were tested using commercially available ELISA kits (Abbott GmBH Diagnostika, Wiesbaden-Delkenheim, Germany). HBeAg and antibodies to HBe (anti-hbe) were measured by ELISA assays (Sanofi Diagnostics, Pasteur, France). Virological assays Residual serum samples were stored at -80 C. HBV DNA was extracted by QIAGEN QIAamp DNA Mini Kit (QIAGEN Inc., Chatsworth, CA, USA) according to the instructions of the manufacturer. HBV DNA was quantified by TaqMan real-time PCR assay as described previously [22,23]. The range of HBV DNA detection was 10 2 10 9 copies/ml with correlation coefficient of the standard curve routinely >0.990. HBV genotyping was determined by restriction fragment length polymorphism and confirmed by direct sequencing in case of doubt in the residual serum samples at the initial visit, as described previously 556 2008 International Medical Press

Pegylated interferon for antiviral failure [24,25]. Lamivudine- and adefovir-resistant mutants were determined by the INNO-LiPA HBV DR2 assay according to the instructions of the manufacturer (Innogenetics N.V., Ghent, Belgium) [26]. Endpoints The primary endpoint of this study was virological response, which was defined as HBV DNA <10,000 copies/ml (plus HBeAg seroconversion in HBeAg-positive patients) at 24 weeks post-treatment (week 72). Secondary endpoints included HBeAg seroconversion, HBV DNA undetectable by PCR, ALT normalization at the end of treatment (week 48) and 24 weeks post-treatment (week 72). The ability of baseline characteristics and residual HBV DNA at various stages of PEG-IFN-α2a treatment to predict 24-week post-treatment virological response was also evaluated. We defined ALT flare during PEG-IFN-α2a treatment as pretreatment baseline level plus ALT increase of >116 IU/l (that is, 2 upper limit of normal) at any time during treatment [17,27]. Statistical analysis On the basis of previous studies of PEG-IFN-α2a among treatment-naive patients, 24-week post-treatment virological response was 32% among HBeAg-positive patients and 43% among HBeAgnegative patients [14,15]. To determine a virological response of up to 35% with half-width 95% confidence interval (CI) of 15%, 39 patients needed to be recruited for this study (INSTAT, GraphPad Software Inc, San Diego, CA, USA). Treatment response was analyzed by intention-to-treat analysis. All patients who were recruited for PEG-IFNα2a treatment were included in the analysis and any patient withdrawal or missing data was regarded as treatment failure. For the analysis of adverse events, all patients who received at least one dose of PEG-IFN-α2a were included. Statistical tests were performed by SPSS (version 11.0, Chicago, IL, USA). Continuous variables were expressed as median (range) due to our small sample size. HBV DNA was logarithmic transformed for analysis. Percentages of patients that achieved key outcome measures were presented as 95% CI by the central limit theorem on binomial distribution. Continuous variables were compared using Mann Whitney U test. Categorical variables were compared using Pearson χ 2 test or Fisher s exact test as appropriate. Logistic regression analysis was used to determine the baseline predictors of virological response at 24 weeks post-treatment. Continuous variables (age, ALT, log 10 HBV DNA) were categorized into binary variables by their respective median values in the logistic regression analysis. Area under the curve (AUC) was used to determine the predictive values of baseline and residual HBV DNA at different time intervals for sustained virological response. Statistical significance was taken as P-value <0.05. All statistical tests were two-sided. Results Clinical characteristics A total of 29 HBeAg-positive and 11 HBeAg-negative chronic hepatitis B patients were recruited in this study (Table 1). Before this study, 36 patients had received one antiviral agent (three patients had emtricitabine and 33 patients had lamivudine) and four patients had received two antiviral agents (three patients had emtricitabine and lamivudine, one patient had lamivudine and adefovir dipivoxil). Two patients had received an additional 6 month conventional interferon-α before the antiviral therapy. The duration of the last antiviral treatment was 92 ±61 (range 48 284) weeks and the duration of off-treatment follow-up before recruitment was 176 ±88 (range 32 409) weeks. During the last antiviral treatment, 20 of the 29 HBeAg-positive patients had undergone HBeAg seroconversion for a median of 41 (range 0 401) weeks and nine patients were persistently HBeAg-positive. Among the 20 patients who had HBeAg seroconversion (and subsequent reversion to positive HBeAg before this study), eight patients had HBV DNA available and four (50%) of them had HBV DNA suppressed to <10,000 copies/ml at the time of treatment cessation. Among 29 HBeAgpositive patients, 22 stopped the last antiviral treatment according to the protocol in clinical trials and the remaining seven patients stopped treatment based on the mutual decision of the physician and the patient. Among the 11 HBeAg-negative patients, five patients were initially HBeAg-positive and developed HBeAg seroconversion (only one patient had end-of-treatment HBV DNA available at 27,200 copies/ml) while six patients were persistently HBeAg-negative (three of six patients who had HBV DNA available at the end of treatment had HBV DNA suppressed to <10,000 copies/ml) during the last antiviral treatment. Among the 11 HBeAg-negative patients, eight had antiviral treatment stopped according to the protocol of the clinical trial. Overall, 14 (10 HBeAg-positive and four HBeAg-negative) patients were tested for the presence of lamivudine-resistant mutations at the end of the last antiviral treatment and three patients had the rtm204i mutation. At baseline, 18 (45%) patients had ALT >2 upper limit of normal. Liver cirrhosis was detected on ultrasound in 13 (33%) patients, but none of the patients underwent a liver biopsy within the 6 months of treatment. Two patients had pre-existing lamivudineresistant mutations (one patient with rtm204m/i and Antiviral Therapy 13:4 557

HL-Y Chan et al. Table 1. Clinical patient characteristics HBeAg-positive HBeAg-negative Overall Characteristic n=29 n =11 n =40 Male, n (%) 23 (79) 9 (82) 32 (80) Median age, years (range) 45 (25 60) 49 (38 61) 45 (25 61) Median body mass index, kg/m 2 (range) 23.8 (17.6 30.9) 25.1 (20.0 29.2) 24.6 (17.6 30.9) Median platelet count, 10 9 cells/l (range) 189.0 (81.0 303.0) 234.0 (136.0 327.0) 192.5 (81.0 327.0) Median international normalized ratio (range) 1.05 (1.01 1.25) 1.07 (1.0 1.19) 1.06 (1.0 1.25) Median albumin, g/l (range) 44.0 (39.0 48.0) 44.0 (39.0 47.0) 44.0 (39.0 48.0) Median bilirubin, μmol/l (range) 14.0 (8.0 28.0) 15.0 (7.0 24.0) 14.0 (7.0 28.0) Median alanine aminotransferase, IU/l (range) 115.0 (52.0 420.0) 110.0 (39.0 248.0) 110.0 (39.0 420.0) Median log 10 HBV DNA, copies/ml (range) 7.5 (5.3 8.6) 6.8 (5.0 7.5) 7.3 (5.0 8.6) HBV genotype B, n 6 5 11 C, n 22 6 18 D, n 1 1 Lamivudine resistance, n (%) 2 (7) 0 (0) 2 (5) Early withdrawal, n (%) 2 (7) 0 (0) 2 (5) HBeAg, hepatitis B virus e antigen; HBV, hepatitis B virus. Table 2. Treatment response at end of treatment and 24 weeks post-treatment (sustained response) End of treatment (week 48) 24 weeks post-treatment (week 72) HBeAg- HBeAg- HBeAg- HBeAg- Patients All n=40 positive n=29 negative n=11 P-value All n=40 positive n=29 negative n=11 P-value Virological 22 (55%) 12 (41%) 10 (91%) 0.002 14 (35%) 8 (28%) 6 (55%) 0.11 response* HBeAg NA 14 (48%) NA NA NA 15 (52%) NA NA seroconversion HBV DNA 25 (63%) 15 (52%) 10 (91%) 0.022 14 (35%) 8 (28%) 6 (55%) 0.11 <10,000 copies/ml Undetectable 16 (40%) 7 (24%) 9 (82%) 0.001 9 (23%) 5 (17%) 4 (36%) 0.20 HBV DNA ALT normal 20 (50%) 12 (41%) 8 (73%) 0.077 25 (63%) 14 (48%) 11 (100%) 0.008 *Virological response defined as HBV DNA <10,000 copies/ml (plus hepatitis B e antigen [HBeAg] seroconversion in HBeAg-positive patients). ALT, alanine aminotransferase; HBV, hepatitis B virus; NA, not applicable. another with rtm204m/i/v). One patient withdrew due to mild alopecia and gum bleeding at week 21 and another patient withdrew at week 24 due to hypothyroidism. A total of 38 patients completed per protocol treatment and follow-up. End-of-treatment response At the end of treatment, 22 (55%; 95% CI 25 57%) patients achieved virological response (HBV DNA <10,000 copies/ml and HBeAg seroconversion for HBeAg-positive patients), including 12 (41%; 95% CI 24 61%) HBeAg-positive patients and 10 (91%; 95% CI 57 100%) HBeAg-negative patients (P=0.002; Table 2). No patients had HBeAg loss alone without the development of anti-hbe antibodies. HBeAg-negative patients responded better than HBeAg-positive patients in terms of the percentage of patients with HBV DNA <10,000 copies/ml (P=0.022) and undetectable HBV DNA (P=0.001). The percentages of patients who achieved normal ALT levels were comparable between HBeAg-positive and HBeAg-negative patients (P=0.077). Post-treatment response At 24 weeks post-treatment, virological response was achieved in 14 (35%; 95% CI 21 52%) patients, including eight (28%; 95% CI 13 48%) HBeAg-positive patients and six (55%; 95% CI 25 82%) HBeAg-negative patients (P=0.11; Table 2). Among the eight HBeAg-positive responders, seven patients also had end-of-treatment virological response while the remaining patient had positive HBeAg and undetectable HBV DNA at the end of treatment. The 558 2008 International Medical Press

Pegylated interferon for antiviral failure Table 3. Baseline predictors of virological response at 24 weeks post-treatment Variable Responder, n (%) Non-responder, n (%) Odds ratio 95% confidence interval P-value Age >45 years 8 (57%) 11 (42%) 1.82 0.49 6.76 0.37 Gender, male 12 (86%) 20 (77%) 1.80 0.31 10.69 0.69 Alanine aminotransferase 5 (36%) 14 (54%) 0.48 0.13 1.82 0.27 >110 IU/l Log 10 HBV DNA (copies/ml) 4 (29%) 16 (62%) 0.25 0.06 1.02 0.05 >7.3 copies/ml HBV genotype C 9 (64%) 19 (73%) 0.66 0.16 2.68 0.72 Lamividine-resistant mutation 1 (7%) 1 (4%) 1.92 0.11 3.30 1.0 Continuous variables (age, alanine aminotransferase, log 10 HBV DNA) were categorized into binary variables by their respective median values in the logistic regression analysis. HBV, hepatitis B virus. Table 4. Prediction of virological response at 24 weeks post-treatment by HBV DNA at different time intervals Log 10 HBV DNA, copies/ml Week Responder (n=14)* Non-responder (n=26)* AUC 95% confidence interval P-value 0 6.9 ±1.0 7.4 ±0.8 0.64 0.45 0.82 0.16 4 5.6 ±1.3 6.3 ±1.4 0.66 0.48 0.83 0.11 8 5.2 ±1.6 5.7 ±1.7 0.61 0.43 0.79 0.27 12 4.7 ±1.6 5.5 ±1.8 0.67 0.50 0.84 0.08 18 4.0 ±1.5 5.4 ±1.9 0.73 0.57 0.88 0.018 24 3.4 ±1.6 5.4 ±2.0 0.76 0.60 0.92 0.006 *Values expressed as median (±SD). AUC, area under the curve; CI, HBV, hepatitis B virus. percentages of HBeAg-positive and HBeAg-negative patients who had HBV DNA <10,000 copies/ml (P=0.11) and undetectable HBV DNA (P=0.20) became comparable 6 months after stopping PEG-IFNα2a. Among the two patients who had pre-existing lamivudine resistant-mutants (both had positive HBeAg at baseline), one (50%) of them achieved sustained virological response. Among the 14 patients who had virological response at week 24 post-treatment, 11 of them reached 1 year post-treatment follow-up at the time of analysis. Out of these 11 patients, eight (73%) remained in virological response at 1 year post-treatment. The remaining three patients (two patients had positive HBeAg and one patient had negative HBeAg at baseline) remained HBeAg-negative but had reactivation of HBV DNA from 622,900 to 2,180,000 copies/ml. Two (5%) patients developed HBsAg loss at the end of treatment and sustained this loss until 24 weeks after treatment. One of them had genotype C HBV infection with positive HBeAg, HBV DNA 8.3 log 10 copies/ml and ALT 196 IU/l at baseline. Another patient had genotype B HBV infection with negative HBeAg, HBV DNA 6.2 log 10 copies/ml and ALT 140 IU/l at baseline. Predictors of 24-week post-treatment virological response Responders to PEG-IFN-α2a treatment tended to have lower HBV DNA at baseline but the difference fell short of statistical significance (Table 3). Otherwise, no baseline variable including age, gender, ALT level, HBeAg status, HBV genotype or pre-existing lamivudine-resistant mutation could predict virological response at 24 weeks post-treatment. Serial HBV DNA measurements suggested that residual HBV DNA at week 24 could best predict post-treatment virological response (Tables 4 and 5). At week 24, a HBV DNA cut-off of 1,000 copies/ml had sensitivity, specificity, positive and negative predictive values for post-treatment virological response of 50%, 85%, 64% and 76%, respectively. At a HBV DNA cut-off of 100,000 copies/ml at week 24, the sensitivity, specificity, positive and negative predictive values for post-treatment virological response were 86%, 62%, 55% and 89%, respectively. ALT flare during PEG-IFN-α2a treatment had a marginal association with virological response at 24 weeks post-treatment. Eight of 14 (57%) patients who had ALT flare versus six of 26 (23%) patients without ALT flare developed virological response at 24 weeks post-treatment (P=0.043). There was no difference in the frequency of ALT flares among HBeAg-positive (11 Antiviral Therapy 13:4 559

HL-Y Chan et al. of 29; 38%) and HBeAg-negative (3 of 8; 38%) patients (P=0.72). Safety Most of the adverse events were mild to moderate (Table 6). Five patients developed serious adverse events: two patients developed feverish reaction requiring hospitalization after the first dose of PEG- IFN-α2a; one patient developed upper gastrointestinal bleeding due to Mallory Weiss syndrome at week 16; one patient was admitted for upper respiratory tract infection at week 56 (post-treatment week 8) and one patient developed hypothyroidism at week 24. The patients who developed hypothyroidism withdrew from study and the thyroid function test results returned to normal on thyroxine replacement therapy. Dosage adjustment was needed for thrombocytopenia (one patient), neutropenia (one patient) and hyperthyroidism (one patient, who was started on carbimazole treatment). Three additional patients needed dosage interruption of PEG-IFN-α2a due to elevated ALT levels at 27 32 weeks of treatment. No patient died or required liver transplantation. Discussion In this study, we have demonstrated that the 24-week post-treatment virological response to PEG-IFN-α2a for patients who failed previous antiviral treatment was 35%. This response rate was comparable to that reported among treatment-naive patients in previous clinical trials [12 15]. Our patients had active liver disease for a mean of 176 months after cessation of their last course of antiviral treatment and two patients also failed a course of conventional interferon treatment. Although we did not have a control group, we believe that spontaneous disease remission could hardly account for the high virological response rate among our patients. In other words, PEG-IFN-α2a can be considered as a viable treatment option among patients who have failed previous antiviral treatment. At the end of treatment, nine of 11 (82%) HBeAgnegative patients had undetectable HBV DNA by PCR. This response was much higher than that achieved by 2-year lamivudine treatment in a placebocontrolled study (26%) and was comparable to that reported in the pivotal trial of PEG-IFN-α2a in HBeAg-negative patients (63%) [9,15]. We must admit a possible overestimation due to our small sample size, but the lower bound of the 95% CI was still 48%. The excellent virological suppression by PEG-IFN-α2a among HBeAg-negative patients in our study might be partly related to the relatively lower baseline HBV DNA than that of the HBeAg-positive patients. Nonetheless, we believe that augmentation of immune response is a very important mechanism for viral clearance in HBeAg-negative patients, among whom immune tolerance is less of a problem than in HBeAg-positive patients. Even among lamivudinetreated patients, a very high maintained response and low rate of drug resistance could be seen in severe acute exacerbation of chronic hepatitis B, which indicated a vigorous host immune clearance [28]. Unfortunately, in this study, approximately half of the patients could not sustain the virological response after stopping PEG-IFN-α2a treatment. Table 5. Performance of different HBV DNA levels at various time intervals of treatment for predicting virological response at 24 weeks post-treatment HBV DNA, Responder Non-responder Positive Negative Likelihood Week log 10 copies/ml (n=14) (n=26) Sensitivity, % Specificity, % predictive value, % predictive value, % ratio 4 <4 3 3 21 88 50 68 1.75 <5 3 5 21 81 38 66 1.11 8 <3 1 2 7 92 33 65 0.88 <4 3 5 21 81 38 66 1.11 <5 4 6 29 77 40 67 1.26 12 <3 2 4 14 85 33 65 0.93 <4 4 6 29 77 40 67 1.26 <5 6 8 43 69 43 69 1.39 18 <3 5 5 36 81 50 70 1.89 <4 7 8 50 69 47 72 1.61 <5 10 9 71 65 53 81 2.02 24 <3 7 4 50 85 64 76 3.33 <4 9 8 64 69 53 78 2.06 <5 12 10 86 62 55 89 2.26 HBV, hepatitis B virus. 560 2008 International Medical Press

Pegylated interferon for antiviral failure Previous studies suggested that lower HBV DNA and higher ALT levels were associated with higher post-treatment virological response with PEG-IFN-α2a treatment [13,14]. However, this phenomenon could not be confirmed in our previous study with long-term post-treatment follow-up [17]. In this study, lower baseline HBV DNA had a tendency for higher 24-week post-treatment virological response but ALT levels did not predict the response. We believe that the response to pegylated interferon-based PEG-IFN-α2a therapy is largely related to the cccdna levels. Patients with lower baseline cccdna level tend to respond better to a combination of PEG-IFN-α2a and lamivudine treatment [19,29]. The use of quantitative serum hepatitis B surface antigen, which has a good correlation with the cccdna level, as a surrogate predictor of response should be evaluated in future clinical trials [10,20,29]. A roadmap model has been proposed by a group of hepatologists to guide the treatment strategy on the basis of on-treatment HBV DNA response [30]. Residual HBV DNA at 6 12 months has been found to predict viral suppression and drug resistance at 2 3 years [31 33]. In chronic hepatitis C, early viral response at week 12 can predict sustained virological clearance in genotype 1 hepatitis C virus infection [34]. In chronic hepatitis B, no on-treatment predictor for sustained response to PEG-IFN-α2a has been evaluated. In this study, we found that HBV DNA at week 24 could predict post-treatment virological response better than other earlier time intervals. However, the predictive ability was only modest with an AUC of only 0.76. The sensitivity for post-treatment virological response was only 86% with HBV DNA >100,000 copies/ml at week 24. If PEG-IFN-α2a is stopped at week 24 for patients with unsatisfactory response, 14% of patients who had a chance to respond would have the treatment prematurely terminated. As we are uncertain whether treatment-naive patients respond differently to our patients who have failed previous antiviral treatment, we cannot extrapolate our results to all chronic hepatitis B patients using PEG-IFN-α2a at this stage. Our study is limited by the small sample size and lack of a control group. As the efficacy of PEG-IFNα2a in chronic hepatitis B has been well established by several large-scale, randomized, controlled trials, we did not aim to evaluate its efficacy versus placebo, which might not be ethical, or another active antiviral treatment in this study [16]. By contrast, this is a pilot, proof-of-concept study that PEG-IFN-α2a is effective among patients who have failed previous antiviral treatment. Therefore, we have a more stringent definition for antiviral failure than the subgroup analysis of previously conducted studies [21]. Unfortunately, some of these patients received per protocol antiviral Table 6. Adverse events (>5% occurrence) and laboratory abnormalities (grade 3 and 4) during the 72-week study period Adverse event n (%) Weight loss <5% 17 (42.5) >10% 7 (17.5) Upper respiratory tract infection 8 (20) Fever 24 (60) Alopecia 14 (35) Right upper abdominal discomfort 7 (17.5) Malaise 28 (70) Headache 17 (42.5) Myalgia 13 (32.5) Decreased appetite 14 (35) Erytherma at injection sites 24 (60) Skin allergy 4 (10) Vomiting or diarrhoea 6 (15) Laboratory abnormalities Elevated alanine aminotransferase 8 (20) Hypophosphataemia 2 (5) Thrombocytopenia 2 (5) Hypothyroidism 1 (2.5) treatment in previous clinical trials and did not fulfil the recommended criteria for treatment cessation of the recent practice guidelines [6 8]. The high rate of HBeAg reversion after the previous antiviral treatment might be related to the inadequate consolidation treatment after HBeAg seroconversion. Our sample size, albeit small, was adequate to show a reasonable sustained virological response to PEG-IFN-α2a even without a control group. In conclusion, PEG-IFN-α2a has satisfactory efficacy to treat chronic hepatitis B patients who have failed previous antiviral treatment. Before retreating these patients with long-term antiviral therapy, a finite course of PEG-IFN-α2a can be considered as an alternative if there is no contraindication. Acknowledgements We thank Roche (Hong Kong) for the supply of PEG- IFN-α2a for this study. We also thank Kelvin KF Tsoi for his invaluable advice on statistics. Disclosure statement Roche (Hong Kong) provided a free supply of PEG- IFN-α2a but no additional financial assistance. The authors had the final responsibility for the study protocol, case report forms, progress of the study, statistical analysis, reporting of data and manuscript submission. The authors had full access to the data files of the study. HL-Y Chan is an advisory board member of Schering Plough, Novartis Phamaceutical and Antiviral Therapy 13:4 561

HL-Y Chan et al. Bristol Myers Squibb. JJ-Y Sung received paid lecture fees from AstraZeneca Hong Kong Limited, GlaxoSmithKline Pharmaceuticals International and the American Society for Gastrointestinal Endoscopy. The other authors declare no conflicts of interest. References 1. Chen CJ, Yang HI, Su J, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 2006; 295:65 73. 2. Chan HL, Tse CH, Mo F, et al. High viral load and hepatitis B virus subgenotype Ce are associated with increased risk of hepatocellular carcinoma. J Clin Oncol 2008; 26:177 182. 3. Song BC, Suh DJ, Lee HC, Chung YH, Lee YS. Hepatitis B e antigen seroconversion after lamivudine therapy is not durable in patients with chronic hepatitis B in Korea. Hepatology 2000; 32:803 806. 4. van Nunen AB, Hansen BE, Suh DJ, et al. Durability of HBeAg seroconversion following antiviral therapy for chronic hepatitis B: relation to type of therapy and pretreatment serum hepatitis B virus DNA and alanine aminotransferase. Gut 2003; 52:420 424. 5. Lee KM Cho SW, Kim SW, Kim HJ, Hahm KB, Kim JH. Effect of virological response on post-treatment durability of lamivudine-induced HBeAg seroconversion. J Viral Hepat 2002; 9:208 212. 6. Lok AS, McMahon BJ. Chronic hepatitis B. Hepatology 2007; 45:507 539. 7. Liaw YF, Leung N, Guan R, et al. Asian-Pacific consensus statement on the management of chronic hepatitis B: a 2005 update. Liver Int 2005; 25:472 489. 8. Keeffe EB, Dieterich DT, Han SH, et al. A treatment of algorithm for the management of chronic hepatitis B virus infection in the United States: An update. Clin Gastroenterol Hepatol 2006; 4:936 962. 9. Chan HL, Wang H, Niu J, Chim AM, Sung JJ. Two-year lamivudine treatment for hepatitis B e antigen-negative chronic hepatitis B: a double-blind, placebo-controlled trial. Antivir Ther 2007; 12:345 353. 10. Werle-Lapostolle B, Bowden S, Locarnini S, et al. Persistence of cccdna during the natural history of chronic hepatitis B and decline during adefovir dipivoxil therapy. Gastroenterology 2004; 126:1750 1758. 11. Sung JJY, Wong ML, Bowden S, et al. Intrahepatic hepatitis B virus covalently closed circular DNA can be a predictor of sustained response to therapy. Gastroenterology 2005; 128:1890 1897. 12. Chan HL, Leung NW, Hui AY, et al. A randomized, controlled trial of combination therapy for chronic hepatitis B: comparing pegylated interferon-α2b and lamivudine with lamivudine alone. Ann Intern Med 2005; 142:240 250. 13. Janssen HLA, van Zonneveld M, Senturk H, et al. Pegylated interferon alfa-2b alone or in combination with lamivudine for HBeAg-positive chronic hepatitis B: a randomized trial. Lancet 2005; 365:123 129. 14. Lau GK, Piratvisuth T, Luo KX, et al. Peginterferon alfa-2a, lamivudine, and the combination for HBeAg-positive chronic hepatitis B. N Engl J Med 2005; 352:2682 2695. 15. Marcellin P, Lau GK, Bonino F, et al. Peginterferon alfa-2a alone, lamivudine alone, and the two in combination in patients with HBeAg-negative chronic hepatitis B. N Engl J Med 2004; 351:1206 1217. 16. Hui AY, Chan HL, Cheung A, Cooksley G, Sung JJ. Treatment of chronic hepatitis B virus infection by pegylated interferon: systematic review of 4 randomized studies. Aliment Pharmacol Ther 2005; 22:519 528. 17. Chan HL, Hui AY, Wong VW, Chim AM, Wong ML, Sung JJ. Long-term follow-up of peginterferon and lamivudine combination treatment in HBeAg-positive chronic hepatitis B. Hepatology 2005; 41:1357 1364. 18. Marcellin P, Bonino F, Lau GK, et al. Suppression of HBV DNA in patients with HBeAg-negative CHB treated with peginterferon alfa-2a (40KD) ± lamivudine: 2-year follow up results. 57th Annual Meeting of the American Association for the Study of Liver Diseases. 27 31 October 2006, Boston, MA, USA. Abstract 972. 19. Chan HL, Wong VW, Chim AM, et al. Virological response to different combination regimes of peginterferon alfa-2b and lamivudine in hepatitis B e antigen positive chronic hepatitis B. Antivir Ther 2007; 12:815 823. 20. Wursthorn K, Lutgehetmann M, Dandri M, et al. Peginterferon alpha-2b plus adefovir induce strong cccdna decline and HBsAg reduction in patients with chronic hepatitis B. Hepatology 2006; 44:675 684. 21. Flink HJ, Hansen BE, Heathcote EJ, et al. Successful treatment with peginterferon alfa-2b of HBeAg-positive HBV non-responders to standard interferon or lamivudine. Am J Gastroenterol 2006; 101:2523 2529. 22. Chan HL, Chui AK, Lau WY, et al. Factors associated with viral breakthrough in lamivudine monoprophylaxis of hepatitis B virus recurrence after liver transplantation. J Med Virol 2002; 68:182 187. 23. Loeb KR, Jerome KR, Goddard J, Huang ML, Cent A, Corey L. High-throughput quantitative analysis of hepatitis B virus DNA in serum using the TaqMan fluorogenic detection system. Hepatology 2000; 32:626 629. 24. Lindh M, Andersson AS, Gusdal A. Genotypes, nt 1858 variants, and geographical origin of hepatitis B virus largescaled analysis using a new genotyping method. J Infect Dis 1997; 175:1285 1293. 25. Chan HL, Tsang SW, Liew CT, et al. Viral genotype and hepatitis B virus DNA levels are correlated with histologic liver damage in HBeAg-negative chronic hepatitis B virus infection. Am J Gastroenterol 2002; 97:406 412. 26. Hussain M, Fung S, Libbrecht E, et al. Sensitive line probe assay that simultaneously detects mutations conveying resistance to lamivudine and adefovir. J Clin Microbiol 2006; 44:1094 1097. 27. Nair S, Perrillo RP. Serum alanine aminotransferase flares during interferon treatment of chronic hepatitis B: Is sustained clearance of HBV DNA dependent on levels of pretreatment viraemia? Hepatology 2001; 34:1021 1026. 28. Chan HL, Wong VW, Hui AY, et al. Long-term lamivudine treatment is associated with a good maintained response in severe acute exacerbation of chronic HBeAg-negative hepatitis B. Antivir Ther 2006; 11:465 471. 29. Chan HL, Wong VW, Tse AM, et al. Serum hepatitis B surface antigen quantitation can reflect hepatitis B virus in the liver and predict treatment response. Clin Gastroenterol Hepatol 2007; 5:1462 1468. 30. Keeffe EB, Zeuzem S, Koff RS, et al. Report of an international workshop: Roadmap for management of patients receiving oral therapy for chronic hepatitis B. Clin Gastroenterol Hepatol 2007; 5:890 897. 31. Lai CL, Gane E, Liaw TF, et al. Maximal early HBV suppression is predictive of optimal virologic and clinical efficacy in nucleoside treated hepatitis B patients: scientific observations from a large multinational trial (The GLOBE Study). 56th Annual Meeting of American Association for the Study of Liver Diseases. 11 15 November 2005, Boston, MA, USA. Abstract 92. 32. Locarnini S, Qi X, Arterburn S, et al. Incidence and predictors of emergence of adefovir resistant HBV during four years of adefovir dipivoxil (ADV) therapy for patients with chronic hepatitis B (CHB). 40th Annual Meeting of European Association for the Study of the Liver. 13 17 April 2005, Paris, France. Abstract 36. 33. Chan HL, Wong VW, Tse CH, et al. Early virological suppression is associated with good maintained response to adefovir dipivoxil in lamivudine resistant chronic hepatitis B. Aliment Pharmacol Ther 2007; 25:891 898. 34. Davis GL. Monitoring of viral levels during therapy of hepatitis C. Hepatology 2002; 36 Suppl 1:S145 S151. Accepted for publication 24 February 2008 562 2008 International Medical Press