Potential conflict of interest: Nothing to report.
The study was financed and run by F. Hoffmann-La Roche, Basel, Switzerland
As there is currently a lack of consensus on the most appropriate dose and duration of peginterferon alfa-2a (PEG-IFNα-2a) therapy in hepatitis B e antigen (HBeAg)-positive patients, the efficacy and safety of either 24 or 48 weeks' duration and 90 μg/week or 180 μg/week doses were compared. HBeAg-positive patients (n = 544; 34% genotype B, 51% genotype C) were randomized to receive PEG-IFNα-2a (2 × 2 factorial design) for 24 or 48 weeks and at 90 μg/week or 180 μg/week and included in the per-protocol population. The primary efficacy endpoint of the noninferiority study was HBeAg seroconversion 6 months posttreatment. The prespecified odds ratio (OR) noninferiority margin was 1.88 with a one-sided significance level of 0.025. The highest rates of HBeAg seroconversion 6 months posttreatment were in the 180/48 arm (36.2% versus 14.1%-25.8% in the other arms). When the dose and duration arms were pooled, the OR for noninferiority of 24 weeks versus 48 weeks was 2.17 (95% confidence interval [CI] 1.43, 3.31; P = 0.749) and for 90 μg versus 180 μg was 1.79 (95% CI 1.18, 2.72; P = 0.410). As the upper limit of the 95% CI of the ORs were >1.88, 24 weeks were inferior to 48 weeks and 90 μg/week was inferior to 180 μg/week. The highest rates of response in the 180/48 arm were achieved by patients with HBsAg <1,500 IU/mL at Week 12 (58%) or Week 24 (57%), whereas patients with HBsAg >20,000 IU/mL did not respond. Adverse events were typical of those associated with PEG-IFNα-2a. Conclusion: Compared with lower doses and shorter durations, the licensed PEG-IFNα-2a treatment regimen (180μg/48 weeks) was the most efficacious and beneficial for HBeAg-positive patients predominantly infected with hepatitis B virus genotypes B or C. (HEPATOLOGY 2011;)
Peginterferon alfa-2a (40 kD; PEG-IFNα-2a) has proven efficacy in hepatitis B e antigen (HBeAg)-positive chronic hepatitis B (CHB)1 and is one of the first-line drugs recommended for CHB by all international treatment guidelines.2-4 Although the licensed dose and duration of PEG-IFNα-2a is 180 μg/week for 48 weeks, there is currently a lack of consensus on the most appropriate dose and duration, which is reflected in varying recommendations in international guidelines. The European Association for the Study of the Liver (EASL) and American Association for the Study of Liver Diseases (AASLD) guidelines recommend 48 weeks of treatment, but EASL guidelines do not provide dose information for any of the antiviral agents recommended2; APASL guidelines recommend 90-180 μg/week for at least 6 months and that patients infected with genotype C are treated longer3 and the AASLD guidelines suggest that 90 μg/week for at least 6 months may be appropriate for HBeAg-positive patients.4
It is likely that this lack of consensus results from varying interpretations of the results from the phase 2 and the phase 3 studies of PEG-IFNα-2a in HBeAg-positive patients. In the phase 2 study (n = 194), patients received PEG-IFNα-2a 90, 180, or 270 μg/week for 24 weeks. A similar rate of HBeAg seroconversion around 35% was achieved by patients receiving the 90-μg/week and 180-μg/week doses.5 In the phase 3 study, patients received a finite 48-week course of PEG-IFNα-2a (180 μg/week) and 32% of patients achieved HBeAg seroconversion 6 months posttreatment.1 This could account for the recommendation that 90 μg/week for 6 months may be sufficient in HBeAg-positive patients.3, 4 As the phase 2 study was not powered to determine differences between doses and neither the phase 2 nor the phase 3 study was designed to identify the most appropriate duration of PEG-IFNα-2a, and cross-study comparisons are not appropriate, a single study comparing different treatment regimens was required.
The aim of the NEPTUNE study was to compare the efficacy and safety of PEG-IFNα-2a given for either 24 or 48 weeks and at doses of either 90 μg/week or 180 μg/week.
ALT, alanine aminotransferase; CHB, chronic hepatitis B; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; OR, odds ratio; PEG-IFN, peginterferon.
Patients and Methods
This was a phase 4, double-blind, randomized, multicenter trial (NCT00435825) with a 2 × 2 factorial design. HBeAg-positive adults (>18 years) who had been HBsAg-positive for more than 6 months with hepatitis B virus (HBV) DNA >100,000 IU/mL as measured by polymerase chain reaction, alanine aminotransferase (ALT) levels 1-10 × upper limit of normal (ULN), and with biopsy-confirmed liver disease consistent with CHB were eligible for the study.
Patients with any of the following were not eligible: antiviral therapy for CHB within 6 months of study initiation; coinfection with hepatitis C, hepatitis delta or human immunodeficiency virus, history or any other evidence of a medical condition associated with chronic liver disease or evidence of decompensated liver disease; pregnant or breast-feeding women; neutrophil count <1,500 cells/mm3, platelet count <90,000 cells/mm3, serum creatinine >1.5 × ULN; evidence of alcohol and/or drug abuse within 1 year of entry; history of severe psychiatric disease, immunologically mediated disease, chronic pulmonary disease, severe cardiac disease, severe seizure disorder or current anticonvulsive use, systemic antineoplastic or immunomodulatory therapy, major organ transplantation, poorly controlled thyroid disease, severe retinopathy, or evidence of an active or suspected cancer or history of malignancy.
Patients were randomized (1:1:1:1) to receive PEG-IFNα-2a (90 μg/week or 180 μg/week for 24 or 48 weeks) by subcutaneous injection once weekly. Randomization was centralized using random permuted block design within each ALT and genotype combination. A patient was enrolled in the study and assigned a patient number from the IVRS Company (ICON) after all protocol eligibility requirements were met. The investigator or designee entered the patient's randomization number in the electronic case report form. It was the investigator's responsibility to ensure that each patient received the assigned treatment. The dose of PEG-IFNα-2a was administered in a double-blind fashion throughout the study and the duration of treatment was double-blind until Week 24. In the case of intolerance, the dose could be reduced in increments to 75%, 50%, and 25% of the full dose. Patients were stratified by ALT level at screening (1-2 × ULN; 2-5 × ULN; 5-10 × ULN) and viral genotype (A, B, D, or C plus other genotypes). Because few patients would be infected with genotypes other than A, B, C, and D, they were pooled and added to the predicted largest genotype group (genotype C).
The primary efficacy parameter was HBeAg seroconversion (loss of HBeAg and presence of anti-HBe) determined 6 months posttreatment. Secondary efficacy parameters included HBeAg loss, HBsAg clearance, HBsAg seroconversion, ALT ≤ULN, HBV DNA <20,000 IU/mL (HBV TaqMan, Roche, lower limit of quantification 29 IU/mL), HBV DNA <2,000 IU/mL, and two triple endpoints: (1) HBeAg seroconversion, ALT normalization, and HBV DNA <20,000 IU/mL, (2) HBeAg seroconversion, ALT normalization, and HBV DNA <2,000 IU/mL 6 months posttreatment. A revision to the protocol introduced efficacy assessments (primary and secondary) at Week 72 of the trial as a secondary endpoint.
Factors Associated with Response.
To investigate the association of response 6 months posttreatment with on-treatment HBsAg levels, which has been identified as being predictive of response in the phase 3 study of PEG-IFNα-2a,6 HBeAg-positive patients receiving PEG-IFNα-2a (180 μg for 48 weeks) with HBsAg data available at baseline and Weeks 12, 24, 48, and 72 were evaluated. Serum HBsAg levels were quantified retrospectively using Roche Elecsys HBsAg II quant assay (detection limit 0.05 IU/mL, Roche Diagnostics, Mannheim, Germany) according to manufacturer's instructions. Rates of HBeAg seroconversion 6 months posttreatment were assessed according to HBsAg cutoff levels at Weeks 12 and 24 identified in the analysis of the phase 3 study (<1,500, 1,500-20,000, and >20,000 IU/mL). The potential association of baseline ALT levels and on-treatment HBsAg levels was also investigated.
Adverse events (AEs) occurring during treatment or up to 8 weeks posttreatment were recorded and graded as mild, moderate, or severe by the investigator. Fatal, life-threatening, and events requiring hospitalization, or persistent or significant disability or incapacity, were considered serious AEs (SAEs) and were collected throughout the treatment and follow-up periods. Laboratory tests were recorded at baseline, weeks 2, 4, 8, and 12 and then at 6-week intervals throughout the treatment period.
Three null hypotheses were tested: (1) no interaction between duration and dose; (2) inferiority of 24 weeks of treatment compared with 48 weeks of treatment; (3) inferiority of 90 μg/week compared with 180 μg/week. If there was no significant interaction between duration and dose, assessing noninferiority of dose and duration by pooling can occur: to test for noninferiority of 24 weeks versus 48 weeks of treatment, the dosage arms (90 μg and 180 μg) were pooled; to test the noninferiority of 90 μg/week versus 180 μg/week, the duration arms (24 weeks and 48 weeks) were pooled.
In the phase 3 study, the HBeAg seroconversion rate 6 months posttreatment was 32%, whereas lamivudine response was 20%.1 The latter was adapted as the limit of noninferiority; thus, the margin was set at 12%, the difference between response rates (32% − 20%).
The noninferiority margin of 12% is equivalent to a margin of 1.88 (0.32/1-0.32 ÷ 0.2/1-0.2) for the odds ratio (OR) of 48-week/24-week response or 180 μg/week/90 μg/week. The primary analysis used the OR to test for noninferiority. If the upper limit of the two-sided 95% confidence interval (CI) of the OR is <1.88, the null hypothesis that 24 weeks is inferior to 48 weeks' treatment or 90 μg/week is inferior to 180 μg/week would be rejected and the one-sided P-value of noninferiority would be significant at less than 0.025, implying that 24 weeks is noninferior to 48 weeks or 90 μg/week is noninferior to 180 μg/week. Allowing for a 10% rate of withdrawal and to achieve a power of 80%, 238 patients were needed in each pooled group (four arms of 131 patients). The total number of patients needed to be randomized was 524.
As this was a noninferiority study, the per-protocol population, which included patients who satisfied all key study inclusion and exclusion criteria and received at least four doses of PEG-IFNα-2a, was the most appropriate for analysis.7
Logistic regression analyses were performed to identify any parameters with a significant effect on HBeAg seroconversion. Prognostic factors were assessed alone or in combination. Stratification factors included in the models were: screening ALT strata, screening genotype strata, and screening ALT strata-by-screening genotype strata. Prognostic factors and interaction terms included in the multivariate models were: sex, age, weight, body mass index, region, baseline log10 HBsAg, baseline log10 HBV DNA, baseline log10 ALT, dose, duration, dose-by-duration, dose-by-screening ALT strata, duration-by-screening ALT strata, dose-by-genotype strata, duration-by-genotype strata, dose-by-screening ALT strata-by-screening genotype strata, duration-by-screening ALT strata-by-screening genotype strata, dose-by-duration-by-screening ALT strata-by-screening genotype strata, sex-by-dose, sex-by-duration, age-by-dose, and age-by-duration.
The study was conducted in accordance to local regulations and the Declaration of Helsinki and all patients provided written informed consent. The study protocol was approved by an Independent Ethics Committee.
A total of 792 patients were screened between January 2007 and December 2008. Patients were randomized to PEG-IFNα-2a (90 μg/24 weeks, n = 141; 180 μg/24 weeks, n = 136; 90 μg/48 weeks, n = 138; 180 μg/48 weeks, n = 136). Of the randomized patients, one patient in the 90 μg/24-week arm and two patients in the 90 μg/48-week arm did not receive medication. Patient demographics and baseline characteristics were similar in all arms (Table 1).
Table 1. Demographic Data and Baseline Characteristics (Intent-to-Treat Population)
90 μg/24 Weeks (N=140)
180 μg/24 Weeks (N=136)
90 μg/48 Weeks (N=136)
180 μg/48 Weeks (N=136)
The genotype C group includes patients with genotype C (n=69, 73, 70, and 70, respectively), E (n=3, 3, 0, and 2, respectively), F (n=1, 0, 2, and 1, respectively), mixed genotypes (n=1, 0, 2, and 1, respectively) or unknown (n=0, 0, 1, and 0, respectively).
For the per-protocol population, patients excluded were those receiving fewer than 4 weeks of medication (n = 5), a lack of serology data (n = 3), no baseline data (n = 1), or not meeting the inclusion criteria (n = 2). Additionally, patients were assigned to treatment duration actually received; thus, 10 patients randomized to 48 weeks' treatment duration but who received less than 28 weeks of PEG-IFNα-2a were included in the per-protocol 24-week duration arms. The per-protocol population consisted of 142, 140, 132, and 130 patients, respectively.
Overall, three, four, six, and nine patients prematurely discontinued PEG-IFNα-2a therapy in the 90 μg/24-week, 180 μg/24-week, 90 μg/48-week, and 180 μg/48-week arms, respectively; five, one, four, and eight patients, respectively, did not complete the 6-month follow-up period. The majority of patients (86%-93%) received the planned treatment duration and the mean doses received were 94%-98% of the total planned dose. Only 2%-7% of patients discontinued treatment.
The majority of patients were infected with genotype B (34%; mean ± standard deviation baseline HBV DNA 7.8 ± 0.87 log10 IU/mL; baseline ALT: 177.7 ± 153.9 U/L) or genotype C (51.0%; baseline HBV DNA 7.6 ± 1.09 log10 IU/mL; baseline ALT: 158.7 ± 126.9 U/L), with only a small number of patients being infected with genotype A (n = 23) or genotype D (n = 39).
The rate of HBeAg seroconversion at 6 months posttreatment was highest in the 180 μg/48-week arm (36.2%; Fig. 1A). HBeAg seroconversion rate at Week 72 remained highest in the 180 μg/48-week arm, even though the response rates in the 90 μg/24-week and 180 μg/24-week arms (equivalent to HBeAg seroconversion rate 1 year posttreatment) increased to 18.3% and 27.1%, respectively (Fig. 1A). There was no interaction between dose and duration (P = 0.9), therefore allowing the main effects of dose and duration to be tested. The upper limit of the two-sided 95% CIs were >1.88, the P-values for noninferiority were nonsignificant for dose and duration, and, therefore, the null hypotheses were retained. This demonstrated that the 24 weeks of therapy were inferior to 48 weeks of therapy and 90 μg/week was inferior to 180 μg/week (Fig. 1B).
In patients achieving HBeAg seroconversion 6 months posttreatment, the median time to HBeAg seroconversion that was sustained up to 6 months posttreatment was 35.6 weeks (95% CI 12.1, 48.1) in the 180 μg/24-week arm compared with 48.1 weeks (95% CI 24.1, 48.3) in the 90 μg/24-week arm. Time to sustained HBeAg seroconversion was 36.1 weeks (95% CI 24.1, 48.4) in the 180 μg/48-week arm compared with 59.7 weeks (95% CI 36.1, 72.1) in the 90 μg/48-week arm.
Highest rates of all secondary endpoint parameters were in the 180 μg/48-week arm (Table 2). This is consistent with the observation in the phase 2 study that HBV DNA decline was more profound in the 180 μg/week group than in the 90 μg/week group.5 Three patients in each of the 48-week arms achieved HBsAg clearance 6 months posttreatment compared with one patient in the 90 μg/24-week arm and none of the patients in the 180 μg/24-week arm. There was also no interaction between dose and duration for any of the parameters. In the majority of cases where hypothesis testing was conducted, 24 weeks of therapy were inferior to 48 weeks of therapy and 90 μg/week was inferior to 180 μg/week. For ALT normalization, 90 μg/week was borderline noninferior to 180 μg/week (upper two-sided 95% CI = 1.78).
Table 2. Response Rates 6 Months Posttreatment in the Four Treatment Groups (Per-Protocol Population)
Response n (%)
90 μg/24 Weeks (N=142)
180 μg/24 Weeks (N=140)
90 μg/48 Weeks (N=132)
180 μg/48 Weeks (N=130)
HBV DNA <2000 IU/mL
HBV DNA <20,000 IU/mL
HBeAg seroconversion, HBV DNA <20,000 IU/mL, ALT normalization
HBeAg seroconversion, HBV DNA <2000 IU/mL, ALT normalization
Response rates were analyzed by genotype. Response rates in the 90 μg/24-week, 180 μg/24-week, 90 μg/48-week, and 180 μg/48-week arms were 0/7, 1/8, 1/3, 1/4, respectively, in genotype A-infected patients and 1/11, 2/6, 2/11, and 0/10, respectively, in genotype D-infected patients. Highest rates of HBeAg seroconversion 6 months posttreatment in patients infected with genotypes B and C were in the 180 μg/48-week arm (42.2% [19/45] and 38.8% [26/67], respectively). Response rates in the other treatment arms are shown in Fig. 1C.
Testing the main effects of dose and duration in genotype B gave ORs of 1.89 (95% CI: 0.98, 3.67; P = 0.505) and 1.44 (95% CI: 0.75, 2.78; P = 0.215). For genotype C, ORs were 2.00 (95% CI: 1.09, 3.69; P = 0.581) and 3.29 (95% CI: 1.76, 6.15; P = 0.96), respectively. As the upper limit of the two-sided 95% CIs were >1.88, the P-values for noninferiority were nonsignificant and, therefore, the null hypotheses were retained. Consequently, 24 weeks of therapy were inferior to 48 weeks of therapy and 90 μg/week was inferior to 180 μg/week in genotypes B and C.
Patients were stratified according to screening ALT levels; in all arms HBeAg seroconversion rate 6 months posttreatment was highest in patients receiving 180 μg/week for 48 weeks: 19% where screening ALT was 1-2 × ULN, 45% for 2-5 × ULN, and 61% for 5-10 × ULN (Fig. 2A).
In the logistic regression analyses, dose, duration, log10 baseline HBsAg, and log10 baseline HBV DNA were the only significant independent predictors of HBeAg seroconversion 6 months posttreatment (P < 0.05 for univariate models that included the stratification and predictor factors; and P < 0.05 in a multivariate model that included all significant predictors and stratification factors).
Factors Associated with Response.
Of the 114 patients (84% of the per-protocol population) in the 180 μg/48-week arm with available HBsAg data, 37.6% achieved HBeAg seroconversion 6 months posttreatment. On-treatment HBsAg decline from baseline was more pronounced in these patients (−1.34 log10 IU/mL at Week 48) than in nonresponders (−0.61 log10 IU/mL at Week 48). HBsAg <1,500 IU/mL was reached by 27% and 40% of patients at Weeks 12 and 24, respectively, and was associated with the highest rates of sustained response (positive predictive values [PPV] of 58% and 57%, respectively; Table 3). PPVs for HBsAg 1,500-20,000 IU/mL at Weeks 12 and 24 were 42% and 35%, respectively. Patients with HBsAg > 20,000 IU/mL at Weeks 12 or 24 did not achieve sustained response (negative predictive value [NPV] 100%). The predictive value of these HBsAg levels was also apparent in the other treatment arms (data not shown). The PPV of HBsAg levels <20,000 IU/mL was further enhanced when combined with baseline ALT levels (Fig. 2B). Overall, 50% (57/114) of patients had ALT >2 × ULN at baseline; of these, 33% (19/57) had HBsAg <1,500 IU/mL at Week 12. This group of patients achieved the highest rates of HBeAg seroconversion 6 months posttreatment (PPV 68%).
Table 3. On-Treatment HBsAg Levels Confirm Prediction of Response with PEG-IFNα-2a 180 μg for 48 Weeks in Phase 3 Study in HBeAg-Positive Patients
The safety population included patients who received at least one dose of study medication and who had at least one postbaseline safety assessment. The type of AEs were similar between arms and consistent with previous experience with PEG-IFNα-2a (Table 4). Laboratory abnormalities were consistent with previous PEG-IFNα-2a experience and were clinically manageable with dose modification. The incidence of AEs, SAEs, AEs leading to dose modification, or treatment withdrawal were low in all treatment arms.
This study shows that the licensed regimen of PEG-IFNα-2a (180 μg/week for 48 weeks) resulted in the highest HBeAg seroconversion rates 6 months posttreatment (36.2%), which is consistent with a rate of 32% seen in the phase 3 study1 and higher than the rates seen in the lower dose and shorter duration arms of the current study. The primary efficacy results demonstrated that 24 weeks were inferior to 48 weeks and 90 μg/week was inferior to 180 μg/week. Although rates of response increased between the 6-month posttreatment and Week 72 timepoint in the 24-week groups (equivalent to 1 year posttreatment), it did not reach the levels achieved in the licensed regimen group 6 months posttreatment.
When the time to sustained HBeAg seroconversion in patients achieving HBeAg seroconversion 6 months posttreatment was analyzed, patients receiving 180 μg/week PEG-IFNα-2a achieved a response earlier than patients receiving 90 μg/week. Data for these analyses were not pooled by duration group because of timing differences in the scheduled visits where HBeAg seroconversion was determined.
The trends from all the secondary endpoint analyses were consistent with the primary efficacy endpoint, the null hypotheses of inferiority being retained, thus indicating the robustness of the dataset.
The majority of patients enrolled in this study were infected with either genotype B or genotype C, which is consistent with most patients being recruited from the Asian Pacific region. Several studies have investigated the correlation between response to interferon-based therapy and infecting genotype and have shown that response is higher in genotype B than in genotype C and that genotypes C and D are more difficult to treat than genotypes A and B.5, 8-11 However, consistent with the phase 3 PEG-IFNα-2a study in HBeAg-positive patients,1 the HBeAg seroconversion rate 6 months posttreatment in the 180 μg/48-week arm in this study was similar in genotypes B and C. In contrast, in the phase 2 6-month PEG-IFNα-2a study the response in genotype B-infected patients was better than genotype C-infected patients.5 These findings suggest that this dose and duration of PEG-IFNα-2a overcomes the difficulty in managing harder-to-treat genotype C patients and that 180 μg and 48 weeks provide the best response in genotype B. When the secondary endpoints were assessed, some differences between genotypes were observed. For example, for the endpoint HBV DNA <2,000 IU/mL, the response rate for the 180 μg/48-week arm in genotype B was higher than in genotype C (47% and 25%, respectively) and similar differences between ALT normalization were observed (64% and 49%, respectively), thereby supporting data seen in other studies.8-11 The study was not powered for subgroup analysis and this should be taken into account when interpreting the data in the subgroups because discrepancies can occur as a result of chance. In addition, baseline characteristics are not necessarily balanced between the treatment arms, which may affect treatment responses. Whereas for the main study population, stratification for ALT strata and genotype ensured balance between treatment arms, this is not necessarily the case in the subgroup analyses. It is therefore important to consider the overall trends in all parameters when interpreting subgroup analyses. With regard to the genotype B and genotype C subanalyses 6 months posttreatment and at Week 72 for both primary and secondary efficacy parameters, the overall trends were consistent with the main results of the study; namely, that the 180 μg/48-week regimen was the most effective in these patients.
Log10 baseline HBV DNA and HBsAg levels were shown to be significant independent predictors for HBeAg seroconversion in this analysis. The importance of baseline HBV DNA levels has been reported previously in both HBeAg-positive12 and HBeAg-negative patients.13 However, baseline ALT levels or ALT strata were not shown to be prognostic factors for response in the current study even though the highest rates of response were achieved in patients with the highest baseline ALT levels. Interestingly, the expectation was that in patients with high baseline ALT levels (>5 × ULN), lower dose and shorter duration of therapy may have been as effective as the licensed regimen; however, in all ALT strata the highest rates of response were achieved in the 180 μg/48-week arm.
In recent years there has been considerable interest in the potential of on-treatment HBsAg levels acting as predictors of posttreatment response to PEG-IFNα-2a.14-16 Previous studies have investigated whether on-treatment HBsAg levels can be used to predict response to peginterferon in HBeAg-positive patients. Sonneveld et al.14 demonstrated that when response was defined as HBeAg loss combined with HBV DNA <10,000 copies/mL, HBeAg-positive patients with no HBsAg decline at Week 12 of therapy had a 97% chance of nonresponse (NPV 97%). However, this observation could not be repeated in the phase 3 study of PEG-IFNα-2a using either the combined endpoint (NPV 82%) or HBeAg seroconversion alone (NPV 69%).15 It has been suggested that differences in study population could explain why the stopping rule proposed by Sonneveld et al.14 could not be validated in the phase 3 analysis.6 The current study confirms an association between on-treatment HBsAg levels and HBeAg seroconversion 6 months posttreatment, validating cutoff levels identified initially in the phase 3 study of PEG-IFNα-2a in HBeAg-positive patients.16 After 6 months of PEG-IFNα-2a treatment, an increasing proportion of patients reached levels of HBsAg <1,500 IU/mL at Week 24 (40%) compared with Week 12 (27%), thus increasing the proportion of patients with an improved chance of responding. The positive predictive value of HBsAg <1,500 IU/mL at Week 12 or Week 24 were further enhanced when combined with baseline ALT levels. Patients with HBsAg levels >20,000 IU/mL by Weeks 12 and 24 are unlikely to achieve sustained response with 48 weeks of PEG-IFNα-2a, so it would be important to explore whether these hard-to-treat patients require extended treatment duration more than 48 weeks or other management strategy in order to achieve a sustained response or if they are nonresponders. This finding may be a step towards a response-guided therapy in the management of CHB.
PEG-IFNα-2a was well tolerated, with high rates of adherence to therapy. The AEs and laboratory abnormalities reported were consistent with previous experience in HBeAg-positive patients1 and were clinically manageable with dose modification. The incidence of AEs and SAEs was low in all treatment arms.
In conclusion, compared with a lower dose or shorter duration, PEG-IFNα-2a 180 μg/week for 48 weeks was the most efficacious regimen, as it was associated with the highest posttreatment response rates. AEs during treatment were manageable with dose modification and did not result in high rates of discontinuation. As neither higher dose nor longer duration affected the safety profile or adherence to therapy, this study confirms that the licensed PEG-IFNα-2a dose and duration are the most efficacious and beneficial for HBeAg-positive patients.
The authors thank Dr. Philip McCloud, Dr. Sandra Thompson, Dr. Nigel Pluck, and Dr. Martyn Corbey for statistical expertise, clinical expertise, and data analysis. We also thank Dr. Liesje Thomas from Elements Communications for providing medical writing assistance supported by F. Hoffmann-La Roche, Basel, Switzerland.