Effect of switching from nucleos(t)ide maintenance therapy to PegIFN alfa-2a in patients with HBeAg-positive chronic hepatitis B: A randomized trial.

AIMS: Induction of a durable viral response is difficult to achieve in patients with chronic hepatitis B (CHB), even from long-term use of a nucleos(t)ide analogue (NA). This study investigated whether switching to peginterferon (PegIFN) alfa-2a after long-term NA therapy induced a durable viral response.
METHODS: Patients with hepatitis B e antigen (HBeAg)-positive CHB who received any NA for at least 72 weeks and had a low level of HBV DNA (≤100 IU/mL) were randomized (1:1) to receive PegIFN alfa-2a (180 μg/week) or NA for 48 weeks. The primary endpoint was change in the hepatitis B surface antigen (HBsAg) titer during antiviral therapy.
RESULTS: We randomized 149 CHB patients to the two groups. Compared to baseline, the HBsAg levels in both groups were not lower at week 12, but were lower after 24, 36, and 48 weeks (all p<0.001). The maximal HBsAg decline in the PegIFN alfa-2a group was at week 36 (0.50±0.88 log10 IU/mL), and this decline was smaller in the NA group (0.08±0.46 log10 IU/mL). The percentage of patients with HBeAg seroconversion at week 48 was also greater in the PegIFN alfa-2a group (15/75 [20.0%] vs. 5/74 [6.8%], p = 0.018). Multivariable analysis indicated the PegIFN alfa-2a group had a greater change in HBeAg seroconversion at week 48 (p = 0.027). Patients had relatively good tolerance to PegIFN alfa-2a therapy.
CONCLUSIONS: CHB patients who switched to PegIFN alfa-2a for 48 weeks had a significantly lower HBsAg titer and increased HBeAg seroconversion relative to those who remained on NA therapy. TRIAL REGISTRATION: (ClinicalTrials.gov; NCT01769833).

Introduction

The use of potent oral nucleos(t)ide analogues (NAs) to treat patients with chronic hepatitis B (CHB) can reduce the rate of hepatitis B virus (HBV) -associated disease progression, including the development of hepatocellular carcinoma and hepatic decompensation, and dramatically improve patient survival [1-4]. However, even after long-term NA treatment of these patients, many of them do not achieve a durable viral response (i.e., a sustained response following NA cessation), such as seroclearance of HB surface antigen (HBsAg), seroconversion of HB e-antigen (HBeAg) with undetectable HBV DNA. Instead, long-term NA treatment can increase the risk of viral resistance. Functional cure, defined as undetectable HBV DNA and sustained loss of detectable HBsAg [5], is associated with favorable clinical outcome, and these patients can discontinue NA treatment without experiencing virological relapse [6-8].

Other research reported that treatment with peginterferon (PegIFN) alfa-2a for 1 year led to a higher rate of HBsAg loss (3–7%), and this increased to 12% at 5 years after discontinuation [9-11]. HBeAg seroconversion is also sustained after discontinuation of PegIFN alfa-2a treatment, reaching 36% at 6 months and 83% at 1 year [12-14]. However, PegIFN alfa-2a is not effective in all patients, and is associated with higher rates of adverse events [15]. Thus, most CHB patients are currently treated with oral NAs rather than PegIFN alfa-2a.

Although efforts are underway to develop novel therapies that can achieve a functional cure of CHB [8, 16], a durable viral response remains difficult to achieve and sustain. Several trials have tested the ability of PegIFN alfa-2a to provide HBeAg seroconversion in HBeAg-positive patients following long term NA treatment. Two studies showed that the concurrent use of lamivudine plus PegIFN alfa-2a did not provide greater efficacy than PegIFN alfa-2a or oral NA alone [12, 17]. In contrast, another study found that the rate of HBeAg seroconversion was higher in patients who switched from entecavir to PegIFN alfa-2a than in those who continued NA monotherapy [18]; however, patients in this previous study were highly selected from a population of patients who were most likely respond to PegIFN alfa-2a therapy.

The objective of the present study was to compare the effects of a 48-week treatment with PegIFN alfa-2a or NA on HBsAg change and HBeAg seroconversion in NA-controlled HBeAg-positive CHB patients.

Methods

Study population

The study protocol was approved by the independent ethics committee of each participating institution, and all included patients provided written informed consent. This study was registered at ClinicalTrials.gov as NCT01769833. This phase III, randomized, open-label trial of CHB patients compared the effect of switching to PegIFN alfa-2a after long-term NA therapy with the effect of continuing NA therapy on the level of HBsAg and HBeAg seroconversion. Patients in the two groups were recruited from four Liver Centers in South Korea between August 2013 and December 2020. Patients were eligible if they were HBeAg-positive before NA treatment, received any NA-regimen except telbivudine for at least 18 months, had an HBV DNA level of 100 IU/mL or less for at least 12 months (the detection limit at study onset in 2012), and had an HBsAg titer of 100 IU/mL or more at enrollment. The other inclusion criteria were male or female over the age of 20 years; alanine aminotransferase (ALT) level less than 10-times the upper limit of normal (ULN); negative urine or serum pregnancy test result within 24 hours before the first administration of the test drug (for women of childbearing age); ability to participate in the treatment and follow-up observations and to follow the research protocol; and receipt of written informed consent from the participant. The exclusion criteria were presence of decompensated cirrhosis (Child score B–C); history of spontaneous bacterial peritonitis, bleeding due to varicose veins, hepatic encephalopathy, or other signs of liver dysfunction; clinical or radiographic evidence suggesting hepatocellular carcinoma; concomitant infection with the hepatitis C virus (HCV) or HIV; other causes of liver disease; pregnancy or lactating; use of an immunomodulatory or immunosuppressant agent within 6 months prior to registration; serious additional disease that could affect the test results (e.g., congestive heart failure, kidney failure, chronic pancreatitis, uncontrolled diabetes, alcoholism, malignant tumors, etc.); receipt of or planning for a liver transplant; history of hypersensitivity to interferon; resistance to the NAs currently being administered; or previous use of telbivudine.

At the beginning of the study, there was a one-to-one random assignment of patients to receive 180 μg/week PegIFN alfa-2a or continue daily oral NA therapy for 48 weeks. A stratified probability scheme was applied for drug allocation and patients were allocated according to a randomization table separately prepared for each participating hospital. To reduce the risk of ALT flares in patients switching to PegIFN alfa-2a, patients randomized to this group continued taking oral NAs for the first 12 weeks of PegIFN alfa-2a treatment. Patients in the PegIFN alfa-2a group who experienced HBV DNA elevation (viral breakthrough), with or without ALT flare, were allowed to restart concomitant oral NA at the discretion of the researchers. In case of viral breakthrough in the NA group, rescue management, such as adding an NA or changing the NA, was performed according to contemporary HBV guidelines based on the mutation profile and patient compliance, and at the discretion of the researchers. Viral breakthrough in the PegIFN alfa-2a group was defined as an increase in HBV DNA to 2000 IU/mL or more during treatment, and viral breakthrough in the NA group was defined as an increase in HBV DNA (IU/mL) by 10-fold or more from its nadir on more than one occasion during NA treatment. PegIFN alfa-2a was not discontinued in case of viral breakthrough. The criteria for dropout were severe adverse effects or withdrawal of consent to participate.

This study was performed in accordance with Good Clinical Practices and the ethical principles of the Declaration of Helsinki.

Efficacy

The primary endpoint was the change in HBsAg quantity (log10 HBsAg) during drug administration in each group. The secondary endpoints were (i) changes in serum HBV DNA level from baseline and the ratio of below 20 IU/mL (the detection limit), below 2000 IU/mL, and below 20,000 IU/mL during drug administration in each group and (ii) change of HBeAg seroconversion and loss during drug administration in each group.

After randomization, patients in each group were evaluated at baseline and every 12 weeks until the end of treatment (week 48). Each visit included measurements of vital signs, a physical examination, routine laboratory measurements (complete blood count, liver biochemistry, hepatitis B serology, serum HBV DNA, and quantitative HBsAg titer), and determination of adverse events. An interim analysis was conducted 24 weeks after trial initiation. Serum concentrations of HBsAg, anti-HBs antibody, HBeAg, and anti-HBe antibody were measured every 3 months using the ARCHITECT i2000SR immunoassay analyzer (Abbott, Illinois, U.S.A.). Routine biochemical tests (serum concentrations of ALT, aspartate aminotransferase [AST], albumin, total bilirubin, and creatinine) were performed using a sequential multiple autoanalyzer. HBV DNA was measured using a real-time PCR assay on a Cobas TaqMan 48 Analyzer (Roche Molecular Diagnostics, Branchburg, NJ, USA), and had a detection limit of 20 IU/mL. HBsAg concentration was quantified using the Elecsys® HBsAg II Quant Assay (Roche Diagnostics, Indianapolis, IN, USA). Adverse events were classified according to the Medical Dictionary for Drug Regulatory Affairs using World Health Organization terminology. A severe adverse event was defined as an event that led to study drug discontinuation.

Statistical analysis

Previous results suggested that the mean ± standard deviation (SD) between-group difference in the change of HBsAg titer would be 0.6 (±1.1) log10 IU/mL [19]. Thus, 72 patients per treatment group would have a statistical power of at least 90% for detecting a between-group difference of 0.6 (±1.1) log10 IU/mL at a significance level of 0.05. Assuming a dropout rate of 10%, the sample size was 80 patients per group. The primary end points of the two groups were compared using an independent samples t-test.

Data are presented as numbers with percentages (categorical variables) or means ± SDs (continuous variables). Differences in patient characteristics were compared using the chi-square test or Fisher’s exact test (categorical variables) or the independent samples t-test or Mann-Whitney’s U test (continuous variables). The Shapiro-Wilk test was used to determine the normality of distributions.

Considering the unbalanced nature of the repeated measurement data, a linear mixed model (LMM) with random intercepts was used for statistical fitting. The LMM model included repeated measures of continuous variables as dependent variables; group, time, and group×time interaction as fixed effects; baseline outcome as a continuous covariate; and study participant as a random effect. To avoid making assumptions about the covariance structure, an unstructured covariance matrix that was allowed to differ between groups for the LMM analysis was used, and the Bonferroni procedure was applied in post-hoc analyses. A generalized estimating equation (GEE) model was used to examine changes in seroconversion from baseline in the two groups. The response variable (seroconversion) was a binary variable (1 if yes and 0 otherwise); the fixed effects were time, group, and time×group interaction; and study participant was a random effect. The model accounted for within-subject correlation in outcomes using the appropriate covariance structure, as determined by the quasi-information criteria (QIC).

Univariate and multivariate logistic regression analyses were performed to identify prognostic factors that were independently related to seroconversion.

Intention to treat (ITT) analysis was performed to evaluate treatment safety and efficacy. For the continuous variables included in the efficacy evaluation, the last observation carried forward (LOCF) method was used to impute missing values related to the treatment period. In the evaluation of the binary efficacy endpoints, missing values were considered as failures; for subjects who restarted concomitant oral NA due to HBV DNA elevation (viral breakthrough), the same rule was applied. Sensitivity analyses were used to assess the robustness of the findings from the primary analysis. In particular, post-hoc sensitivity analyses on the primary endpoint were provided, with presentation of the corresponding results. All statistical analyses were carried out using SPSS version 26.0, and a p value less than 0.05 was considered statistically significant.

Results

Baseline characteristics of patients

We screened 174 patients for eligibility, and randomized 149 patients, 75 to the PegIFN alfa-2a group and 74 to the NA group (Table 1). Overall, the mean age was 46.26 ± 10.68 years, 75.8% of the patients were male, and the two groups had similar clinical characteristics at baseline. The most common antiviral agent in both groups was entecavir, followed by lamivudine plus adefovir. Seven patients in the PegIFN alfa-2a group and 5 in the NA group dropped out before week 48; five patients the PegIFN alfa-2a group dropped out due to adverse events considered to be related to this drug (Fig 1).

Table 1

Baseline characteristics of patients in the two groups*.

		Group
Variable	Overall (n = 149)	PegIFNα-2a (n = 75)	NA (n = 74)
Age (years)	46.26±10.68	45.29±9.39	47.24±11.82
Gender
Male	113 (75.8%)	60 (80.0%)	53 (71.6%)
Body mass index, kg/m 2	23.82±3.24	23.67±3.28	24.04±3.20
Liver cirrhosis	17 (11.4%)	6 (8.0%)	11 (14.9%)
Duration of previous NA treatment (months)	72.89±40.16	77.36±45.48	68.37±33.64
HBsAg (log 10 IU/mL)	3.50±0.53	3.50±0.55	3.49±0.51
HBV DNA (log 10 IU/mL)	0.07±0.34	0.06±0.32	0.09±0.37
AST (U/L)	24.67±8.68	23.85±8.38	25.49±8.95
ALT (U/L)	26.97±20.66	26.12±22.60	27.82±18.63
Previous antiviral agents
Entecavir	66 (44.3%)	28 (37.3%)	38 (51.4%)
Lamivudine	13 (8.7%)	4 (5.3%)	9 (12.2%)
Tenofovir	9 (6.0%)	2 (2.7%)	7 (9.5%)
Lamivudine + Adefovir	20 (13.4%)	12 (16.0%)	8 (10.8%)
Entecavir + Adefovir	16 (10.7%)	14 (18.7%)	2 (2.7%)
Entecavir + Tenofovir	14 (9.4%)	11 (14.7%)	3 (4.1%)
Lamivudine + Tenofovir	11 (7.4%)	4 (5.3%)	7 (4.1%)

*Data are presented as mean±SD or number (%).

NA, nucleos(t)ide analogues; PegIFNα-2a, peginterferon α-2a; HBsAg, Hepatis B surface Antigen; AST, aspartate transaminase; ALT, alanine transaminase.

Fig 1

CONSORT flow diagram.

PegIFN alfa-2a, peginterferon alfa-2a.

Change of HBsAg titer during 48 weeks of treatment

Baseline measurements (Fig 2A and S1 Table) indicated the mean log10 HBsAg level was similar in the PegIFN alfa-2a group (3.50±0.55 log10 IU/mL) and the NA group (3.49±0.51 log10 IU/mL). At week 12, these levels were not different from baseline (p = 0.103), but they were lower than baseline in both groups after week 24 (all p < 0.001). Notably, the amount of HBsAg decline was different between the groups (p = 0.033), and the interaction between group and time was significant (p = 0.033; S1 Table). Post-hoc analyses indicated a significant decline of HBsAg after week 24 in the PegIFNα-2a group, and a statistically significant but smaller decrease in the NA group at these times. These results indicated that the decline of the HBsAg titer was much greater in the PegIFN alfa-2a group than in the NA group (Fig 2B).

Fig 2

Changes of HBsAg level (A), HBsAg reduction (B), HBeAg seroconversion (C) and HBeAg loss (D) during treatment between groups. HBsAg, hepatitis B surface antigen; HBeAg, hepatitis B e Antigen; PegIFN alfa-2a, peginterferon alfa-2a. * p < .05, ** p < .01, *** p < .001.

In comparing PegIFNα-2a with NA, we also performed an ITT analysis using LOCF for missing data, ITT analysis without LOCF, and a complete case analysis ignoring the missing data (S2 Table). A complete case analysis is a less conservative approach than ITT analysis. These results showed the effects of PegIFNα-2a on HBsAg were the same in all three analyses.

Serologic response during 48 weeks of treatment

The percentage of patients with HBeAg seroconversion at week 48 was higher in the PegIFN alfa-2a group (15/75 [20.0%] vs. 5/74, [6.8%], p = 0.018; Fig 2C). The percentage of patients with seroconversion also increased over time (p < 0.001) and differed between the two groups (p = 0.015, S1 Table). Post-hoc analyses indicated that the significant improvement in seroconversion was related to PegIFNα-2a treatment. The percentage of patients with seroconversion was significantly greater at week 36 week than at week 12 in the PegIFNα-2a group, but not in the NA group.

There were increases in the percentages of patients with loss of HBeAg during the four assessment times (all p < 0.001), but there were no differences between the two groups (p = 0.283; S1 Table and Fig 2D). One patient in the PegIFN alfa-2a group had HBsAg at loss week 48, but no patients in the NA group had HBsAg loss.

HBV DNA elevation during 48 weeks of treatment

Baseline measurements (Fig 3 and S3 Table) indicated the mean HBV DNA level was similar in the PegIFN alfa-2a group (0.06±0.32 log10 IU/mL) and the NA group (0.09±0.37log10 IU/mL). From week 12, HBV DNA level was much greater in the PegIFN alfa-2a group than in the NA group (all p < 0.001). During the 48 weeks of treatment, the percentage of patients in the PegIFN alfa-2a group who newly developed an HBV DNA level of 2000 IU/mL or more was 1.3% at week 12, 24% at week 24, 12% at week 36, and 8% at week 48 (Table 2). The percentage of patients in the NA group with an HBV DNA level of 2000 IU/mL or more was 1.4% at week 12, and 0% at all other times. Analysis of HBV DNA levels of 20 IU/mL or more and of 20,000 IU/mL or more indicated similar between-group differences (Table 2).

Fig 3

Changes in serum HBV DNA levels in the two groups.

PegIFN alfa-2a, peginterferon alfa-2a. ** p < .01, *** p < .001.

Table 2

HBV DNA elevation during 48 weeks of treatment.

	Group			Group
Variable	PegIFNα-2a (n = 75) *	NA (n = 74) *	p	PegIFNα-2a (n = 75) ¶	NA (n = 74) ¶	p
HBV DNA≥2,000 IU/mL
baseline	0 (0.0)	0 (0.0)	1.000	0 (0.0)	0 (0.0)	1.000
12 weeks	1 (1.3)	1 (1.4)	1.000	1 (1.3)	1 (1.4)	1.000
24 weeks	18 (24.0)	0 (0.0)	<0.001	18 (24.0)	0 (0.0)	<0.001
36 weeks	9 (12.0)	0 (0.0)	0.003	21 (28.0)	0 (0.0)	<0.001
48 weeks	6 (8.0)	0 (0.0)	0.028	24 (32.0)	0 (0.0)	<0.001
HBV DNA≥20 IU/mL
baseline	3 (4.0)	4 (5.4)	0.719	3 (4.0)	4 (5.4)	0.719
12 weeks	13 (17.3)	3 (4.1)	0.026	15 (20.0)	4 (5.4)	0.012
24 weeks	30 (40.0)	1 (1.4)	<0.001	38 (50.6)	4 (5.4)	<0.001
36 weeks	10 (13.3)	3 (4.1)	0.045	38 (50.6)	5 (6.7)	<0.001
48 weeks	6 (8.0)	1 (1.4)	0.116	42 (56.0)	3 (4.1)	<0.001
HBV DNA≥20,000 IU/mL
baseline	0 (0.0)	0 (0.0)	1.000	0 (0.0)	0 (0.0)	1.000
12 weeks	0 (0.0)	1 (1.4)	0.497	0 (0.0)	1 (1.4)	1.000
24 weeks	13 (17.3)	0 (0.0)	<0.001	13 (17.3)	0 (0.0)	0.001
36 weeks	3 (4.0)	0 (0.0)	0.245	11 (14.7)	0 (0.0)	0.001
48 weeks	10 (13.3)	0 (0.0)	0.001	19 (25.3)	0 (0.0)	<0.001

*Values indicate only patients who newly developed HBV DNA elevation at each time point; patients who developed HBV DNA elevation at a previous time point are not included.

¶ Values indicate the proportion of patients with HBV DNA level >20 IU/mL, 2000 IU/mL and 20,000 IU/mL at each time point, excluding cases with follow-up loss or with rescue NA retreatment.

NA, nucleos(t)ide analogues; PegIFNα-2a, peginterferon α-2a.

In addition, among patients in the PegIFN alfa-2a group who experienced HBV DNA elevation (mostly above 2000 IU/mL), 1.3% started concomitant NA at week 12, 9.3% started concomitant NA at week 24, 4% started concomitant NA at week 36, and 32% started concomitant NA at week 48 (Table 3).

Table 3

Retreatment with nucleos(t)ides analogues in the PegIFN alfa-2a group.

NA retreatment	PegIFNα-2a (n = 75)†	Incidence*	95% CI
baseline	0 (0.0%)	0.0%	0.0–0.0%
12 weeks	1 (1.3%)	1.3%	0.0–3.8%
24 weeks	7 (9.3%)	10.7%	3.6–17.8%
36 weeks	3 (4.0%)	14.7%	6.7–22.7%
48 weeks	24 (32.0%)	46.7%	35.3–58.1%

† Values indicate noncumulative numbers and percentages.

*Cumulative percentage.

NA, nucleos(t)ide analogues; PegIFNα-2a, peginterferon α-2a; CI, confidence interval.

ALT level during 48 weeks of treatment

Analysis of adverse events indicated 77 patients experienced 86 events of ALT elevation (62 in the PegIFN alfa-2a group and 24 in the NA group) during the 48 weeks of treatment (Table 4). Among these 86 events, 15 in the PegIFN alfa-2a group and 1 in the NA group appeared to be associated with viral breakthrough, based on the transient HBV DNA elevation that followed ALT elevation. Analysis of these 16 incidents indicated the ALT elevation in the PegIFN alfa-2a group was 1–5×ULN (12 incidents), 5–10×ULN (2 incidents), and more than 10×ULN (1 incident); the one incident in the NA group was an elevation of more than 10×ULN. Forty-seven incidents of ALT elevation in the PegIFN alfa-2a group occurred within 3 months after beginning PegIFN alfa-2a in patients who had sustained HBV DNA suppression due to 3 months of concomitant NA treatment. Analysis of these 47 incidents indicated the ALT elevation was 1–5×ULN (45 incidents) and 5–10×ULN (2 incidents). Twenty-three incidents of ALT elevation in the NA group were not associated with virologic rebound or breakthrough, and the elevation in these 23 incidents was 1–5×ULN.

Table 4

Rates of adverse events, dose modification, and withdrawal*.

Adverse event	PegIFN alfa-2a (n = 75)	NA (n = 74)	P
Discontinuation
For safety reasons¶	10 (13.3)	0 (0.0)	<0.001
For other reasons	2 (2.7)	5 (6.8)	0.276
Dose modification
Total	5 (6.7)	0 (0.0)	0.058
Adverse events	0 (0.0)	0 (0.0)	-
Laboratory abnormality	5 (6.7)	0 (0.0)	0.058
≥1 adverse event	64 (85.3)	28 (37.8)	<0.001
≥1 serious adverse event	3(4.0)	2 (1.4)	1.000
Death	0 (0.0)	0 (0.0)	-
ALT flare due to CHB exacerbation†	3 (18.7)	1 (1.4)	0.620
Lung cancer	0 (0.0)	1 (1.4)	0.497
Maximum ALT level
<ULN	24 (32.0)	50 (67.6)	<0.001
1–5×ULN	56 (74.6)	23 (31.1)	<0.001
5–10×ULN	5 (6.7)	0 (0.0)	0.058
>10×ULN	1 (1.3)	1 (1.4)	1.000

*Values indicate non-cumulative numbers and percentages.

¶ Decreased visual acuity (n = 1), pneumonia (n = 1), depression (n = 2), insomnia (n = 2), shortness of breath (n = 1), general ache (n = 1), weight loss (n = 1), itching sensation (n = 1)

† Defined as ALT>5×ULN.

CHB, chronic hepatitis B; ALT, alanine transaminase; ULN, upper limit of normal; PegIFNα-2a, peginterferon α-2a; NA, nucleos(t)ide analogues

Safety

The rates of adverse events were significantly higher in the PegIFN alfa-2a group than in the NA group but the rate of serious adverse events were similar between the two groups (Table 4). Dose modification and discontinuation due to safety issues during treatment only occurred in the PegIFN alfa-2a group. Ten patients in the PegIFN alfa-2a group discontinued treatment for safety reasons; 5 patients permanently discontinued this treatment due to depression (n = 2), decreased visual acuity (n = 1), weight loss (n = 1), or itching sensation (n = 1). Six patients in the PegIFN alfa-2a group experienced ALT flares (>5×ULN), three within 12 weeks after beginning PegIFN alfa-2a treatment, and spontaneous resolution without dose reduction or discontinuation occurred in all 3 patients. The ALT flares in the other 3 patients, including 1 with a very high ALT level (>10×ULN), occurred after HBV DNA elevation; resolution occurred in all 3 patients after the addition of NA with concomitant PegIFN alfa-2a. One patient in the NA group who received 0.5 mg/day entecavir experienced an ALT flare that occurred after virologic breakthrough at week 12. This ALT flare resolved after addition of tenofovir (300 mg/day).

The rates of neutropenia and thrombocytopenia during antiviral treatment were higher in the PegIFN alfa-2a group (Table 5). Most of these incidents in the PegIFN alfa-2a group resolved after dose reduction, except for one patient who had concomitant pneumonia. This patient, who developed pneumonia with grade-3 neutropenia (520 cells/mm3) after week 24, discontinued PegIFN alfa-2a for 1 week, and resumed treatment after recovery from the pneumonia. This patient completed 48 weeks of PegIFN alfa-2a treatment without further adverse events.

Table 5

Incidence of adverse events*.

	PegIFN alfa-2a (n = 75)	NA (n = 74)
Adverse event leading to discontinuation
Weight loss	1 (1.3)	0
Insomnia	2 (2.6)	0
General ache	1 (1.3)	0
Depression	2 (2.6)	0
Shortness of breath	1 (1.3)	0
Decreased visual acuity	1 (1.3)	0
Pneumonia	1 (1.3)	0
Itching sensation	1 (1.3)	0
Any adverse events
ALT elevation	62 (82.6)	24 (32)
Flu-like symptom	23 ()	0
Neutropenia	9 (12)	0
Alopecia	5 (6.7)	0
Headache	2 (2.6)	0
Fatigue	4 (5.3)	0
Dyspepsia	2 (5.3)	0
Itching sensation	3 (5.3)	0
Depression	1 (1.3)	0
Acute gastroenteritis	0	1 (1.3)
Hypertension	3 (2.6)	1 (1.3)
Diabetes	0	1 (1.3)
Hypothyroidism	1 (1.3)	0
Ecchymosis	1 (1.3)	0
Chest discomfort	1 (1.3)	0
Thrombocytopenia	2 (2.6)	0
Gingiva pain	1 (1.3)	0
Pulmonary tuberculosis	0	1 (1.3)
Upper respiratory infection	2 (2.6)	0
Eczema	1 (1.3)	0

*Values are given as noncumulative number and percentage.

PegIFNα-2a, peginterferon α-2a; NA, nucleos(t)ide analogues

Discussion

This prospective, randomized, controlled study enrolled patients who were HBeAg-positive and did not experience HBeAg loss or seroconversion despite prior long-term NA treatment. We compared the on-treatment efficacy and safety of patients who switched to PegIFN alfa-2a with patients who continued NA for 48 weeks. Although several previous studies evaluated the effects of the combined use of NA and PegIFN alfa-2a, the optimal combination regimen remains unknown. Early trials indicated that de novo concurrent combinations did provide clinical benefit [12, 17], whereas addition of or switching to PegIFN alfa-2a therapy provided greater efficacy than NA alone in patients who had suppressed HBV DNA following NA treatment [18, 20]. However, these studies were performed in highly selected patients and reported efficacy in populations that had specific characteristics. The present study investigated the effect of switching to PegIFN alfa-2a by patients who received various NA maintenance therapies, but remained HBeAg-positive. Therefore, the results of our study are likely more relevant for CHB patients in actual clinical settings.

The results of this study showed that, compared with continuing NA for 48 weeks, switching to PegIFN alfa-2a for 48 weeks significantly reduced the HBsAg titer and increased HBeAg seroconversion in patients who achieved virological suppression following oral NA treatment. We used change in HBsAg concentration from baseline as the primary endpoint, because HBsAg concentration correlates with the level of intrahepatic covalently closed circular HBV DNA in HBeAg-positive patients [21]. The aim of this study was to evaluate the extent of HBsAg decline due to PegIFN alfa-2a therapy for a definite duration in patients who had CHB, were HBeAg-positive, and previously received long-term NA therapy. A significant decline of HBsAg can induce a durable viral response and ultimately a functional cure. We found that the maximal mean decline in HBsAg level from baseline to week 48 was 0.50 ± 0.88 log10 IU/mL in the PegIFN alfa-2a group, but only 0.08 ± 0.46 log10 IU/mL in NA group, a statistically significant difference. Another study reported that the mean decline in HBsAg level during 48 weeks of treatment with PegIFN alfa-2a plus entecavir was about 1.0 log10 IU/mL, and that this decline persisted during follow-up when patients received entecavir alone [20]. However, most Korean CHB patients are infected with HBV genotype C, which has a poorer response to PegIFN alfa-2a than other genotypes [22, 23]. Furthermore, all of our enrolled patients were HBeAg-positive, despite long term NA treatment (mean duration > 5 years), making them a difficult-to-treat population. These factors may explain the smaller decline of HBsAg level in our patients.

At week 48, the rate of HBeAg seroconversion in our PegIFN alfa-2a group was 20%, but it was only 6.8% in our NA group (p = 0.018). A previous study that also examined the effect of switching therapy showed that the HBeAg seroconversion rate was 14.9 to 21% in a PegIFN alfa-2a group [18, 20]. Another study of treatment-naive HBeAg-positive CHB patients reported the HBeAg seroconversion rate was 29 to 32% after 48 weeks of PegIFN alfa-2a treatment and 12 to 22% after 1 year of various oral NA therapies [24]. Our study population consisted of difficult-to-treat patients (low HBeAg seroconversion rate following NA), and the rate of HBeAg seroconversion in our patients was comparable to those of previous studies. However, head-to-head comparisons are not possible due to differences in study design and population characteristics.

There were similar rates of adverse events in our study and previous studies. Although most of the patients in our PegIFN alfa-2a group finished the 48-week course of treatment without severe adverse events, there were 10 severe adverse events, and 5 patients required permanent discontinuation due to safety issues. In addition, ALT elevation was more common in our PegIFN alfa-2a group (62/75 vs. 24/74). Among the 62 events of ALT elevation in our PegIFN alfa-2a group, 47 were associated with immunologic reactions to PegIFN alfa-2a. Previous research reported these ALT flares occurred in patients who responded to PegIFN alfa-2a, and were associated with immunologic clearance of HBV [25]. In the present study, ALT elevation in the PegIFN alfa-2a group was associated with a decline of HBsAg at week 48 (S4 Table), but not with HBeAg seroconversion at that time. The incidence of ALT elevation associated with breakthrough in our PegIFN alfa-2a group was significantly higher than reported in previous studies [18]. In addition, the rate of viral breakthrough was also slightly higher in our study, perhaps because all of our patients were HBeAg-positive at the time of enrollment and NA treatment was stopped after 3 months of PegIFN alfa-2a treatment. However, NA add-on treatment led to good control of viral breakthrough, and PegIFN alfa-2a was continued until week 48 without interruption.

This study had several limitations. HBV genotype could not be determined because the HBV DNA levels at baseline were too low. Thus, the effect of genotype on response to PegIFN alfa-2a could not be assessed. However, 99~100% of Korean patients infected with HBV are infected with genotype C [22], which is associated with more aggressive liver disease and poorer response to antiviral therapy than genotype B [23, 26, 27]. This might explain the smaller decline of HBsAg titer in our patients. Second, adding on PegIFN alfa-2a (rather than switching) could lead to higher response rates in similar populations without viral breakthrough, suggesting that inclusion of a PegIFN alfa-2a add-on group may have provided better information. Because this study was designed to investigate whether PegIFN alfa-2a alone could induce durable responses in patients with well-suppressed HBV DNA following long term NA treatment, we analyzed PegIFN alfa-2a switching rather than add-on. Third, the time point for rescue NA treatment in both groups could not be predefined because the need for adding NA as a rescue medication was not predictable. Fewer than half of our patients had restarted NA treatment by week 48. However, the impact on mean HBsAg level would be minor, because most of them restarted at week 48. Furthermore, adding NA as recue therapy did not affect mean HBsAg level, as shown in the S5 Table. Fourth, we collected 2 years of follow-up data after the 48 weeks of PegIFN alfa-2a switching therapy, as shown in the protocol. However, not all of the follow-up data were available at the time of writing. As a result, the secondary endpoint in this study was not exactly the same as described in the protocol. We will present these data soon. Fifth, because the primary endpoint was a continuous variable and not a dichotomous variable, it was difficult to account for loss-to-follow-up in the primary efficacy analysis. Thus, we used the LOCF approach for missing values in the continuous variables included in the efficacy evaluation. In addition, we performed sensitivity analysis to assess the robustness of the findings from the primary analysis. Finally, patients in this study had been treated with various types of oral NA and more than 10% of patients were treated with lamivudine-based regimens that are no longer used as first line NA. However, the use of previous types of oral NA did not affect the rate of viral breakthrough or HBeAg seroconversion (S7 Table) or the level of HBsAg decline (S4 Table).

In conclusion, this study of CHB patients with suppressed HBV DNA but with HBeAg positivity following NA treatment showed that switching to 48 weeks of PegIFN alfa-2a treatment, rather than continuing NA therapy, significantly reduced the HBsAg titer and increased rate of HBeAg seroconversion. However, HBV DNA was frequently elevated in patients who received PegIFN alfa-2a, thus warranting use of NA as an add-on treatment.

CONSORT 2010 checklist of information to include when reporting a randomised trial*.

(DOC)

Click here for additional data file.

Outcome variables at each assessment time in the two groups.

(DOCX)

Click here for additional data file.

Outcomes at each assessment time in the two groups based on ITT with LOCF, ITT without OCF, and complete case analysis.

(DOCX)

Click here for additional data file.

(DOCX)

Click here for additional data file.

Predictors of a decline in HBsAg by 0.5 log10 IU/mL or more at week-48 in all 149 patients.

(DOCX)

Click here for additional data file.

HBsAg levels in the 75 patients who received PegIFN alfa-2a alone or PegIFN alfa-2a with NA retreatment.

(DOCX)

Click here for additional data file.

Achievement of two different combined end-points by patients in the two groups.

(DOCX)

Click here for additional data file.

Prognostic factors associated with HBeAg seroconversion at week-48 in all 149 patients.

(DOCX)

Click here for additional data file.

Achievement of two different end-points by patients in the two groups.

(DOCX)

Click here for additional data file.

Changes of HBsAg level (A), HBsAg reduction (B), and HBeAg seroconversion (C) in patients with ALT elevation due to viral breakthrough or due to use of PegIFN alfa-2a. HBsAg, Hepatis B surface Antigen; ALT, alanine aminotransferase; HBeAg, hepatitis B e Antigen; PegIFNα-2a, peginterferon α-2a.

(TIF)

Click here for additional data file.

Change in ALT level (A) during treatment for the PegIFN and NA groups and changes in HBsAg level (B) and ALT level (C) according to HBeAg seroconversion at week 48. HBsAg, Hepatis B surface Antigen; ALT, alanine aminotransferase; NA, nucleos(t)ide analogues; PegIFNα-2a, peginterferon α-2a. * p < .05, *** p < .001.

(TIF)

Click here for additional data file.

(XLSX)

Click here for additional data file.

(PDF)

Click here for additional data file.

(PDF)

Click here for additional data file.

21 Feb 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Jee-Fu Huang, M.D., Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for submitting your clinical trial to PLOS ONE and for providing the name of the registry and the registration number. The information in the registry entry suggests that your trial was registered after patient recruitment began. PLOS ONE strongly encourages authors to register all trials before recruiting the first participant in a study.

As per the journal’s editorial policy, please include in the Methods section of your paper:

1) your reasons for your delay in registering this study (after enrolment of participants started);

2) confirmation that all related trials are registered by stating: “The authors confirm that all ongoing and related trials for this drug/intervention are registered”.

3. Thank you for stating the following in the Competing Interests section:

I have read the journal's policy and the authors of this manuscript have the following competing interests: [JH received a grant from Roche Korea, HYW, WYT, HJL, WJC, JGP, YJP, YRL, SYP, YOK and JSH report no conflicts of interest.].

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

4. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

5. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The present study aims to investigate change in HBsAg titer in chronic hepatitis B patients switching to PegIFN alfa-2a after long term NA therapy. This is a well design, randomized trial. However, this issue had been well studied before and the present study does not provide any novel finding. Comments to this study list below.

1. The inclusion criteria included undetectable HBV viral load, which was defined as HBV DNA ≤100 IU/mL, for at least 12 months. HBV DNA of 100 IU/mL as undetectable is relatively too high. A lower level of HBV DNA would be more indicated. Besides, the efficacy analysis of undetectable HBV viral load was defined as HBV DNA ≤20 IU/mL. (Line 204) Can authors explain why a different definition of undetectable HBV viral load? And I suggest a uniform definition for this.

2. As we know, treatment response may develop even after stopping PegIFN therapy. Did authors analyze post-IFN response, like change of HBsAg titer/HBsAg loss after stopping PegIFN?

3. A large proportion of patients were treated with lamivudine-based regimens which was no more used as first line NA. This would be one of the major limitations. And for those developed viral breakthrough, were they developed YMDD mutation?

4. Table 1 - Please check HBV DNA data and confirm data accuracy.

5. How many patients had HBsAg loss? And whether a difference of HBsAg loss between 2 groups?

6. Table 4 – at 48 weeks, only 6 (8.0%) of patients in PegIFN group had HBV DNA ≥20 and 2000 IU/mL, but 10 (13.3%) patients had HBV DNA ≥20000 IU/mL. Is the data correct? Authors should check the data again.

7. Table 6 - Table format needs modification for a better reading.

8. Table 7 – the number of ALT elevation in PegIFN group was 62 (82.6%), but the same ALT elevation showed 51 (68.0%) in table 6. Authors should explain this inconsistent data and confirm which one is correct.

Reviewer #2: Woo et al compared the HBsAg levels and decline and HBeAg seroconversion rates in patients with HBeAg-positive CHB who had been virally suppressed with nucleos(t)ide analogues (NAs) and switched to PegIFN alfa-2a or continued NA for another 48 weeks in a prospective, randomized multi-center study. They demonstrated that PegIFN alfa-2a treatment yielded a significantly greater decline in HBsAg levels and a significantly higher rate of HBeAg seroconversion both at 24, 36 and 48 weeks of treatment. Patients had relatively good tolerance to PegIFN alfa-2a treatment. This study provides valuable data on the extent of HBsAg decline and HBeAg seroconversion rate in patients with HBeAg-positive CHB who started NA therapy and already achieved HBeAg seroconversion with undetectable HBV DNA and switch to PegIFN alfa-2a with an aim to achieving HBsAg loss. There are several issues the authors need to address to improve the scientific merit of the manuscript.

1. Page 7, line 160: undetectable HBV DNA defined as HBV DNA ≤100 IU/mL versus Page 9, line 204: HBV DNA <20 IU/mL. Please clarify.

2. Page 7, line 168- Page 8, line 180 on “exclusion criteria”: some of the sentences look like being copied from the informed consent form. Please correct.

3. Table 2: Are HBsAg levels and reductions at 12, 24, 36 and 48 weeks after switching to PegIFN alfa-2a derived from all 75 patients? Because a significant proportion of patients received NA retreatment (1.3%, 9.3%, 4.0% and 32% at 12, 24, 36 and 48 weeks, respectively, Table 5), it is worthwhile to further address whether there is a difference in HBsAg levels at various time points between patients who received PegIFN alfa-2a alone and those who received NA retreatment. This information may unveil the impact of NA retreatment on HBsAg decline in case of viral breakthrough. Please also discuss this finding as appropriate.

4. Table 4: At 48 weeks of PegIFN alfa-2a treatment, there were 6 (8.0%) patients with HBV DNA ≥20 IU/mL or ≥2000 IU/mL. Why were there 10 (13.3%) patients with HBV DNA ≥20000 IU/mL? Please clarify.

5. Table 5: 24 (32%) patients received NA retreatment at 48 weeks of PegIFN alfa-2a treatment. Why? What are the indications?

6. Page 20, line 364- Page 21, line 373: 15 incidents of ALT elevation occurred following viral breakthrough whereas 47 incidents of ALT elevation occurred during the first 3 months of PegIFN alfa-2a treatment while the patients were still receiving NA therapy. Please compare HBsAg levels and reductions and HBeAg seroconversion rates across treatment course between these two subgroups of patients and present the data at least in a supplementary figure, similar to Figure 2.

7. Table 6: 23 (31.1%) patients had maximal ALT levels 1-5x ULN in the NA group, which was not attributed to viral breakthrough. What are the causes of these ALT elevations? Please clarify and discuss.

8. Some typos and grammar mistakes. This manuscript needs English editing.

Reviewer #3: This is an important RCT that provides the information that switching to PegIFN from NA is beneficial for HBeAg-positive patients who received NA treatment for 72 weeks at least. The switch group has a higher HBeAg seroconversion rate and a more prominent HBsAg reduction.

I have a few comments as follows

1. The English writing could be improved. For example, the authors do not need to emphasize “significant” when doing some comparison. It is clear to achieve a statistic difference when the p-value is present. In addition, the following sentence present in the exclusion criteria is inappropriate “if you have been given an immunomodulatory/immunosuppressant within 6 months prior to registration “

2. It is better to report the data of combined endpoint, for example HBeAg loss + viral load < 2000 IU/mL

3. Previous data has shown that HBsAg level <100 IU/mL is associated with higher HBsAg loss in patients with spontaneous HBeAg seroconversion.1 Please define HBeAg seroconversion + HBsAg <100 IU/mL as another combined endpoint and present the data about the patient number achieving this endpoint in each group.

Reference

1. Tseng TC, Liu CJ, Su TH, et al. Serum hepatitis B surface antigen levels predict surface antigen loss in hepatitis B e antigen seroconverters. Gastroenterology 2011;141:517-525 e512.

Reviewer #4:

# Major Issues

## Introduction

The concept of *durable* viral response appears twice in the first two sentences of the abstract and elsewhere in the text yet is *not* defined. This must be corrected. What do the authors mean by a "durable" viral response? Is there a standard or accepted definition for this phenomenon?

## Methods

The description of the inclusion and exclusion criteria need to be revised. There are many inconsistencies with tenses (some verbs use the present tense while others use the past tense), person (some statements use the second person and others use the third), voice (some statements use the passive voice while others use the active voice), and clause structure (some a clauses while others are complete sentences). The authors need to be consistent and apply grammatical rules appropriately.

*IMPORTANT* Lines 181-196 describe the use of medication to reduce ALT flares and in the event of viral breakthrough. However, this only applies to those receiving PegIFN. This is in direct contradiction to the design stated in the objectives (Lines 130-132) where PegIFN was compared to NA for 48 weeks. Formally, the authors are NOT testing PegIFN versus NA, but PegIFN plus NA for 12 weeks then PegIFN with NA as rescue medication for 36 weeks versus NA alone for 48 weeks. You can see how the second description is quite unlike the first. The authors need to be transparent about the two arms. This has to be clear from the title and abstract.

*IMPORTANT* Why is there no commensurate rescue management for viral breakthrough in those receiving NA alone?

*IMPORTANT* Why are the definitions of viral breakthrough dependent on the medication group? What justification is there for this?

*IMPORTANT* In Line 195, the authors state two terms -- viral breakthrough and viral rebound. Only the former is defined. The latter needs to be defined, too.

*IMPORTANT* There is a major discrepancy between the primary outcome as stated in the protocol compared to the outcome defined in the manuscript. In the protocol, the primary endpoint is defined as "Primary endpoint: Changes in HBsAg quantity (log10 HBsAg) during administration of antiviral

agents in each group".

*IMPORTANT* There appears to be discrepancies between the secondary endpoints as stated in the protocol and the ones reported here:

1. The protocol states that secondary endpoint 1 is changes from baseline in serum HBV DNA levels and HBV DNA non-detection rates and HBV DNA <20 IU/mL during administration of antiviral agents in each group and follow-up, comparison of the ratio of HBV DNA <2,000 IU/mL during follow-up and administration of antiviral agents of each group and comparison of the ratio of HBV DNA <20,000 IU/mL during the administration of antiviral agents of each group and follow-up. In contrast, the present manuscript (Lines 204-205) only report the proportions of MBV DNA <20, <2,000 and >20,000 IU/mL.

2. The protocol states that secondary endpoint 2 is HBeAg seroconversion rate and loss rate, in contrast to the manuscript, which only reports HBeAg seroconversion.

3. The manuscript identifies secondary endpoint 3 as HBsAg loss and HBsAg seroconversion at the end of treatment. This appears nowhere in the protocol. The planned secondary endpoint is actually HBsAg loss and HBsAg seroconversion 1 and 2 years after administration/termination of antiviral agents in each group.

4. Secondary endpoint 4 in the protocol is not included in this manuscript.

5. This manuscript includes an examination of predictors of seroconversion, which does not appear in the protocol.

The authors MUST adhere to the protocol. All endpoints must be presented as planned. All analyses that do not appear in the protocol are necessary post-hoc and must be deleted or identified as such. All deviations from the protocol MUST be identified and justified. This is non-negotiable.

*IMPORTANT* The description of the safety endpoints is quite inadequate. Significantly, the authors claim that "serious" adverse events were a criterion for early dropout of participants, but the term "serious" is not defined (although it is in the protocol).

*IMPORTANT* The authors use the "last observation carried forward" (LOCF) method to impute missing values. The LOCF method has been shown many. many times to be of dubious statistical validity. The use of this method has been discouraged due to the biases that may arise. The authors have adopted the method outside of the parameters of the protocol. Thus, the LOCF method should NOT be used here. The authors must perform the analysis without the LOCF method. If they wish to apply the LOCF method, then they can do so under sensitivity analysis. This is non-negotiable.

*IMPORTANT* Both the linear mixed models and the general estimating equation models must be described in sufficient detail to enable replication. The authors need to enhance their description of these methods as important information is missing.

*IMPORTANT* The ethics approval provided by the authors allowed only for the recruitment of 144 patients. In contrast, the authors randomised 149 patients, or five more than they were allowed. The authors need to provide evidence that this modification was submitted to the ethics committee and permission was obtained.

## Results

Lines 281-282: "All patients were... and at randomisation." This sentence is redundant because this status is required for study entry. This can be deleted.

The section headed "Predictors of HBeAg seroconversion at week 48" should be deleted. This is NOT an objective stated in the protocol and is entirely post-hoc.

## Table 1

Line 289 should be deleted and the information should be included in the stub.

The column of p-values should be omitted. It is highly inappropriate to apply formal statistical tests for baseline characteristics in a randomised controlled trial as this is meaningless.

## Table 2

The information here is better presented as a graph.

Line 305 should be deleted and the information should be included in the stub.

Line 313 should be deleted.

*IMPORTANT* The results of the linear mixed model MUST adjust for the baseline level of the outcome. It is not clear if this is the case.

## Table 4

The subheadings are wrong. They must be changed.

## Table 5

Standard deviations should be provided, not standard errors. Alternatively, 95% confidence intervals can be provided in place of standard errors.

## Table 6

The footnote explaining that the chi-square test could not be used is wrong. This is because the data presented are wrong. The authors need to provide the number of patients in each group who experienced at least one adverse event. This number cannot exceed the total number of participants. This must be corrected.

# Minor Issues

The authors need to define all abbreviations and acronyms on first use. HBsAg is defined, but HBeAg is not, for example. They need to check the manuscript for similar occurrences.

The demonstrative pronoun in Line 130 is vague. I read it to mean reference 18. Please improve the sentence to reduce confusion.

The compound adjective in Line 132 is missing a hyphen, as are those in Lines 158-159 and elsewhere. Please review and modify.

What is the limit of detection of HBeAg? That is to say, at what titer of HBeAg is a patient considered HBeAg-positive?

Line 238: replace "numeric" with "continuous"

Line 247: replace "observed" with "observation"

# Recommendation

The clinical trial described in the manuscript contains many unexplained and important deviations from the protocol. I am unable to support the approval of this manuscript for publication in the journal until these issues are corrected.

Reviewer #5: This study by Heo J et al provides important information for how the peg-IFN may benefit on the the HBeAg positive CHB patients who remained HBeAg positive after 5-6 years of NA treatment. This is a well conduct study.

There are few questions remained to be answered before acceptance:

1. When the ALT elevation occurred in the peg-IFN arm? The on-treatment ALT flare is worth to be noted especially when the higher HBeAg seroconversion in the peg-IFN group. Have these patients checked with the autoimmune hepatitis markers to exclude the possibility of IFN related autoimmune issue? It would be great if the ALT kinetics be provided in addition to HBsAg kinetics as well.

2. What's the reason for ALT elevation in the NA treated arm? It is quite surprising that the proportion for ALT > 1X ULN in the NA arn is as high as 32.4%!

3. How many proportion of these two arms patients had reach HBsAg quantification level < 100 by the end of the study? How many proportion reach HBsAg rapid decline by the end of the study, which defined as HBsAg reduction > 0.5 log10IU/mL per year? Since these two endpoints may be the sentinal predictor for subsequent HBsAg loss, it is important to know whether the peg-IFN arm have the higher chance for functional cure.

4. What's the outcome of the dropped out patients? It is also interesting to know what happened when these treated 5-6 years patients be discontinued from antiviral therapy.

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

21 Apr 2022

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

As indicated, our manuscript meets PLOS ONE's style requirements.

2. Thank you for submitting your clinical trial to PLOS ONE and for providing the name of the registry and the registration number. The information in the registry entry suggests that your trial was registered after patient recruitment began. PLOS ONE strongly encourages authors to register all trials before recruiting the first participant in a study.

As per the journal’s editorial policy, please include in the Methods section of your paper:

1) your reasons for your delay in registering this study (after enrolment of participants started);

This study was planned at January 2012 and registered to clinicaltrial.gov at January 2013 after IRB approval. Patient enrollment started from August 2013.

2) confirmation that all related trials are registered by stating: “The authors confirm that all ongoing and related trials for this drug/intervention are registered”.

We confirmed this statement.

3. Thank you for stating the following in the Competing Interests section:

I have read the journal's policy and the authors of this manuscript have the following competing interests: [JH received a grant from Roche Korea, HYW, WYT, HJL, WJC, JGP, YJP, YRL, SYP, YOK and JSH report no conflicts of interest.].

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

We have included an updated competing interests statement in the cover letter.

4. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

We understand and will update the data availability statement.

5. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

As indicated, we provided these data as supplementary information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The present study aims to investigate change in HBsAg titer in chronic hepatitis B patients switching to PegIFN alfa-2a after long term NA therapy. This is a well design, randomized trial. However, this issue had been well studied before and the present study does not provide any novel finding. Comments to this study list below.

1. The inclusion criteria included undetectable HBV viral load, which was defined as HBV DNA ≤100 IU/mL, for at least 12 months. HBV DNA of 100 IU/mL as undetectable is relatively too high. A lower level of HBV DNA would be more indicated. Besides, the efficacy analysis of undetectable HBV viral load was defined as HBV DNA ≤20 IU/mL. (Line 204) Can authors explain why a different definition of undetectable HBV viral load? And I suggest a uniform definition for this.

Thank you for the comment. At the beginning of this trial in 2012, the detection limit of HBV DNA was 100 IU/mL, but the detection limit at present is 10 or 20 IU/mL. We therefore revised the Methods to say “... had an HBV DNA level of 100 IU/mL or less for at least 12 months (the detection limit at study onset in 2012).” We considered the limit of detection to be 20 IU/mL.

2. As we know, treatment response may develop even after stopping PegIFN therapy. Did authors analyze post-IFN response, like change of HBsAg titer/HBsAg loss after stopping PegIFN?

This is a good point. We collected patient data for 2 years after recording the IFN response, but these results are not yet available. We plan to present these data soon.

3. A large proportion of patients were treated with lamivudine-based regimens which was no more used as first line NA. This would be one of the major limitations. And for those developed viral breakthrough, were they developed YMDD mutation?

Patients randomized to the PegIFN alfa-2a group continued taking oral NAs for the first 12 weeks of PegIFN alfa-2a treatment. After that, patients in PegIFN alfa-2a stopped taking oral NAs. The rate of viral breakthrough in the PegIFN alfa-2a group was similar in those who initially used different oral NAs (entecavir: 32.1%, 9/28; tenofovir: 50%, 2/4; lamivudine: 50%, 1/2; lamivudine+adefovir: 50%, 6/12; entecavir+adefovir: 35.7%, 5/14; entecavir+tenofovir: 63.7%, 7/11; lamivudine+tenofovir: 100%, 4/4; P = 0.176) One patient who received lamivudine developed viral breakthrough at week-36, had a YMDD mutation, and received concomitant tenofovir. In patients randomized to the oral NA group, viral breakthrough did not occur in patients using lamivudine alone, and only occurred in 1 patient who received entecavir (0.5 mg).

4. Table 1 - Please check HBV DNA data and confirm data accuracy.

The data are accurate and expressed as log10(IU/mL) of HBV DNA.

5. How many patients had HBsAg loss? And whether a difference of HBsAg loss between 2 groups?

One patient in the PegIFN alfa-2a group had HBsAg loss at week-48. No patients in NA group had HBsAg loss. We added this information to the Results.

6. Table 4 – at 48 weeks, only 6 (8.0%) of patients in PegIFN group had HBV DNA ≥20 and 2000 IU/mL, but 10 (13.3%) patients had HBV DNA ≥20000 IU/mL. Is the data correct? Authors should check the data again.

This is a good point. The data are correct. Six patients who had HBV DNA above 2000 IU/mL but less than 20,000 IU/mL at week-24 and week-36 developed HBV DNA above 20,000 IU/mL at week-48. Rescue oral antiviral therapy for breakthrough was delayed in these patients.

7. Table 6 - Table format needs modification for a better reading.

We revised the Table (now Table 4) as indicated.

8. Table 7 – the number of ALT elevation in PegIFN group was 62 (82.6%), but the same ALT elevation showed 51 (68.0%) in table 6. Authors should explain this inconsistent data and confirm which one is correct.

This is good point. The number 51 was wrong in Table 6, so we made the correction. The number of adverse events was also presented in Table 7 (now Table 5), so we revised Table 6 (now Table 4) by deleting the last column of this table.

Reviewer #2: Woo et al compared the HBsAg levels and decline and HBeAg seroconversion rates in patients with HBeAg-positive CHB who had been virally suppressed with nucleos(t)ide analogues (NAs) and switched to PegIFN alfa-2a or continued NA for another 48 weeks in a prospective, randomized multi-center study. They demonstrated that PegIFN alfa-2a treatment yielded a significantly greater decline in HBsAg levels and a significantly higher rate of HBeAg seroconversion both at 24, 36 and 48 weeks of treatment. Patients had relatively good tolerance to PegIFN alfa-2a treatment. This study provides valuable data on the extent of HBsAg decline and HBeAg seroconversion rate in patients with HBeAg-positive CHB who started NA therapy and already achieved HBeAg seroconversion with undetectable HBV DNA and switch to PegIFN alfa-2a with an aim to achieving HBsAg loss. There are several issues the authors need to address to improve the scientific merit of the manuscript.

1. Page 7, line 160: undetectable HBV DNA defined as HBV DNA ≤100 IU/mL versus Page 9, line 204: HBV DNA <20 IU/mL. Please clarify.

Thank you for the excellent comment, which was also noted by Review #1. At the beginning of this trial in 2012, the detection limit of HBV DNA was 100 IU/mL, but the detection limit at present is 10 or 20 IU/mL. We therefore revised the Methods to say “... had an HBV DNA level of 100 IU/mL or less for at least 12 months (the detection limit at study onset in 2012).” We considered the limit of detection to be 20 IU/mL..

2. Page 7, line 168- Page 8, line 180 on “exclusion criteria”: some of the sentences look like being copied from the informed consent form. Please correct.

As indicated, we modified the exclusion criteria accordingly.

3. Table 2: Are HBsAg levels and reductions at 12, 24, 36 and 48 weeks after switching to PegIFN alfa-2a derived from all 75 patients? Because a significant proportion of patients received NA retreatment (1.3%, 9.3%, 4.0% and 32% at 12, 24, 36 and 48 weeks, respectively, Table 5), it is worthwhile to further address whether there is a difference in HBsAg levels at various time points between patients who received PegIFN alfa-2a alone and those who received NA retreatment. This information may unveil the impact of NA retreatment on HBsAg decline in case of viral breakthrough. Please also discuss this finding as appropriate.

This is a good point. As indicated, HBsAg level and reduction at 12, 24, 36, and 48 weeks were derived from all 75 patients. We presented data about HBsAg level at various times for patients who received PegIFN alfa-2a alone and those who received NA retreatment in a S5 Table.

4. Table 4: At 48 weeks of PegIFN alfa-2a treatment, there were 6 (8.0%) patients with HBV DNA ≥20 IU/mL or ≥2000 IU/mL. Why were there 10 (13.3%) patients with HBV DNA ≥20000 IU/mL? Please clarify.

This is a good point, which was also noted by Reviewer #1. The data are correct. Six patients who had HBV DNA above 2000 IU/mL but less than 20,000 IU/mL at week-24 and week-36 developed HBV DNA above 20,000 IU/mL at week-48. Rescue oral antiviral therapy for breakthrough was delayed in these patients.

5. Table 5: 24 (32%) patients received NA retreatment at 48 weeks of PegIFN alfa-2a treatment. Why? What are the indications?

The indications for restarting NA treatment for patients in the PegIFN alfa-2a group was provided in the Methods: “Patients in the PegIFN alfa-2a group who experienced HBV DNA elevation (viral breakthrough), with or without ALT flare, were allowed to restart concomitant oral NA at the discretion of the researchers. Viral breakthrough in the PegIFN alfa-2a group was defined as an increase in HBV DNA to 2000 IU/mL or more during treatment.”

6. Page 20, line 364- Page 21, line 373: 15 incidents of ALT elevation occurred following viral breakthrough whereas 47 incidents of ALT elevation occurred during the first 3 months of PegIFN alfa-2a treatment while the patients were still receiving NA therapy. Please compare HBsAg levels and reductions and HBeAg seroconversion rates across treatment course between these two subgroups of patients and present the data at least in a supplementary figure, similar to Figure 2.

There was some overlap of patients (n=12) among those with ALT elevation due to viral breakthrough and ALT elevation during the first 3 months of PegIFN alfa-2a treatment. Therefore, we cannot perform a statistical comparison of these groups. Instead, we presented the HBsAg level, HBsAg reduction, and HBeAg seroconversion during treatment in these two subgroups of patients as a S1 Fig.

7. Table 6: 23 (31.1%) patients had maximal ALT levels 1-5x ULN in the NA group, which was not attributed to viral breakthrough. What are the causes of these ALT elevations? Please clarify and discuss.

Although the number of patients with ALT elevation in the NA group was not small, the ALT elevation was mild and not persistent. The causes of mild ALT elevation in the NA group are nonspecific, and may be due to use of a cold medication, excessive exercise, or excessive work.

8. Some typos and grammar mistakes. This manuscript needs English editing.

As indicated, English editing was performed.

Reviewer #3: This is an important RCT that provides the information that switching to PegIFN from NA is beneficial for HBeAg-positive patients who received NA treatment for 72 weeks at least. The switch group has a higher HBeAg seroconversion rate and a more prominent HBsAg reduction.

I have a few comments as follows

1. The English writing could be improved. For example, the authors do not need to emphasize “significant” when doing some comparison. It is clear to achieve a statistic difference when the p-value is present. In addition, the following sentence present in the exclusion criteria is inappropriate “if you have been given an immunomodulatory/immunosuppressant within 6 months prior to registration “

As indicated, English editing was performed.

2. It is better to report the data of combined endpoint, for example HBeAg loss + viral load < 2000 IU/mL

We presented these data in a S6 Table because this is not in the protocol.

3. Previous data has shown that HBsAg level <100 IU/mL is associated with higher HBsAg loss in patients with spontaneous HBeAg seroconversion.1 Please define HBeAg seroconversion + HBsAg <100 IU/mL as another combined endpoint and present the data about the patient number achieving this endpoint in each group.

We presented these data in a S6 Table because this is not in the protocol. However, the number of patients with HBeAg seroconversion plus HBsAg below 100 IU/mL was very small, and these patients did not show HBsAg loss until week-48.

Reference

1. Tseng TC, Liu CJ, Su TH, et al. Serum hepatitis B surface antigen levels predict surface antigen loss in hepatitis B e antigen seroconverters. Gastroenterology 2011;141:517-525 e512.

Reviewer #4:

# Major Issues

## Introduction

The concept of *durable* viral response appears twice in the first two sentences of the abstract and elsewhere in the text yet is *not* defined. This must be corrected. What do the authors mean by a "durable" viral response? Is there a standard or accepted definition for this phenomenon?

Durable viral response means sustained off-treatment response following NA cessation, such as HBs seroclearance and HBeAg seroconversion with undetectable HBV DNA. As indicated, we defined durable viral response in the Introduction as follows: “However, even after long-term NA treatment of these patients, many of them do not achieve a durable viral response (i.e., a sustained response following NA cessation), such as seroclearance of HB surface antigen (HBsAg), seroconversion of HB e-antigen (HBeAg), and undetectable HBV DNA.”

Reference: Liaw YF, Kao JH, Piratvisuth T, Chan HL, Chien RN, Liu CJ, et al. Asian-Pacific consensus statement on the management of chronic hepatitis B: a 2012 update. Hepatol Int. 2012;6:531–561.

## Methods

The description of the inclusion and exclusion criteria need to be revised. There are many inconsistencies with tenses (some verbs use the present tense while others use the past tense), person (some statements use the second person and others use the third), voice (some statements use the passive voice while others use the active voice), and clause structure (some a clauses while others are complete sentences). The authors need to be consistent and apply grammatical rules appropriately.

As indicated, English editing was performed.

*IMPORTANT* Lines 181-196 describe the use of medication to reduce ALT flares and in the event of viral breakthrough. However, this only applies to those receiving PegIFN. This is in direct contradiction to the design stated in the objectives (Lines 130-132) where PegIFN was compared to NA for 48 weeks. Formally, the authors are NOT testing PegIFN versus NA, but PegIFN plus NA for 12 weeks then PegIFN with NA as rescue medication for 36 weeks versus NA alone for 48 weeks. You can see how the second description is quite unlike the first. The authors need to be transparent about the two arms. This has to be clear from the title and abstract.

This is a very good point. However, the main purpose of this study was to evaluate the effect of switching from NA to PegIFN alfa-2a for 48 weeks compared to continuing NA. Adding NA for first 12 weeks was performed to reduce the risk of ALT flare that may be caused by switching to PegIFN alfa-2a, as described in the protocol. In the previous study, NA was also added for first 8 weeks.

The need for adding NA as a rescue medication is not predictable, and only 11 patients needed retreatment prior to week-48; the other 24 patients started NA retreatment at week-48 (after the end of PegIFN alfa-2a treatment). Therefore, we conclude that adding NA as recue therapy did not affect the results, as shown in the S5 Table. Although this is an important detail, adding this information to the Title or Abstract could create confusion.

*IMPORTANT* Why is there no commensurate rescue management for viral breakthrough in those receiving NA alone?

In case of viral breakthrough in the NA group, recue management such as adding a new NA or switching to another NA, was performed according to contemporary HBV guidelines, and considering the mutation profile and patient compliance at the discretion of the researchers. We added this in the Methods.

*IMPORTANT* Why are the definitions of viral breakthrough dependent on the medication group? What justification is there for this?

This is because virologic response of PegIFN alfa-2a is different from virologic response of NA. For PegIFN alfa-2a, virologic response is defined as a decrease in serum HBV DNA to below 2000 IU/mL after 6 months and at the end of therapy. For NA, virologic response is defined as a decrease in serum HBV DNA to an undetectable level based on a real-time PCR assay.

*IMPORTANT* In Line 195, the authors state two terms -- viral breakthrough and viral rebound. Only the former is defined. The latter needs to be defined, too.

This is a good point. Viral rebound does not apply in this study, so we removed this text.

*IMPORTANT* There is a major discrepancy between the primary outcome as stated in the protocol compared to the outcome defined in the manuscript. In the protocol, the primary endpoint is defined as "Primary endpoint: Changes in HBsAg quantity (log10 HBsAg) during administration of antiviral agents in each group".

We have described primary endpoint of this study as the change in log10 HBsAg titer during antiviral therapy. We modified the text to say: “The primary endpoint was the change in HBsAg quantity (log10 HBsAg) during drug administration in each group. ”

*IMPORTANT* There appears to be discrepancies between the secondary endpoints as stated in the protocol and the ones reported here:

1. The protocol states that secondary endpoint 1 is changes from baseline in serum HBV DNA levels and HBV DNA non-detection rates and HBV DNA <20 IU/mL during administration of antiviral agents in each group and follow-up, comparison of the ratio of HBV DNA <2,000 IU/mL during follow-up and administration of antiviral agents of each group and comparison of the ratio of HBV DNA <20,000 IU/mL during the administration of antiviral agents of each group and follow-up. In contrast, the present manuscript (Lines 204-205) only report the proportions of MBV DNA <20, <2,000 and >20,000 IU/mL.

This is a very good point. As indicated, this secondary endpoint was not exactly same as described in the protocol because we do not yet have all the follow-up data. Therefore, we deleted the term “follow-up”. We presented the change of HBV DNA level from baseline during 48 weeks of administration of antiviral agents in each group. Therefore, we corrected this text as follows: “The secondary endpoints were (i) changes in serum HBV DNA level from baseline and the ratio of below 20 IU/mL (the detection limit), below 2000 IU/mL, and below 20,000 IU/mL during drug administration in each group and (ii) change of HBeAg seroconversion and loss during drug administration in each group.

2. The protocol states that secondary endpoint 2 is HBeAg seroconversion rate and loss rate, in contrast to the manuscript, which only reports HBeAg seroconversion.

As indicated, above, we corrected this text.

3. The manuscript identifies secondary endpoint 3 as HBsAg loss and HBsAg seroconversion at the end of treatment. This appears nowhere in the protocol. The planned secondary endpoint is actually HBsAg loss and HBsAg seroconversion 1 and 2 years after administration/termination of antiviral agents in each group.

As described above, we do not yet have all the follow up data. Therefore, we deleted this secondary endpoint.

4. Secondary endpoint 4 in the protocol is not included in this manuscript.

As described above, we do not yet have all the follow up data, so we did not include this secondary endpoint.

5. This manuscript includes an examination of predictors of seroconversion, which does not appear in the protocol.

This was a post-hoc analysis. Although it was not included as an endpoint, it is essential to analyze predictors associated with endpoints. We moved this text to the supplementary section (S7 Table).

The authors MUST adhere to the protocol. All endpoints must be presented as planned. All analyses that do not appear in the protocol are necessary post-hoc and must be deleted or identified as such. All deviations from the protocol MUST be identified and justified. This is non-negotiable.

*IMPORTANT* The description of the safety endpoints is quite inadequate. Significantly, the authors claim that "serious" adverse events were a criterion for early dropout of participants, but the term "serious" is not defined (although it is in the protocol).

This is a very good point. As indicated, it is not serious adverse event but severe adverse event. We added following text in the Methods: “A severe adverse event was defined as an event that led to study drug discontinuation.” And we modified the Table 6 (now Table 4) according to the definition of serious adverse event in the protocol

*IMPORTANT* The authors use the "last observation carried forward" (LOCF) method to impute missing values. The LOCF method has been shown many. many times to be of dubious statistical validity. The use of this method has been discouraged due to the biases that may arise. The authors have adopted the method outside of the parameters of the protocol. Thus, the LOCF method should NOT be used here. The authors must perform the analysis without the LOCF method. If they wish to apply the LOCF method, then they can do so under sensitivity analysis. This is non-negotiable.

Because the primary endpoint was not a dichotomous variable, it is difficult to account for loss-to-follow up in the PegIFN group, such discontinuation of this drug due to adverse events. Thus, we used the LOCF approach. Instead, we have performed sensitivity analysis, as described in the Methods and presented in a supplementary table.

*IMPORTANT* Both the linear mixed models and the general estimating equation models must be described in sufficient detail to enable replication. The authors need to enhance their description of these methods as important information is missing.

As indicated, we modified the Methods section accordingly.

*IMPORTANT* The ethics approval provided by the authors allowed only for the recruitment of 144 patients. In contrast, the authors randomised 149 patients, or five more than they were allowed. The authors need to provide evidence that this modification was submitted to the ethics committee and permission was obtained.

We reported this to the IRB and received permission to enroll 5 more patients.

## Results

Lines 281-282: "All patients were... and at randomisation." This sentence is redundant because this status is required for study entry. This can be deleted.

As indicated, this sentence was deleted.

The section headed "Predictors of HBeAg seroconversion at week 48" should be deleted. This is NOT an objective stated in the protocol and is entirely post-hoc.

This was a post-hoc analysis. Although it was not included as an endpoint, it is essential to analyze predictors associated with the endpoint. We moved this section to the supplementary section (S7 Table).

## Table 1

Line 289 should be deleted and the information should be included in the stub.

We deleted the line and added the following text to the table: “Data are presented as mean±SD or number (%).”

The column of p-values should be omitted. It is highly inappropriate to apply formal statistical tests for baseline characteristics in a randomised controlled trial as this is meaningless.

As indicated, the column of p-values was removed.

## Table 2

The information here is better presented as a graph.

As indicated, this table was presented as a Fig. 2.

Line 305 should be deleted and the information should be included in the stub.

We deleted this line and added the following to the table: “Data are presented as mean±SD or number (%).”

Line 313 should be deleted.

As indicated, this line was deleted.

*IMPORTANT* The results of the linear mixed model MUST adjust for the baseline level of the outcome. It is not clear if this is the case.

As indicated, we explained this in the Methods as follows: “The LMM model included repeated measures of numeric variables as dependent variables; group, time, and group×time interaction as fixed effects; baseline outcome as a continuous covariate; and study participant as a random effect.”

## Table 4

The subheadings are wrong. They must be changed.

We changed the subheadings accordingly.

## Table 5

Standard deviations should be provided, not standard errors. Alternatively,

We have provided 95% confidence intervals in place of standard errors.

## Table 6

The footnote explaining that the chi-square test could not be used is wrong. This is because the data presented are wrong. The authors need to provide the number of patients in each group who experienced at least one adverse event. This number cannot exceed the total number of participants. This must be corrected.

This is a very good point. We did not calculate the number, but the incidence. We corrected the Table 6 (now Table 4) accordingly.

# Minor Issues

The authors need to define all abbreviations and acronyms on first use. HBsAg is defined, but HBeAg is not, for example. They need to check the manuscript for similar occurrences.

As indicated, we defined all abbreviations and acronyms on first use.

The demonstrative pronoun in Line 130 is vague. I read it to mean reference 18. Please improve the sentence to reduce confusion.

As indicated, we changed this expression.

The compound adjective in Line 132 is missing a hyphen, as are those in Lines 158-159 and elsewhere. Please review and modify.

As indicated, we revised these texts.

What is the limit of detection of HBeAg? That is to say, at what titer of HBeAg is a patient considered HBeAg-positive?

The ARCHITECT HBeAg assay detection limit is 0.5 PEI U/mL.

Line 238: replace "numeric" with "continuous"

As indicated, we changed the word.

Line 247: replace "observed" with "observation"

As indicated, we changed the word.

# Recommendation

The clinical trial described in the manuscript contains many unexplained and important deviations from the protocol. I am unable to support the approval of this manuscript for publication in the journal until these issues are corrected.

Reviewer #5: This study by Heo J et al provides important information for how the peg-IFN may benefit on the the HBeAg positive CHB patients who remained HBeAg positive after 5-6 years of NA treatment. This is a well conduct study.

There are few questions remained to be answered before acceptance:

1. When the ALT elevation occurred in the peg-IFN arm? The on-treatment ALT flare is worth to be noted especially when the higher HBeAg seroconversion in the peg-IFN group. Have these patients checked with the autoimmune hepatitis markers to exclude the possibility of IFN related autoimmune issue? It would be great if the ALT kinetics be provided in addition to HBsAg kinetics as well.

We excluded patients with other liver diseases such as autoimmune hepatitis at the time of inclusion and at the time of ALT flare. We also provided HBsAg kinetics and ALT kinetics according to HBeAg seroconversion at week 48 in a S2 Fig From baseline to week 48, HBsAg level was significantly lower in patients who achieved HBeAg seroconversion at week 48 (all p<0.05). ALT tended to be higher in patients who achieved HBeAg seroconversion at week 48, but there was no statistically significant difference.

2. What's the reason for ALT elevation in the NA treated arm? It is quite surprising that the proportion for ALT > 1X ULN in the NA arn is as high as 32.4%!

Although the number of patients with ALT elevation in the NA group was not small, the ALT elevations were mild and not persistent. The possible causes of a mild ALT elevation in the NA group may use of a cold medication or due to excessive exercise or excessive work.

3. How many proportion of these two arms patients had reach HBsAg quantification level < 100 by the end of the study? How many proportion reach HBsAg rapid decline by the end of the study, which defined as HBsAg reduction > 0.5 log10IU/mL per year? Since these two endpoints may be the sentinal predictor for subsequent HBsAg loss, it is important to know whether the peg-IFN arm have the higher chance for functional cure.

In PegIFN alfa-2a group, 26.7% of patients had an HBsAg reduction of at least 0.5 log10 IU/mL at week-48, but this number was only 2.7% in the NA group. Eight patients in PegIFN alfa-2a group had an HBsAg titer below 100 IU/mL at week-48, but this number was only 2 in the NA group. We provided these data in a S8 Table.

4. What's the outcome of the dropped out patients? It is also interesting to know what happened when these treated 5-6 years patients be discontinued from antiviral therapy.

We do not know the outcomes of all patients who dropped out, because some of them did not return to our study center. Most of them were still receiving nucleos(t)ide therapy because there was no evidence of HBsAg seroclearance.

11 May 2022

8 Jun 2022

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

As indicated, we have reviewed our reference list and confirm that it is complete and correct.

Additional Editor Comments (if provided):

Please respond sufficiently to the comments of the Reviewer 5.

Reviewer #5: The authors response make me quite confused than the original drafting.

1. I would suggest to add the limitation in their response to the reviewer into the limitation paragraph.

We have added the limitations mentioned in our response to the reviewer to the limitation paragraph in the manuscript.

2. As for their response to the reviewer 1 point 6 and reviewer 2 point 4, I think the discordance of the patients number and the statement makes me really confused about what happened actually to these patients. Would it be possible to make it clear? (Please find a native English speaker to check the fluency before re-submitting the response to the reviewers).

We are very sorry for the confusion. We did not provide sufficient explanation about Table 4 (now Table 2) in our first response to the reviewer. Table 4 (now Table 2) contained information only on cases with HBV DNA elevation that was newly developed at each time point during the study period. The data did not include cases where HBV DNA elevation developed at a previous time point. Therefore, the number of patients with HBV DNA>20,000 IU/mL could be larger than that of patients with HBV DNA>20 IU/mL or >2,000 IU/mL if the rescue antiviral therapy were delayed. We added the above explanation to the footnote of the table and added an additional column to the table to avoid such confusion. That additional column indicated the proportion of patients with HBV DNA level >20 IU/mL, 2,000 IU/mL and 20,000 IU/mL at each time point, excluding cases lost to follow-up or with rescue NA retreatment. This new column shows that the number of patients is more concordant across each HBV DNA level at each time point.

## We also add the answer to previous comment 1 by reviewer #5

We added data on ALT kinetics during treatment for the PegIFN and NA groups to S2 Fig. Elevated ALT was observed from week 12 to week 48 in the PegIFN group, and the values for this group were significantly higher than those of the NA group (all p<0.001).

Submitted filename: answer to reviewer(1).docx

Click here for additional data file.

16 Jun 2022

Effect of switching from nucleos(t)ide maintenance therapy to PegIFN alfa-2a in patients with HBeAg-positive chronic hepatitis B: A randomized trial

PONE-D-22-01751R2

Dear Dr. Heo,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Jee-Fu Huang, M.D., Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

14 Jul 2022

PONE-D-22-01751R2

Effect of switching from nucleos(t)ide maintenance therapy to PegIFN alfa-2a in patients with HBeAg-positive chronic hepatitis B: A randomized trial

Dear Dr. Heo:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Jee-Fu Huang

Academic Editor

PLOS ONE